GAMBLING AND PUBLIC HEALTH IN GREENLAND – A
LARGE INDIGENOUS POPULATION IN TRANSITION
A study of gambling behavior and problem gambling in relation to social transition, addictive behaviors and health
among Greenland Inuit.
PhD Thesis
Christina Viskum Lytken Larsen
Centre for Health Research in Greenland
January 2014
PhD Thesis
Centre for Health Research in Greenland, National Institute of Public Health, Faculty of Health
Science, University of Southern Denmark
GAMBLING AND PUBLIC HEALTH IN GREENLAND – A LARGE INDIGENOUS
POPULATION IN TRANSITION
A study of gambling behavior and problem gambling in relation to social transition, addictive
behaviors and health among Greenland Inuit.
Christina Viskum Lytken Larsen
January 2014
3
ACKNOWLEDGEMENTS
Without question I am indebted to my husband Malik (I have no words!), my daughter Selma and the
twins (at the time still unborn) who supported me through a busy summer and fall of 2012 and all
the other busy times...
This goes for my closest family as well: My brothers, Thomas and Søren, and their very cool women,
Eva and Nikoline, for moral support, baby sitting and good fun. My parents, Mogens and Marianne,
for intellectual support, lots of love and heroic visits during the last months before deadline, which
included fun times for Selma, emptying the dish washer again and again, doing laundry and tidying
up (no need to end these visits by the way…).
Also, I would like to thank:
•
•
•
•
Cecilia, Inger, Sus, Charlotte, Anni, Marit, Isabelle and Ingelise at the Center for Health
Research in Greenland. For fun times, support and great discussions.
My colleagues at the Institute for Nursing and Health Science, University of Greenland, and
my former colleagues at the Department of Social Work (ISI), University of Greenland, for
teaching me about health and people in Greenland even before I came into the research
business.
The students I had the pleasure to meet at the University of Greenland. I am sure they have
taught me more, than I could actually teach them.
The people at the National Institute of Public Health for an inspiring working environment
and many relevant discussions.
I owe a special thanks to my supervisors, Professor Peter Bjerregaard and PhD Tine Curtis, who have
guided me through many analyses, discussions and methodological considerations during these
years. It has been an inspiring learning process and I look forward to continue our discussions in the
future.
Finally I would like to dedicate this thesis to my beloved and respected grandfather Poul Viskum, a
pioneering medical doctor in his field, who passed away while I was finishing my PhD studies. Among
the many inspiring accomplishments of my grandfather, were medical visits to towns in Midwest
Greenland during the 1950s with the Danish ship Heimdal. These visits were the foundation of many
important discussions along with all the other great discussions we have had during the past 30
years. I will miss you!
4
Outline of thesis
Paper I: Larsen CVL, Curtis T, Bjerregaard P. Gambling behavior and problem gambling reflecting
social transition and traumatic childhood events among Greenland Inuit - a cross-sectional study in a
large indigenous population undergoing rapid change; Journal of Gambling Studies, 2013, 29 (4): 73348.
Paper II: Larsen CVL, Curtis T, Bjerregaard P. Harmful alcohol use and frequent use of marijuana
among lifetime problem gamblers and the prevalence of cross-addictive behaviors among Greenland
Inuit – evidence from the cross-sectional Inuit Health in Transition Greenland Survey 2006-2010;
International Journal of Circumpolar Health, 2013, 72:19551. doi: 10.3402/ijch.v72i0.19551.
Paper III: Larsen CVL, Curtis T, Bjerregaard P. Health status and health behavior associated with
frequent gambling and problem gambling in a large indigenous population – A cross-sectional study
based on The Inuit Health in Transition Greenland Survey 2006-2010. Article in review November
2013
Academic supervisors:
Professor Peter Bjerregaard, MD, DMSs. Centre for Health Research in Greenland, National Institute
of Public Health, University of Southern Denmark.
Professor Tine Curtis, PhD, MA Sociology. Local Government Denmark and National Institute of Public
Health, University of Southern Denmark.
5
List of abbreviations
BBGS
Brief Biosocial Gambling Screen
CI
Confidence Interval
DSM-III
Third edition of the Diagnostic and Statistical Manual
DSM-III-R
Third revised edition of the Diagnostic and Statistical Manual
DSM-IV
Fourth edition of the Diagnostic and Statistical Manual
DSM-IV-R
Fourth revised edition of the Diagnostic and Statistical Manual
DSM-V
Fifth edition of the Diagnostic and Statistical Manual
ICD-9
WHO’s ninth edition of the International Classification of Diseases
ICD-10
WHO’s tenth edition of the International Classification of Diseases
IHIT
Inuit Health in Transition Study
NODS
National Opinion Research Center DSM-IV Screen for Gambling Problems
NODS CLiP
Short version of National Opinion Research Center DSM-IV Screen
OR
Odds ratio
SOGS-R
Revised South Oaks Gambling Screen
6
Contents
ACKNOWLEDGEMENTS ........................................................................................................................... 4
Outline of thesis ...................................................................................................................................... 5
List of abbreviations ................................................................................................................................ 6
1. INTRODUCTION ................................................................................................................................... 9
1.1 Objective ..................................................................................................................................... 10
2. Background........................................................................................................................................ 11
2.1 Concepts and definitions ............................................................................................................. 11
Gambling activities and gambling behavior. ................................................................................. 11
Pathological gambling – a definition ............................................................................................. 12
Problem gambling – widening the concept of pathological gambling .......................................... 13
2.2 Towards a public health perspective on gambling ...................................................................... 14
2.3 The role of gambling and problem gambling in a population undergoing social transition ....... 15
3. MATERIAL AND METHODS ................................................................................................................ 19
3.1 Inuit health in transition across the Arctic .................................................................................. 19
3.2 Sample and participation ............................................................................................................ 21
3.3 Data Collection ............................................................................................................................ 24
3.4 Outcomes and exposures ............................................................................................................ 24
Past year gambling behavior ......................................................................................................... 24
Past year and lifetime problem gambling ..................................................................................... 25
Sociodemographic variables.......................................................................................................... 27
Social transition ............................................................................................................................. 27
Traumatic events during childhood............................................................................................... 28
Addictive behaviors ....................................................................................................................... 28
Health status ................................................................................................................................. 29
Health behavior ............................................................................................................................. 29
3.5 Data analysis and statistical methods ......................................................................................... 30
3.6 Validity and completeness of data source .................................................................................. 30
3.7 Ethical considerations ................................................................................................................. 31
4. RESULTS ............................................................................................................................................. 33
4.1 Overview of results...................................................................................................................... 33
4.2 Paper I.......................................................................................................................................... 33
4.3 Paper II......................................................................................................................................... 34
7
4.4 Paper III........................................................................................................................................ 35
5. DISCUSSION ....................................................................................................................................... 37
5.1 Main findings ............................................................................................................................... 37
Main findings in perspective to other studies............................................................................... 37
5.2 Methodological considerations ................................................................................................... 37
Selection bias................................................................................................................................. 37
Information bias ............................................................................................................................ 38
Confounding and intermediate factors ......................................................................................... 40
Type-I and type II-errors ................................................................................................................ 41
Lessons to learn about measuring problem gambling .................................................................. 41
5.3 Understanding the nature of problem gambling ........................................................................ 42
5.4 Problem gambling in the light of the ongoing social transition .................................................. 44
6. CONCLUSION ..................................................................................................................................... 47
7. IMPLICATIONS ................................................................................................................................... 49
7.1 For research................................................................................................................................. 49
7.2 For public health.......................................................................................................................... 50
8. SUMMARY ......................................................................................................................................... 51
9. RESUMÉ ............................................................................................................................................. 52
10. EQIKKAANEQ ................................................................................................................................... 53
11. REFERENCES .................................................................................................................................... 55
APPENDIX 1 ........................................................................................................................................... 63
APPENDIX 2 ........................................................................................................................................... 67
APPENDIX 3 ........................................................................................................................................... 71
PAPERS I-III ............................................................................................................................................ 73
8
1. INTRODUCTION
During recent years gambling in general and pathological gambling in particular has become a much
debated subject in Greenland. The economic and social problems for individuals, families and
communities which extensive gambling can lead to are of great concern. At the start of the project,
knowledge in the field was limited to a countrywide telephone survey and a qualitative survey
conducted by the Department of Health in Greenland (Paarisa) in 2005 among local public health
coordinators and more knowledge was needed (1). The health coordinators had pointed out
extensive gambling on bingo and card games as the biggest gambling related problem in the
Greenlandic communities and especially bingo was suspected to lead to the neglect of children
because it is often played both during the day and during the evening, where children need their
parents’ attention (1). The yearly revenue from gambling products sold in Greenland at Danske Spil who controls the lotteries in Greenland and Denmark – had increased with 50% from DKK 47.8 to
DKK 71.6 million in the years 1999-2007, indicating an increase in gambling activities. In addition to
this, the revenue from gambling on the privately owned slot machines was estimated to be DKK 80.3
million in 2004 alone (1). The revenue from bingo and gambling on cards and dice in private settings
are unknown.
An increase in gambling activities and the concern for pathological gambling is by no means limited
to the Greenlandic context, but a much more global concern following the expansion of gambling
opportunities online and liberalization of the gambling legislation seen in many western countries
these years. However indigenous populations and ethnic minorities are considered to be especially
vulnerable to gambling problems based on a higher prevalence of pathological gambling among
these groups compared to the general populations within the same regions (2-5), but little is known
of why it is so. Following Shaffer et al. “the study of incidence among vulnerable and resilient
populations is a road yet to be taken” (4), p. 504. This is also a road where determinants of gambling
problems should be studied more carefully. It is likely that the marginalization many of these
populations experience in regions with other majority populations is a part of the explanation. In
Greenland, Inuit constitute the majority of the population and the present study therefore provides a
unique opportunity to investigate problem gambling in a large indigenous population, that is a
majority in own country.
An increase in social pathologies such as violence, suicide, alcohol and substance use is a key feature
shared by indigenous populations in the circumpolar region undergoing social transition (6;7). In
Greenland, the high proportion of the adult population who experienced alcohol problems in their
childhood home and a similar high prevalence of persons who were sexually abused as children are
examples of such pathologies. An overall motivation for the study was thus to investigate, if
pathological gambling should be added to this list of social pathologies among Inuit in Greenland,
through a research based understanding of the role of gambling and gambling problems in Greenland
today in the light of the ongoing social transition.
9
1.1 Objective
The objective of the thesis was to investigate the prevalence of gambling behavior and problem
gambling in a representative sample of Greenland Inuit and its association with social transition,
addictive behaviors and health.
The overall objective was investigated based on three research questions answered in Paper I-III:
I.
II.
III.
How prevalent are different types of gambling and problem gambling among Inuit in
Greenland and how is gambling behavior and problem gambling associated with social
transition and traumatic events during childhood? [PAPER I]
How is lifetime problem gambling associated with harmful alcohol use and frequent use
of marijuana and does the association vary according to sociodemographic
characteristics? [PAPER II]
How is health status and health behavior associated with different types of frequent
gambling and lifetime problem gambling? [PAPER III]
10
2. Background
2.1 Concepts and definitions
Gambling activities and gambling behavior.
Gambling is a common leisure activity in most of the world and can be traced far back in history to
archaeologists’ uncovering of primitive dice found in caves dating back to 3500 BCE (8). Gambling
opportunities are available on the internet, television, radio, in casinos, in bars, diners and various
gambling venues. You can gamble at home or you can go out and gamble. Gambling activity naturally
presupposes gambling opportunities, and most studies have found a significant association between
access and availability of gambling on one hand, and the prevalence of gambling on the other.
Typically the introduction of new gambling opportunities such as the opening of a casino will lead to
an increase in gambling activities, but there is also evidence, that some form of adaptation can take
place over time and the prevalence will stabilize (8;9).
In the studies for Papers I-III, the term gambling activities refer to different types of available
gambling opportunities such as the lotteries and bingo. Gambling behavior refers to the type of
gambling activities people engage in and how often they gamble. There is no historic documentation
of gambling in Greenland and the possible existence and role of gambling among Inuit back in time is
therefore unknown (1). Most gambling activities present in Greenland today have been adopted
from the Danish gambling market and thus resemble gambling opportunities in Denmark and more
generally Scandinavia. Common gambling activities in Greenland are the lottery, bingo, slot
machines/electronic gaming machines (EGMs) and card/dice games.
The lottery is a general term for several different types of games such as lotto, scratch tickets, betting
on sports and other numbers and knowledge games all controlled by the national lottery in Denmark,
Danske Spil. The lottery is available at the local stores in all towns and in some villages. In addition
lottery gambling from Danske Spil is available online. Gambling on the lottery is regulated by Danish
law, but Greenlandic law applies to the distribution of a proportion of the revenue from the products
sold by Danske Spil in Greenland (10). Some of which are spent on sports, youth and cultural
activities.
Bingo is played all over Greenland both over the radio and in local community halls and sports
centers. Most bingo games are organized by local radio stations and local associations for which the
revenue serves as the primary source of income. It is also common for parents to organize bingo
games to raise money for their children’s school trips and excursions. The amount of organized bingo
games varies from place to place. In most towns and villages bingo is available on a daily basis. In
some places bingo is played several times a day given the availability over different local radio
stations in the area in addition to games held in the local halls and sports centers. Only public bingo
games require permission from the police. Bingo games arranged by the local associations for
members are not registered. A taxation of the winnings is required by law (11).
Slot machines are located at the bars and grills, but in towns only. No permissions for slot machines
have been granted to villages by the municipal councils. The slot machines are owned by private
investors and generate a yearly revenue of up to 1 million DKK (1). The revenues from slot machines
serve as a main source of income for the local bars and grills. Taxation is regulated by law (12;13).
11
Gambling on cards/dice represent a grey area since gambling on these types of games is prohibited
by law. Cards and dice games presumably take place in private settings, but no official records
document the extent of gambling in this area due to its illegal status. Despite the lack of
documentation, it is well known within Greenland that card games are played extensively on the East
coast, where the traditional social gatherings ‘Kaffemik’ are often turned into gambling tournaments,
when coffee and cake has been consumed.
People can engage in gambling in very different ways and the role of gambling can vary significantly
between communities. Gambling on the lottery can be limited to buying a scratch ticket once in a
while or a weekly lotto coupon, when shopping for groceries. Others participate in a weekly bingo
game with friends or family as a social event or play cards during the lunch break at work. These are
examples of unproblematic gambling behaviors common in modern Greenland and for most people
gambling in itself is an unproblematic activity which one can choose to engage in and maybe even
win some money. However, for a minority the gambling develops into an extensive activity with
negative consequences for the person who gambles, and his or her surroundings. This situation is
defined as pathological gambling.
Pathological gambling – a definition
The concept of pathological gambling is rooted in the psychiatric discipline and was first listed as
diagnose in 1980 in the third edition of the Diagnostic and Statistical Manual (DSM-III) published by
the American Psychiatric Association and in WHO’s ninth edition of the International Classification of
Diseases (ICD) in 1977. In the DSM-III pathological gambling was initially defined as an impulse
control disorder. This diagnose has been revised twice, first with the revised DSM-III in 1987 (14) and
later in the fourth edition of the Diagnostic and Statistical Manual from 1994 (DSM-IV), which was
revised in 2000 (DSM IV-R) (15). The revisions from DSM III to IV were of substantial character and
the criteria for pathological gambling have been linked to addictive behaviors such as alcohol and
drug dependence based on empirical findings, although it is still listed as an impulse disorder (4;16).
Currently the definition of pathological gambling is under discussion again in preparations of the fifth
edition of the Diagnostic and Statistical Manual (DSM V).
In DSM-IV-R, pathological gambling is listed under ‘Impulse-Control Disorders Not Elsewhere
Classified’ along with disorders such as Kleptomania and Pyromania:
“The essential feature of Pathological Gambling is persistent and recurrent
maladaptive gambling behavior (Criterion A) that disrupts personal, family, or
vocational pursuits. The diagnosis is not made if the gambling behavior is
better accounted for by a Manic Episode (Criterion B).” (15), p.671.
The following ten criteria (defined as the DSM-IV-R criteria) for pathological gambling are listed in the
DSM IV-R. Fulfillment of five or more of these is diagnosed as pathological gambling in a clinical
context.
12
Table1. The DSM-IV-R criteria for pathological gambling
Preoccupied
Tolerance
Loss of control
Withdrawal
Escape
Chasing
Lying
Illegal acts
Risked significant relationship
Bailout
The individual may be preoccupied with gambling
(e.g., relieving past gambling experiences, planning
the next gambling venture, or thinking of ways to get
money with which to gamble).
Most individuals with Pathological Gambling say that
they are seeking “action” (an aroused, euphoric state)
or excitement even more than money. Increasingly
larger bets, or greater risks, may be needed to
continue to produce the desired level of excitement.
Individuals with Pathological Gambling often continue
to gamble despite repeated efforts to control, cut
back, or stop the behavior.
There may be restlessness or irritability when
attempting to cut down or stop gambling.
The individual may gamble as a way of escaping from
problems or to relieve a dysphoric mood (e.g., feelings
of helplessness, guilt, anxiety, depression).
A pattern of “chasing” one’s losses may develop, with
an urgent need to keep gambling (often with larger
bets or the taking of greater risks) to undo a loss or
series of losses. The individual may abandon his or her
gambling strategy and try to win back losses all at
once. Although all gamblers may chase for short
periods, it is the long-term chase that is more
characteristic of individuals with Pathological
Gambling.
The individual may lie to family members, therapists,
or others to conceal the extent of involvement with
gambling.
When the individual’s borrowing resources are
strained, the person may resort to antisocial behavior
(e.g., forgery, fraud, theft, or embezzlement) to obtain
money.
The individual may have jeopardized or lost a
significant relationship, job, or educational or career
opportunity because of gambling.
The individual may also engage in “bailout” behavior,
turning to family or others for help with a desperate
financial situation that was caused by gambling.
Most of the screening tools used in prevalence surveys and epidemiological research of pathological
gambling are based on these DSM-IV-R criteria or the earlier revised DSM-III-R criteria, but operate
with different levels of threshold compared to the clinical screening in order to catch the less severe
gambling problems along with more serious ones. Pathological gambling is considered a chronic
disorder in the DSM-IV-R with a natural evolution from a less to a more severe state over time.
Problem gambling – widening the concept of pathological gambling
In our study, we use the term problem gambling, which is typically applied in the epidemiological
field of gambling research as a wider concept of pathological gambling, when measured at a
population level (8). Problem gambling is used to define problematic gambling behavior that may not
qualify as pathological gambling if assessed in a clinical context, which is of relevance in order to get
13
an overview of how prevalent more or less severe gambling problems are in a population and thus
the need for early prevention and treatment. More specifically problem gambling is used to define a
subclinical state of pathological gambling. Internationally the prevalence varies from country to
country with a past year prevalence of problem and pathological gambling ranging from less than 2%
in the Scandinavian countries (17), and 3.5% in the United States and Canada (4) to 4-5% in several
Asian countries (18). The widening of the concept is especially relevant in a public health perspective,
where the goal is to target prevention, either towards the general population or towards specific
groups, who may be more exposed or vulnerable to the problem.
2.2 Towards a public health perspective on gambling
Considering gambling and problem gambling in a public health perspective is a relatively new
approach following in the footsteps of the increase in epidemiological surveys estimating national
rates for gambling behavior and especially problem and pathological gambling. The public health
perspective considers gambling from a societal and population health point of view with a focus on
how gambling and problems related to gambling are affecting not just individuals, but families,
communities and societies. Furthermore the public health perspective opens up for considerations
regarding the influence and significance of structural conditions in a society, which transcends the
individuals’ access to a healthy choice regarding gambling.
It is important to recognize that a public health approach to gambling is but one perspective among
many. In Table 2 different examples of how gambling is viewed are illustrated. Although these are
taken from a Canadian context, they reflect the general perspectives in the public debate in
Greenland. Statement H represents the public health perspective promoted by Korn and others (1921), who argue in favor of a well-informed evidence-based policy for access to gambling, preventive
strategies towards gambling related problems and treatment programs targeted vulnerable
population groups.
Table 2. Different frames for understanding public policy on gambling proposed by Korn, Gibbens and Azmier
(19), p. 237
A. Gambling is a matter of individual freedom; apart from addressing a legitimate concern with crime and the
protection of minors, governments should not restrict how people spend their after-tax income.
B. Gambling is a recreational activity, a form of entertainment.
C. Gambling is a major source of public revenue; one rendered all the more appealing to governments because
it can be portrayed as a form of voluntary taxation.
D. Gambling is an important tool for economic development through increased tourism and employment, one
that may be particularly attractive to Aboriginal communities.
E. Gambling addiction is an individual rather than social pathology, and therefore should be treated within a
medical model much like other mental disorders.
F. Gambling is a cultural artifact that is more deeply embedded in some cultures than it is in others.
G. Gambling is a way to escape the class constraints of Canadian society, allowing winners to leap with a single
bound into the ranks of the wealthy.
H. Gambling is seen within the context of public accountability, public responsibility, and public health. Because
gambling is in the public domain in Canada, there is an incumbent responsibility for political leaders to be
informed about the costs and benefits of gambling, and to be held publicly accountable for their policy choices.
14
2.3 The role of gambling and problem gambling in a population undergoing
social transition
Greenland was colonized by Denmark in 1721 and reforms of the Greenlandic infrastructure, industry
and welfare during the 20th century were based on Danish administrative systems. In the early 1900
climate warming brought the Atlantic cod to the west coast of Greenland and cod fishing became a
major source of cash income. This initiated a movement away from the hunting of seals and whales
towards a modern fishing industry. People moved to larger and fewer towns. In 1953 Greenland was
given the status of a county in the Kingdom of Denmark. This development brought fundamental
reforms with it, and G-50 and G-60 were implemented with the purpose of creating a modern
infrastructure to handle the new fishing industry, which was planned to be a main source of revenue
for the modern Greenlandic society. During these years extensive infrastructural development
occurred, which transformed Greenland from a traditional hunting society into a modern economy,
where most people depended on earning wages (22). When the cod disappeared in the 1960s the
shrimp arrived in the Disko Bay area in North Greenland. This lead to a substantial growth in this
region until the shrimp disappeared from the coastal waters in the 1990. These changes illustrate
how population growth and societal changes are closely linked to local conditions such as the
availability of fish or shrimp and how rapid these conditions can change. Many communities are still
very vulnerable to these changes because the availability of jobs and income from fishing rely on the
availability of certain species (23).
In 1979 Greenland was granted Home Rule Government and in 2009 Self-rule. Today there are still
close ties between Denmark and Greenland. In less than 100 years Greenland has developed from a
traditional subsistence based economy to a modern society. The rapid transition places immense
social and cultural demands on both individuals and communities (23-25). Despite its large
geographical size, the total population of Greenland is only about 57,000 of whom 90% are ethnic
Greenlanders (Inuit). Genetically, Greenlanders are Inuit (Eskimos) with a mixture of European,
mainly Scandinavian genes. They are genetically and culturally closely related to the Inuit/Iñupiat in
Canada and Alaska and, somewhat more distantly, to the Yupiit of Alaska and Siberia (22). Greenland
has a total of 80 communities all located along the coast divided into towns and villages. A town is
defined historically as the largest community in each of 17 districts. In 2010, the population of the
towns varied between 469 to 5,460 and 15,469 in the capital Nuuk while that of villages varied from
less than 10 to around 550. In the towns are located district school(s), health centre or hospital,
church, district administration and main shops. These institutions are absent or present to a much
smaller extent in villages. Village schools are usually limited to 7th grade, after which children have to
move to the nearest town in order to complete their schooling. High schools are located in two
towns in South and North Greenland respectively and in the capital. The University of Greenland is
also located in the capital of Nuuk.
15
Figure 1. Social transition in Greenland. Changes in the percentage of the total population
th
living in villages in Greenland during the 20 century.
The number of persons living in villages in Greenland has been stable around 10,000 since the
beginning of the 20th century, but the relative distribution between villages and towns has changed
dramatically. During the first half of the 20th century around 80% were living in villages, but this
changed around the year 1950 towards the present situation. Today the majority of the population in
Greenland lives in towns, more than 25% live in the capital of Nuuk, and only 16% live in villages
(22;26). There are also significant differences between the socioeconomic status of inhabitants in
towns and villages. In villages the people are in average younger, have a lower income and
unemployment rates are higher compared to the towns. In 2007 the average annual income in towns
was 179,000 DKK compared to only 98,000 DKK in villages (27).
The isolated populations in villages and smaller towns face large challenges with limited
opportunities. In contrast, life in the capital reflects contemporary Scandinavian lifestyle with a wide
range of educational, occupational and recreational possibilities. Typically a more traditional lifestyle
is found in the villages compared to the towns and especially the capital. These differences reflect
different stages of social transition in Greenland today, where the traditional Inuit lifestyle of hunting
and fishing in villages and smaller towns coexist with the lifestyle of high-educated professionals in
larger towns and the capital.
There are also great differences between the different regions in Greenland. The majority of the
population is concentrated on the south central west coast. Only about 3,500 persons live on the
East Coast and only around 1,000 in the far North (Thule). There are also historic differences
according to time of colonization. East Greenland was not colonized until the late 19th century and
North Greenland not until the late 20th century and thus 200-300 years later than the initial
16
colonization of southern and central Greenland by Hans Egede in 1721 (22). The historic and
geographic separation is still reflected in modern Greenland with significant differences between
North and East Greenland compared to Midwest and South Greenland. The living conditions in the
isolated North and East are harsh and access to education and occupation is very limited compared
to the rest of Greenland.
Figure 2. Map of Greenland
17
As touched briefly upon in the introduction of the thesis, an increase in social pathologies is a key
feature shared by indigenous populations undergoing rapid changes. The high rates of alcohol
consumption in the 70s and 80s (28) as well as the high prevalence of mental health problems (29)
and suicides have been linked to the ongoing transition (30;31). Different explanations focus on the
stress of rapid social change, acculturation and the inadequacy of traditional conflict resolution
behaviors in the new, more urbanized environments (24;30;31). Social transition and its potential
influence on health is an important framework for the study of gambling behavior and problem
gambling among Greenland Inuit in the present thesis.
18
3. MATERIAL AND METHODS
3.1 Inuit health in transition across the Arctic
The present thesis is based on data from the Inuit Health in Transition Greenland Survey which is a
part of an international cross-arctic collaboration: The Inuit Health in Transition Study (IHIT). IHIT was
established with the aim of creating a longitudinal cohort study among the Inuit in Greenland,
Canada (Nunavik and Nunavut) and Alaska. The studies in Greenland and Canada (Nunavik in
particular) follow similar protocols and a lot of work was put into ensuring comparability between
the surveys across regions. The purpose of the collaboration was to start a longitudinal study of the
interaction between the environment and genetic factors on the health and disease pattern of the
Inuit in Greenland, Canada and Alaska. Thus the data collection in Greenland is part of an
international study with data collection in several villages and cities in all three countries. The project
is expected to contribute to a better understanding of the health effects of the transition from a
traditional lifestyle to a modern, industrialized life, which takes place in most present day developing
countries. The present thesis’ focus on problem gambling is one example this. Others are studies of
dietary patterns (32;33), obesity (34), diabetes (35), cardiovascular risk factors (36) and physical
activity (37). The first round of data collections were carried out between 2004 and 2010 across the
different regions. A follow-up has been scheduled in 2014 for both Nunavik and Greenland. In
Greenland the follow-up will also include data from registers.
Figure 3. Map of Greenland with study communities for the
Inuit Health in Transition – Greenland Survey 2005-2010
The Inuit Health in Transition Greenland survey was carried out in 9 towns and 13 villages from 20052010 in Greenland (38). Questions about gambling were included in the self-administered
19
questionnaire from 2006 and onwards and was therefore only carried out in 8 of the 9 towns but in
all villages. In 2 of the 8 towns, the questions about gambling were included in a second visit carried
out to obtain the desired participation rate. The list of towns included: Nuuk, Qaqortoq, Aasiaat,
Maniitsoq, Narsaq, Upernavik, Tasiilaq and Qaanaaq. The Inuit Health in Transition Greenland survey
also included Qasigiannguit. The list of villages included: Eqalugaarsuit, Narsarmiit, Aappilattoq,
Atammik, Napasoq, Kullorsuaq, Innaarsuit, Aappilattoq, Kuummiut, Tiniteqilaaq, Siorapaluk,
Qeqertat, Moriussaq (Figure 3).
During the five-year period, the study was carried out during summer in some communities and
during winter in others. The logistics of the study did not permit data collection throughout the year
in each location or the collection of all data in one season. Data was collected by a team of local
persons responsible for the recruitment of participants, a supervisor, one of two laboratory
technicians, 2-4 interviewers, and two clinical assistants. The interviews were conducted in both
Greenlandic and Danish according to the choice of the participant. A total of 81 persons assisted with
data collection and the processing of data.
Data collection in the Arctic is characterized by very unique circumstances because most destinations
can be difficult to access depending on the weather and season of the year. There are no roads
between towns or villages in Greenland, thus towns except Upernavik, Tasiilaq and Qaanaaq were
visited by public transport (flight). These towns and all the villages were visited on three expeditions
by a chartered boat (M/S Kisaq, figure 4). Kisaq can sail in almost all weather conditions and has
accommodation for 12 passengers and large volumes of equipment. It is an unambiguously more
convenient alternative to public transport, chartered helicopter or local boat charter.
Figure 4. M/S Kisaq
20
3.2 Sample and participation
Participants were selected as a stratified random sample of adults aged 18 years and older, born in
Greenland or Denmark. Population lists from the central population register were used to initially
specify the sample. Faroese people or persons born in other countries were not included. Greenland
was divided into 12 strata based on geography (South west coast; Central west coast; North west
coast; East Greenland; North Greenland) and community size (towns with ≥ 2000 inhabitants; towns
with < 2000 inhabitants; villages). From each of these strata one or more towns and 2-3 villages were
selected for the study as being representative of the stratum with regard to living conditions. A
random sample was drawn from the central population register to obtain around 300 participants
from each town; this number represents the practical limit for a research team during a 4-6 weeks’
visit. Villages were chosen at random in the strata and in the selected villages all adults were invited
to participate. Individuals in the sample were contacted in writing with an invitation to participate.
Information about the study and examination procedures was given, and the recipients were asked
to inform the investigators by letter or phone whether or not they wanted to participate. The
samples were revised locally with information about who were not actually living in the community
at the time of the examination. Neighbors and the municipality office (in the villages) were good
sources of information.
Ethnicity as ‘Greenland Inuit’ or ‘Dane’ was determined at enrolment based on the primary language
of the participant and self-identification. Only one ethnicity was allowed for each participant. The
following concerns Greenlanders only. Participation ranged from 83.3% in the village of Aappilattoq
to 55.2% in the village of Napasoq. According to community size, the participation was 61.4% in
Nuuk, 65.1% in other large towns, 69.9% in small towns and 68.5% in the villages (p<0.001).
Participation rates also varied by age and sex. Women more often participated than men and
particularly young men were under-represented. The reasons for non-participation are seen from
flow chart 1. There were certain differences between the communities; in particular in the capital,
Nuuk, many persons indicated lack of time as the reason for not wanting to participate (17% of the
non-participants compared with 2% in the rest of the communities).
21
The Inuit Health in Transition – Greenland Survey 2005-2010
N=6,015
Stratified random sample of adult inhabitants in Greenland (18+) born in Greenland
(‘Greenland Inuit’) or Denmark (‘Danes’). Drawn from the central population register
N=5010
Revised sample
N=1005
Reduction of initial sample:
Moved (744)
Excluded for logistic reasons (94)
Pregnant (60)
Deceased (54)
Unknown in the community (40)
Other reasons (13)
N=350
Exclusion of Danes
N=1552
Non-participation:
Did not want to participate (796)
Illness or disability (107)
Hunting, fishing or mining (68)
Out of town for other reasons (53)
Other reasons (23)
No contact (505)
N=4660
Revised sample ‘Greenland Inuit’
N=3108
Participants in clinical examinations and interview
Figure 5. Flow chart 1
We know that persons with serious illness or disability are over represented among the nonparticipants as well as those who tend to move often, and we suspect that socially exposed persons,
alcohol abusers and persons who frequently go in and out of jobs and the unemployed likewise are
over represented among the non-participants. It was the impression of the interviewers that there
was a distinct downwards social trend from the beginning to the end of data collection in a town. In
some towns it could be demonstrated that during the first week of the study 10% of those who had
made an appointment did not show up, while during the last week of the study as many as 26% did
not show up (p<0.001).
A social bias in the participation was confirmed by information about income obtained from Statistics
Greenland. The Inuit participants in the study had an average personal income in 2005-2010 of DKK
161,000 while those who were refused to participate had an income of DKK 152,000, those who
were excluded from the sample DKK 141,000 and those who couldn’t be contacted DKK 134,000
(p<0.001; adjusted for age and sex). For disposable household income per person only those who
could not be contacted differed from the rest.
Revised sample for the Gambling Study
The revised sample for 2006-2010 included 3,892 Greenland Inuit. A total of 2,454 persons
participated in the general survey (63%) and 2,189 persons filled out the self-administered
questionnaire (56%). Only those who filled out the self-administered questionnaire were included in
the present study of gambling behavior and problem gambling. In total 2,012 of these 2,189
22
participants (92%) answered at least one question about gambling. The analyses were based on the
2189 participants, who had the possibility to answer questions about gambling.
The study of gambling behavior and problem gambling among Greenland Inuit 2006-2010
Inclusion of gambling
questions from 2006
N=4660
Revised sample ‘Greenland Inuit’ 2005-2010
N=768
Reduction of revised sample:
Exclusion of revised sample for
Qasigiannguit visited in 2005
(413)
st
Exclusion of participants from 1
visit to Aasiaat in 2005 (135)
st
Exclusion of participants from 1
visit to Qaqortoq in 2005 (220)
N=3892
Revised sample ‘Greenland Inuit’ 2006-2010
N=2454
Participants in clinical examinations and interview
N=1438
Non-participation
Did not want to participate (758)
Illness or disability (91)
Hunting, fishing or mining (68)
Out of town for other reasons (45)
Other reasons (26)
No contact (448)
N=265
Non-participation in selfadministered questionnaire
N=2189
Participants who filled out the self-administered
questionnaire
N=205
Missing data on gambling
behavior
N=1984
Gambling behavior
N=1542
Problem gambling
Answered at least one question
regarding gambling behavior past
year (the lottery, bingo, slot
machines, cards/dice)
Answered one or both questions
in the lie/bet screen
Figure 6. Flow chart 2
23
N=647
Missing data on the lie/bet screen
3.3 Data Collection
The participants were informed about the arrival of the team and about the investigation by a
personal letter and they were after the arrival of the team contacted by the person responsible for
recruitment. On the day of the investigation the participants were asked to show up at an appointed
time, fasting (i.e. at least 8 hours without eating or drinking). They were further informed about the
investigation and signed an informed consent. From the time of arrival 2.5-3 hours went by. The
participants were interviewed (40 min.), filled in a self-administered questionnaire, had various
clinical tests performed and were issued with an Actiheart device to monitor physical activity for a 14 days' monitoring of heart rate and movements. At the end of the session, participants were
informed about the results of the investigation and were invited to ask questions. When the
Actiheart device was returned, a compensation of DKK 200 was paid to each participant.
3.4 Outcomes and exposures
Past year gambling behavior
The questions about past year gambling behavior measured in the self-administered questionnaire
regarded four types of gambling; i.e. the lotteries, bingo, cards/dice and slot machines (Figure 7).
Respondents answered how often they had gambled within last year, how much time, they spent
each time they gambled and how much money they spent on gambling during last month on each of
the four types of gambling. The questions regarding past year gambling behavior are identical with
questions included in the Inuit Health in Transition Nunavik Survey, with the exception of the
question concerning slot machines, which was not included in Nunavik. Only information on past
year gambling and gambling frequency has been used in this thesis. Information on time and money
spent on each type of gambling has not yet been analyzed, but will be included in future comparative
studies with data from Nunavik.
24
Figure 7. Questions regarding past year gambling behavior in the self-administered questionnaire
Past year and lifetime problem gambling
Our perception of problem gambling is based on the assumption that extensive gambling can have a
damaging influence on your social life and employment as well as your health and economic status,
as it is defined in DSM-IV. Following this perspective, it is important to stress the public health
perspective underlying this study, which implies an understanding of problems as structural rather
than individual. Problem gambling was measured through four questions in the self-administered
questionnaire. Two of these questions were combined in a short screen (the lie/bet questionnaire)
used in all three papers, while the remaining two questions were used as separate one-item
measures of problem gambling included in Paper III. The short screen and the additional one-item
measures are described below.
25
Figure 8. Questions regarding problem gambling in the self-administered questionnaire
The lie/bet questionnaire
The lie/bet questionnaire was originally suggested by Johnson and Hammer (39) and later validated
in both treatment (40) and community samples (41). The lie/bet questionnaire represents a twoquestion short version of the 10 DSM-IV criteria (15) for screening pathological gambling.
Respondents were asked whether they had lied to friends and family about their gambling activities,
and whether they had felt a need to increase bets. Questions are shown in Figure 8 (question 23-24).
Both questions were posed regarding past year and previously in life. When these two are combined,
it represents a measure for lifetime problem gambling, i.e. problem gambling that has occurred at
some point in life whether it is a current and/or previous condition. Persons who answered one of
the two questions, but not the other, were kept in the new variable for problem gambling, because
one positive answer on either of the two questions qualify you to be considered as a problem
gambler. This reduced the number of missing in the final variable. A rather large proportion of
individuals indicated they had gambled during past year, when asked about specific types of games,
but answered they never gambled in the following questions concerning gambling problems. These
were not accepted as non-gamblers, which reduced the total of non-gamblers.
International gambling studies have typically reported both past year and lifetime prevalence. The
lie/bet screen has been found valid to identify lifetime problem gamblers in a community sample,
defined as those who responded positively to five or more of the 10 DSM-IV criteria combined with
those who only responded positively to three or four of these criteria (41). The short screen does not
qualify to distinguish between pathological and problem gambling. It is important to stress, that the
26
purpose of screening for problem gambling in a large health survey is to investigate an overall
prevalence of how widely gambling is affecting a population, which is by no means comparable to a
thorough clinical assessment of pathological gambling based on the DSM-IV-R criteria.
Self-rated problem gambling
Self-rated problem gambling was included as an additional one-item measure. Respondents were
asked whether they themselves felt they had a gambling problem or had been told so by others (see
question 25 in Figure 8. The purpose of including a self-rated measure for problem gambling was to
see how well participants’ own perception corresponded with the lie/bet screen. The question was
taken from the revised South Oaks Gambling Screen (SOGS-R) originally developed by Leisure and
Blume (42;43). The SOGS-R has been the main instrument used to study the prevalence of problem
and pathological gambling, although recently abandoned for newly developed screens for
epidemiological research (44;45) such as the 9-item Problem Gambling Severity Index (PGSI) (46), the
17-item National Opinion Research Center DSM-IV Screen for Gambling Problems (NODS) (47) also in
a 3-item version called NODS CLiP (48), and the 3-item Brief Biosocial Gambling Screen (BBGS) (49).
Too much time and money spent on gambling
The second one-item measure of problem gambling regarded whether participants felt they had ever
spent too much time or money on gambling during past year or previously (question 22, Figure 8).
This item was also included in the Inuit Health in Transition Nunavik Survey, but has not yet been
analyzed. We have not been able to combine the data from the two surveys yet either, but this will
be done in a future study.
Sociodemographic variables
Residence at age 10 was obtained from the interview and recoded into village or town. Family job
type was determined from questions about job title of participant and spouse and recoded into
hunters/fishermen and others. Formal education was determined from questions about highest
school education attained and further vocational or academic education and recoded into primary
school/high school only, short vocational education (less than three years), and longer
vocational/academic education. Age was divided into four groups; 18-24, 25-34, 35-59 and 60+ years.
Place of residence was divided into the capital of Nuuk, villages and towns (also used as a measure of
social transition).
Social transition
Two measures of social transition were included.
Place of residence
The first measure was place of residence, which was divided into villages, towns and the capital
because there are substantial differences in lifestyle and living conditions between these places in
general.
Level of involvement in the ongoing social transition
In order to supplement this measure we combined current place of residence and childhood
residence with formal education and family job type in a second measure that has previously been
used to document changes in cardiovascular risk factors and physical activity among Greenland Inuit
(36;37). Six categories of social transition were defined as (A) hunters and fishermen in villages; (B)
other inhabitants of villages; (C) blue collar migrants (inhabitants of towns, with no vocational
27
education, having lived in villages at age 10); (D) other blue collar participants (inhabitants of towns,
with no vocational education, having lived in towns at age 10); (E) intermediate (inhabitants of
towns, with short vocational education); and (F) professionals (inhabitants of towns, with longer
vocational or academic education). In order not to misclassify participants who had not yet finished
their education and to minimize the proportion of participants outside the work force this measure
only included those aged 25–64 years.
Traumatic events during childhood
Alcohol related problems in childhood home and sexual abuse during childhood were included as
traumatic childhood events. Participants were asked through the self-administered questionnaire if
there were alcohol problems in their childhood home and whether anyone had forced any kind of
sexual activity upon them as a child (before the age of 13).
Addictive behaviors
Harmful alcohol use (CAGE-C)
Harmful alcohol use was measured by the modified CAGE-test: CAGE-C. It is a simple screening tool
suited for identifying alcohol problems in populations with a high prevalence of at-risk drinkers. The
original CAGE test was based on a four-item questionnaire and measured harmful alcohol use in a
lifetime perspective (50). However the validity of the original questionnaire outside a clinical context
has been questioned (51) and the sensitivity of the test has been criticized in several studies (52;53).
The six-item questionnaire CAGE-C (Table 3) was suggested by Zierau et al. in 2005 (54) and validated
against a diagnostic interview based on ICD-10 (55) and DSM-III R (14) criteria. The questionnaire has
been used to assess harmful alcohol intake among Greenland Inuit in an earlier study (56). The
modified CAGE-test measures harmful alcohol use in a past year perspective and includes a question
regarding the number of days per week of alcohol use and a question concerning alcohol intake on
weekdays outside meals. CAGE positives were defined by a positive answer in two or more of
question 1-4 and 6 or one positive answer in question 1-4 and 6 in addition to alcohol intake on four
or more days per week.
Table 3. The CAGE-C questionnaire
1
Have you, within the past year, felt that you should cut down on your drinking? Yes/No
2
Have people, within the past year, annoyed you by criticizing your drinking? Yes/No
3
Have you, within the past year, felt bad or guilty about your drinking? Yes/No
4
Have you, within the past year, from time to time had a drink first thing in the morning to steady your
nerves or to get rid of a hangover (eye opener?) Yes/No
Have many days per week, do you drink alcohol?
5
0-1 day
6
2 days
3 days
4 days
5 days
6 days
7 days
Do you drink alcohol on weekdays outside mealtimes? Yes/No
Frequent use of marijuana
Frequent use of marijuana during past year was measured through two questions in the selfadministered questionnaire. Participants were asked if they had ever smoked marijuana. Those who
answered ‘yes, once or a few times’ or ‘yes, several times’ were asked how often they had smoked
marijuana during past year. Those who had tried to smoke marijuana and additionally answered they
28
had smoked at least 1-3 times a month during past year were categorized as past year frequent users
of marijuana. Those who had never tried to smoke and those who had smoked less than once a
month during past year, were categorized together as non-frequent users/non users.
Health status
Five different measures of health status were included in Paper III. These were self-rated health, two
different mental symptoms, long-standing illness and obesity.
Self-rated health
Self-rated health was measured in the interviewer-based questionnaire by asking respondents to rate
their own health within five categories: Excellent, good, fair, poor, very poor. In the analyses selfrated health was dichotomized into an excellent/good vs. fair/poor/very poor self-rated health.
Mental symptoms
Mental symptoms were measured in the interview questionnaire, by asking respondents whether
they had been bothered a little, a lot or not at all by (a) anxiety, nervousness, agitation, or fear
(hereafter anxiety); or (b) feeling depressed or unhappy (hereafter depression), in the two weeks
preceding the interview. Dichotomous variables were created to identify people reporting feeling
anxious vs. those who did not, and people reporting feeling depressed vs. those who did not.
Long-standing illness
Long-standing illness was assessed through interview by asking: Do you suffer from any long-standing
illness, after-effect of an injury, handicap or other long-standing disorder? Yes/no. If a person
answered yes, they were asked to specify their illness. Only the dichotomized variable based on
yes/no was used in this study.
Obesity
Height and weight were measured with the participants stripped to their underwear and socks (9).
Weight was measured on a standard electronic clinical scale. Body Mass Index (BMI) was calculated
based on these measurements (kg/m2) and dichotomized according to a BMI above or below 30.
Participants with a BMI above 30 were classified as obese (16).
Health behavior
Six different measures of health behavior were included in Paper III. These were daily smoking,
average daily sedentary time, an unhealthy diet, use of health services during past three months,
harmful alcohol use and frequent use of marijuana.
Daily smoking
Information about daily smoking was obtained from the interview questionnaire.
Sedentary time
Sedentary time was measured through the interview questionnaire based on two questions
regarding time spent on sitting during weekdays and weekend respectively based on the past seven
days from the International Physical Activity Questionnaire (IPAQ) (long version) adapted for living
conditions in Greenland (17;18). The variable was dichotomized using the median in order to
compare the 50% with the most sedentary time to the 50% with the least sedentary time (>3 hours
vs. <3 hours daily).
29
Unhealthy diet
Information on diet was obtained by an interviewer-administered Food Frequency Questionnaire
(FFQ) with portion sizes. The FFQ was developed from information obtained through a 24-hour
dietary recall. Questions were asked about 67 food items, including 23 local and 44 imported items
(9). An unhealthy diet was defined as 25 E% or more from soda pop, fast food, snacks, sweets and
sugar added to coffee or tea (19).
Use of health services during past three months
Information regarding use of health services was obtained from the interview questionnaire.
Consultations with a doctor, nurse or healthcare assistant at the local health center or hospital
during the three months preceding the interview was registered as having made use of health
services during the past three months.
3.5 Data analysis and statistical methods
All analyses were performed using IBM Statistics 19.0. Descriptive statistics of prevalence estimates
according to different sociodemographic variables were performed in all three papers. Chi-squared
test was used to determine if observed differences between groups were significant (57). The
significance level was set at 0.05. To obtain national prevalence estimates, data were weighted for
sampling strata by the weight procedure of SPSS. This was done to compensate for the oversampling
which took place in certain strata and the variation in participation rate according to age and gender.
Most prevalence estimates in Papers I-III were stratified for gender and when relevant also age group
and place of residence. Prevalence estimates were also weighted for regional and age differences in
the sample. Logistic regression was performed in all three papers and the results were reported as
odds ratios (OR) with a 95% confidence interval (57). All outcome variables were constructed as
binary variables and all exposures and additional variables that were included in the analyses were
treated as categorical variables. An overall analytical model of the associations investigated across
papers can be found in appendix 3. Kappa statistics (58) were used to test the agreement between
the different measures of problem gambling in Paper III.
All logistic regressions were either stratified or adjusted according to gender and age group, because
significant differences in gambling behavior and problem gambling between men and women and
different age groups have been documented in several prevalence studies. Place of residence, formal
education and alcohol problems in childhood were also adjusted for when relevant. These choices
are discussed under methodological considerations.
3.6 Validity and completeness of data source
Data was double entered by different persons and validated in EpiData (http://www.epidata.dk/).
The data files were subsequently imported into the SAS package and combined with results of blood
analyses and clinical procedures. The validity of data was checked against permitted values and
logical errors. Analyses were performed with SAS v. 9.1 or higher and with SPSS v. 15.0 or higher.
30
3.7 Ethical considerations
The study was ethically approved by the Commission for Scientific Research in Greenland.
Participants gave their written consent after being informed about the study orally and in writing.
Data are treated with confidentiality and participants are guaranteed complete anonymity. This can
be a challenge when dealing with small communities, where identities can be obvious even though
names are kept confidential. Therefore the anonymity of individuals is considered carefully before
results are published.
31
32
4. RESULTS
4.1 Overview of results
PAPER
I
TITLE
Gambling behavior and problem
gambling reflecting social transition
and traumatic childhood events among
Greenland Inuit - a cross-sectional
study in a large indigenous population
undergoing rapid change
II
Harmful alcohol use and frequent
use of marijuana among lifetime
problem gamblers and the prevalence
of cross-addictive behaviors among
Greenland Inuit – evidence from the
cross-sectional Inuit Health in
Transition Greenland Survey 20062010.
III
Health status and health behavior
associated with frequent gambling and
problem gambling in a large
indigenous population – A crosssectional study based on The Inuit
Health in Transition Greenland Survey
2006-2010.
MAIN FINDINGS
The study found a comparably high prevalence of lifetime
problem gambling among Greenland Inuit which adds problem
gambling to the list of social pathologies in Greenland. A
significant association between lifetime problem gambling, social
transition and traumatic childhood events suggests people
caught between tradition and modern ways of life are more
vulnerable to gambling problems.
The study showed that among more than half of the lifetime
problem gamblers, the gambling problems were combined with
either a harmful alcohol use (past year), a frequent use of
marijuana (past year) or both. The study also found that one or
more of these addictive behaviors were present among half of
the men, one third of the women and in almost half of the
households with children. This indicates that many families in
Greenland are presently affected negatively by alcohol, gambling
and marijuana. These findings suggest that pathological gambling
should be included systematically in future public health
strategies in addition to the present focus on alcohol and
marijuana.
The study found that frequent gambling and lifetime problem
gambling were associated with a poorer health status and
unhealthy behavior in general. Variation according to different
gambling activities and different measures of problem gambling
suggests that the associations between gambling and health is a
complex process closely related to living conditions and
socioeconomic differences among Greenland Inuit, and thus an
important issue for public health, in order to reduce social
inequalities in health.
4.2 Paper I
PAPER I showed the majority of Greenland Inuit (80%) had engaged in at least one of four types of
gambling during past year and gender and age differences applied. The lottery was played by four
out of five followed by bingo, cards/dice and slot machines. Bingo was more common in villages and
towns compared to the capital. People in villages gambled more frequently on cards/dice compared
to the capital and in towns while the lottery and slot machines were much more common in the
capital and in towns. This indicated a mix of patterns between modern games popular in
contemporary Scandinavia and games found to be traditionally popular among Inuit in Nunavik and
First Nations and thus reflects the ongoing social transition in Greenland.
Regarding problem gambling Paper I showed the past year prevalence of problem gambling to be 4%
among men and 3% percent among women. The lifetime prevalence of problem gambling was 16%
among men and 10% among women (p<0.0001). The analyses showed lifetime problem gambling
33
was associated with social transition. This was reflected in the higher prevalence of lifetime problem
gambling in towns (19%) compared to Nuuk (11%) and villages (12%) (men only, p=0.020) and a
lower odds ratios for lifetime problem gambling among both hunters and fishermen in villages as
well as among professionals in towns compared to other villagers, blue collar workers in towns and
persons with a short vocational education in towns (p=0.023). Growing up with alcohol problems in
your childhood home (p=0.001/p=0.002) and having been exposed to sexual abuse in childhood
(women only, p=0.030) further increased the odds ratio of lifetime problem gambling. Figure 9 shows
how different birth cohorts experience different prevalence rates of lifetime problem gambling,
which follows the same pattern as alcohol problems in childhood home and sexual abuse before the
age of 13.
40
35
30
25
20
15
10
5
0
-1954
1955-1959
1960-1964
1965-1969
Often experienced alcohol problems in childhood home
1970-1974
1975-1979
Lifetime problem gambling
1980-1985
1985-
Sexually abused as a child
Figure 9. Prevalence of alcohol problems in childhood home, lifetime problem gambling and sexual abuse in
childhood according to birth cohort. Unadjusted percent, weighted for regional differences in the sample.
Greenland Inuit 2006-2010. Alcohol problems in childhood home (N=1985, p<0.0001); Lifetime problem
gambling (N=1542, p<0.0001); Sexual abuse (N=1359, p<0.0001).
4.3 Paper II
Analyses in PAPER II showed that the odds ratio for a harmful alcohol use (men only) and a frequent
use of marijuana were significantly higher among lifetime problem gamblers compared to nonproblem gamblers/non-gamblers. One or more addictive behaviors were present among more than
half of the men (53%) and among one third of the women (37%) (Figure 10). Among lifetime problem
gamblers, the gambling problems were more often than not combined with a harmful alcohol use, a
frequent use of marijuana or both – especially among men. The prevalence of lifetime gambling
combined with alcohol and/or marijuana in a cross-addictive behavior was 12.2% among men and
3.7% among women.
Among lifetime problem gamblers the prevalence of a harmful alcohol use increased with aged,
while the prevalence of a harmful alcohol use decreased with aged among the non-problem
gamblers.
In total the prevalence of one or more addictive behaviors in households with children was 44%. The
prevalence of only one addictive behavior in households with children was similar to households
without children (32% vs. 28%), but the prevalence of two or more addictive behaviors was slightly
34
higher in households without children compared to those with children (19% vs. 12%) (p<0.0001).
The odds ratio for finding one or more addictive behaviors in a household was slightly lower for
households with children compared to households without children (OR=0.76; C.I. 95%: 0.60-0.97).
Figure 10. Prevalence of lifetime problem gambling, frequent use of marijuana, harmful alcohol use and the
different combinations between the three as well as no addictive behavior. Men (left), N=579, missing=404.
Women (right), N=605, missing=601. The total prevalence of each of the three addictive behaviors as well as no
addictive behavior is written in bold figures. The prevalence of only one addictive behavior or the combination
of two or three addictive behaviors is written in ( ). Unadjusted percent weighted for regional and age
differences in the sample. Greenland Inuit 2006-2010.
4.4 Paper III
PAPER III investigated gambling behavior and problem gambling in a general public health
perspective by looking at health behavior and health status among frequent gamblers and lifetime
problem gamblers. The analyses showed significant associations between selected measures of
health status and health behavior with both gambling behavior and lifetime problem gambling. More
frequent gamblers and problem gamblers experienced mental symptoms, smoked daily, enjoyed an
unhealthy diet, reported of a poorer self-rated health, a higher prevalence of long-standing illness
and had made more use of health services during the past three months compared to those who
gambled less than once a month and non-problem gamblers respectively.
Exceptions were obesity and sedentary time during the day, where the pattern was reversed;
meaning frequent gamblers and lifetime problem gamblers spent less time on sedentary activities
during the day and were less obese. Associations varied according to gender, type of gambling and
how problem gambling was measured.
35
36
5. DISCUSSION
5.1 Main findings
The objective of the thesis was to investigate the prevalence of gambling behavior and problem
gambling in a representative sample of Greenland Inuit and its association with social transition,
addictive behaviors and health. In short terms, Papers I-III found a high prevalence of especially
lifetime problem gambling among Greenland Inuit; a significant association between social transition
with both gambling behavior and lifetime problem gambling; a prevalent overlap between lifetime
problem gambling and other addictive behaviors (harmful alcohol use and frequent use of
marijuana); and a generally poorer health status and unhealthier behavior among lifetime problem
gamblers compared to non-problem gamblers and those who never gambled.
Main findings in perspective to other studies
The prevalence of lifetime problem gambling among Greenland Inuit is high compared to national
prevalence estimates from Scandinavia (17;59-61), New Zealand (3;62) and the United States and
Canada (4;63), but similar high rates have been found among aboriginal populations in North
America (5) and New Zealand (3). The results contribute with new knowledge with regards to
understanding the high prevalence rates among indigenous populations in the light of the ongoing
social transition.
The high prevalence of other addictive behaviors among problem gamblers compared to nonproblem gamblers is in line with international population based studies of co-morbid behavior
reviewed by Lorrains et al. (64). However the prevalence of a harmful alcohol use among male
lifetime problem gamblers seems relatively high among Greenland Inuit compared to findings in
other population based studies.
The empirical literature regarding problem gambling and health beyond mental disorders, smoking,
substance and alcohol use is limited, but the existing studies found lifetime pathological gambling to
be associated with several medical disorders, poor lifestyle choices and worse quality of life (65;66).
Our results supplement these findings and contribute with new knowledge regarding the
associations between health and frequent gambling.
5.2 Methodological considerations
Selection bias
Selection bias can occur when the study procedures or the methods of sampling for the study cause a
systematic exclusion of individuals who were theoretically eligible for the study (67).
As described, participation in the survey required the individual participant to show up at a given
destination, at a given time and go through 2.5-3 hours of health examinations, interview and finally
fill out a self-administered questionnaire. These procedures are likely to create selection bias
because people with few economic and social resources typically will not have the energy to
participate. Also they can be harder to reach by mail or telephone. In general, persons with serious
illness or disability are over represented among non-participants in populations-based surveys as are
37
those who tend to move often. It is likely that also socially exposed persons, persons with a
substance use or an alcohol use disorder as well as the unemployed are over-represented among the
non-participants. Analyses of register based income information showed that the personal income of
participants was higher than among non-participants. This confirms the social bias known from most
studies. For the study of problem gambling this is particularly relevant since especially problem
gamblers with the most severe gambling problems could be expected to belong to a group of socially
exposed persons, who do not participate in surveys. This can have resulted in an underestimation of
the prevalence of both past year and lifetime problem gambling in the study. Stratification and
adjustments were used in the statistical models to minimize this bias.
The variation in participation rates across the country and between villages, towns and the capital
presents a bias for the countrywide estimates. So does the stratification procedure of the random
sample, which means that some regions have proportionately more participants despite their small
percentage of the total population. This applied to North Greenland (the Thule Area) and Midwest
Greenland. Because of logistic challenges it would be almost impossible and very expensive to base
the population survey on an unstratified sample. This bias is compensated for through weighting
procedures to ensure the extern validity and representativity of the results.
In addition to the potential bias caused by sampling method and data collections, the response rate
for questions used as key variables in the analyses represent an important source of bias. Analyses
showed that response rates to questions about gambling behavior and problem gambling varied
according to gender, age group, place of residence and formal education. This was also the case for
most of the included measures of health behavior and health status. The odds ratios for having
answered the questions regarding the included measures were higher among men than women and
among inhabitants in the capital compared with people from towns and villages. Furthermore, the
odds ratio decreased with age and increased with level of education. Only response to the question
regarding use of health services during the past three months went in a different direction with a
decrease in the odds ratio for having answered this question with increasing level of formal
education. Consequently, there is an additional social bias in the analyses, which can cause residual
confounding. Stratification and adjustment for these variables has been applied to the possible
extent, but there are limits as to how much a small sample can be stratified and adjusted due to the
risk of type-II errors (discussed below). Another way to deal with the social gradient in the response
rates could be to weight data further according to this gradient, but this could potentially be a
greater source of misinformation compared to the missing data. Consequently the results must be
interpreted with this social bias in response rates in mind.
Information bias
Information bias occurs if there are errors in the applied measurements, which can lead to the
misclassification of individuals. This is a central issue to discuss in relation to the lie/bet screen used
to identify problem gamblers in this study, because there is always a risk of including false positives
and excluding false negatives from a prevalence estimate. This has to do with the sensitivity, i.e. the
probability that the test correctly classifies those with the given condition as positives, and the
specificity, i.e. the probability that the test classifies people without the given condition as negatives
(67).
38
Three validation studies have been carried out for the lie/bet screen with the DSM-IV criteria as the
gold standard. Johnson et al. (39) found a sensitivity of the lie/bet screen of 0.99 and a specificity of
0.91 in their initial validation study of the screen based on men only and a sensitivity of 1.00 and a
specificity of 0.85 in a follow-up study which also included women (40). Both of these studies were
based on pathological gamblers recruited from Gamblers Anonymous matched with controls. The
third validation study carried out by Götestam et al. (41) was based on a population study and found
a sensitivity of 0.92 and a specificity of 0.96. Although Johnson et al. found the lie/bet screen to be
useful for screening for pathological gambling; this conclusion was based on a study population of
pathological gamblers who met the diagnostic criteria of DSM-IV-R. Based on the validation of the
lie/bet screen in a population sample, Götestam et al. concluded that the short screen could not be
used to discriminate between problem and pathological gambling in a normal population, but was a
useful tool for screening for problem gambling including the group of more severe pathological
gamblers at the population level. The assessment of problem gambling among Greenland Inuit was
based on this strategy, to ensure the validity for screening at the population level.
Götestam et al. (41) reported a 0.54% prevalence of lifetime problem gambling based on the lie/bet
screen in a community sample of adults (18+) in Norway. This prevalence is a little lower than the
prevalence found by Lund and Nordlund (59) in Norway in the same survey year based on the NODS
as a screening tool suggesting the lie/bet screen did not overestimate the lifetime prevalence in the
Norwegian population despite its brief character. Furthermore the Danish Health Interview Survey
from 2005 (68) included the lie/bet screen and found a lifetime prevalence for problem gambling of
2.6% among participants aged 16+ (unpublished results). This prevalence was a little higher than the
one found in the Danish Gambling Survey by Bonke and Borregaard (61) mentioned above, where
the SOGS-R identified 1.7% lifetime problem gamblers and the NODS identified 0.7% lifetime
problem gamblers among adults (18+). Both studies were based on a representative population
samples. This suggests that the lie/bet screen slightly overestimated the lifetime prevalence in the
Danish population compared to both NODS and SOGS-R. There is a risk that one is categorized as a
lifetime problem gambler based on symptoms, which might have occurred one at a time over several
years, but not at the same time (4). However, since one symptom is enough to be categorized as a
problem gambler based on the lie/bet screen, this is considered to be less of an issue in this study.
The sensible character of questions regarding people’s gambling problems represents another
potential source of information bias. Although the questions were posed in a self-administered
questionnaire to respect the privacy of the respondents, it is a known phenomenon that people tend
to underestimate their own bad habits and are influenced by social norms of what is right and wrong,
when responding to questions in a survey. Furthermore - according to professionals working with
gambling treatment - pathological gamblers are often in denial about their gambling related
problems. The same potential risk of information bias applies to the questions about drinking and
smoking marijuana included in this study.
Another source of information bias is attributed to an imprecise recall of the information retrieved
from participants in the study (67). Because questions regarding problem gambling were asked for
past year and earlier in life the answers might be subject to recall bias, particularly if potential
gambling problems were less severe. People tend to forget the problems they have had in the past
and they might seem less severe with time. Participants might also be influenced by a social
desirability to respond according to what they think the researcher wants to hear, or according to
39
social and cultural norms. Again this can have led to an underestimation of the prevalence found in
this study.
Confounding and intermediate factors
The risk of confounding from variables that are not included as exposures in a study, but which can
lead to a mixing of effects because they influence the relationship between exposure and outcome,
are important to consider in any epidemiological study. Rothman and Greenland (67) defines three
criteria for a confounding factor:
“1. A confounding factor must be an extraneous risk factor for the disease; 2. A
confounding factor must be associated with the exposure under study in the
source population (the population at risk from which the cases are derived); 3.
A confounding factor must not be affected by the exposure or the disease. In
particular, it cannot be an intermediate step in the causal path between the
exposure and the disease.” (67)p. 132
Given the complexity of addictive behaviors or disorders such as problem gambling, it is difficult to
point to specific confounders that are not in some way thought to be part of the explanation and
thus rather have a modifying effect as intermediate variables in the relationship between exposure
and outcome, cf. the third criteria for a confounding factor.
All analyses were stratified or adjusted for gender and age group. These variables were not
considered as confounders, but rather as a part of the explanation, i.e. intermediate factors in a
causal path between exposure and outcome (67). In addition, most of the analyses were adjusted for
place of residence, except when this would result in too few individuals in one or more of the created
subgroups. Place of residence was in general not considered as a confounder, but rather as an
intermediate factor given the differences in living conditions. This was relevant when investigating
different health outcomes in relation to gambling behavior and problem gambling (Paper III),
because there are great differences in access to health service, healthy foods, physical activity, etc.
between villages, towns and the capital. However, in the analysis of how different types of weekly
gambling activities were associated with past year problem gambling, place of residence was
considered a confounder because the availability of gambling activities depends on where you live
(Paper I). This was the case with regards to slot machines, which are only available in towns and the
capital, but not in villages.
In Paper II, alcohol problems in childhood home were included in the analyses of the association
between lifetime problem gambling and other addictive behaviors, because growing up in a family
with problems related to addictive behavior is known to be a contributory cause of addictive
behavior later in life. Again, this was not considered a confounder, but an intermediate factor.
In Paper III, formal education was included in the stratified analyses in addition to age group and
place of residence, based on the well-known association between educational level and people’s
health status and health behavior. In general this association means that a higher level of education
is associated with a healthier life. Formal education is therefore treated as an intermediate factor
and not a confounder.
40
The analysis of the association between the level of involvement in the ongoing social transition and
lifetime problem gambling in Paper I is an example of a complex variable that was constructed on the
basis of different sociodemographic information as well as childhood conditions. As a consequence it
would not make sense to adjust separately for these variables since they were already included
indirectly as exposures in the analyses.
Type-I and type II-errors
An incorrect rejection of the null-hypothesis, stating that there is no difference between the exposed
and unexposed, is called a type-I error, and the opposite scenario where the null-hypothesis is false,
but not rejected, is called a type-II error (67). Type-II errors are particularly relevant to consider,
when working with a relatively small sample size as it is typically the case in research regarding
indigenous populations in the Circumpolar North. Stratification and the inclusion of additional
variables to ensure proper adjustment often renders very few individuals in each strata. Type-II error
then occurs when an association between a given exposure and outcome is missed, because the level
of significance for a chi-2 test might not meet the conventional criteria of p<0.05 or because the 95%
CI is too large. This occurs because the effect is small and/or observations are too few as well as from
different sources of bias (67). As an example there are only two female ‘past year problem gamblers’
from Nuuk and eight male ‘past year problem gamblers’ from villages in the sample. As a
consequence, most analyses were carried out for lifetime problem gamblers rather than past year
problem gamblers and stratification and adjustment was limited to the variables that were
considered the most important.
Lessons to learn about measuring problem gambling
A general lack of testing and validation of screens used in epidemiological surveys combined with a
confusion of concepts creates a weak fundament for comparative research (4;16). The definition of
pathological gambling in the majority of the methodological screens used to investigate problem and
pathological gambling in epidemiological research has been imported from the DSM-IV-R criteria, but
the different levels of gambling problems have been defined for epidemiological use without
investigating the validity of the DMS-IV-R criteria in a population context outside the clinical setting.
Instead of working in one direction with the aim of developing a gold standard, several new screens
have been introduced, many of which have not been validated. As a consequence few prevalence
surveys are directly comparable (4;8). Despite this, the prevalence of pathological gambling across
countries and methods is remarkably similar. Based on this observation it has been suggested, that
the most severe gambling problems are caught despite differences in screening tools (4). However,
there is less consistency when it comes to the less severe gambling problems.
Given the brief character, the lie/bet screen is likely to cover milder gambling problems along with
more severe problems compared to a more comprehensive assessment such as the revised South
Oaks Gambling Screen (SOGS-R) (69) or the NORC DSM Screen for Gambling Problems (NODS) (16).
This is considered a weakness of the screen. Although the validation studies of the lie/bet screen
displayed satisfying levels of sensitivity and specificity, a short screen with a reduced number of
questions will inevitably give a result based on less information, and thereby a potentially less valid
result. However, short screens are an important tool if gambling is to be investigated in larger
population based health surveys because these surveys usually include several topics, which
challenges the limit for duration, as to how much time people can be expected to spend on
participating in a survey.
41
The NODS CLiP (70) and the Brief Biosocial Gambling Screen (BBGS) (71) are examples of the most
recent short screens for problem and pathological gambling targeted population based research, and
they have shown excellent sensitivity and specificity in the North American population samples. The
NODS CLiP is a three-item short version of the NODS and measures gambling problems in a lifetime
perspective. In comparison to the lie/bet screen it also includes the question about lying. The two
other questions concern, whether a person has tried to cut down or control gambling activities, and
whether a person has been preoccupied with thinking about gambling experiences or planning future
gambling activities. A positive answer to one of these three items indicates likely problem or
pathological gambling. The three-item BBGS was developed from the National Epidemiological
Survey on Alcohol and Related Conditions (72) and measures past year problem or pathological
gambling. In comparison to the lie/bet screen, the BBGS also includes a question about having lied to
friends or family. The two additional questions concern, symptoms of withdrawal when trying to cut
down on gambling, and borrowing money due to financial problems caused by gambling. A positive
answer to one of the three questions is considered as possible pathological gambling with a need for
clinical evaluation.
Having lied to friends and family is included in all three short screens, but other than that, all three
screens include different criteria from the DSM-IV-R. This in itself suggests that there is no obvious
gold standard in the short screening tools, since different samples have led to the development of
different short screens. To enhance the use of the existing short screens and to improve their validity
across samples, the short screens should be cross-validated and tested in more than one population,
including more vulnerable populations such as ethnic minority groups and indigenous populations.
It should also be carefully considered whether or not the DSM-IV criteria should be the gold standard
to validate the short screens against or if other aspects besides the ten DSM-IV criteria such as the
negative impact on communities and families could be included when screening for gambling
problems at the population level. This is especially relevant from a public health point of view.
Lesieur (73) has argued, that most epidemiological surveys seriously underestimate the extent of
problem and pathological gambling. This is due to the above stated lack of validity for many screens
as well as failure to include institutionalized and other specialized populations, and a lack of strategy
to handle the level of denial among problem gamblers (74). The critique still seems relevant, and
should be included considered in future methodological studies.
5.3 Understanding the nature of problem gambling
To provide a better understanding of the development of problem gambling the Pathways Model
suggested by Blaszczynski and Nower (75) was discussed in Paper I and III. The Pathways Model
argues that problem gamblers are a heterogeneous group with different backgrounds for developing
gambling problems and proposes three different pathways: a) One pathway covers problem
gamblers with no severe psychological problems prior to their extensive gambling, but where alcohol
abuse, depression and anxiety might appear as a result of their gambling; b) a second pathway
covers a group of problem gamblers that are characterized by an emotional vulnerability due to for
example childhood disturbance. This vulnerability is present before the gambling problems and
gambling is used to modify affective states or meet specific needs; c) the third pathway covers a
42
group of highly disturbed individuals with substantial psychological features such as impulsivity and
antisocial personality disorder.
The high prevalence of a harmful alcohol use and frequent use of marijuana among lifetime problem
gamblers found in Paper II should be considered in relation to the first pathway (a), which indicates
that substance abuse and alcohol disorders can be a consequence of problem gambling. It would
provide an important insight to investigate, if this was the case among the relatively large percentage
of persons who reported of lifetime problem gambling in a combination with a harmful alcohol use
and/or a frequent use of marijuana. Given the cross-sectional character of our data, it was not
possible to investigate, what came first, and how the etiology of pathological gambling could be
understood in relation to alcohol and substance use disorders. Few studies contribute with
knowledge in this field, and the results are based on retrospective questions asking participants to
remember their age of onset for different addictive behaviors (28).
The association between traumatic events during childhood and lifetime problem gambling, found in
Paper I, suggests that the second (b) and the third (c) pathway to problem gambling through
emotional vulnerability and mental health problems is a relevant perspective. These pathways should
be considered in further investigations, to gain a better understanding of the high prevalence of
lifetime problem gambling among Greenland Inuit. The high prevalence of people, who experienced
alcohol problems in their childhood home and were sexually abused during their childhood, can have
led to such emotional vulnerability in a large part of the population. This is particularly important to
consider for the birth cohorts born in the 1970s, where the prevalence of traumatic childhood events
has been found to be highest. Traumatic childhood events was found to be associated with mental
health problems among indigenous populations in the North in other studies (29;76;77), and a high
prevalence of mental health disorders among Greenland Inuit (29) adds to the argument, of why a
pathway to problem gambling through emotional vulnerability and mental health problems should
be investigated further.
However none of the pathways, discussed by Blaszczynski and Nower (75), were based on empirical
findings among indigenous populations, among whom the sociocultural and societal changes have
been documented as important determinant of both physical and mental health (24). The
associations found in Paper III between gambling and a poorer health as well as unhealthy behavior,
suggests that gambling problems could be a result of an even more complex process creating
inequalities in health that are related to living conditions and socioeconomic differences among
Greenland Inuit. This perspective follows the social epidemiological approach regarding social
determinants of health, which states that health follows a social gradient meaning the higher the
social position, the better the health (78-84). The focus of epidemiological research should in this
perspective be the causes of causes, such as the living conditions and the circumstances that
influence peoples’ quality of life and access to a healthy lifestyle. These causes are defined as social
determinants of health. Following this line of thought, public health research in Greenland should
focus on the social determinants of a poor health and unhealthy behavior among Greenland Inuit
and investigate how frequent gambling behavior and gambling problems are linked to social
determinants.
Although gambling is defined as a chronic and progressive disorder in the DSM-IV used as the gold
standard in most epidemiological research on gambling, the few prospective studies that have been
43
conducted, suggest pathological gambling is an intermittent problem rather than a chronic disease
and especially people with less severe gambling problems (not meeting clinical DSM-IV-R criteria)
have been found to recover from gambling problems on their own rather than moving on to
pathological gambling as the inevitable last stage (4;9;73;85). Thus symptoms of disordered gambling
can move in two directions: worse or better (4). The high one-year remission rate found among
lifetime problem gamblers in Paper I supports these observations. Among 74% of the men and 71%
of the women, who reported of gambling problems at some point in their life, did not report of any
past year problems. International studies consistently show that the past year prevalence of
gambling problems is about two-thirds of the lifetime prevalence, which has been suggested to
indicate a recovery rate of one third (8).
In a follow-up study in 1998 (62), Abbott et al. re-interviewed the pathological and problem gamblers
found in the first national gambling survey in New Zealand in 1991. At follow-up, most of the
problem gamblers were no longer problem gamblers, and only a minority had developed a more
serious problem and was classified as pathological gamblers. Slutske et al. (86) found problem
gambling to be transitory and episodic at the individual level rather than chronic, although the
prevalence at the aggregate level remained stable in a four-wave longitudinal study. Finally, a recent
review (87) of five prospective studies of gambling behavior found no evidence to support the
assumption that individuals cannot recover from disordered gambling. Neither did the review find
evidence supporting the assumption that individuals, who have more severe gambling problems, are
less likely to improve their condition compared to individuals, who experience less severe gambling
problems. Based on these conclusions across studies, the ongoing discussions of how to define
pathological gambling in the upcoming fifth version of the Diagnostic and Statistical Manual of
Mental Disorders (planned for publication in 2013) seem highly relevant for the understanding of and
conceptualization of problem and pathological gambling in future epidemiological studies.
The proposed changes imply a new categorization of pathological gambling as a behavioral addiction;
changing the name into ‘Disordered Gambling’; eliminating the criterion of ‘illegal acts’ and reducing
the threshold for criteria to discriminate between pathological vs. non-pathological from five to four
(88;89). Similar discussions were carried out in the 1990s during the preparations of the DSM-IV. A
literature review carried out by Lesieur and Rosenthal (90) then found little data on the validity and
the classification accuracy of the DSM-III-R criteria, which are almost similar to the current DSM-IV-R
criteria. A discussion of the validity and the relevance of the new DSM-V criteria for population-based
research should be discussed simultaneously in the field of epidemiological gambling studies.
5.4 Problem gambling in the light of the ongoing social transition
The analyses of gambling behavior and problem gambling in relation to the ongoing social transition
in Greenland showed lower odds ratios for lifetime problem gambling among both hunters and
fishermen in villages as well as among professionals in towns (Paper I). The measure of different
levels of social transition was based on information regarding the participant’s current place of
residence and childhood residence combined with formal education and family job type. These
differences in lifetime problem gambling between hunters and fishermen in villages compared to
other villagers; and between professionals in towns compared to the three less educated or unskilled
groups in towns, could indicate that there are important social differences between different groups
44
based on childhood conditions combined with place of residence, education and occupation in adult
life. Social differences within the population of Greenland Inuit seem relevant for the understanding
of the prevalence of problem gambling and health inequalities in general based on the documented
importance of social determinants on health (80-83).
The challenge is to gain a better understanding of how social transition might influence the
constitution of differences between social groups in modern Greenland. The lower rates of problem
gambling found among professionals in towns in this study (Paper I) and earlier findings of a better
mental health among groups that have integrated successfully into modern Greenlandic society
(24;77) could indicate the importance of education combined with living in the capital or a larger
town with access to a wider range of opportunities as a successful strategy towards social
integration. Social integration is generally considered an important determinant of health in social
epidemiological theory (91-95) and the lack of social integration can be attributed to an imbalance in
society often seen in relation to fundamental societal changes. The concept of social integration has
also been discussed in earlier studies among indigenous peoples in the circumpolar North in relation
to acculturative stress, as an explanation of the relationship between cultural changes and mental
health (23;24;77;96). The theory of acculturative stress proposed by Berry and Kim (97) defines four
different modes of acculturation: integration, separation, assimilation and marginalization resulting
from how the importance of maintain one’s own cultural identity and the relationship with other
groups is perceived. Integration is considered to be associated with the lowest levels of acculturative
stress, but all four modes seem to imply a negative outcome of cultural changes. However the theory
has not be empirically verified among Greenland Inuit, and it has been suggested that acculturation
as defined by Berry, plays a lesser role for mental health among Greenland Inuit (77;98).
The similar low rates of lifetime problem gambling among families of hunters and fishermen in
villages could in this perspective be explained by their social integration in a more traditional Inuit
lifestyle that is still prevalent in some villages and smaller towns. Hunters and fishermen are naturally
integrated in village life where traditional Inuit lifestyle is still prevalent as are professionals in towns
where life reflects the modern Greenlandic Society. In contrast village life is more challenging if you
are not hunting and fishing because very few educational and occupational opportunities are
available. So is the life in a larger town, if you are not skilled or well educated, because expectations
and demands for qualifications rise with the level of specialization in modern society.
People in towns in Greenland might generally be considered to have the most difficult position in the
ongoing social transition because they are caught in the middle of the more traditional life in the
villages and the modern life in the capital, but the differences between social groups within villages
and towns defined in relation to the ongoing social transition, indicate an even more complex
process, which seem to depend on a combination of educational and occupational skills combined
with local living conditions and opportunities. The construction of a variable that measures the
individual’s level of involvement in the ongoing social transition has tried to incorporate these
differences.
Although the causality of the association is unclear, it seems as if people caught between modern
and traditional lifestyle are more vulnerable to problem gambling because they are neither
integrated in the traditional or the modern Greenlandic society, which to some extent coexist. The
measure for social transition applied in this study has not been used to investigate differences in
45
health in other Inuit populations. Although mental symptoms were included in one study of
cardiovascular risk factors in relation to social transition (36) the measures has not yet been applied
systematically in the study of social pathologies among Greenland Inuit other than the present study
of problem gambling. However, measuring the level of involvement in the ongoing social transition
as a way to capture social differences between groups seems to be a relevant approach for further
investigations of health inequalities among Greenland Inuit. This measure has the potential to
capture even more aspects of the possible social determinants of health compared to the common
indicators of socioeconomic position (99-103). Common indicators such as level of education,
occupation and income are without doubt important, and have been demonstrated to influence
health among Greenland Inuit (24), but the differences between life in a village compared to a town,
or life in East Greenland compared to life in South and Midwest Greenland, childhood conditions and
family history are not directly reflected in these indicators.
Due to the rapid changes, the social, health and living conditions can also vary fundamentally from
generation to generation. This was illustrated by the significant differences in the prevalence of
traumatic events during childhood according to birth cohorts (Paper I). The different living
conditions, possibilities and demands that characterize the life of different generations and the
changes over the lifecourse are important perspectives to include when trying to measure social
differences in a population undergoing social transition. This line of thought agrees with recent
discussions in social epidemiology with a focus on the lifecourse (99;100;104). It has been argued,
that the understanding of socioeconomic patterns in adult health risks to be severely limited, if we
ignore how exposures and resources may cascade and accumulate over the lifecourse to effect adult
health (101). An understanding of the complexity underlying differences between social groups and
birth cohorts among Inuit in Greenland is a prerequisite for future investigations of the social
determinants of poor health and unhealthy behavior among Greenland Inuit and how gambling
behavior, problem gambling and other social pathologies are linked to these determinants in the
light of the ongoing social transition.
46
6. CONCLUSION
Gambling is a part of everyday life in Greenland, and the majority of Greenland Inuit engaged in
gambling during past year. The patterns of gambling behavior reflected a preference for both
modern games that are popular in contemporary Scandinavia (the lotteries), as well as games found
to be traditionally popular among Inuit in Nunavik and First Nations (bingo, cards/dice). Men
gambled more on the lotteries, slot machines and cards/dice than women, while women gambled
more on bingo than men. Problem gambling had occurred at some point in life for 10% women and
16% men, but only about 4% had experienced these problems during past year.
Lifetime problem gambling was associated with the level of social transition, illustrated by a lower
prevalence of lifetime problem gambling among hunters and fishermen in villages and among
professionals in towns, compared to other groups. Having experienced traumatic events during
childhood also increased the odds ratio for lifetime problem gambling.
Lifetime problem gambling was more often than not combined with a harmful use of alcohol,
frequent use of marijuana or both. Among 53% men and 37% women and in almost half of all
households with children, one or more addictive behaviors were present. The health status and
health behavior among frequent gamblers and lifetime problem gamblers was generally poorer and
less healthy, compared to non-problem gamblers and those who never gambled.
The association between social transition and lifetime problem gambling may suggest, that people
caught between the traditional and modern lifestyles, which co-exist in Greenland, are more
vulnerable to gambling problems. Social integration in either a traditional life in the villages or a
modern life in the capital or larger towns could contribute to the understanding of differences
according to level of social transition, while emotional vulnerability caused by traumatic events
during childhood could be a potential pathway for gambling problems among Greenland Inuit. A
poorer health and less healthy behavior among frequent gamblers and lifetime problem gamblers
further suggests that gambling problems could be linked to more fundamental inequalities in health
among Greenland Inuit. However the causality of these associations cannot be established based on
the cross-sectional data in this thesis and should be investigated in future prospective studies.
47
48
7. IMPLICATIONS
7.1 For research
The findings in this thesis indicate that the level of social transition and traumatic events during
childhood could be important aspects of understanding the prevalence and distribution of problem
gambling among Greenland Inuit. Prospective studies are needed to investigate the causal
relationship of social transition, traumatic childhood events and problem gambling. The study of
incidence among vulnerable populations is still a road yet to be taken.
Prospective and register-based studies are also needed to investigate the causal relationship in the
overlap found in this thesis between problem gambling, harmful alcohol use and frequent use of
marijuana. This would lead to a better understanding of what comes first, and how the prevention
and treatment of one might prevent other addictive behaviors.
The association between frequent gambling, problem gambling and health found in this thesis should
be investigated in future register-based and prospective studies in order to identify the social
determinants of inequalities in health and how these are linked to the prevalence of problem
gambling and frequent gambling.
Future qualitative studies should investigate, what motivates people to gamble in Greenland and
what contributes to the increase or decrease in gambling frequency? We need to consider, what
alternative activities are available on a local basis. A review of the literature regarding gambling
among the North American aboriginal populations indicated lack of access to leisure activities in the
reservations as an important motivating factor for engaging in gambling activities (5). This could also
be the case in smaller communities in Greenland and should be investigated further. If the majority
of a smaller community engages in gambling it might be an important part of the social life. Different
needs and motivations according to age and gender would be interesting perspectives to include.
Short screens for measuring problem at the population level are an important tool in epidemiological
research, but the short screens need to be validated in different populations and more research
should be put into investigating, how people understand and respond to the questions based on the
DSM-IV-R criteria. From a public health point of view, it is important to consider whether the clinical
criteria for pathological gambling are the most relevant or whether we need a revised
methodological framework that can capture the consequences for families and children affected by
gambling problems.
There is a need to develop a valid and meaningful paradigm for social position among Inuit in
Greenland and more generally among northern indigenous populations for use in epidemiological
research of change in health conditions. Such measures should build on a lifecourse perspective and
include socioeconomic variables, intergenerational differences and local living conditions, and
incorporate people’s position in relation to the ongoing social transition. This work can build on the
existing measure for social transition used in this thesis, but should also include qualitative studies
and longitudinal data in order to gain a better understanding of how social transition is experienced
over time by different social groups across different communities and regions in Greenland.
49
Future research should also contribute to a better understanding of why undergoing rapid social
transition has led to an increase in social pathologies across indigenous populations in order to
prevent a further increase of problem gambling as well as alcohol use disorder, substance use
disorders and mental health problems in the future. This research should look into why some
population groups seem caught between traditional and modern way of life, while others integrate
well. This would contribute to the understanding of the causes behind the causes and social
determinants of both mental and physical health, as well as the importance of social integration (as
defined in the social epidemiological literature) for health among Greenland Inuit, with relevance for
other indigenous populations in the North.
7.2 For public health
Problem gambling should be included systematically in public health programs and more specifically
in strategies for treatment and prevention targeted the field of addictive behaviors, which at present
has only included focus on alcohol and marijuana among adults.
Treatment strategies should incorporate a diverse understanding of problem gambling as well as
other addictive behaviors, because more and more evidence points towards the heterogeneity of
these groups, and thus different needs should be met in order for prevention and treatment to have
an impact. This includes the perspective of different pathways to problem gambling, poor health, and
unhealthy behavior as well as other addictive behaviors given the associations with social transition
and traumatic events during childhood. The higher prevalence of problem gambling among those,
who often experienced alcohol problems in their childhood home and were sexually abused are
examples of traumatic events that should be taken into account in treatment strategies. The
prevalence of combined addictive behaviors is another example.
One or more addictive behaviors were found in almost half of all families with children. Prevention
targeted families seems crucial to increase the quality of life among future generations and reduce
the prevalence of social pathologies. Improving the lives of families in Greenland is already a high
priority in public health strategies, but prevention and treatment of gambling problems have not
been included systematically in these strategies yet.
The association between alcohol, hash and marijuana suggest that health professionals and social
workers should consider including assessment for all three addictive behaviors when working with
people in need of treatment.
50
8. SUMMARY
During recent years gambling behavior and pathological gambling has become a much-debated
subject in Greenland. The economic and social problems for individuals, families and communities,
which extensive gambling can lead to are of great concern. In less than 100 years Greenland has
developed from a traditional subsistence based economy to a modern society. The rapid transition
places immense social and cultural demands on both individuals and communities. An increase in
social pathologies such as violence, suicide, alcohol and substance use is a key feature shared by
indigenous populations undergoing similar social transition and represent major challenges for public
health in Greenland. More knowledge was needed to determine the role of gambling in Greenland.
The objective of the thesis was to investigate the prevalence of gambling behavior and problem
gambling in a representative sample of Greenland Inuit and its association with social transition,
addictive behaviors and health. The overall objective was investigated in three Papers.
Data were derived from the Inuit Health in Transition Greenland Survey carried out in 9 towns and 13
villages from 2005-2010 in Greenland. The survey combined clinical examinations with an
interviewer based questionnaire and a self-administered questionnaire and obtained an overall
participation rate of 64.9%.
The main findings across papers were i) a high prevalence of lifetime problem gambling among
Greenland Inuit (men 16%; women 10%); ii) a significant association between gambling behavior and
problem gambling with social transition and traumatic events during childhood; iii) a significant cooccurrence of problem gambling with other addictive behaviors and iv) significant associations
between selected measures of health status and health behavior with both gambling behavior and
lifetime problem gambling.
Based on the main findings it is argued that problem gambling can be added to the list of social
pathologies in Greenland, and it is suggested that people caught between tradition and modern ways
of life are more vulnerable to gambling problems. Furthermore it is argued that problem gambling
should be included systematically in future public health strategies, treatment programs and
interventions in Greenland because there is a prevalent overlap between lifetime problem gambling
with a harmful alcohol use and frequent use of marijuana. Finally it is suggested, that the association
between frequent gambling, problem gambling and health could indicate that gambling and problem
gambling among Greenland Inuit is linked to a complex process closely related to social determinants
of health, and thus an important issue for public health intervention, in order to reduce social
inequalities in health.
Implications for research and public health are discussed in relation to the main findings.
51
9. RESUMÉ
Gennem de seneste år har pengespil og ludomani været et tilbagevendende emne i den offentlige
debat i Grønland. Der har især været fokus på de økonomiske og sociale problemer som ludomani
kan føre med sig. Både for den enkelte, men også for familien og lokalsamfundet. På mindre en 100
år har Grønland udviklet sig fra et traditionelt fangersamfund til et moderne samfund. Den hurtige
udvikling stiller store sociale og kulturelle krav til den enkelte og til samfundet. En øget forekomst af
de såkaldte sociale patologier så som vold, selvmord, alkohol- og hashmisbrug er et vilkår, der deles
af oprindelige befolkninger i social transition. Disse forhold udgør en stor udfordring for
folkesundheden i Grønland, og der var behov for mere viden om hvilken rolle pengespil og ludomani
spillede i denne sammenhæng.
Afhandlingen havde til formål at undersøge forekomsten af pengespil samt ludomani og
sammenhængen med social transition, misbrugsadfærd og helbred i et repræsentativt udsnit af
grønlændere. Dette formål er blevet undersøgt gennem tre videnskabelige artikler.
Afhandlingen er baseret på data fra Befolkningsundersøgelsen i Grønland 2005-2010 (Inuit Health in
Transition Greenland Survey). Data er indsamlet i 9 byer og 13 bygder i forskellige dele af Grønland
og omfattede kliniske undersøgelser, et interviewerbaseret spørgeskema samt et selvudfyldt
spørgeskema. Den overordnede deltagerprocent var 64.9%.
Afhandlingen viser, i)at forekomsten af livstids problem spil blandt grønlændere er høj (mænd 16%;
kvinder 10%); ii) at der er sammenhæng mellem pengespil og problem spil med social transition; iii)
at der er et forholdsvist stort overlap med andre former for misbrugsadfærd blandt problemspillere;
iv) samt en generel forekomst af dårligere helbred og usunde vaner blandt dem, der spiller ofte og
problemspillere.
Baseret på disse resultater argumenteres der i afhandlingen for, at problemspil kan tilføjes til listen
over sociale patologier i Grønland, ligesom det kunne tyde på, at folk der er fanget mellem en
traditionel og en moderne livsstil er mere sårbare overfor problemer med pengespil. Der
argumenteres desuden for, at ludomani og problemspil bør inkluderes systematisk i
folkesundhedsstrategier, behandlingsindsatser og interventioner fremover, fordi sammenfaldet
mellem problemspil og andre former for misbrugsadfærd er relativt stort. En høj forekomst af
misbrug i husstande med børn gør det ligeledes nødvendigt at målrettet disse indsatser mod udsatte
familier. Afslutningsvist argumenteres der for, at sammenhængen mellem hyppigt pengespil og
problemspil med mere generelle helbredsforhold kan være tegn på, at forekomsten af problemspil er
knyttet til en række grundlæggende levevilkår og sociale forhold, der indvirker på sundheden. Dette
understreger vigtigheden af at inddrage problemspil i kommende folkesundhedsprogrammer med
henblik på at reducere den sociale ulighed i sundhed i Grønland.
Resultaternes betydning for folkesundhedsområdet samt behovet for mere viden diskuteres i
afhandlingen.
52
10. EQIKKAANEQ
Ukiuni kingulliunerusuni aningaasanoorajunneq aningaasanoornermullu nakkaassimaneq Kalaallit
Nunaanni oqallisaallattaajualerpoq. Aningaasanoornermut nakkaassimanerup inunnut ataasikaanut,
ilaqutariinnut najukkamilu innuttaasunut, aningaasaqarnikkut inuttullu atukkatigut sunniutai,
pingaarnertut qitiutinneqarput. Ukiut untritillit inortut ingerlanerini Kalaallit Nunaat piniartuunermiit
inooriaatsimut nutaamut ikaarsaarpoq. Inuiaqatigiit taama sukkatigisumik ineriartornerat, kulturikkut
inuttullu inuunermut- inunnut ataasiakkaanut inuiaqatigiinnullu- annertuumik
piumasaqaatitaqarpoq. Inuiaqatigiit nappaataannik taaneqartartut soorlu nakuuserneq, imminut
toqunneq, imigassamik hashimillu atornerluineq annertuumik atugaasut nunallu inooqqavini allani
aamma atuuttut, peqqissutsimut annertuumik unammillernartuupput.
Ilisimatusaatitut allaaserisap: 1) aningaasanoorajunnerup inuuneq tamaat isigalugu kalaallit
akornanni annertuumik atugaanera (angutit 16%-ii; arnat 10 %-ii); 2) aningaasanoorajunnerup
atornerluinernik allanik ilaqartinneqartarnera; 3)kiisalu peqqinnginnerup piqqinnanngitsunillu
ileqqoqarnerup aningaasanoorajuttuni aningaasanoornermullu nakkaassimasuni nalinginnaasuunera
takutippaa.
Ilisimatusaatitut allaaserisap inerneri tunngavigalugit, aningaasanoornermut nakkaassimanerup
Kalaallit Nunaanni inuiaqatigiit nappaataat pillugit nalunaarsukkanut ilanngunneqarsinnaanera
tunngavilersorneqarpoq, taamatuttaaq inuit qangatut inooriaatsip inooriaatsillu nutaap akornanni
unittoortut aningaasanoorajunnermik ajornartorsiuteqalernissamut
sanngiiffeqarnerusorinarsinnaapput. Aningaasanoornermut nakkaassimanerup atornerluinerillu allat
ataqatigiinnerat annertungaatsiarmat, siunissami aningaasanoorajunnerup aningaasanoornermullu
nakkaassimanerup innuttaasut peqqissuunissaat pillugu suliniutinut, katsorsaanernut
allannguiniarlunilu akuliunnernut ilaatinneqartariaqarnissaat tunngavilersorneqarportaaq.
Ilisimatusaatitut allaaserisaq 2005-imiit 2010-mut Kalaallit Nunaanni innuttaasut peqqissusaannik
misissuisitsinermi paasissutissanik tunngaveqarpoq (Inuit Health in Transition Greenland Survey).
Paasissutissat Kalaallit Nunaanni nunap immikkoortuini assigiinngitsuni- illoqarfinni qulaaluani
nunaqarfinnilu 13-ini inersimasut akornanni- katersorneqartut, misissortinnerupput apersuilluni
immersuinerullutik namminerlu apeqqutinik akiuilluni immersuinerullutik. Ataatsimut katillugit
peqataasut 64,9 %-iupput.
53
54
11. REFERENCES
(1) PAARISA - Department of Health GG. Gambling and gambling patterns [In Danish]. Nuuk:
PAARISA - Government of Greenland; 2006.
(2) Wardle H, Moody A, Spence S, Orford J, Volberg R, Jotangia D. British Gambling Prevalence
Survey 2010. Prepared for The Gambling Commission. London: National Centre for Social
Research.; 2011.
(3) Abbott M, Volberg R. Taking the pulse on gambling and problem gambling in New Zealand: A
report on phase one of the 1999 national prevalence survey. Report Number Three of the
New Zealand Gaming Survey. Wellington: The Department of Internal Affairs; 2000.
(4) Shaffer HJ, Labrie RA, Laplante DA, Nelson SE, Stanton MV. The road less travelled: moving
from distribution to determinants in the study of gambling epidemiology. Canadian Journal
of Phychiatry 2004 Aug;49(8):504-16.
(5) Wardman D, el-Guebaly N, Hodgins D. Problem and pathological gambling in North American
Aboriginal populations: a review of the empirical literature. Journal of Gambling Studies
2001;17(2):81-100.
(6) Bjerregaard P, Young K, Dewailly E, Ebbesson SOE. Indigenous health n the Arctic: an
overview of the circumpolar Inuit population. Scandinavian Journal of Public Health
2002;32:390-5.
(7) Kruse J, Poppel B, Abryutina L, Duhaime G. Survey of Living Conditions in the Arctic: What did
we learn? Anchorage: Institute of Social and Economic Research, University of Alaska
Anchorage; 2007.
(8) Hodgins DC, Stea JN, Grant JE. Gambling disorders. Lancet 2011 Nov 26;378(9806):1874-84.
(9) Shaffer HJ, Hall MN. The natural history of Gambling and Drinking Problems Among Casino
Employees. The Journal of Social Psychology 2002;142(4):405-24.
(10) Landstingslov nr. 13 af 11. november 2000 om fordeling af Tips og lottomidler, Government
of Greenland, (2000).
(11) Landstingslov nr. 7 af 31. oktober 1987 om lotteriafgifter, Government of Greenland, (1987).
(12) Landstingslov nr. 19 af 30. oktober 1992 om indførselsafgifter (afgift på spillemaskiner),
Government of Greenland, (1992).
(13) Landstingslov nr. 3 af 17. maj 1990 om afgift på automatspil, Government of Greenland,
(1990).
(14) American Psychiatric Association. Diagnostic and Statistical Manual of mental disorders.
Revised 3rd edition ed. Washington DC: American Psychiatric Association; 1987.
(15) American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders Text Revision. 4th Edition ed. 2000.
55
(16) Gerstein DR, Arwood RA, Hoffmann J, Larison C, Engelman L, Murphy S. Gambling impact and
behavior study. Report to the National Gambling Impact Study Commission. Chicago:
National Opinion Research Center at the University of Chicago.; 1999.
(17) Bonke J, Borregaard K. The prevalence of problematic gambling behaviour: a Scandinavian
comparison. Scandinavian Journal of Public Health 2009 Aug;37(6):654-60.
(18) Wardle H, Sproston K, Orford J, Erens B, Griffiths M, Constantine R, et al. British Gambling
Prevalence Survey 2007. London: National Centre for Social Research; 2007.
(19) Korn D, Gibbins R, Azmier J. Framing public policy towards a public health paradigm for
gambling. Journal of Gambling Studies 2003;19(2):235-56.
(20) Korn DA. Examining Gambling Issues from a public health perspective. 1-18. 23-6-2002.
(21) Shaffer HJ, Korn DA. Gambling and related mental disorders: a public health analysis. Annual
Review of Public Health 2002;23:171-212.
(22) Young TK, Bjerregaard P. Health Transitions in Arctic Populations. Toronto: University of
Toronto Press; 2008.
(23) Berner J, Furgal C, Bjerregaard P, Bradley M, Curtis T, De Fabo E, et al. ACIA, 2005. Arctic
Climate Impact Assessement. New York: Cambridge University Press; 2005.
(24) Curtis T, Kvernmo S, Bjerregaard P. Changing living conditions, life style and health. Journal of
Circumpolar Health 2005;64(5):442-50.
(25) Lynge I, Munk-Jørgensen P, Lynge Pedersen A, Mulvad G, Bjerregaard P. Mental health
among patients in the Greenlandic health system [In Danish]. Copenhagen: SIFs Writings on
Greenland no. 15. National Institute of Public Health; 2003.
(26) Statistics Greenland. Population in Greenland. 2012. 12-10-2012.
(27) Statistics Greenland. Income in Greenland. 2012. 12-10-2012.
(28) Statistics Greenland. Import of alcohol to Greenland 1975-2001 [In Danish]. Nuuk: Statistics
Greenland; 2002.
(29) Lynge I, Munk-Jørgensen P, Lynge Pedersen A, Mulvad G, Bjerregaard P. Common mental
health disorders among patients in primary health care in Greenland. International Journal of
Circumpolar Health 2004;63(Supplements 2):377-83.
(30) Lehti V, Niemel+ñ S, Hoven C, Mandell D, Sourander A. Mental health, substance use and
suicidal behaviour among young indigenous people in the Arctic: A systematic review. Social
Science & Medicine 2009 Oct;69(8):1194-203.
(31) Kirmayer LJ, Brass GM, Tait CL. The mental health of Aboriginal peoples: transformations of
identity and community. Canadian Journal of Phychiatry 2000 Sep;45(7):607-16.
56
(32) Bjerregaard P, Jeppesen C. Inuit dietary patterns in modern Greenland. International Journal
of Circumpolar Health 2010;69(1):13-24.
(33) Jeppesen C, Jørgensen ME, Bjerregaard P. Assessement of consumption of marine food in
Greenland by a food frequency questinnaire and biomarkers. Journal of Circumpolar Health
2012;71(doi:10.3402/ijch.v71i0.18361).
(34) Young TK, Bjerregaard P, Dewailly E, Risica PM, Jørgensen ME, Ebbesson SOE. Prevalence of
Obesity and Its Metabolic Correlates Among the Circumpolar Inuit in 3 Countries. American
Journal of Public Health 2007;97(4):691-5.
(35) Jørgensen ME, Borch-Johnsen K, Bjerregaard P. Diabetes in Greenland and its relationship
with urbanization. Diabetes Medicine 2012;29:755-60.
(36) Bjerregaard P, Dahl-Petersen IK. How well does social variation mirror secular change in
prevalence of cardiovascular risk factors in a country in transition? Amerian Journal of
Human Biology 2011 Nov;23(6):774-9.
(37) Dahl-Petersen IK, Jorgensen ME, Bjerregaard P. Physical activity patterns in Greenland: a
country in transition. Scandinavian Journal of Public Health 2011 Nov;39(7):678-86.
(38) Bjerregaard P. Inuit Health in Transition - Greenland Survey 2005-2010. Population sample
and survey methods. 2nd revised edition. Copenhagen: National Institute of Public Health;
2011.
(39) Johnson EE, Hammer, R. The Lie/Bet Questionnaire for Screening Pathological Gamblers.
Psychological Reports 1997;80:83-8.
(40) Johnson EE, Hammer RM. The Lie/Bet Questionnaire for Screening Pathological Gamblers: A
Follow-Up Study. Psychological Reports 1998;83:1219-24.
(41) Götestam GK, Johansson A, Wenzel HG, Simonsen IE. Validation of the Lie/Bet Screen for
Pathological Gambling on Two Normal Population Data Sets. Psychological Reports
2004;95:1009-13.
(42) Lesieur HR, Blume SB. The South Oaks Gambling Screen (SOGS): A New Instrument for the
Identification of Pathological Gamblers. American Journal of Psychiatry 1987;144(9):1184-8.
(43) Lesieur HR, Blume SB. Revising the South Oaks Gambling Screen in different settings. Journal
of Gambling Studies 1993;9(3):213-23.
(44) Korn D, Gibbins R, Azmier J. Framing Public Policy Towards a Public Health Paradigm for
Gambling. Journal of Gambling Studies 2003;19(2):235-56.
(45) Hodgins DC, Stea JN, Grant JE. Gambling disorders. Lancet 2011 Nov 26;378:1874-84.
(46) Ferris J, Wynne H. The Canadian Problem Gambling Index: Final Report. Canadian Centre on
Substance Abuse (CCSA); 2001 Feb. Report No.: Final .
(47) Gerstein DR, arwood RA, Hoffmann J, Larison C, Engelman L, Murphy S, et al. Gambling
impact and behavior study. Report to the National Gambling Impact Study Commission.
National Opinion Research Center at the University of Chicago; 1999 Apr.
57
(48) Toce-Gerstein M, Gerstein DR, Volberg RA. The NODS-CLiP: A Rapid Screen for Adult
Pathological and Problem Gambling. Journal of Gambling Studies 2009 Dec;25(4):541-55.
(49) Gebauer L, LaBrie R, Shaffer HJ. Optimizing DSM-IV-TR Classification Accuracy: A Brief
Biosocial Screen for Detecting Current Gambling Disorders Among Gamblers in the General
Household Population. Canadian Journal of Psychiatry-Revue Canadienne de Psychiatrie 2010
Feb;55(2):82-90.
(50) Mayfield D, McLeod G, Hall P. The CAGE questionnaire: Validation of a new alcoholism
screening instrument. American Journal of Psychiatry 1974;131(10):1121-3.
(51) Bisson J, Nadeau L, Demers A. The validity of the CAGE scale to screen for heavy drinking and
drinking problems in a general population survey. Addiction 1999;94:715-22.
(52) McCusker MT, Basquille J, Khwaja M, Murray-Lyon IM, Catalan J. Hazardous and harmful
drinking: a comparison of the AUDIT and CAGE screening questionnaires. The Quarterly
journal of medicine 2002;95:591-5.
(53) MacKenzie DM, Langa A, Brown TM. Identifying hazardous or harmful alcohol use in medical
admissions: a comparison of AUDIT, CAGE and BRIEF MAST. Alcohol and Alcoholism
1996;31:591-9.
(54) Zierau F, Hardt J, Henriksen H, Holm SS, Jørring S, Melsen T, et al. Validation of a selfadministered modified CAGE test (CAGE-C) in a somatic hospital ward: Comparison with
biochemical markers. Scandinavian Journal of Clinical Laboratory Investigation 2005;65:61522.
(55) World Health Organization. The ICD-10 classification of mental and behavioral disorders.
Clinical descriptions and diagnostic guidelines. Geneva: World Health Organization; 1992.
(56) Madsen MH, Grønbæk M, Bjerregaard P, Becker U. Urbanization, migration and alcohol use
in a population of Greenland Inuit. International Journal of Circumpolar Health 2005;64:23445.
(57) Kirkwood BR, Sterne JAC. Essential Medical Statistics. Second Edition ed. Oxford: Blackwell
Science; 2003.
(58) Gordis L. Epidemiology. Foruth edition ed. The United States of America: Saunders Elsevier;
2009.
(59) Lund I, Nordlund S. Gambling and gambling problems (In Norwegian). Norwegian Institute for
Alcohol and Drug Research (SIRUS); 2003.
(60) SWELOGS. The Swedish longitudinal gambling study 2008/2009. Main results from the
population study [In Swedish]. Sweden: Swedish National Institute of Public Health; 2011.
(61) Bonke J, Borregaard K. Pathological gambling in Denmark. The prevalence of gambling and
problem gambling [In Danish]. Copenhagen: The Danish National Centre for Social Research;
2006.
(62) Abbott M, Williams MM, Volberg R. A prospective study of Problem and Regular Nonproblem
Gamblers Living in the Community. Substance Use & Misuse 2004;39(6):855-84.
58
(63) Welte J, Barnes G, Wieczorek W, Tidwell M-C, Parker J. Alcohol and Gambling Pathology
among U.S. Adults: Prevalence, Demographic Patterns and Comorbidity. Journal of Studies on
Alcohol and Drugs 2001;62(5):706-12.
(64) Lorrains FK, Cowlishaw S, Thomas SA. Prevalence of comorbid disorders in problem and
pathological gambling: systematic review and meta-analysis of population surveys. Addiction
2012;106:490-8.
(65) Morasco BJ, Petrzak RH, Blanco C, Grant BF, Hasin D, Petry NM. Health problems and medical
utilization associated with gambling disorders: results from the National Epidemiological
Survey on Alcohol and Related Conditions. Psychosomatic Medicine 2006;68(6):976-84.
(66) Black DW, Shaw M, McCormick B, Allen J. Pathological gambling: relationship to obesity, selfreported chronic medical conditions, poor lifestyle choices, and impaired quality of life.
Comprehensive Psychiatry 2012;In press.
(67) Rothman KJ, Greenland S, Lash TL. Modern Epidemiology. Third edition ed. Philadelphia:
Lippincott Williams & Wilkins; 2008.
(68) Ekholm O, Hesse U, Davidsen M, Kjøller M. The study design and characteristics of the Danish
national health interview surveys. Scandinavian Journal of Public Health 2009;37(7):758-65.
(69) Lesieur HR, Blume SB. Revising the South Oaks Gambling Screen in different settings. Journal
of Gambling Studies 1993;9(3):213-23.
(70) Toce-Gerstein M, Gerstein DR, Volberg R. The NODS-CLiP: a rapid screen for adult
pathological and problem gambling. Journal of Gambling Studies 2009;25:541-55.
(71) Gebauer L, LaBrie R, Shaffer HJ. Optimizing DSM-IV-TR classification accuracy: a brief
biosocial screen for detecting current gambling disorders among gamblers in the general
household population. Canadian Journal of Psychiatry 2010;55:82-90.
(72) Petry NM, Stinson FS, Grant BF. Comorbidity of DSM-IV pathological gambling and other
psychiatric disorders: results from the National Epidemiological Survey on Alcohol and
Related Conditions. Journal of Clinical Psychiatry 2005;66:564-74.
(73) Lesieur HR. Epidemiological Surveys of Pathological Gambling: Critique and Suggestions for
Modification. Journal of Gambling Studies 1994;10(4):385-98.
(74) Rockloff MJ, Ehrich J, Themessl-Huber M, Evans LG. Validation of One Item Screen for
Problem Gambling. Journal of Gambling Studies 2011;27:701-7.
(75) Blaszczynski A, Nower L. A pathways model of problem and pathological gambling. Addiction
2002 May;97(5):487-99.
(76) Curtis T, Larsen FB, Helweg-Larsen K, Bjerregaard P. Violence, sexual abuse and health in
Greenland. International Journal of Circumpolar Health 2002;61(2):110-22.
(77) Bjerregaard P, Curtis T. Cultural change and mental health in Greenland. The association of
childhood conditions, language and urbanization with mental health and suicidal thoughts
among the Inuit of Greenland. Social Science and Medicine 2002;(54):-33.
59
(78) Commission on Social Determinants of Health. Closing the gap in a generation: health equity
through action on the social determinants of health. Final report of the Commission on Social
Determinants of Health. Geneva: World Health Organization; 2008.
(79) Berkman L, Kawachi I. Social Epidemiology. New York: Oxford University Press; 2000.
(80) Marmot M, Wilkinson RG. Social Determinants of Health. Second edition ed. New York:
Oxford University Press; 2011.
(81) Marmot M. Mulitilevel Approaches to Understanding Social Determinants. In: Berkman L,
Kawachi I, editors. Social Epidemiology.New York: Oxford University Press; 2000. p. 349-67.
(82) Marmot M. Social determinants of health inequalities. Lancet 2005;(365):1099-104.
(83) Diderichsen F, Evans T, Whitehead M. The Social Basis of Disparities in Health. In: Evans T,
Whitehead M, Diderichsen F, Bhuiya A, Wirth M, editors. Challenging inequities in health.
From ethics to action.New York: Oxford University Press; 2001. p. 13-23.
(84) Davey SG, Krieger N. Tackling health inequalities. British Medical Journal 2008;337:529-30.
(85) Abbott M. What do we know about gambling and problem gambling in New Zealand? Report
Number Seven of the New Zealand Gaming Survey. Wellington: The Department of Internal
Affairs; 2001.
(86) Slutske WS, Jackson KM, Sher KJ. The natural history of problem gambling from age 18 to 29.
Journal of Abnormal Psychology 2003;112:263-74.
(87) Laplante DA, Nelson SE, Labrie RA, Shaffer HJ. Stability and Progression of Disordered
Gambling: Lessons From Longitudinal Studies. Canadian Journal of Phychiatry 2008;53(1):5260.
(88) Potenza M. Should addictive disorders include non-substance-related conditions? Addiction
2006;101:142-51.
(89) Mitzner GB, Whelan JP, Meyers AW. comments from the Trenches: Proposed Changes to the
DSM-I Classification of Pathological Gambling. Journal of Gambling Studies 2011;27:517-21.
(90) Lesieur HR, Rosenthal RJ. Pathological gambling: A Review of the Literature (Prepared for the
American Psychiatric Association Task Forse on DSM-IV Committee on Disorders of Impulse
Control Not Elswhere Classified. Journal of Gambling Studies 1991;7(1):3-39.
(91) Berkman L, Kawachi I. A Historical Framework for Social Epidemiology. In: Berkman L,
Kawachi I, editors. Social Epidemiology.New York: Oxford University Press; 2000. p. 3-12.
(92) Berkman L, Glass T. Social Integrationl, Social Networks, Social Support, and Health. In:
Berkman L, Kawachi I, editors. Social Epidemiology.New York: Oxford University Press; 2000.
p. 137-73.
(93) Kawachi I, Berkman L. Social Cohesion, Social Capital, and Health. In: Berkman L, Kawachi I,
editors. Social Epidemiology.New York: Oxford University Press; 2000. p. 174-90.
(94) Durkheim E. Le Suicide. Étude de Sociologie. Paris: Presses Universitaire de France; 1897.
60
(95) Durkheim E. De la division du travail social. Paris: Les Presses universitaires de France ; 1893.
(96) Berry JW. Acculturation and adaptation: Health consequences of culture contact among
circumpolar peoples. Arctic Medical Research 1990;49:142-50.
(97) Berry JW, Kim U. Acculturation and mental health. In: Dasen J, Berry JW, Sartorius N, editors.
Health and cross-cultural psychology.London: Sage; 1988. p. 207-36.
(98) Koch MW, Bjerregaard P, Curtis T. Acculturation and mental health - empirical verification of
J. W. Berry's model of acculturative stress. International Journal of Circumpolar Health
2004;63(Suppl 2):371-6.
(99) Galobardes B, Shaw M, Lawlor DA, Lynch JW, Davey SG. Indicators of socioeconomic position
(part 1). Journal of Epidemiology and Community Health 2006;60(1):7-12.
(100) Galobardes B, Shaw M, Lawlor DA, Lynch JW, Davey SG. Indicators of socioeconomic position
(part 2). Journal of Epidemiology and Community Health 2006;60(2):95-101.
(101) Lynch J, Kaplan G. Socioeconomic Position. In: Berkman L, Kawachi I, editors. Social
Epidemiology.New York: Oxford University Press; 2000. p. 13-35.
(102) Macintyre S, McKay L, Der G, Hiscock R. Socio-economic position and health: what you
observe depends on how you measure it. Journal of Public Health Medicine 2003;25(4):28894.
(103) Macintyre S, Ellaway A, Cummins S. Place effects on health: how can we conceptualise,
operationalise and measure them? Social Science & Medicine 2002;55(1):125-39.
(104) Lynch J, Smith GD. A life course approach to chronic disease epidemiology. Annual Review of
Public Health 2005;26:1-35.
61
62
APPENDIX 1
ID:
Dette spørgeskema indeholder en række personlige spørgsmål om private forhold, som mange
mennesker helst vil holde for sig selv. Når du er færdig med at besvare spørgsmålene, lægger
du skemaet i en konvolut og lukker den. Konvolutten bliver først åbnet af den forsker, der er
ansvarlig for undersøgelsen. Oplysningerne bliver behandlet fortroligt, og der vil ikke blive
offentliggjort resultater, som kan føres tilbage til dig.
Du svarer på spørgsmålene ved at sætte kryds ud for det svar, du synes passer bedst. Nogle
steder kan du evt. sætte flere kryds. Hvis der er spørgsmål, som du ikke bryder dig om at
svare på, så spring det over og gå videre til næste spørgsmål.
63
De følgende spørgsmål handler om hvor meget tid og hvor mange penge,
du bruger på at spille bingo, Lotto, kort eller terninger og lignende
18.
Hvor ofte har du spillet bingo inden for det sidste år?
hver dag eller næsten hver dag ................................................
3-6 gange om ugen ..................................................................
1-2 gange om ugen ..................................................................
1-3 gange om måneden ............................................................
sjældnere ..................................................................................
aldrig ........................................................................................
1
2
3
4
5
6
Hvor lang tid bruger du normalt på det, hver gang du spiller bingo?
_____ timer
Hvor mange penge bruger du normalt på en måned på at spille bingo?
kr.
19.
Hvor ofte inden for det seneste år har du spillet Tips, Lotto, Skrabespil, Joker
eller lignende spil?
hver dag eller næsten hver dag ................................................
3-6 gange om ugen ..................................................................
1-2 gange om ugen ..................................................................
1-3 gange om måneden ............................................................
sjældnere ..................................................................................
aldrig ........................................................................................
1
2
3
4
5
6
Hvor mange penge bruger du normalt på en måned på disse typer af spil?
kr.
64
20.
Hvor ofte inden for det seneste år har du spillet på spillemaskiner (enarmede
tyveknægte)?
hver dag eller næsten hver dag ................................................
3-6 gange om ugen ..................................................................
1-2 gange om ugen ..................................................................
1-3 gange om måneden ............................................................
sjældnere ..................................................................................
aldrig ........................................................................................
1
2
3
4
5
6
Hvor lang tid bruger du normalt på det, hver gang du spiller på
spillemaskiner?
_____ timer
Hvor mange penge bruger du normalt på en måned på at spille på
spillemaskiner?
kr.
21.
Hvor ofte inden for det seneste år har du spillet kort, terninger eller andre
lignende spil om penge?
hver dag eller næsten hver dag ................................................
3-6 gange om ugen ..................................................................
1-2 gange om ugen ..................................................................
1-3 gange om måneden ............................................................
sjældnere ..................................................................................
aldrig ........................................................................................
1
2
3
4
5
6
Hvor lang tid bruger du normalt på det, hver gang du spiller kort, terninger
eller lignende?
_____ timer
Hvor mange penge bruger du normalt på en måned på disse typer af spil?
kr.
65
22.
Har du nogensinde følt, at du har brugt for mange penge eller for megen tid
på spil?
ja, inden for det seneste år .......................................................
ja, tidligere ...............................................................................
nej ............................................................................................
jeg spiller aldrig .......................................................................
23.
Har du nogensinde løjet (for familiemedlemmer, venner, kollegaer eller
lærere) om, hvor meget du spiller, hvor meget du har tabt, eller hvor stor din
spillegæld er?
ja, inden for det seneste år .......................................................
ja, tidligere ...............................................................................
nej ............................................................................................
24.
1
2
3
Har du nogensinde haft behov for at spille med større og større indsatser (for
at opnå den samme følelse af spænding)?
ja, inden for det seneste år .......................................................
ja, tidligere ...............................................................................
nej ............................................................................................
25.
1
2
3
 4 → gå til spm. 26
1
2
3
Har du nogensinde selv tænkt, at du havde et spilleproblem, eller fået at vide
af andre, at du havde et spilleproblem?
ja, inden for det seneste år .......................................................
ja, tidligere ...............................................................................
nej ............................................................................................
66
1
2
3
APPENDIX 2
ID:
Apersuiffissiaq una inuppassuit namminneerlutik pigerusutaannik inunnut tunngasunik apeqqutinik
arlalinnik imaqarpoq. Apeqqutit akineri naammassigukkit immersuiffissaq allakkat puussaannut
ikeriarlugu taanna matussavat. Allakkat puuat misissuinermut akisussaalluni ilisimatusartumit
aatsaat ammarneqassaaq. Paasissutissat matoqqasumik suliarineqassapput, inernernillu ilinnut
sammitinneqarsinnaasunik tamanut tusarliisoqarnavianngilaq.
Akissutinut ilinnut tulluutsinnerusavit nalaasigut nalunaaqutsiillutit apeqqutit akissavatit. Apeqqutit
ilaanni arlalinnik nalunaaqutsiisarsinnaavutit. Apeqqutit ilaat akerusunngisatit alluinnarlugit
apeqqummut tullermut ingerlaqqiinnassaatit.
67
Apeqqutit tulliuttut piffissaq qanoq annertutigisoq
aningaasanoornermut, bingo, Lotto, nallukattarnermut assigisaannullu
atuisarnermut tunngasuupput
18.
Ukiup kingulliup iluani qanoq akulikitsigisumik binkortarpit ?
ullut tamaasa ullulluunniit tamangajaasa ..................................
sapaatip akunneranut 3-6-riarlunga ..........................................
sapaatip akunneranut 1-2-riarlunga ..........................................
qaammammut 1-3-riarlunga .....................................................
akuttunerusumik .......................................................................
naamerluinnaq ..........................................................................
1
2
3
4
5
6
Piffissaq qanoq annertutigisoq naliginnaasumik bingonermut atortarpiuk?
nalunaaquttap akunneri _____
Aningaasat qanoq amerlatigisut naliginnaasumik bingornermut atortarpigit?
kr.
19.
Ukiup kingulliup iluani qanoq akulikitsigisumik Lotto, Tips, kiliugassat
(skrabespil), Jokeri assigisaanillu pisisarpit?
ullut tamaasa ullulluunniit tamangajaasa ..................................
sapaatip akunneranut 3-6-riarlunga ..........................................
sapaatip akunneranut 1-2-riarlunga ..........................................
qaammammut 1-3-riarlunga .....................................................
akuttunerusumik .......................................................................
naamerluinnaq ..........................................................................
1
2
3
4
5
6
Aningaasat qanoq amerlatigisut naliginnaasumik eqquiniutinut
taakkununnga atortarpigit?
kr.
68
20.
Ukiup kingulliup iluani qanoq akulikitsigisumik maskiina aningaasanoorut
atortarpiuk?
ullut tamaasa ullulluunniit tamangajaasa ..................................
sapaatip akunneranut 3-6-riarlunga ..........................................
sapaatip akunneranut 1-2-riarlunga ..........................................
qaammammut 1-3-riarlunga .....................................................
akuttunerusumik .......................................................................
naamerluinnaq ..........................................................................
1
2
3
4
5
6
Piffissaq qanoq annertutigisoq naliginnaasumik maskiinanut
aningaasanoorutinut atortarpiuk?
nalunaaquttap akunneri _____
Aningaasat qanoq amerlatigisut naliginnaasumik maskiinanut
aningaasanoorutinut atortarpigit?
kr.
21.
Ukiup kingulliup iluani qanoq akulikitsigisumik nallukattat, pinnguaatillu
assigisaat atorlugit aningaasanoortarpit?
ullut tamaasa ullulluunniit tamangajaasa ..................................
sapaatip akunneranut 3-6-riarlunga ..........................................
sapaatip akunneranut 1-2-riarlunga ..........................................
qaammammut 1-3-riarlunga .....................................................
akuttunerusumik .......................................................................
naamerluinnaq ..........................................................................
1
2
3
4
5
6
Piffissaq qanoq annertutigisoq naliginnaasumik nallukattarlutit
assigisaannullu atortarpiuk?
nalunaaquttap akunneri _____
Aningaasat qanoq amerlatigisut naliginnaasumik qaammatip ingerlanerani
nallukattarlutit assigisaanullu atortarpigit?
kr.
69
22.
Qaqugukkulluunniit aningaasat amerlavallaartut imaluunniit piffissaq
annertuvallaartoq aningaasanoortarnermut atorsimallugit misigisarsimavit?
aap, ukiup kingulliup ingerlanerani ..........................................
aap, siornatigut .........................................................................
naamik .....................................................................................
aningaasanoormeq ajorpunga ..................................................
23.
Qaqugukkulluunnit sallusimavit (Ilaquttannut, ikinngutinnut, suleqatinnut
imaluunniit ilinniartitsisunnut) aningaasanoornermut tunngatillugu qanoq
annaasaqarsimallutit, imaluunniit qanoq akiitsut annertutiginerinut?
aap, ukiup kingulliup ingerlanerani ..........................................
aap, siornatigut .........................................................................
naamik .....................................................................................
24.
1
2
3
Qaqqugukkulluunnniit immeraanermut akiliut annertusisariaqarsimallugu
pisariaqartissimaviuk (pissanganassusia misigerusullugu)?
aap, ukiup kingulliup ingerlanerani ..........................................
aap, siornatigut..........................................................................
naamik .....................................................................................
25.
1
2
3
 4 → apeqq 26 nuugit
1
2
3
Qaqugukkulluunniit eqqarsaatigisimaviuk aningaasanoortarnermut
tunngatillugu ajornartorsiuteqarlutit, imaluunniit allanit
aningaasanoortarnermut tunngatillugu ajornartorsiuteqartutit
oqarfigineqarsimavit?
aap, ukiup kingulliup ingerlanerani ..........................................
aap, siornatigut .........................................................................
naamik .....................................................................................
70
1
2
3
APPENDIX 3
SOCIAL
TRANSITION
PROBLEM GAMBLING
ADDICTIVE BEHAVIORS
(Paper II)
GAMBLING BEHAVIOR
HEALTH BEHAVIOR &
HEALTH STATUS (Paper III)
TRAUMATIC
EVENTS IN
CHILDHOOD
(Paper I)
AGE GROUP / GENDER/PLACE RESIDENCE
71
72
PAPERS I-III
73
74
PAPER I
Larsen CVL, Curtis T, Bjerregaard P. Gambling behavior and problem gambling reflecting social transition
and traumatic childhood events among Greenland Inuit - a cross-sectional study in a large indigenous
population undergoing rapid change; Journal of Gambling Studies, 2013, 29 (4): 733-48.
J Gambl Stud
DOI 10.1007/s10899-012-9337-6
ORIGINAL PAPER
Gambling Behavior and Problem Gambling Reflecting
Social Transition and Traumatic Childhood Events
Among Greenland Inuit: A Cross-Sectional Study
in a Large Indigenous Population Undergoing
Rapid Change
Christina Viskum Lytken Larsen • Tine Curtis • Peter Bjerregaard
Ó Springer Science+Business Media New York 2012
Abstract An increase in social pathologies is a key feature in indigenous populations
undergoing transition. The Greenland Inuit are a large indigenous population constituting a
majority in their own country, which makes it possible to investigate differences within the
population. This led us to study gambling behavior and problem gambling among
Greenland Inuit in relation to the ongoing social transition and traumatic events during
childhood. A large representative cross-sectional study was conducted among Greenland
Inuit (n = 2,189). Data was collected among adults (18?) in 9 towns and 13 villages in
Greenland from 2005 to 2010. Problem gambling, gambling behavior and traumatic
childhood events were measured through a self-administered questionnaire. The lie/bet
screen was used to identify past year and lifetime problem gambling. Social transition was
measured as place of residence and a combination of residence, education and occupation.
The lifetime prevalence of problem gambling was 16 % among men and 10 % among
women (p \ 0.0001); and higher in towns (19 %) compared to the capital of Nuuk (11 %)
and in villages (12 %) (men only, p = 0.020). Lifetime problem gambling was associated
with social transition (p = 0.023), alcohol problems in childhood home (p = 0.001/
p = 0.002) and sexual abuse in childhood (women only, p = 0.030). A comparably high
prevalence of lifetime problem gambling among Greenland Inuit adds problem gambling
to the list of social pathologies in Greenland. A significant association between lifetime
problem gambling, social transition and traumatic childhood events suggests people caught
between tradition and modern ways of life are more vulnerable to gambling problems.
C. V. L. Larsen (&) P. Bjerregaard
National Institute of Public Health, University of Southern Denmark, Oester Farimagsgade 5A,
2, 1353 Copenhagen, Denmark
e-mail: [email protected]
P. Bjerregaard
e-mail: [email protected]
T. Curtis
Local Government Denmark, Copenhagen, Denmark
e-mail: [email protected]
123
J Gambl Stud
Keywords Problem gambling Gambling behavior Inuit Indigenous health Social transition
Introduction
During recent years gambling behavior and pathological gambling has become a muchdebated subject in Greenland. The economic and social problems for individuals, families
and communities, which extensive gambling can lead to are of great concern (Paarisa
2006). This Study represents the first population-based survey regarding gambling
behavior and problem gambling conducted among Inuit in Greenland.
In only 60 years Greenland has developed from a traditional subsistence based economy
into a modern society. The rapid transition places immense social and cultural demands on
both individuals and communities (Berner et al. 2005; Bjerregaard and Curtis 2002). An
increase in social pathologies such as violence, suicide, alcohol and substance use is a key
feature shared by indigenous populations undergoing similar social transition (Bjerregaard
et al. 2004). In Greenland, the high proportion of the adult population who experienced
alcohol problems in their childhood home and a similar high prevalence of persons who
were sexually abused, as children are examples of such pathologies. Especially the generations growing up in the 70s and 80s, when the consumption of alcohol was at its highest
(Statistics Greenland 2002), have been exposed to traumatic events during childhood.
Despite its large geographical size, the total population of Greenland is only about 57,000
of whom 90 % are ethnic Greenlanders (Inuit). Genetically, Greenlanders are Inuit
(Eskimos) with a mixture of European, mainly Scandinavian genes. They are genetically
and culturally closely related to the Inuit/Iñupiat in Canada and Alaska and, somewhat
more distantly, to the Yupiit of Alaska and Siberia (Young and Bjerregard 2008).
Greenland was colonized by Denmark in 1721 and reforms of the Greenlandic infrastructure, industry and welfare during the 50s and 60s were based on Danish administrative
systems. In 1979 Greenland established its own Home Rule Government and in 2009 Selfrule was assumed. Today there are still close ties between Denmark and Greenland.
Greenland has a total of 80 communities all located along the coast divided into towns
and villages. A town is defined historically as the largest community in each of 17 districts.
In 2010, the population of the towns varied between 469 to 5,460 and 15,469 in the capital
Nuuk while that of villages varied from less than 10 to around 550. In the towns are located
district school(s), health centre or hospital, church, district administration and main shops.
These institutions are absent or present to a much smaller extent in villages. The isolated
populations in villages and smaller towns face the largest challenges with limited opportunities. In contrast, life in the capital reflects contemporary Scandinavian lifestyle with a
wide range of educational, occupational and recreational possibilities (Berner et al. 2005).
Typically a more traditional lifestyle is found in the villages compared to the towns and
especially the capital. These differences reflect different stages of social transition in
Greenland today where the traditional Inuit lifestyle of hunting and fishing in villages and
smaller towns coexist with the lifestyle of well-educated professionals in larger towns and
the capital.
Public health professionals have pointed to a negative impact from bingo and card
games on families and communities. In particular bingo has been suspected to lead to
neglect of children because it is frequently played in the evening, where children especially
need their parents’ attention (Paarisa 2006). Bingo is played all over Greenland through the
123
J Gambl Stud
radio and in local community halls and sports centers. Local radio stations, societies and
sport clubs for which the revenue serves as the primary source of income organize most
bingo games. In most towns and villages bingo is available on a daily basis and some
places bingo is played several times a day. Gambling on cards/dice presumably takes place
in private settings, but no official records document the extent of gambling in this area,
because it is illegal. Despite the lack of evidence, it is well known within Greenland that
card games are played extensively on the East coast, where the traditional social gatherings
‘Kaffemik’ are often turned into gambling tournaments.
It seems bingo and card games are more popular among Inuit in the Arctic compared to
their non-indigenous neighbors. The Inuit Health in Transition Study conducted in
Nunavik, Canada, found 23 % men and 50 % women aged 15? had played bingo during
past year and 24 % had gambled on cards/dice (Muckle et al. 2007). Furthermore, a status
report regarding gambling problems in First Nations and Inuit Communities of Québec,
found bingo to be one of the most popular games among both Inuit and Cree (Papineau
2010). In comparison only 9 % among the non-Inuit population of Québec played bingo
past year and less than 1 % had gambled on cards (Muckle et al. 2007). In Scandinavia the
past year prevalence for both bingo and card games was similarly low in recent national
surveys (Lund and Nordlund 2003; Bonke and Borregaard 2006; SWELOGS 2011).
Several studies have found a higher prevalence of pathological gambling among
indigenous populations and ethnic minorities compared to the general population within
the same region (Abbott and Volberg 2000; Shaffer et al. 2004; Wardle et al. 2011). Most
often these populations are smaller marginalized groups. Inuit constitute the majority of the
population in Greenland and the present study therefore provides a unique opportunity to
investigate problem gambling in a large indigenous population that is not marginalized.
The increased public health awareness regarding gambling problems and a general
increase in social pathologies in Greenland during the last 60 years led us to study gambling behavior and problem gambling among Inuit in relation to social transition. We
hypothesized gambling behavior and problem gambling is associated with community
specific living conditions and the level of involvement in the ongoing social transition.
Further we hypothesized the highly prevalent traumatic events during childhood, such as
alcohol problems in childhood home and sexual abuse, are associated with problem
gambling (Blaszczynski and Nower 2002; Hodgins et al. 2010).
The objectives of the study were thus to analyze (a) the prevalence of gambling
behavior and problem gambling and (b) its association with social transition and traumatic
events during childhood.
Materials and Methods
Study Design, Setting and Participants
Data were derived from the cross-sectional Inuit Health in Transition Greenland Survey
2005–2010—a population based general health study among adults (18?). A full methodological report of the study is published elsewhere (Bjerregaard 2011). Greenland was
divided into 12 strata based on geography and community size. From each region a number
of towns and villages were selected for the study. A random sample was drawn from the
central population register to obtain around 300 participants from towns included. In the
selected villages, everyone was invited to participate. Data were collected by structured
interviews and self-administered questionnaires conducted in both Greenlandic and Danish
123
J Gambl Stud
according to the choice of the participant. The local hospitals or health center assisted with
procuring facilities for the interviews. A total of 81 persons assisted with data collection
and processing data.
Questions about gambling were included in the self-administered questionnaire from 2006
and onwards. The sample for 2006–2010 included 3,893 Greenland Inuit. A total of 2,451
persons participated in the general survey (63 %) and 2,189 persons filled out the selfadministered questionnaire (56 %). The analyses were based on these 2,189 participants. In
total 2012 of the 2,189 participants (92 %) answered at least one question about gambling.
Gambling Behavior and Problem Gambling
Respondents were asked how often they gambled during past year, how much time they
spent each time they gambled and how much money they spent on gambling during past
month regarding the lottery, bingo, cards/dice and slot machines.
Problem gambling was measured using the lie/bet screen (Johnson and Hammer 1997),
which has been validated in both treatment (Johnson and Hammer 1998) and community
samples (Götestam et al. 2004). It is a two questions version of the 10 DSM-IV criteria for
pathological gambling (American Psychiatric Association 2000). According to DSM-IV, a
positive response to 5 or more of the 10 criteria is characterized as pathological gambling
while three-four positive answers is defined as problem gambling. The lie/bet screen has
been found valid to identify lifetime problem gamblers in a community sample, defined as
those who responded positively to 5 or more of the 10 DSM-IV criteria combined with
those who only responded positively to three-four of these criteria (Götestam et al. 2004).
The short screen does not qualify to distinguish between pathological and problem gambling and it is important to stress, that the purpose of screening for problem gambling in a
large health survey is to investigate an overall prevalence of how widely gambling is
affecting a population, which is by no means comparable to a thorough clinical assessment
of pathological gambling based on the DSM-IV criteria. The questions in the lie/bet screen
are shown in Table 1. Positive answers to one or both questions for past year and previously were combined into a measure of lifetime problem gambling. We also included a
question about self-rated problem gambling to see how well participants’ own perception
corresponded with the lie/bet screen.
Sociodemographic Variables
Residence at age 10 was obtained from the interview and recoded into village or town.
Family job type was determined from questions about job title of participant and spouse
and recoded into hunters/fishermen and others. Formal education was determined from
questions about highest school education attained and further vocational or academic
education and recoded into primary school/high school only, short vocational education
(less than 3 years), and longer vocational/academic education. Age was divided into four
groups; 18–24, 25–34, 35–59 and 60? years.
Social Transition
Two measures of social transition were included. First measure was place of residence
based on the differences in lifestyle and living conditions found between villages, towns
and the capital. In order to supplement this measure we combined current place of
123
J Gambl Stud
Table 1 Overview of key variables in the study. Frequency, percent and missing data. Greenland Inuit
2006–2010
Frequency
(pct.)
Total
N
Missing
data
2,189
Gender
Men
Women
983 (44.9)
1,206 (55.1)
Age groups
18–24
278 (12.7)
25–34
385 (17.6)
35–59
1,273 (58.2)
60?
253 (11.6)
Problem gambling
Lie/bet screen Q1: Have you ever lied (to family members, friends,
co-workers or teachers) about how much you gamble, how much
you have lost or how large your debt is?
Yes, past year
37 (2.4)
Yes, previously
100 (6.6)
No/never gamble
Yes, past year
22 (1.5)
Yes, previously
73 (5.0)
1,456
733
1,500
689
1,361 (93.5)
Selfrated problem gambling: Have you ever felt you had a gambling
problem or were told so by others?
Yes, past year
39 (2.6)
Yes, previously
74 (4.9)
No/never gamble
671
1,381 (91.0)
Lie/bet screen Q2: Have you ever felt the need to bet more and more
money (to reach the same level of exitement)?
No/never gamble
1,518
1,387 (92.4)
Social transition
Place of residence
2,189
Nuuk
419 (19.1)
Town
1,110 (50.7)
Village
660 (30.2)
Level of involvement in social transition (25–64 years)
Hunters/fishermen in villages
1,750
25
1,985
204
161 (9.2)
Other villagers
374 (21.4)
Blue collar migrants in towns
172 (9.8)
Other blue collar in towns
381 (21.8)
Intermediate in towns
468 (26.7)
Professionals in towns
194 (11.1)
Traumatic childhood events
Alcohol problems in childhood home
No, never
822 (41.4)
Yes, occasionaly
798 (40.2)
Yes, often
365 (18.4)
123
J Gambl Stud
Table 1 continued
Frequency
(pct.)
Sexually abused as a child
Yes
No
N
Missing
data
1,359
830
322 (23.7)
1,037 (76.3)
Response to Q1 in the lie/bet screen was associated with gender (p \ 0.0001), age group (p \ 0.0001) and
formal education (p = 0.0001). Response to Q2 in the lie/bet screen was associated with gender
(p \ 0.0001), age group (p \ 0.0001) and formal education (p \ 0.0001). Response to question of selfrated
problem gambling was associated with gender (p \ 0.0001), age groups (p \ 0.0001) and formal education
(p = 0.003). Response to question about alcohol problem in childhood home was associated with gender
(p \ 0.0001), age groups (p \ 0.0001) and formal education (p \ 0.0001). Response to question about
sexual abuse as a child was associated with gender (p \ 0.0001), age group (p \ 0.0001), formal education
(p \ 0.0001) and place of residence (p \ 0.0001)
residence and childhood residence with formal education and family job type in a second
measure that has previously been used to document changes in mental health and health
behavior among Greenland Inuit (Bjerregaard and Dahl-Petersen 2011; Dahl-Petersen and
Bjerregaard 2011). Six categories of social transition were defined as (A) hunters and
fishermen in villages; (B) other inhabitants of villages; (C) blue collar migrants (inhabitants of towns, with no vocational education, having lived in villages at age 10); (D) other
blue collar participants (inhabitants of towns, with no vocational education, having lived in
towns at age 10); (E) intermediate (inhabitants of towns, with short vocational education);
and (F) professionals (inhabitants of towns, with longer vocational or academic education).
In order not to misclassify participants who had not yet finished their education and to
minimize the proportion of participants outside the work force this measure only included
those aged 25–64 years.
Traumatic Childhood Events
Alcohol related problems in childhood home and sexual abuse during childhood were
included as traumatic childhood events. Participants were asked through the self-administered questionnaire if there were alcohol problems in their childhood home and whether
anyone had forced any kind of sexual activity upon them as a child (before the age of 13).
Statistical Analyses
Statistical analyses were performed in IBM SPSS Statistics 19. Dichotomous variables
were analyzed by bivariate logistic regression. In case of interaction between gender and
outcome, analyses were performed separately for men and women. Differences in prevalence according to gender, age and social transition were tested using Pearson’s Chi2-test.
Due to the sampling procedure, oversampling took place in certain strata while the participation rate was not similar according to age and gender. Accordingly, data were
weighted for sampling strata by the weight procedure of SPSS in order to obtain national
prevalence estimates.
123
J Gambl Stud
Ethical Review
The study was ethically approved by the Commission for Scientific Research in Greenland.
Participants gave their written consent after being informed about the study orally and in
writing.
Results
Characteristics of the study population according to key variables and analyses of response
rates for specific questions are shown in Table 1. The majority of Greenland Inuit (80 %)
engaged in at least one of four types of gambling past year and gender and age differences
applied. The lottery was played by four out of five followed by bingo, cards/dice and slot
machines (Table 2). A higher percentage of men compared to women had gambled on the
lottery, cards/dice and slot machines during past year while more women than men had
played bingo. Gambling behavior was associated with social transition measured as place
of residence. Bingo was more common and played more frequently in villages and towns
compared to the capital. People in villages gambled more frequently on cards/dice compared to the capital and in towns while the lottery and slot machines were much more
common in the capital and in towns. Response to questions about gambling behavior varied
according to sociodemographic differences. In general more men than women and more
inhabitants from the capital compared to towns and villages had answered the questions.
Further the odds ratio for answering decreased with age and increased with level of
education (see note, Table 1). Gambling at least once a week on slot machines and cards/
dice was associated with past year problem gambling for both men and women, while
playing bingo was not (Table 3).
Using the lie/bet screen we found a past year prevalence for problem gambling of four
percent among men and three percent among women. The lifetime prevalence was 16 %
among men and 10 % among women (Table 4). The odds ratios for answering the questions in the lie/bet screen were higher among men than women and odds ratios for
answering decreased with age and increased with level of education (see note, Table 1). A
rather large percentage who reported gambling problems at some point in their life, did not
report of any past year problems. At an overall level this was the case for 74 % among men
and 71 % among women (results not tabulated). Among women the 1-year remission rate
among those aged 18–34 years was 80–86 % compared to 64 % among those aged 35–59.
Among men the 1-year remission rates among the 25–59 years old was 74–78 % compared
to a lower 55 % in the youngest age group (18–24 years).
Lifetime problem gambling was significantly associated with social transition and
traumatic childhood events (Table 5). Hunters and fishermen in villages (A) and professionals (F) displayed the lowest odds ratio for lifetime problem gambling compared to
other villagers and inhabitants of towns (B-E). These two groups represent each their end
of the social transition scale. Among men, the odds ratio for lifetime problem gambling
was higher in towns compared to the capital and in villages.
Among those who often experienced alcohol problems in their childhood home the odds
ratio for lifetime problem gambling was 3–4 times higher compared to those who never
experienced such problems in their childhood home. Among women, being sexually
abused as a child was also associated with lifetime problem gambling. When analyzed
according to birth cohort the prevalence of lifetime problem gambling followed the pattern
123
J Gambl Stud
Table 2 Prevalence of past year gambling* (at least once during past year) and frequent gambling* (at least
once a month during past year) analyzed by type of gambling according to age and place of residence,
stratified by gender. Percent. Greenland Inuit 2006–2010
Lottery
Bingo
Cards and dice
Slot machines
(Men = 827;
women = 924;
missing = 438)
(Men = 848;
women = 1,040;
missing = 301)
(Men = 807;
women = 924;
missing = 510)
(Men = 786;
women = 893;
missing = 458)
Past year
Monthly
Past year
Past year
Past year
18–24 years
41.1
19.6
23.6
3.6
30.9
10.9
32.4
9.3
25–34 years
74.3
38.2
37.0
16.3
44.1
15.6
27.3
9.1
5.9
Monthly
Monthly
Monthly
Men
Age groups
35–59 years
80.5
44.2
41.8
20.4
32.1
8.3
25.4
60? years
54.9
35.3
29.7
16.8
16.5
3.3
9.8
6.5
0.011
0.011
0.454
p value
\0.0001
\0.0001
0.001
\0.0001 \0.0001
Place of residence
Nuuk
78.5
43.9
16.5
4.9
31.3
10.6
31.7
8.2
Towns
68.3
37.3
43.5
22.6
34.6
10.9
21.6
7.3
8.7
45.0
15.9
37.8
20.6
6.9
Villages
p value
All Men
39.7
\0.0001
\0.0001 \0.0001
\0.0001
0.358
\0.0001 \0.0001
9.9
25.6
1.6
0.003
69.6
37.5
36.4
16.2
32.5
7.1
18–24
47.4
13.6
41.8
16.4
27.6
9.8
17.1
0.7
25–34
58.1
21.1
50.7
23.5
26.7
10.3
11.1
1.6
Women
Age groups
35–59
68.5
25.4
56.2
33.0
24.0
8.3
15.1
4.4
60?
51.4
25.4
57.6
43.0
20.6
11.8
5.7
3.8
0.103
0.062a
p value
\0.0001
0.018
\0.0001
\0.0001
0.612
0.718
Place of residence
Nuuk
75.8
28.3
33.7
12.8
19.3
6.7
21.6
Towns
59.6
22.2
61.0
37.1
23.7
9.1
9.8
2.0
Villages
21.4
4.3
67.8
28.0
43.6
19.0
3.2
0.5
p value
All women
\0.0001
60.8
\0.0001 \0.0001
22.2
51.9
\0.0001 \0.0001
29.1
24.3
\0.0001 \0.0001
8.9
13.7
5.7
0.004
2.8
The numbers reported for age groups are weighted for regional differences in the sample. The numbers
reported according to place of residence are weighted for age differences in the sample. The overall
prevalence for men and women are weighted for both regional and age differences in the sample
* Response regarding gambling on the lottery was associated with gender (p \ 0.0001), age group
(p \ 0.0001), level of education (p \ 0.0001) and place of residence (p \ 0.0001). Response regarding
playing bingo was associated with age group (p \ 0.0001), level of education (p \ 0.0001) and place of
residence (p = 0.005). Response regarding gambling on cards/dice was associated with gender
(p \ 0.0001), age group (p \ 0.0001), level of education (p \ 0.0001) and place of residence (p \ 0.0001).
Response regarding frequency of gambling on slot machines was associated with gender (p \ 0.0001), age
groups (p \ 0.0001); Level of education (p \ 0.0001); Place of residence (p \ 0.0001)
a
Expected count less than 5
123
J Gambl Stud
Table 3 Past year problem gambling according to weekly gambling activity during past year (at least once
a week) among men and women. Adjusted for age and place of residence. Odds Ratio (OR) with 95 %
Confidence Interval (C.I.). Greenland Inuit 2006–2010
Weekly gambling activity during past year
Men
OR [C.I. 95 %]
Women
p value
OR [C.I. 95 %]
p value
The lottery (men = 827; women = 924)
Less than once a week
(ref.)
At least once a week
1.03 [0.28–3.84]
(ref.)
0.964
9.25 [2.37–36.00]
0.837
2.04 [0.60–6.94]
0.001
Bingo (men = 848; women = 1,040)
Less than once a week
(ref.)
At least once a week
0.87 [0.22–3.40]
(ref.)
0.106
Cards/dice (men = 807; women = 924)
Less than once a week
(ref.)
(ref.)
At least once a week
10.27 [3.44–30.64] \0.0001
7.16 [1.73–29.65]
0.007
Slot machines (men = 786; women = 893)
Less than once a week
(ref.)
At least once a week
5.94 [1.45–24.29]
(ref.)
0.013
20.11 [1.76–229.75]
0.016
of both alcohol problems in childhood home and sexual abuse during childhood with the
highest prevalence found for the generations born in the 70s (Figure 1).
Compared to the lifetime prevalence of problem gambling identified by the lie/bet
screen fewer persons felt they had ever had a gambling problem or had been told so by
others (7.5 %, Table 1). No difference was found between men and women.
Discussion
Gambling behavior and problem gambling among Greenland Inuit were associated with
social transition measured as residence in the capital, town or village and significant
differences according to age and gender were found. This association was further elaborated by a significant association between lifetime problem gambling and social transition
measured as place of residence combined with childhood residence in town (including the
capital) or village as well as formal education and family job. Lifetime problem gambling
was also associated with traumatic events during childhood. The prevalence for past year
gambling was higher among men than women for all types of gambling but bingo. This
corresponded well with a higher prevalence of problem gambling among men compared to
women although only significant for lifetime problem gambling. Also a rather high 1-year
remission rate was found among lifetime problem gamblers suggesting that one does not
necessarily have a gambling problem forever.
Compared to findings from the Scandinavian countries Denmark, Sweden and Norway
the prevalence of both past year and lifetime problem gambling among Inuit in Greenland
was 4–6 times higher (Bonke and Borregaard 2009). However these Scandinavian prevalence surveys used the revised South Oaks Gambling Screen (SOGS-R) (Lesieur and
Blume 1993) and the NORC DSM Screen for Gambling Problems (NODS) (Gerstein et al.
1999) to identify past year and lifetime problem gamblers and the results are therefore not
123
J Gambl Stud
Table 4 Prevalence of past year and lifetime problem gambling (PG) among men and women* according
to age and social transition (place of residence). p value based on Chi2-test. Unadjusted percent. Greenland
Inuit 2006–2010
% past year PG (freq.)
p value
% lifetime PG (freq.)
p value
Men
Age group
18–24
9.5 (9)
21.1 (20)
25–34
5.6 (7)
21.8 (27)
35–59
3.0 (13)
60?
1.2 (1)
13.8 (59)
0.013
11.8 (10)
0.053
Social transition (place of residence)
Nuuk
0.8 (1)
Towns
5.3 (21)
Villages
3.7 (8)
Men total
11.4 (14)
19.3 (76)
0.081
12.0 (26)
4.1 (30)
15.8 (116)
18–24
1.7 (2)
8.6 (10)
25–34
1.2 (2)
0.020
Women
Age group
35–59
4 (19)
60?
0 (0)
9.3 (15)
11.2 (53)
0.104
3.3 (2)
0.249
Social transition (place of residence)
Nuuk
1.1 (2)
8.4 (15)
Town
3.4 (14)
10.5 (43)
Villages
3.2 (7)
Women total
2.8 (23)
0.286
10.0 (22)
0.735
9.9 (80)
Men = 733; Women = 809
* Overall difference between men and women for past year problem gambling (Chi2-test; p = 0.179) and
lifetime problem gambling (Chi2-test; p \ 0.001)
directly comparable. Götestam et al. (2004) reported a 0.54 % prevalence of lifetime
problem gambling based on the lie/bet screen in a community sample of adults (18?) in
Norway. This prevalence is actually a little lower than the prevalence found by Lund and
Nordlund (2003) in Norway in the same survey year based on the NODS as a screening
tool suggesting the lie/bet screen did not overestimate the lifetime prevalence in the
Norwegian population despite its brief character. Furthermore the Danish Health Interview
Survey from 2005 (Ekholm et al. 2009) included the lie/bet screen and found a lifetime
prevalence for problem gambling of 2.6 % among those aged 16? (unpublished results).
This prevalence was a little higher than the one found in the Danish Gambling Survey by
Bonke and Borregaard (2006) mentioned above, where the SOGS-R identified 1.7 %
lifetime problem gamblers and the NODS identified 0.7 % lifetime problem gamblers
among adults (18?). This suggests the lie/bet screen slightly overestimated the lifetime
prevalence in the Danish population compared to both NODS and SOGS-R. Despite
methodological differences, the lifetime prevalence found among Greenland Inuit seems
high. Given the many similarities in gambling availability and legislation between Denmark and Greenland, it is alarming to find a markedly higher lifetime prevalence of
123
J Gambl Stud
Table 5 Lifetime problem gambling according to social transition and traumatic events during childhood.
Adjusted analyses. Odds ratio (OR) with 95 % Confidence Intervals (C.I.). Greenland Inuit 2006–2010
Men
OR [C.I.]
Women
p value
OR [C.I.]
Total
p value
OR [C.I.]
p value
Social transition
Place of residence (Men = 733;
Women = 809)
0.695
0.021
Nuuk
(ref.)
Town
1.86
[1.01–3.44]
0.047
1.31
[0.71–2.42]
(ref.)
0.396
Villages
1.05
[0.52–2.10]
0.891
1.23
[0.62–2.49]
0.532
Level of involvement in social
transition (25–64 years only,
N = 1,237)
0.023
Hunters/fishermen in villages
(ref.)
Other villagers
1.88
[0.83–4.25]
0.132
Blue collar migrants in towns
1.60
[0.62–4.10]
0.329
Other blue collar in towns
2.56
[1.16–5.64]
0.020
Intermediate in towns
2.22
[1.02–4.85]
0.046
Professionals in towns
0.81
[0.29–2.23]
0.679
Traumatic events during childhood
(men = 549; women = 509)
Sexual abuse during childhood
No
(ref.)
Yes
1.53
[0.80–2.90]
Alcohol problems in childhood home
(ref.)
0.196
2.02
[1.07–3.81]
0.001
0.030
0.002
Never
(ref.)
(ref.)
Occasionally
1.27
[0.70–2.31]
0.432
1.42
[0.66–3.08]
0.374
Often
3.12
[1.65–6.20]
0.001
3.64
[1.68–7.91]
0.001
Model 1 included place of residence stratified by gender and adjusted for age group. Model 2 included level of involvement
in social transition adjusted for gender and age group. No interaction was found between gender and level of involvement
in social transition and the analysis was thus carried out for men and women together. Model 3 included sexual abuse in
childhood and alcohol problems in childhood stratified by gender and adjusted for age group, level of education and place
of residence
Bold values indicate p \ 0.05
problem gambling among Greenland Inuit. A higher prevalence of problem gambling
among ethnic minorities and indigenous populations compared to their non-indigenous
neighbors has been found in other studies (Abbott and Volberg 2000; Shaffer et al. 2004;
Wardle et al. 2011). However there is an important difference between a high prevalence of
problem gambling in a smaller often marginalized indigenous population constituting a
minority in a larger population compared to a high prevalence in an indigenous population
that constitute the majority in their own country, which is the case for Inuit in Greenland.
123
J Gambl Stud
40
35
30
25
20
15
10
5
0
-1954
1955-1959
1960-1964
1965-1969
1970-1974
1975-1979
1980-1985
1985-
Often experienced alcohol problems in childhood home
Lifetime problem gambling
Sexually abused as a child
Fig. 1 Prevalence of alcohol problems in childhood home, lifetime problem gambling and sexual abuse in
childhood according to birth cohort. Unadjusted percent. Greenland Inuit 2006–2010. Chi2-test for
difference between birth cohorts: Alcohol problems (p \ 0.0001) (N = 1,985); Lifetime problem gambling
(p = 0.017) (N = 1,542); Sexual abuse (p \ 0.0001) (N = 1,359)
The associations found with social transition and childhood conditions thus suggest that the
high prevalence could be related to the ongoing rapid change and the social consequences
of such changes rather than marginalization.
The association between gambling behavior and social transition was reflected in a
preference for the lottery and slot machines in the capital and for bingo and cards/dice
in towns and villages. This indicates a mix of patterns between modern games popular in
contemporary Scandinavia and games found to be traditionally popular among Inuit in
Nunavik and First Nations (Muckle et al. 2007; Papineau 2010). The preference for bingo,
which was played by 60–70 % of the women in towns and villages, but only by 32 % of the
women in the capital, illustrated this pattern. Despite the large influence bingo presumably
has on daily life, no association between weekly gambling on bingo and past year problem
gambling was found. It is however important to keep in mind, that the lie/bet screen and other
screens developed to identify problem gamblers might not be suited to identify the potential
social consequences of gambling in families and communities. In contrast to bingo weekly
gambling on cards/dice and slot machines was significantly associated with higher odds ratios
for past year problem gambling. These games typically involve excitement and are fundamentally different from bingo and most forms of the lottery. The distinct characteristics of the
different types of gambling might explain the differences in associations between weekly
gambling and past year problem gambling. Although the association between weekly gambling on slot machines and past year problem gambling is in accordance with a general
consideration that electronic gaming is the most addictive form of gambling, the empirical
evidence supporting this perception is inconclusive (Dowling et al. 2005).
The association between problem gambling and social transition was reflected in a
higher prevalence of lifetime problem gambling in towns compared to villages and the
capital among men and lower odds ratios for lifetime problem gambling among both
hunters and fishermen in villages as well as among professionals in towns compared to
other villagers, blue collar workers in towns and persons with a short vocational education
123
J Gambl Stud
in towns. People in towns in Greenland might generally be considered to have the most
difficult position in the ongoing social transition because they are caught in the middle of
the more traditional life in the villages and the modern life in the capital. Differences in
odds ratios for lifetime problem gambling between hunters and fishermen in villages
compared to other villagers and between professionals in towns compared to the three less
educated or unskilled groups in towns add to the understanding of the complex association
between social transition and gambling problems among Greenland Inuit. Hunters and
fishermen are naturally integrated in village life where traditional Inuit lifestyle is still
prevalent as are professionals in towns where life reflects the modern Greenlandic Society.
In contrast village life is more challenging if you are not hunting and fishing because very
few educational and occupational opportunities are available. So is life in a larger town if
you are not skilled or well educated as expectations and demands for qualifications rise
with the level of specialization in modern society.
The prevalence of alcohol problems in childhood, sexual abuse and lifetime problem
gambling analyzed according to birth cohorts illustrated how the generation born in the 70s
experienced the highest prevalence across all three outcomes. This generation grew up
during the 70s and the 80s when the alcohol consumption in Greenland was at its highest
(Statistics Greenland 2002). These were also decades where extensive infrastructural
development occurred, which transformed Greenland from a traditional hunting society
into a modern economy where most people depended on earning wages (Young and
Bjerregaard 2008). Based on these observations the high prevalence of alcohol problems in
childhood home and sexual abuse in childhood during these years seems linked to social
transition. Life style changes, social change, and changes in society and the environment
have been documented as major determinants of health among the northern indigenous
populations (Bjerregaard et al. 2004; Young and Bjerregaard 2008). Many causes of social
pathologies such as alcohol misuse have been postulated at the individual, family and
community level and different explanations focus on the stress of rapid social change and
the inadequacy of traditional conflict resolution behaviors in the new, more urbanized
environments (Kirmayer et al. 2000; Lethi et al. 2009). However a previous study among
Greenland Inuit has suggested that successful integration into the modern Greenlandic
society might explain why some groups within the population have better mental health
compared to others (Bjerregaard and Curtis 2002).
Other studies have found an association between traumatic childhood events and lifetime problem gambling—especially among pathological gamblers in treatment (Hodgins
et al. 2010; Kausch et al. 2006; Petry and Steinberg 2005; Wardman et al. 2001). Childhood disturbance as a cause of emotional vulnerability has been suggested to constitute a
potential pathway for the development of problem and pathological gambling (Blaszczynski and Nower 2002). Given the high prevalence of alcohol problems in families,
violence and sexual abuse in most circumpolar indigenous communities (Kruse et al.
2007), targeted prevention programs concerning problem gambling should take the complexity of these issues into account.
Future research should look into why some population groups are caught between
traditional and modern way of life while others integrate well and thereby contribute to the
understanding of causes of both mental and physical health inequalities in indigenous
populations. Prospective studies are needed to investigate whether causality between
traumatic childhood events and problem gambling later in life can be established. Finally
we need to gain a better understanding of why undergoing rapid social transition has led to
an increase in social pathologies across indigenous populations in order to prevent a further
increase in the future.
123
J Gambl Stud
A particular strength of the study is the large representative sample of an indigenous
population, which is a majority in its country. This allowed us to investigate differences
according to social transition within the population. Another strength is the various topics
included in the general health survey that provided us with the opportunity to include
traumatic childhood events in the study.
A weakness of the study is the social gradient in the response to questions about
problem gambling and traumatic childhood events. It is likely this has led to an underestimation of the prevalence. We found the odds ratio for answering the lie/bet screen was
significantly higher among men, the younger age groups and those with an academic
degree. We also checked to see if it made a difference whether the respondents had chosen
to answer the questions in Danish or Greenlandic, but this was not the case. Based on the
existing data we were not able to further asses why a relatively large proportion of the
participants chose not to answer the lie/bet screen or chose to answer one of the two
questions but not the other. The number of missing data on the lie/bet screen is larger than
the number of missing data on the questions regarding different types of gambling, which
suggests people are more likely to answer questions about gambling activities rather than
gambling problems. Whether this has to do with participants not wanting to answer
questions about gambling problems they might have or with the fact that it is tempting to
skip a question in a long questionnaire if it seems irrelevant to you personally, should be
investigated through future qualitative methodological studies. Such insights would be
valuable for future prevalence studies.
Another weakness is the fact that the lie/bet screen has only been used in few other
studies. This makes it difficult to compare our results with other prevalence studies.
Furthermore comparisons of the lie/bet screen with peoples own perception suggests either
a high level of denial among lifetime problem gamblers or a questionable validity of the
short screen. It is likely the lie/bet screen captures very mild gambling problems along with
more serious ones. The large proportion of lifetime problem gamblers who did not report of
problems with in the year preceding the survey further supports this assumption suggesting
some of the problems measured by the screen could be brief and passing by within a short
timeframe. This methodological issue should be further investigated.
Acknowledgments The authors are grateful to everyone who was willing to take their time and participate
in the Inuit Health in Transition Greenland Survey. The study is a part of CVLL’s PhD Study funded by The
Danish Council for Independent Research | Social Sciences.
Conflict of interest The authors declare that there are no conflicts of interest.
References
Abbott, M. V. & Volberg, R. A. (2000). Taking the pulse on gambling and problem gambling in New
Zealand: A report on phase one of the 1999 national prevalence survey. Report number three of the
New Zealand gaming survey. The Department of Internal Affairs, 2000. http://www.dia.govt.nz/pub
forms.nsf/URL/TakingthePulse.pdf/$file/TakingthePulse.pdf. Accessed July 30, 2012.
American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders—text
revision (4th ed.). Arlington: American Psychiatric Association.
Berner, J., Furgal, C., Bjerregaard, P., Bradley, M., Curtis, T., De Fabo, E., et al. (2005). Human health. In
ACIA 2005, Arctic climate impact assessment (pp. 863–906). New York: Cambridge University Press.
Bjerregaard, P. (2011). Inuit health in transition—Greenland Survey 2005–2010. Population sample and
survey methods. 2nd revised edition. SIF writings on Greenland vol. 19. Copenhagen: National
123
J Gambl Stud
Institute of Public Health. http://www.si-folkesundhed.dk/upload/inuit_health_in_transition_greenland_
methods_5_2nd_revision.pdf. Accessed July 30, 2012.
Bjerregaard, P., & Curtis, T. (2002). Cultural change and mental health in Greenland. The association of
childhood conditions, language and urbanization with mental health and suicidal thoughts among the
Inuit of Greenland. Social Science and Medicine, 54, 33–48.
Bjerregaard, P., & Dahl-Petersen, I. K. (2011). How well does social variation mirror secular change in
prevention of cardiovascular risk factors in a country in transition? American Journal of Human
Biology, 23, 774–779.
Bjerregaard, P., Young, T. K., Dewailly, E., & Ebbesson, S. O. E. (2004). Indigenous health in the Arctic:
An overview of the circumpolar Inuit population. Scandinavian Journal of Public Health, 32, 390–395.
Blaszczynski, A., & Nower, L. (2002). A pathways model of problem and pathological gambling. Addiction,
97, 487–499.
Bonke, J. & Borregaard, K. (2006). Pathological gambling in Denmark. The prevalence of gambling
and problem gambling. Report (in Danish). The Danish National Center for Social Research.
http://www.sfi.dk/rapportoplysninger-4681.aspx?Action=1&NewsId=117&currentPage=4&PID=9267.
Accessed July 30, 2012.
Bonke, J., & Borregaard, K. (2009). The prevalence of problematic gambling behavior: A Scandinavian
comparison. Scandinavian Journal of Public Health, 37(6), 654–660.
Dahl-Petersen, I. K., & Bjerregaard, P. (2011). Physical activity patterns in Greenland: A country in
transition. Scandinavian Journal of Public Health, 39, 678–686.
Dowling, N., Smith, D., & Thomas, T. (2005). Electronic gaming machines: Are they the ‘crack-cocaine’ of
gambling? Addiction, 100, 33–45.
Ekholm, O., Hesse, U., Davidsen, M., & Kjøller, M. (2009). The study design and characteristics of the
Danish national health interview surveys. Scandinavian Journal of Public Health, 37(7), 758–765.
Gerstein, D. R., Arwood R. A., Hoffmann, J., Larison, C., Engelman, L., Murphy, S., et al. (1999). Gambling
impact and behavior study. Report to the National Gambling Impact Study Commission. National
Opinion Research Center at the University of Chicago. http://govinfo.library.unt.edu/ngisc/reports/
fullrpt.html. Accessed September 16, 2012.
Götestam, G. K., Johansson, A., Wenzel, H. G., & Simonsen, I.-E. (2004). Validation of the lie/bet screen
for pathological gambling on two normal population data sets. Psychological Reports, 95, 1009–1013.
Hodgins, D. C., Schopflocher, D. P., el Guebaly, N., Casey, D. M., Smith, G. J., Williams, R. J., et al. (2010).
The association between childhood maltreatment and gambling problems in a community sample of
adult men and women. Psychology of Addictive Behaviors, 24(3), 548–554.
Johnson, E. E., & Hammer, R. (1997). The lie/bet questionnaire for screening pathological gamblers.
Psychological Reports, 80, 83–88.
Johnson, E. E., & Hammer, R. M. (1998). The lie/bet questionnaire for screening pathological gamblers: a
follow-up study. Psychological Reports, 83, 1219–1224.
Kausch, O., Rugle, L., & Rowland, D. Y. (2006). Lifetime histories of trauma among pathological gamblers.
American Journal on Addictions, 15(1), 35–43.
Kirmayer, L. J., Brass, G. M., & Tait, C. L. (2000). The mental health of aboriginal peoples. Transformations of identity and community. Canadian Journal of Psychiatry-Revue/Canadienne de Psychiatrie, 45, 607–616.
Kruse, J., Poppel, B., Abryutina, L., Duhaime, G. (2007). Survey of living conditions in the Artic: What did
we learn? Institute of Social and Economic Research, University of Alaska Anchorage. http://elders.
uaa.alaska.edu/reports/SLiCA_07.pdf. Accessed July 30, 2012.
Lesieur, H. R., & Blume, S. B. (1993). Revising the south oaks gambling screen in different settings.
Journal of Gambling Studies, 9(3), 213–223.
Lethi, V., Niemelä, S., Hoven, C., & Sourander, A. (2009). Mental health, substance use and suicidal
behavior among young indigenous people in the Arctic: A systematic review. Social Science and
Medicine, 69, 1194–1203.
Lund, I., & Nordlund, S. (2003). Gambling and gambling problems. Report (in Norwegian). Norwegian
Institute for Alcohol and Drug Research (SIRUS). http://www.sirus.no/filestore/Import_vedlegg/
sirusrap.2.03.pdf. Accessed July 30, 2012.
Muckle, G., Boucher, O., Laflamme, D., Chevalier, S., Rochette, L. (2007). Alcohol, drug use and gambling
among the Inuit of Nunavik: Epidemiological Profile. Report. Institut national de santé publique du
Québec. Nunavik Regional Board of Health and Social Services. http://www.scribd.com/doc/5993
6832/657-Nunavik-Qanuippita-Alcool-Drogues-Gambling. Accessed July 30, 2012.
Paarisa (2006). Gambling and gambling patterns. Report (in Danish). Department of Health, Government of
Greenland. http://www.peqqik.gl/Sundhed/Spil_og_spillevaner.aspx. Accessed July 30, 2012.
123
J Gambl Stud
Papineau, E. (2010). Gambling problems in first nations and Inuit Communities of Québec: A brief status
report. Québec, Direction du développement des individus et des communautés. Institut national de
santé publique Québec. http://www.inspq.qc.ca/pdf/publications/1072_ProblJeuPremNationsVillages
Inuits_VA.pdf. Accessed July 30, 2012.
Petry, N. M., & Steinberg, K. L. (2005). Childhood maltreatment in male and female treatment-seeking
pathological gamblers. Psychology of Addictive Behaviors 19(2), 226–229.
Shaffer, H. J., LaBrie, R. A., LaPlante, D. A., Nelson, S. E., & Stanton, M. V. (2004). The road less
travelled: Moving from distribution to determinants in the study of gambling epidemiology. Canadian
Journal of Psychiatry, 49(8), 504–516.
Statistics Greenland (2002). Import of alcohol to Greenland 1975–2001. Report (in Danish). http://
www.stat.gl/dialog/main.asp?lang=da&version=2001&link=AL&subthemecode=p1&colcode=p. Accessed
July 30, 2012.
SWELOGS (2011). The Swedish longitudinal gambling study 2008/2009. Main results from the population
study. Report (in Swedish). The Swedish Institute of National Public Health. http://www.
fhi.se/PageFiles/10965/R2010-23-Spel-om-pengar-o-spelproblem.pdf. Accessed July 30, 2012.
Wardle, H., Moody, A., Spence, S., Orford, J., Volberg, R. A., Jotangia, D., et al. (2011). British gambling
prevalence survey 2010. National Centre for Social Research, United Kingdom. Prepared for The
Gambling Commission. http://www.gamblingcommission.gov.uk/PDF/British%20Gambling%20Prevalence
%20Survey%202010.pdf. Accessed July 30, 2012.
Wardman, D., el-Guebaly, N., & Hodgins, D. C. (2001). Problem and pathological gambling in North
American aboriginal populations: A review of the empirical literature. Journal of Gambling Studies,
17, 81–100.
Young, T. K., & Bjerregaard, P. (2008). Health transitions in Arctic populations. Toronto: University of
Toronto Press.
123
PAPER II
Larsen CVL, Curtis T, Bjerregaard P. Harmful alcohol use and frequent use of marijuana among lifetime
problem gamblers and the prevalence of cross-addictive behaviors among Greenland Inuit – evidence from
the cross-sectional Inuit Health in Transition Greenland Survey 2006-2010; International Journal of
Circumpolar Health, 2013, 72:19551. doi: 10.3402/ijch.v72i0.19551.
ORIGINAL RESEARCH ARTICLE
æ
Harmful alcohol use and frequent use of
marijuana among lifetime problem
gamblers and the prevalence of crossaddictive behaviour among Greenland
Inuit: evidence from the cross-sectional
Inuit health in transition Greenland
survey 20062010
Christina Viskum Lytken Larsen1*, Tine Curtis2 and
Peter Bjerregaard1
1
National Institute of Public Health, University of Southern Denmark, Denmark; 2Local Government Denmark,
Denmark
Background and objectives. Public health research has pointed to alcohol and substance abuse as the most
significant public health challenges in Greenland with the negative impact on families and communities that
entail, but few studies have investigated the role of problem gambling as addictive behaviour among Inuit.
The objectives of the present study were to investigate (a) the association between lifetime problem gambling
and harmful alcohol use as well as frequent use of marijuana and (b) the prevalence of cross-addictive
behaviour among Greenland Inuit.
Design. A representative cross-sectional study among Greenland Inuit (n 2,189). Data was collected among
adults (18) in 8 towns and 13 villages in Greenland from 20062010. Lifetime problem gambling, harmful
alcohol use and frequent use of marijuana were measured through a self-administered questionnaire.
Results. The odds ratio for harmful alcohol use and frequent use of marijuana was significantly higher among
lifetime problem gamblers compared to non-problem gamblers/non-gamblers. One or more addictive
behaviours were present among more than half of the men (53%) and one third of the women (37%), and
the co-occurrence of lifetime problem gambling with either harmful alcohol use, frequent use of marijuana or
both was found among 12.2% of men and 3.7% of women. The prevalence of one or more addictive
behaviours was 44% in households with children.
Conclusions. For lifetime problem gamblers, the gambling problems were more often than not combined
with harmful alcohol use, frequent use of marijuana or both especially among men. The high prevalence
of addictive behaviours in households with children indicates that many families are presently affected
negatively by alcohol, gambling and marijuana. This suggests that pathological gambling should be included
systematically in future public health strategies, treatment programs and interventions in Greenland.
Keywords: problem gambling; indigenous health; social pathologies; addictive behaviour; Inuit
Received: 15 August 2012; Revised: 24 December 2012; Accepted: 24 December 2012; Published: 15 March 2013
ublic health research has pointed to alcohol and
substance abuse as the most significant public
health challenge in Greenland. The negative consequences of especially alcohol problems reach far into
the lives of communities, families and children. It is
well known that use of alcohol and marijuana is often
P
combined, but few studies have investigated how gambling problems might be included in a pattern of crossaddictive behaviours among Greenland Inuit.
In recent years, pathological gambling has become
an issue of increasing concern in Greenland (1). For the
majority, gambling is a harmless activity, but for some
Int J Circumpolar Health 2013. # 2013 Christina Viskum Lytken Larsen et al. This is an Open Access article distributed under the terms of the Creative
Commons Attribution-Noncommercial 3.0 Unported License (http://creativecommons.org/licenses/by-nc/3.0/), permitting all non-commercial use,
distribution, and reproduction in any medium, provided the original work is properly cited.
Citation: Int J Circumpolar Health 2013, 72: 19551 - http://dx.doi.org/10.3402/ijch.v72i0.19551
1
(page number not for citation purpose)
Christina Viskum Lytken Larsen et al.
gambling becomes a disorder with a negative impact on
their personal lives, families and the communities they
live in (2). Little research has been conducted in this area
among Greenland Inuit, which might explain why public
health programs and intervention strategies have not yet
taken pathological gambling into account. In the Inuit
Health in Transition Greenland survey 20062010,
we found the lifetime prevalence for problem gambling
to be 9% among women and 16% among men, and the
gambling patterns reflected a mix of modern games
popular in contemporary Scandinavia and games found
to be traditionally popular among Inuit in Nunavik
and First Nations (3). The high prevalence of lifetime
problem gambling reflects the increase in social pathologies Greenland has experienced during the past 60 years.
The rapid social transition Greenland Inuit have gone
through has been accompanied by a high prevalence of
addictive behaviours (4).
Although the average alcohol intake has decreased
since the 1980s, where the alcohol consumption was at
its highest (5), a drinking pattern dominated by binge
drinking persists as a major cause of social and health
problems in Greenland. Today, Greenland is still struggling with the negative consequences for the generations
who grew up with alcohol problems in their childhood
home such as suicidal behaviour and a high prevalence
of violence and sexual abuse (6,7), and, in recent years,
increased political attention has been given to improving the lives of children and families. The use of
hard drugs has never been a problem in Greenland.
Given the geographical isolation, the police have managed to prevent drugs, other than marijuana, from
entering Greenland. The issue of substance abuse
among Greenland Inuit this primarily regards the use
of marijuana.
Several international studies have found problem
gambling to be associated with alcohol and substance
use as well as mental disorders, but most of these studies
have been conducted in a clinical context among pathological gamblers in treatment, and their findings cannot
easily be generalised to problem gamblers in the general
population. A recent review evaluated the prevalence of
comorbid disorders among problem and pathological
gamblers in population surveys and pointed to an elevated
prevalence of both substance and alcohol use disorders
among problem and pathological gamblers compared to
non-problem gamblers across studies (8). However, none
of these studies had been conducted among indigenous
populations, where the prevalence of addictive behaviours
is high.
The objectives of the present study were, thus, to
investigate (a) the association between lifetime problem
gambling and harmful alcohol use as well as frequent
use of marijuana and (b) the prevalence of crossaddictive behaviour among Greenland Inuit. We hypothe-
2
(page number not for citation purpose)
sised that lifetime problem gambling was associated
with a higher prevalence of harmful alcohol use and
frequent use of marijuana, compared to the prevalence
among non-problem gamblers and those who never
gambled.
Material and methods
Data and sample size
Data was derived from the cross-sectional Inuit Health
in Transition Greenland Survey 20052010. The study
was a population-based general health study among
adults (18). A full methodological report of the study
is published elsewhere (9). Greenland was divided into
12 strata based on geography and community size. From
each region, a number of towns and villages were selected
for the study. A random sample was drawn from the
central population register to obtain around 300 participants from the towns included. In the selected villages,
everyone was invited to participate. Data were collected
by an interviewer-based questionnaire combined with
a self-administered questionnaire containing the more
sensitive and personal questions, which were believed not
to be suited for the face-to-face interview. The questions
regarding problem gambling, alcohol use and use of
marijuana were all included in the self-administered
questionnaire. Questions about gambling were included
in the self-administered questionnaire from 2006 and
onwards. The 20062010 sample included 3,893 Greenland Inuit. A total of 2,451 persons participated in the
general survey (63%), and 2,189 persons filled out the
self-administered questionnaire (56%). The analyses were
based on these 2,189 participants. In total, 2,012 of the
2,189 participants (92%) answered at least one question
about gambling.
Despite Greenland’s large geographical size, its total
population is only about 57,000, of whom 90% are
ethnic Greenlanders (Inuit). Genetically, Greenlanders
are Inuit (Eskimos) with a mixture of European, mainly
Scandinavian, genes. They are closely related to the Inuit/
Iñupiat in Canada and Alaska and, somewhat more
distantly, to the Yupiit of Alaska and Siberia (10).
Lifetime problem gambling, harmful alcohol use
and frequent use of marijuana
The questions about pathological gambling in this study
were a small part of a larger health survey, which forced us
to use a short screen. Problem gambling was measured
using the Lie/Bet Questionnaire (see Table II) originally
suggested by Johnson and Hammer (11) and later validated by both Johnson and Hammer (12) and Götestam
et al. (13). The Lie/Bet Questionnaire is a two-question
version of the 10 DSM-IV criteria for pathological gambling (2). According to DSM-IV, a positive response to 5 or
more of the 10 criteria is characterised as pathological
Citation: Int J Circumpolar Health 2013, 72: 19551 - http://dx.doi.org/10.3402/ijch.v72i0.19551
Cross-addictive behaviour among Greenland Inuit
gambling, while three to four positive answers are defined
as problem gambling. The lie/bet screen has been found
valid to identify lifetime problem gamblers, defined as
those who responded positively to 5 or more of the 10 DSMIV criteria in a combined group with those who responded
positively to only three to four of these criteria (13).
The short screen cannot distinguish between pathological
and problem gambling. Respondents were asked whether
or not they had lied to friends and family about their
gambling activities, and whether or not they had felt a
need to increase bets in the past year or previously in life.
Positive answers for the past year and previously were
combined for a prevalence of lifetime problem gambling,
that is problem gambling that had occurred at some point
in life. Analyses are based on lifetime gamblers as opposed
to past year gamblers in order include enough individuals
to be able to analyse subgroups in the sample. Nonproblem gamblers and those who never gamble are
collapsed into one group in the analyses. When mentioned in the text, they are called non-problem gamblers,
although they include a small group who never gamble.
Harmful alcohol use was measured by the modified
CAGE-test: CAGE-C. It is a simple screening tool suited
for identifying alcohol problems in populations with a
high prevalence of at-risk drinkers. The original CAGE
test was based on a 4-item questionnaire and measured
harmful alcohol use in a lifetime perspective (14).
However, the validity of the original questionnaire outside a clinical context has been questioned (15), and the
sensitivity of the test has been criticised in several studies
(16,17). The 6-item questionnaire CAGE-C was suggested by Zierau et al. in 2005 (18) and validated against
a diagnostic interview based on ICD-10 (19) and DSMIII R (20) criteria. The questionnaire (Fig. 1) has been
used to assess harmful alcohol intake among Greenland
Inuit in an earlier study (21). The modified CAGE-test
was included in the self-administered questionnaire. It
measures harmful alcohol use in a past year perspective
and includes a question regarding the number of days per
week of alcohol use and a question concerning alcohol
intake on weekdays outside meals. CAGE positives were
defined by a positive answer in two or more of questions
14 and 6, or one positive answer in question 14 and 6,
in addition to alcohol intake on 4 or more days per week.
Frequent use of marijuana in the past year was
measured through two questions in the self-administered
questionnaire. Participants were asked if they had ever
smoked marijuana. Those who answered ‘‘yes, once or a
few times’’ or ‘‘yes, several times’’ were asked how often
they had smoked marijuana in the past year. Those who
had tried smoking marijuana and additionally answered
that they had smoked at least one to three times a month
in the past year were categorised as past-year frequent
users of marijuana. Those who had never tried to smoke
and those who had smoked less than once a month in the
past year were categorised together as non-frequent
users/non-users.
Socio-demographic variables
Age, place of residence and formal education were
included as socio-demographic variables. To ensure
enough observations, age groups were collapsed into
two groups of 1834 years and 35 years. Formal
education was determined from questions about highest
school education attained and further vocational or
academic education, and recoded into primary school/
high school only, short vocational education (less than 3
years), and longer vocational/academic education. Information on the socio-demographic variables was obtained
through the interviewer-based questionnaire.
Place of residence was divided between the capital
Nuuk, towns and villages. Greenland has a total of 80
communities all located on the coast. A town is defined
historically as the largest community in each of 17
districts. In 2010, the population of Nuuk was 15,469,
the population of the towns varied between 469 and
5,460, while that of the villages varied from less than 10
to around 550. In the towns are located district school(s),
a health centre or hospital, church, district administration and main shops. These institutions are absent or
present to a much smaller extent in the villages. The
isolated populations in the villages and smaller towns
face the largest challenges with limited job and recreational opportunities. In contrast to these, the capital,
Nuuk, with its more than 15,000 inhabitants is a modern
town reflecting contemporary Scandinavian lifestyle with
a wide range of educational, occupational and recreational possibilities (22).
Alcohol problems in the childhood home
Traumatic events during childhood have been found
to be associated with problem gambling in previous
studies (23). Therefore, logistic regressions were adjusted
for alcohol problems in childhood in the analyses of the
association between lifetime problem gambling, harmful
alcohol use and frequent use of marijuana. Participants
were asked through the self-administered questionnaire if
there were alcohol problems in their childhood home
(never; yes, occasionally; yes, often).
Statistical analyses
Differences between problem gamblers and non-problem
gamblers regarding harmful alcohol use and frequent use
of marijuana according to gender, age group and place
of residence were tested with chi-square test. Prevalence
estimates were weighted for regional and age differences
in the sample. Analyses were done by logistic regression
stratified for gender. Results are reported as odds ratios
(OR) with a 95% confidence interval (CI). Both adjusted
and unadjusted ORs are reported.
Citation: Int J Circumpolar Health 2013, 72: 19551 - http://dx.doi.org/10.3402/ijch.v72i0.19551
3
(page number not for citation purpose)
Christina Viskum Lytken Larsen et al.
Ethical review
The study was ethically approved by the Commission for
Scientific Research in Greenland. Participants gave their
written consent after being informed about the study
orally and in writing.
group (pB0.0001) and place of residence (pB0.0001). In
general, the odds ratio for having answered the questions
was higher among men than women, and among inhabitants from the capital Nuuk compared to inhabitants
from towns and villages. The odds ratio decreased with
age and increased with level of education.
Results
Socio-demographic variation in response
The variables included in the study are displayed in
Table I. Initially the response rate for the key variables
included was analysed for associations with sociodemographic variables (results not tabulated). Response
to questions regarding problem gambling was associated
with gender (p B0.0001), formal education (p B0.0001)
and age group (p B0.0001). Response to questions
about harmful alcohol use was associated with gender
(p0.012), formal education (p0.004), age group
(pB0.0001) and place of residence (pB0.0001). Response
to questions about use of marijuana was associated with
gender (pB0.0001), formal education (pB0.0001), age
Table I. Overview of variables included. Frequency, valid
percent, N, missing data and socio-demographic associations
in response rates for the variables regarding problem gambling,
alcohol use and marijuana
Frequency (%)
Total
Men
Women
Age groups
1824 years
0
2,189
0
2,189
0
278 (12.7)
385 (17.6)
1,273 (58.2)
253 (11.6)
Place of residence
Town
Village
2,189
1,206 (55.1)
2534 years
Nuuk
Missing
983 (44.9)
3559 years
60 N
419 (19.1)
1,110 (50.7)
660 (30.2)
Lifetime problem gambling
1,542
647
1,599
590
1,833
356
(Lie/bet screen)
Yes
196 (12.7)
No
1,346 (87.3)
Harmful alcohol use past
year (CAGE-C)
Yes
No
521 (32.6)
1,078 (67.4)
Frequent use of marijuana
past year
Yes
250 (13.6)
No
1,583 (86.4)
Response rates were analysed by logistic regression for
associations with gender, age, place of residence and level of
education. Greenland Inuit 20062010.
4
(page number not for citation purpose)
Harmful alcohol use
Lifetime problem gambling was significantly associated
with a harmful use of alcohol among men, but not
among women (Table II), and the odds ratio for harmful
alcohol use was twice as high for male lifetime problem
gamblers as for non-problem gamblers (both unadjusted
and adjusted). The prevalence of harmful alcohol use
among lifetime problem gamblers compared to nonproblem gamblers was significantly higher for those
aged 35. Among lifetime problem gamblers, the prevalence of harmful alcohol use increased with age, while
decreasing with age among the non-problem gamblers
(men and women). Harmful alcohol use was more
prevalent among lifetime problem gamblers compared
to non-problem gamblers in towns and villages, but not in
the capital (men only).
Frequent use of marijuana
Lifetime problem gambling was significantly associated
with a frequent use of marijuana (Table II). Women
displayed an odds ratio for frequent use of marijuana
almost 3 times higher among lifetime problem gamblers
compared to non-problem gamblers (both unadjusted
and adjusted). The prevalence of frequent use of marijuana was significantly higher among lifetime problem
gamblers in both age groups compared to non-problem
gamblers. There was a significantly higher prevalence
of frequent use of marijuana among lifetime problem
gamblers compared to non-problem gamblers in the
capital and towns, but not in the villages.
Cross-addictive behaviours
In Fig. 2, the different combinations of lifetime problem gambling, harmful alcohol use and frequent use
of marijuana were analysed for men and women
(pB0.0001). A total of 8 different categories were constructed, that is 1) a category of no addictive behaviour,
where neither lifetime problem gambling, frequent use
of marijuana or harmful alcohol use were present;
2) lifetime problem gambling only; 3) harmful alcohol
use only; 4) frequent use of marijuana only; 5) lifetime
problem gambling combined with frequent use of
marijuana; 6) lifetime problem gambling combined
with harmful alcohol use; 7) frequent use of marijuana
combined with harmful alcohol use; and, finally, 8)
lifetime problem gambling combined with both frequent
use of marijuana and harmful alcohol use.
Fifty three percent of men and 37% of women were
either lifetime problem gamblers, used marijuana
Citation: Int J Circumpolar Health 2013, 72: 19551 - http://dx.doi.org/10.3402/ijch.v72i0.19551
Cross-addictive behaviour among Greenland Inuit
Table II. Prevalence of harmful alcohol use and frequent use of marijuana in past year among lifetime problem gamblers compared to
non-problem gamblers/non-gamblers according to gender, age and place of residence
Harmful alcohol use past year (CAGE-C)
Frequent use of marijuana past year
Non-problem
Lifetime problem
gamblers
gamblers/
non-gamblers
Unadjusted
2.24 (1.443.48)
Adjusted for age and place of
2.26 (1.453.54)
2.14 (1.353.41)
Non-problem
p
Lifetime problem
gamblers
gamblers/
non-gamblers
p
1 (ref.)
B0.0001
1.74 (1.072.83)
1 (ref.)
0.027
1 (ref.)
B0.0001
1.73 (1.062.85)
1 (ref.)
0.030
1 (ref.)
0.001
1.56 (0.942.59)
1 (ref.)
0.089
34.4% (178)
0.001
16.8% (94)
B0.0001
Men
Odds ratio (95% CI)
residence
Adjusted for age, place of
residence and alcohol problems
in childhood home
Percent (N) All men
49.6% (54)
40.6% (27)
Age groups
1834 years
34.1% (17)
45.3% (57)
0.207
40.0% (11)
20.4% (30)
0.007
58.7% (37)
29.1% (121)
B0.0001
40.3% (16)
15.2% (41)
B0.0001
40.0% (5)
27.2% (28)
0.303*
53.3% (34)
73.1% (15)
36.4% (91)
41.7% (59)
0.008
0.005
1.58 (0.902.77)
1.59 (0.902.81)
1 (ref.)
1 (ref.)
0.115
0.109
2.75 (1.445.25)
2.98 (1.545.79)
1 (ref.)
1 (ref.)
0.002
0.001
1.26 (0.702.26)
1 (ref.)
0.442
2.69 (1.355.34)
1 (ref.)
0.005
24.3% (156)
0.189
28.0% (14)
11.9% (59)
0.001
41.2% (7)
23.3% (7)
19.5% (30)
8.0% (29)
0.034*
0.001*
Nuuk
44.8% (10)
17.7% (60)
0.001
Town
25.9% (45)
13.2% (171)
0.001
Village
7.0% (20)
9.2% (103)
0.631*
35 years
Place of residence
Nuuk
Town
Village
Women
Odds ratio (95% CI)
Unadjusted
Adjusted for age and place of
residence
Adjusted for age, place of
residence and alcohol problems
in childhood home
Percent (N) All women
34.1% (21)
Age groups
1834 years
21.4% (6)
30.5% (66)
39.4% (15)
23.0% (90)
0.474
0.035
Nuuk
36.4% (5)
21.4% (32)
0.256*
Town
39.1% (11)
26.9% (80)
0.210
Village
26.7% (5)
28.4% (44)
0.890*
35 years
Place of residence
Total
Percent (N) Place of residence
*Expected count less than 5.
Analysis of frequent use of marijuana according to place of residence was not stratified by gender because this resulted in too few
individuals in each group. Greenland Inuit 20062010. Logistic regressions were stratified according to gender. Unadjusted and adjusted
odds ratios (OR) with 95% confidence intervals (CI).
p50.05 values are present in bold.
frequently, and/or had harmful alcohol use, or different
combinations of the three. Harmful alcohol use was the
most prevalent addictive behaviour among both men
and women. Harmful alcohol use and frequent use of
marijuana co-occurred with lifetime problem gambling
among 12.2% of the men and 3.7% of the women.
In comparison, harmful alcohol use and frequent use of
marijuana co-occurred among 12.6% of men and 6.5%
of women. Cross-addictive behaviour was found among
20.4% of men and 8.6% of women. For 4.4% of men
Citation: Int J Circumpolar Health 2013, 72: 19551 - http://dx.doi.org/10.3402/ijch.v72i0.19551
5
(page number not for citation purpose)
Christina Viskum Lytken Larsen et al.
The lie/bet screen questionnaire: Positive answers for past year and previously were combined for a
prevalence of lifetime problem gambling, i.e. problem gambling that has occurred at some point in life.
1
Have you ever lied (to family members, friends, co-workers or teachers) about how much you gamble,
how much you have lost or how large your debt is? Yes, past year/ Yes, previously/ No/ Never gamble
2
Have you ever felt the need to bet more and more money (to reach the same level of exitement)? Yes,
past year/ Yes, previously/ No/ Never gamble
The CAGE-C test questionnaire: A positive result was defined as two or more positive answers in questions
1-4 and 6, or one positive answer in questions1-4 and 6 in addition to alcohol intake on 4 or more days per
week
1
Have you, within the past year, felt that you should cut down on your drinking? Yes/No
2
Have people, within the past year, annoyed you by criticizing your drinking? Yes/No
3
Have you, within the past year, felt bad or guilty about your drinking? Yes/No
4
Have you, within the past year, from time to time had a drink first thing in the morning to steady your
nerves or to get rid of a hangover (eye opener?) Yes/No
5
Have many days per week, do you drink alcohol?
0 days
6
1 day
2 days
3 days
4 days
5 days
6 days
Do you drink alcohol on weekdays outside mealtimes? Yes/No
Fig. 1. Questionnaires for the Lie/Bet screen and the CAGE-C test used in the Inuit Health in Transition Greenland Survey 20062010.
and 1.5% of women all three addictive behaviours were
present.
Addictive behaviours in households with children
In total, the prevalence of one or more addictive
behaviours in households with children was 44%. The
prevalence of only one addictive behaviour in households
with children was similar to households without children
(32% vs. 28%), while the prevalence of two or more
addictive behaviours was slightly higher in households
without children compared to those with children (19%
vs. 12%) (p 0.001, results not tabulated). The odds
ratio for finding one or more addictive behaviours in a
household was slightly lower in households with children
compared to households without children (OR 0.76, CI
95%: 0.600.97; adjusted for age group and gender).
Discussion
Harmful alcohol use and frequent use of marijuana
were both associated with lifetime problem gambling
and the prevalence of these addictive behaviours was
significantly higher among lifetime problem gamblers
compared to non-problem gamblers. Among problem
gamblers the harmful use of alcohol increased with age
while the opposite pattern characterised the non-problem
gamblers. For lifetime problem gamblers, the gambling
problems were more often than not combined with harmful alcohol use, a frequent use of marijuana or both especially among men. More than half of the men and
one third of the women displayed one or more addictive
6
(page number not for citation purpose)
behaviours. This was also the case in almost half of the
households with children.
Our findings regarding co-morbid behaviour are in
agreement with international findings from other population-based studies. Lorains et al. (8) found a weighted
mean estimate of 29.8% for alcohol use disorder among
combined groups of problem and pathological gamblers
across 3 population-based studies, although none of these
were based on indigenous populations. Two of these
studies used the original CAGE questionnaire and found
a total prevalence for alcohol use disorder among
problem gamblers of 14% in 2005 (24,25). Considering
that the original version of CAGE measures harmful
alcohol use in a lifetime perspective, while CAGE-C
applies for the past year, the prevalence of harmful use of
alcohol among lifetime problem gamblers in Greenland
seems high especially for men.
The combination of alcohol and marijuana in a crossaddictive behaviour is well-known among public health
professionals in Greenland, but gambling problems have
not systematically been included in strategies for prevention in this field. Our study shows, that the co-occurrence
of harmful alcohol use and frequent use of marijuana
with lifetime problem gambling is similar to the cooccurrence of harmful alcohol use and frequent use of
marijuana. Problem gambling was measured in a lifetime
perspective. This means that gambling problems may not
be an issue any more, but despite this, there is a prevalent
overlap with alcohol problems and frequent use of
marijuana in the past year. Based on these observations,
Citation: Int J Circumpolar Health 2013, 72: 19551 - http://dx.doi.org/10.3402/ijch.v72i0.19551
Cross-addictive behaviour among Greenland Inuit
Fig. 2. Prevalence of lifetime problem gambling, frequent use of marijuana in the past year, harmful alcohol use in the past year and
the different combinations between the three as well as no addictive behaviour. The total prevalence of each of the three addictive
behaviours as well as no addictive behaviour is written in bold figures. The prevalence of one addictive behaviour only, or the
combination of two or three addictive behaviours, is written in (). Unadjusted percent, weighted for regional and age differences in the
sample. Men (left), N 579, missing 404. Women (right), N605, missing 601. Greenland Inuit 20062010.
it seems important to consider gambling problems as a
part of other addictive behaviours, even though these
problems may not be current.
The high prevalence of one or more addictive behaviours in households with children suggests that many
families and children in Greenland are affected by
alcohol, gambling and marijuana. A report from the
national program for early intervention in pregnant
families in Greenland, which had 15% of all families
with newborns from 20072010 enrolled, showed that all
of these families were struggling with problems related
to addictive behaviour (26). The negative influence of
alcohol was also stated by the participants in the Inuit
Health in Transition Greenland Survey, who rated
alcohol to be the most important theme together with
violence, when asked to rate the importance of the themes
included in Greenland’s first public health program
‘‘Inuuneritta 20072012’’. This evaluation was further
supported by results from the Survey of Living Conditions, where participants rated unemployment, alcohol
and drug abuse, suicide, violence in the family and sexual
abuse as serious social and health-related issues within
the Greenlandic communities (27). An increase in social
pathologies, such as addictive behaviours, is not only a
challenge for Greenland, but a key indicator for indigenous people in the Arctic undergoing a similar transition
(4). These issues need to be addressed through comprehensive public health strategies and preventive measures
taking the complexity and the fundamental challenges
related to sociocultural changes in the Greenlandic
society into account.
Targeting interventions and treatment for pathological
gamblers is a challenge because very few actually seek
treatment for their disorder. A large US survey suggested
that only 712% of the pathological gamblers seek
treatment (28), and Kessler et al. (29) found that none
of the problem gamblers in their study had ever received
treatment for gambling problems, while half of them were
treated for other mental disorders. This corresponds well
with the experience from Greenland, where very few have
been treated for pathological gambling. In consequence,
there seems to be an important task for health professionals and social workers of screening for gambling
problems when working with alcohol and substance use
disorders.
Mental health disorders, such as bipolar disorder,
major depression and antisocial personality disorders,
have been found to be highly associated with problem
gambling (8). However, mental health disorders could
not be diagnosed on the basis of the health questionnaire
in the Greenland Survey and, therefore, these were not
included in the analyses, although they are likely to
contribute to the prevalence of addictive behaviours.
Given a seemingly high prevalence of mental disorders
among Greenland Inuit (30,31), this association is
important to consider when public health interventions
and treatment are planned in practice. A study based
on patients in the Greenlandic health care system
Citation: Int J Circumpolar Health 2013, 72: 19551 - http://dx.doi.org/10.3402/ijch.v72i0.19551
7
(page number not for citation purpose)
Christina Viskum Lytken Larsen et al.
found a prevalence of at least one ICD-10 diagnose
among 49% of the patients, with anxiety, somatisation
disorder, dysthymia and depression as the most prevalent diagnoses. However these results were based
on a small population and cannot be generalised to
the broad population. More knowledge about the prevalence and incidence of specific mental disorders among
Greenland Inuit is needed and should take the possible
link between mental disorders and problem gambling
into consideration.
Future research should look into the aetiology of
pathological gambling in relation to alcohol and substance use disorders, mental health disorders and their
interrelatedness. Few studies have tried to establish what
comes first based on retrospective questions asking
participants to remember their age of onset for different
addictive behaviours (29). Prospective studies are needed
in order to ensure a better understanding of what comes
first and how the prevention and treatment of one
disorder might help to prevent other addictive behaviours. Furthermore, research should investigate the
causes behind the increase in social pathologies such as
addictive behaviours, which have been found to be a key
feature in indigenous populations undergoing a similar
transition as Greenland Inuit (4).
It is considered a strength of the study that different
types of addictive behaviours are measured in the same
survey. This has provided us with the opportunity to
investigate cross-addictive behaviour among lifetime
problem gamblers and the prevalence of different combinations of addictive behaviours. The number of problem
gamblers in our sample is relatively large, which has made
it possible to study the cross-addictive behaviours according to socio-demographic characteristics.
It is an important weakness of the study that the
response to key variables followed a social gradient.
When working with addictive behaviours and socially
exposed groups, the missing data is expected to lead to an
underestimated prevalence, because these groups are hard
to reach through surveys. However, it should be noted
that the prevalence found when combining the three
different variables for addictive behaviours in this study
is slightly higher, compared to the national prevalence
estimates, when calculated separately. This is caused by
the combination of missing data across the 3 measures,
when they are combined in one variable, which reduces
the ‘‘N’’. Furthermore, it is possible that gambling and
alcohol problems are under-reported in the study, due to
the general stigma attached to addictive behaviours.
Based on the existing data, we were not able to further
asses a possible under-reporting regarding these subjects,
but the number of missing data on the lie/bet screen was
larger than the number of missing data on questions
regarding different types of gambling (not included in
this study). This suggests that people are more likely to
8
(page number not for citation purpose)
answer questions about gambling activities, rather than
gambling problems. Whether this has to do with participants not wanting to answer questions about gambling
problems they might have, or with the fact that it is
tempting to skip a couple of questions in a long
questionnaire if they seem irrelevant to you personally,
should be investigated through future qualitative methodological studies.
Finally, the lack of comparable measures for mental
health disorders in the study may lead to a less comprehensive understanding of the prevalence and combinations of problem gambling, harmful use of alcohol and
frequent use of marijuana.
Conflict of interest and funding
The authors have not received any funding or benefits from
industry or elsewhere to conduct this study.
References
1. PAARISA Department of Health GG. Gambling and gambling patterns [in Danish]. Nuuk: PAARISA Government of
Greenland; 2006.
2. American Psychiatric Association. Diagnostic and statistical
manual of mental disorders text revision. 4th edition.
Washington, DC: American Psychiatric Association; 2000.
3. Larsen CVL, Curtis T, Bjerregaard P. Gambling behavior and
problem gambling reflecting social transition and traumatic
childhood events among Greenland Inuit a cross-sectional
study in a large indigenous population. J Gambl Stud 2012.
doi: 10.1007/s10899-012-9337-6.
4. Bjerregaard P, Young K, Dewailly E, Ebbesson SOE. Indigenous health in the Arctic: an overview of the circumpolar Inuit
population. Scand J Public Health. 2002;32:3905.
5. Statistics Greenland. 2002. Import of alcohol to Greenland
19752001. Report (in Danish). Nuuk: Statistics Greenland.
6. Bjerregaard P, Curtis T. Cultural change and mental health in
Greenland. The association of childhood conditions, language
and urbanization with mental health and suicidal thoughts
among the Inuit of Greenland. Soc Sci Med. 2002;54:33.
7. Curtis T, Larsen FB, Helweg-Larsen K, Bjerregaard P.
Violence, sexual abuse and health in Greenland. Int J
Circumpolar Health. 2002;61:11022.
8. Lorrains FK, Cowlishaw S, Thomas SA. Prevalence of
comorbid disorders in problem and pathological gambling:
systematic review and meta-analysis of population surveys.
Addiction. 2012;106:4908.
9. Bjerregaard P. Inuit health in transition Greenland survey
20052010. Population sample and survey methods. 2nd
revised edition. Copenhagen: National Institute of Public
Health; 2011 [cited 2012 Aug 15]. Available from: http://
www.si-folkesundhed.dk/upload/inuit_health_in_transition_
greenland_methods_5_2nd_revision_002.pdf
10. Young TK, Bjerregaard P. Health transitions in Arctic
populations. Toronto: University of Toronto Press; 2008.
11. Johnson EE, Hammer R. The Lie/Bet Questionnaire for
screening pathological gamblers. Psychol Rep. 1997;80:838.
12. Johnson EE, Hammer RM. The Lie/Bet Questionnaire for
screening pathological gamblers: a follow-up study. Psychol
Rep. 1998;83:121924.
13. Götestam GK, Johansson A, Wenzel HG, Simonsen IE.
Validation of the lie/bet screen for pathological gambling
Citation: Int J Circumpolar Health 2013, 72: 19551 - http://dx.doi.org/10.3402/ijch.v72i0.19551
Cross-addictive behaviour among Greenland Inuit
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
on two normal population data sets. Psychol Rep. 2004;95:
100913.
Mayfield D, McLeod G, Hall P. The CAGE questionnaire:
validation of a new alcoholism screening instrument. Am J
Psychiatry. 1974;131:11213.
Bisson J, Nadeau L, Demers A. The validity of the CAGE
scale to screen for heavy drinking and drinking problems in a
general population survey. Addiction. 1999;94:71522.
McCusker MT, Basquille J, Khwaja M, Murray-Lyon IM,
Catalan J. Hazardous and harmful drinking: a comparison of
the AUDIT and CAGE screening questionnaires. Q J Med.
2002;95:5915.
MacKenzie DM, Langa A, Brown TM. Identifying hazardous
or harmful alcohol use in medical admissions: a comparison
of AUDIT, CAGE and BRIEF MAST. Alcohol Alcohol.
1996;31:5919.
Zierau F, Hardt J, Henriksen H, Holm SS, Jørring S, Melsen T,
et al. Validation of a self-administered modified CAGE test
(CAGE-C) in a somatic hospital ward: comparison with
biochemical markers. Scand J Clin Lab Inves. 2005;65:61522.
World Health Organization. 1992. The ICD-10 classification of
mental and behavioral disorders. Clinical descriptions and
diagnostic guidelines. Geneva: World Health Organization.
American Psychiatric Association. Diagnostic and statistical
manual of mental disorders. Revised 3rd edition. Washington,
DC: American Psychiatric Association; 1987.
Madsen MH, Grønbæk M, Bjerregaard P, Becker U.
Urbanization, migration and alcohol use in a population of
Greenland Inuit. Int J Circumpolar Health. 2005;64:23445.
Berner J, Furgal C, Bjerregaard P, Bradley M, Curtis T,
De Fabo E, et al. ACIA, 2005. Arctic climate impact
assessment. New York: Cambridge University Press; 2005.
Hodgins DC, Schopflocher DP, el-Guebaly N, Casey DM,
Smith GJ, Williams RJ, et al. The association between childhood maltreatment and gambling problems in a community
sample of adult men and women. Psychol Addict Behav.
2010;24:54854.
Bondolfi G, Osiek C, Ferrero F. Prevalence estimates of
pathological gambling in Switzerland. Acta Psychiatr Scand.
2000;101:4735.
25. Bondolfi G, Jermann F, Ferrero F, Zullino D, Osiek C.
Prevalence of pathological gambling in Switzerland after the
opening of casinos and the introduction of new preventive
legislation. Acta Psychiatr Scand. 2008;117:2369.
26. PAARISA Government of Greenland. Mid-term evaluation
and status report 2010. Evaluation report for ‘Early intervention in pregnant families’ [Danish only]. Nuuk: PAARISA Government of Greenland; 2011 [cited 2012 Aug 15]. Available
from: http://www.paarisa.gl/media/18450/midtvejsevaluering_
2010_tidlig_indsats.pdf
27. Kruse J, Poppel B, Abryutina L, Duhaime G. Survey of living
conditions in the Arctic: what did we learn? Anchorage:
Institute of Social and Economic Research, University of
Alaska Anchorage; 2007.
28. Slutske WS. Natural recovery and treatment-seeking in pathological gambling: results of two US national surveys. Am J
Psychiatry. 2006;163:297302.
29. Kessler RC, Hwang R, LaBrie M, Petukhova M, Sampson
NA, Winters KC, et al. DSM-IV pathological gambling in the
National Comorbidity Survey Replication. Psychol Med.
2008;38:135160.
30. Lynge I, Munk-Jørgensen P, Lynge Pedersen A, Mulvad G,
Bjerregaard P. Common mental health disorders among
patients in primary health care in Greenland. Int J Circumpolar Health. 2004;63(Suppl 2):37783.
31. Lynge I, Munk-Jørgensen P, Lynge Pedersen A, Mulvad G,
Bjerregaard P. Mental health among patients in the Greenlandic health system [report in Danish]. Copenhagen: National
Institute of Public Health; 2003. Report No.: SIFs Writings on
Greenland no. 15.
*Christina Viskum Lytken Larsen
National Institute of Public Health
University of Southern Denmark
Oester Farimagsgade 5A, 2
1353 Copenhagen K, Denmark
Tel: 45 65507070
Fax: 45 39208010
Email: [email protected]
Citation: Int J Circumpolar Health 2013, 72: 19551 - http://dx.doi.org/10.3402/ijch.v72i0.19551
9
(page number not for citation purpose)
PAPER III
Larsen CVL, Curtis T, Bjerregaard P. Health status and health behavior associated with frequent gambling
and problem gambling in a large indigenous population – A cross-sectional study based on The Inuit Health
in Transition Greenland Survey 2006-2010. Article in review November 2013
Health status and health behavior associated with frequent gambling and
problem gambling in a large indigenous population – A cross-sectional
study based on The Inuit Health in Transition Greenland Survey 20062010.
AUTHORS, AFFILIATION AND ADRESSES:
Authors
Christina V L Larsen1, Tine Curtis2 and Peter Bjerregaard1
Institutions
1
National Institute of Public Health, University of Southern Denmark, Denmark
2
Local Government Denmark, Denmark
E-mails:
Christina V L Larsen: [email protected]
Peter Bjerregaard: [email protected]
Tine Curtis: [email protected]
Correspondence:
Christina Viskum Lytken Larsen
PhD student, MA in Sociology
National Institute of Public Health
University of Southern Denmark
Oester Farimagsgade 5A, 2.
E-mail: [email protected]/ Fax: +45 39208010
1353 Copenhagen K, Denmark
Tel: +45 65507070
RUNNING HEAD:
Gambling and health among Greenland Inuit
DECLARATION OF INTEREST:
None
WORD COUNT: 2,997
ABSTRACT
Background Despite a movement towards a public health paradigm for gambling, few studies have
investigated pathological gambling in a more general health perspective, and to an even lesser extent
gambling behavior has been included in the analytical framework. Indigenous populations undergoing rapid
social transition have been found to be vulnerable to gambling problems, but knowledge in this field is
needed to develop an evidence based public health approach to gambling activities and pathological
gambling. The objective of the study was to investigate health status and health behavior among frequent
gamblers and lifetime problem gamblers among Greenland Inuit.
Methods A representative cross-sectional study was conducted among Greenland Inuit (18+) (n=2189) in 9
towns and 13 villages in Greenland from 2005-2010. Frequent gambling, lifetime problem gambling, health
status and health behavior were measured through clinical examinations, an interview based questionnaire
and a self-administered questionnaire.
Results Frequent gambling and lifetime problem gambling was associated with a poorer health status and
unhealthy behavior. Frequent gamblers and problem gamblers experienced a higher prevalence of mental
symptoms, long-standing illness and a poorer self-rated health. Daily smoking, a harmful use of alcohol, a
frequent use of marijuana, an unhealthy diet and use of health services was also more prevalent compared
to those who gambled less than once a month and non-problem gamblers.
Conclusion Variation according to gambling activity and different measures of problem gambling suggests a
complex process related to living conditions and socioeconomic differences among Greenland Inuit. Future
studies should focus on the social determinants of health and gambling problems among indigenous
populations undergoing rapid social transition.
INTRODUCTION
Pathological gambling has traditionally been an issue of the psychiatric discipline and analyses of
comorbidity have focused on mental disorders and addictive behavior. During the last two decades
increased attention in public health has been paid to the importance of including gambling behavior and
pathological gambling in a public health frame work (1-3). Despite a movement towards a public health
paradigm for gambling, few studies have investigated pathological gambling in a more general health
perspective and as a result little is known about associations with health status and health behavior (4). To
an even lesser extent gambling behavior has been included in the analytical framework, although it is likely
that engaging in different types of gambling activities on a frequent basis may be associated with health
given specific patterns and ways of life that characterize different gambling activities.
Indigenous populations experience a higher prevalence of gambling problems compared to their nonindigenous counterparts (5-7). The rapid transition among indigenous populations in the Circumpolar North
has also been accompanied by an epidemiological transition from infectious diseases to chronic diseases
(8). The changing lifestyles and health conditions as well as an increase in social pathologies among
indigenous populations in the Circumpolar North call for more knowledge on possible associations between
on one hand frequent gambling and gambling problems, and on the other, health in general among these
populations. Gambling was therefore included as a topic in the Inuit Health in Transition Survey covering
both Inuit in Nunavik (Canada) and Greenland (9;10).
This study is based on results from the Inuit Health in Transition Greenland Survey. Given a high prevalence
of lifetime problem gambling among Greenland Inuit (11), the survey represents a unique opportunity to
analyze gambling behavior in a broader health perspective relevant for public health planning with focus on
indigenous populations. The objective of the study was to investigate the general health status and health
behavior according to frequent gambling and lifetime problem gambling in a representative sample of
Greenland Inuit.
METHODS AND MATERIAL
Study design, setting and participants
Data were derived from the cross-sectional Inuit Health in Transition Greenland Survey 2005-2010 - a
population based general health study among a representative sample of adults (18+). A full
methodological report of the study is published elsewhere (9). Data was collected as clinical examinations
as well as through a structured interview conducted by the interviewers in either Greenlandic of Danish
based on the choice of the participant, and finally through a self-administered questionnaire that was filled
out by the participant at the location after the clinical examinations and interview. Questions about
gambling were included in the self-administered questionnaire from 2006 and onwards. The sample for
2006-2010 included 3,893 Greenland Inuit. A total of 2,451 persons participated in the general survey (63%)
and 2,189 persons filled out the self-administered questionnaire (56%). The analyses were based on these
2,189 participants. In total 2,012 of the 2,189 participants (92%) answered at least one question about
gambling.
Greenland has a total of 80 communities all located along the coast divided into towns and villages. A town
is defined historically as the largest community in each of 17 districts. In 2010, the population of the towns
varied between 469 to 5,460 and 15,469 in the capital Nuuk while that of villages varied from less than 10
to around 550. In the towns are located district school(s), health center or hospital, church, district
administration and main shops. These institutions are absent or present to a much smaller extent in
villages. Despite its large geographical size, the total population of Greenland is only about 57,000 of whom
90% are ethnic Greenlanders (Inuit). Genetically, Greenlanders are Inuit (Eskimos) with a mixture of
European, mainly Scandinavian genes. Greenlanders are both genetically and culturally closely related to
the Inuit/Iñupiat in Canada and Alaska and, somewhat more distantly, to the Yupiit of Alaska and Siberia
(8).
Gambling behavior and problem gambling
Frequent gambling was measured through the self-administered questionnaire. Respondents were asked
how often they gambled during past year regarding the lottery, bingo, cards/dice and slot machines and
answers were dichotomized into gambling once a month or more vs. less than once a month for each type
of activity. Problem gambling was measured using the lie/bet screen (12), which has been validated in both
treatment (13) and community samples (14). It is a two questions version of the 10 DSM-IV criteria for
pathological gambling (15). The lie/bet screen has been found valid to identify lifetime problem gamblers in
a community sample, defined as those who responded positively to five or more of the 10 DSM-IV criteria
combined with those who only responded positively to three-four of these criteria (14). The short screen
does not qualify to distinguish between pathological and problem gambling. Positive answers to one or
both questions for past year and previously were combined into a measure of lifetime problem gambling.
Two additional questions were included to measures problem gambling. Respondents were asked a) if they
felt they had spent too much time or money on gambling during past year or previously and b) whether
they themselves had ever considered they had a gambling problem or had been told so by others during
past year or previously. Positive answers for past year and previously were combined into a measure of
lifetime problem gambling for each of these variables.
Health status
Self-rated health was measured in the interview questionnaire by asking respondents to rate their own
health within five categories: Excellent, good, fair, poor, very poor. Answers were dichotomized into
excellent/good vs. fair/poor/very poor. Mental symptoms were measured in the interview questionnaire,
by asking respondents whether they had been bothered a little, a lot or not at all by (a) anxiety,
nervousness, agitation, or fear (hereafter anxiety); or (b) feeling depressed or unhappy (hereafter
depression), in the two weeks preceding the interview. Dichotomous variables were created to identify
people reporting feeling anxious vs. those who did not, and people reporting feeling depressed vs. those
who did not. Long-standing illness was assessed through interview by asking: Do you suffer from any longstanding illness, after-effect of an injury, handicap or other long-standing disorder? Yes/no. If a person
answered yes, they were asked to specify their illness. Only the dichotomized variable based on yes/no was
used in this study. Finally height and weight were measured with the participants stripped to their
underwear and socks (9). Weight was measured on a standard electronic clinical scale. Body Mass Index
(BMI) was calculated based on these measurements (kg/m2) and dichotomized according to a BMI above or
below 30. Participants with a BMI of 30+ were classified as obese (16).
Health behavior
Information about daily smoking was obtained from the interview questionnaire. Harmful alcohol use was
measured by the modified CAGE-test, CAGE-C, which is a simple screening tool suited for identifying alcohol
problems in populations with a high prevalence of at-risk drinkers. The six-item questionnaire CAGE-C was
suggested by Zierau et al. in 2005 (17) and validated against a diagnostic interview based on ICD-10 (18)
and DSM-III R (19) criteria. The questionnaire has been used to assess harmful alcohol intake among
Greenland Inuit in an earlier study (20). Frequent use of marijuana during past year was measured through
the self-administered questionnaire. Those who had tried to smoke marijuana and additionally answered
they had smoked at least 1-3 times a month during past year were categorized as past year frequent users
of marijuana vs. non-frequent users/non users. Sedentary time was measured through the interview
questionnaire based on two questions regarding time spent on sitting during weekdays and weekend
respectively based on the past seven days from the International Physical Activity Questionnaire (IPAQ)
adapted for living conditions in Greenland (17;18). The variable was dichotomized using the median in
order to compare the 50% with the most sedentary time to the 50% with the least sedentary time (>3 hours
vs. <3 hours daily). Information on unhealthy diet was obtained by an interviewer-administered Food
Frequency Questionnaire (FFQ) with portion sizes. The FFQ was developed from information obtained
through a 24-hour dietary recall. Questions were asked about 67 food items, including 23 local and 44
imported items (9). An unhealthy diet was defined as 25 E% or more from soda pop, fast food, snacks,
sweets and sugar added to coffee or tea (19). Information regarding use of health services was obtained
from the interview questionnaire. Consultations with a doctor, nurse or healthcare assistant at the local
health center or hospital during the three months preceding the interview was registered as having made
use of health services during the past three months.
Sociodemographic characteristics
Age was divided into four groups; 18-24, 25-34, 35-59 and 60+ years. Place of residence was divided into
villages, towns and the capital. Formal education was included as a measure of socioeconomic position and
determined from questions in the interviewer based questionnaire about highest school education attained
and further vocational or academic education and recoded into primary school/high school only, short
vocational education (less than 3 years), and longer vocational/academic education.
Statistical analyses
Statistical analyses were performed in IBM SPSS Statistics 19. Due to the sampling procedure, oversampling
took place in certain strata while the participation rate was not similar according to age and gender.
Accordingly, data were weighted for sampling strata by the weight procedure of SPSS in order to obtain
representative prevalence estimates. All variables regarding health behavior and health status were
dichotomous and analyzed by logistic regression. Regressions were performed separately for men and
women for one health outcome at a time and adjusted for age group (treated as a categorical variable),
formal education and place of residence.
Ethical review
The study was ethically approved by the Commission for Scientific Research in Greenland. Participants gave
their written consent after being informed about the study orally and in writing.
[INSERT TABLE 1]
RESULTS
Key variables are shown in Table 1. Frequent gambling on the lottery, bingo as well as cards/dice was
associated with several measures of both health status and health behavior, but most associations were
only significant for either men or women (Table 2). In general frequent gambling was associated with a
poorer health status or an unhealthier behavior compared to those who never gambled or gambled less
than once a month, but significant differences characterized different types of gambling activities. Daily
smoking was almost two-three times higher among frequent players of bingo, cards/dice and slot machines
– especially among women – while the overall prevalence of daily smoking among those who gambled
frequently on the lottery was a little lower compared to those who did not. A harmful use of alcohol was
also less prevalent among male lottery players and among bingo players. Only poor self-rated health and
use of health services were not significantly associated with any types of frequent gambling.
[INSERT TABLE 2]
All measures of health status and health behavior except sedentary time were significantly associated with
lifetime problem gambling, but some variation applied according to gender and measure of problem
gambling (Table 3). Lifetime problem gamblers had a poorer health and an unhealthier behavior compared
to non-problem gamblers with obesity as the exception. Women who reported having spent too much time
or money on gambling were less obese than non-problem gamblers.
Symptoms of anxiety and depression during the past two weeks, daily smoking and harmful use of alcohol
were significantly more prevalent among lifetime problem gamblers no matter how problem gambling was
measured. Respondents’ perception of having spent too much time or money on gambling was associated
with most of the included outcomes.
[INSERT TABLE 3]
DISCUSSION
Frequent gambling and lifetime problem gambling were associated with a poorer health and unhealthy
behavior. Associations varied according to type of gambling and how problem gambling was measured with
significant gender differences. More frequent gamblers and lifetime problem gamblers experienced mental
symptoms, long-standing illness, a poorer self-rated health, smoked daily, had a harmful use of alcohol and
smoked marijuana frequently, enjoyed an unhealthy diet and had made more use of health services
compared to those who gambled less frequently or never gambled as well as non-problem gamblers.
Exceptions were depression, daily smoking, harmful alcohol use and sedentary time during the day among
those who gambled frequently on the lotteries and bingo, because fewer were depressed, sedentary,
smoked daily and were obese compared to non-frequent gamblers.
The associations found between problem gambling and mental symptoms, smoking, harmful alcohol use
and frequent use of marijuana are consistent with international findings from population based surveys
(20). The empirical literature regarding problem gambling and health beyond mental disorders, smoking,
substance and alcohol use is limited, but the existing studies found lifetime pathological gambling to be
associated with several medical disorders, poor lifestyle choices and worse quality of life (21;22). Our
results thus support these findings, although they are not directly comparable because they are based on
different populations and different measures of gambling problems. The positive health outcomes
associated with frequent gambling could indicate a social gradient in preference for type of gambling and
problem gambling. The lotteries are played more among the highest educated in Greenland (unpublished
results), which would explain why the unadjusted analyses show a lower prevalence of symptoms of
depression and daily smoking among those who gamble frequently on the lottery. Lifetime problem
gamblers were found to be less obese, which could be explained by a higher prevalence of problem
gamblers in less privileged groups. Obesity among Greenland Inuit follows a social gradient according to
wealth, meaning the wealthier the more obese (unpublished results). The lower OR for obesity among
lifetime problem gamblers was significant for women even after adjusting for formal education, suggesting
the association is even more complex than the inequalities in health often found in relation to different
levels of education.
Due to the nature of a cross-sectional survey, our data does not provide any information regarding
causality between health and gambling. The Pathways Model of problem gambling suggested by
Blaszczynski and Nower (23) argues that problem gamblers are a heterogeneous group and proposes three
different pathways. The first pathway covers problem gamblers with no severe psychological problems
prior to their extensive gambling, but where alcohol abuse, depression and anxiety might appear as a result
of their gambling; the second pathway covers a group where emotional vulnerability due to for example
childhood disturbance is present prior to the gambling problems. This group uses gambling to modify
affective states or meet specific needs. The third pathway is constituted by highly disturbed individuals with
substantial psychological features such as impulsivity and antisocial personality disorder. Given a high
prevalence of traumatic events during childhood and the documented association with mental health
among Greenland Inuit (24-26) a pathway to problem gambling through emotional vulnerability is a
relevant perspective. However none of the pathways are based on empirical findings among indigenous
populations where sociocultural and societal changes are important determinants of both physical and
mental health (27).
The associations between gambling and a poorer health as well as unhealthy behavior are likely to be a
result of an even more complex process related to living conditions and socioeconomic differences among
Greenland Inuit. This perspective follows the theoretical approach proposed by Marmot and Wilkinson
regarding social determinants of health (28), which states that health follows a social gradient meaning the
higher the social position, the better the health. The focus of epidemiological research should be the causes
of causes such as different behavioral determinants of biological markers and the living conditions and
circumstances that influence peoples’ access to a healthy lifestyle. These causes are defined as social
determinants of health. Following this line of thought future research should focus on the social
determinants of poor health and unhealthy life among Greenland Inuit and how gambling behavior and
problem gambling are linked to these determinants.
Because of its brief character, the lie/bet screen is likely to cover very mild gambling problems along with
more severe ones compared to more comprehensive screens (29;30). However, it is an important finding
that associations with health are significant despite this methodological issue. It is interesting that
associations between problem gambling and health are significant even when using one-item questions of
participants’ own perception to assess lifetime problem gambling. Although these one-item measures
based on peoples own perception of their gambling problems cannot be used to diagnose pathological
gamblers in a clinical context, self-perceived health measures are an important source of information from
a public health point of view. It is well known, that self-rated health is a significant predictor for mortality
(31;32). The negative health outcomes associated with participants feeling they spent too much time or
money on gambling or who rate themselves as problem gamblers could serve as an important source of
information for health professionals and social workers regarding when to intervene.
A particular strength of the study are the various topics included in the general health survey and the
representative sample of a large indigenous population, which provides us with new insight regarding
health and gambling in a vulnerable population.
A weakness of the study is the cross-sectional design which makes it impossible to establish the causality of
the associations. Also, there is a risk of selection bias due to sociodemographic differences in the missing
data. Another weakness is the fact that the lie/bet screen has only been used in few other studies. This
makes it difficult to compare our results with other studies.
More epidemiological as well as qualitative research is needed to understand the relationship between
health and gambling among Greenland Inuit and other indigenous populations. Prospective and register
based studies are needed to investigate causality in the relationship between health and gambling and to
identify the possible underlying social determinants.
CONCLUSION
Frequent gambling and lifetime problem gambling was associated with a poorer health and unhealthy
behavior. Variation according to gambling activity and different measures of problem gambling suggests a
complex phenomenon related to living conditions and socioeconomic differences among Greenland Inuit.
Future studies should focus on the social determinants of health and gambling problems among vulnerable
populations.
REFERENCES
(1) Korn D, Gibbins R, Azmier J. Framing public policy towards a public health paradigm for gambling.
Journal of Gambling Studies 2003;19(2):235-56.
(2) Korn DA. Examining Gambling Issues from a public health perspective. 1-18. 23-6-2002.
(3) Shaffer HJ, Korn DA. Gambling and related mental disorders: a public health analysis. Annual
Review of Public Health 2002;23:171-212.
(4) Griffiths M. Betting your life on it. British Medical Journal 2004 Nov 6;329(7474):1055-6.
(5) Abbott M, Volberg R. Taking the pulse on gambling and problem gambling in New Zealand: A report
on phase one of the 1999 national prevalence survey. Report Number Three of the New Zealand
Gaming Survey. Wellington: The Department of Internal Affairs; 2000.
(6) Wardman D, el-Guebaly N, Hodgins D. Problem and pathological gambling in North American
Aboriginal populations: a review of the empirical literature. Journal of Gambling Studies
2001;17(2):81-100.
(7) Bjerregaard P, Young K, Dewailly E, Ebbesson SOE. Indigenous health n the Arctic: an overview of
the circumpolar Inuit population. Scandinavian Journal of Public Health 2002;32:390-5.
(8) Young TK, Bjerregaard P. Health Transitions in Arctic Populations. Toronto: University of Toronto
Press; 2008.
(9) Bjerregaard P. Inuit Health in Transition - Greenland Survey 2005-2010. Population sample and
survey methods. 2nd revised edition. Copenhagen: National Institute of Public Health; 2011.
(10) Munckle G, Boucher O, Laflamme D, Chevalier S, Rochette L. Alcohol, drug use and gambling among
the Inuit of Nunavik: Epidemiological Profile. Québec: Institut national de santé publique du
Québec. Nunavik Regional Board of Health and Social Services; 2007.
(11) Larsen CVL, Curtis T, Bjerregaard P. Gambling behavior and problem gambling reflecting social
transition and traumatic childhood events among Greenland Inuit - a cross-sectional study in a large
indigenous population. Journal of Gambling Studies 2012.
(12) Johnson EE, Hammer, R. The Lie/Bet Questionnaire for Screening Pathological Gamblers.
Psychological Reports 1997;80:83-8.
(13) Johnson EE, Hammer RM. The Lie/Bet Questionnaire for Screening Pathological Gamblers: A
Follow-Up Study. Psychological Reports 1998;83:1219-24.
(14) Götestam GK, Johansson A, Wenzel HG, Simonsen IE. Validation of the Lie/Bet Screen for
Pathological Gambling on Two Normal Population Data Sets. Psychological Reports 2004;95:100913.
(15) American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders - Text
Revision. 4th Edition ed. 2000.
(16) WHO. Preventing and managing the global epidemic. Report of a WHO consultation. Geneva: WHO;
2000.
(17) Dahl-Petersen IK, Jorgensen ME, Bjerregaard P. Physical activity patterns in Greenland: a country in
transition. Scandinavian Journal of Public Health 2011 Nov;39(7):678-86.
(18) International Physical Activity Questionnaire. http://www.ipaq.ki.se/ipaq.htm. 2012. 23-9-2012.
(19) Bjerregaard P, Jeppesen C. Inuit dietary patterns in modern Greenland. International Journal of
Circumpolar Health 2010;69(1):13-24.
(20) Lorrains FK, Cowlishaw S, Thomas SA. Prevalence of comorbid disorders in problem and
pathological gambling: systematic review and meta-analysis of population surveys. Addiction
2012;106:490-8.
(21) Morasco BJ, Petrzak RH, Blanco C, Grant BF, Hasin D, Petry NM. Health problems and medical
utilization associated with gambling disorders: results from the National Epidemiological Survey on
Alcohol and Related Conditions. Psychosomatic Medicine 2006;68(6):976-84.
(22) Black DW, Shaw M, McCormick B, Allen J. Pathological gambling: relationship to obesity, selfreported chronic medical conditions, poor lifestyle choices, and impaired quality of life.
Comprehensive Psychiatry 2012;In press.
(23) Blaszczynski A, Nower L. A pathways model of problem and pathological gambling. Addiction 2002
May;97(5):487-99.
(24) Lynge I, Munk-Jørgensen P, Lynge Pedersen A, Mulvad G, Bjerregaard P. Common mental health
disorders among patients in primary health care in Greenland. International Journal of Circumpolar
Health 2004;63(Supplements 2):377-83.
(25) Bjerregaard P, Curtis T. Cultural change and mental health in Greenland. The association of
childhood conditions, language and urbanization with mental health and suicidal thoughts among
the Inuit of Greenland. Social Science and Medicine 2002;(54):-33.
(26) Curtis T, Larsen FB, Helweg-Larsen K, Bjerregaard P. Violence, sexual abuse and health in
Greenland. International Journal of Circumpolar Health 2002;61(2):110-22.
(27) Curtis T, Kvernmo S, Bjerregaard P. Changing living conditions, life style and health. Journal of
Circumpolar Health 2005;64(5):442-50.
(28) Marmot M, Wilkinson RG. Social Determinants of Health. Second edition ed. New York: Oxford
University Press; 2011.
(29) Lesieur HR, Blume SB. Revising the South Oaks Gambling Screen in different settings. Journal of
Gambling Studies 1993;9(3):213-23.
(30) Gerstein DR, Arwood RA, Hoffmann J, Larison C, Engelman L, Murphy S. Gambling impact and
behavior study. Report to the National Gambling Impact Study Commission. Chicago: National
Opinion Research Center at the University of Chicago.; 1999.
(31) DeSalvo KB, Bloser N, Reynolds K, He J, Muntner P. Mortality prediction with a single general selfrated health question. A meta-analysis. Journal of Gen Intern Medicine 2006;21:267-75.
(32) Idler EL, Benyamini Y. Self-rated health and mortality: a review of twenty-seven community studies.
Journal of Health and Social Behavior 1997;38:21-37.
Table 1. Overview of key variables included in the study. Frequency, valid percent and missing data. Greenland Inuit 2006-2010 (N=2189).
Frequency
(Pct.)
Sociodemographic characteristica
N
Missing
data
2189
0
Gender
Men
Women
Frequency (Pct.)
Health status
Anxiety
983 (44.9)
1206
(55.1)
2189
0
Yes
730 (33.4)
No
1456 (66.6)
18-24
278 (12.7)
Depression
25-34
Yes
605 (27.7)
35-59
385 (17.6)
1273
(58.2)
No
1576 (72.3)
60+
253 (11.6)
Place of residence
Long-standing illness
Yes
777 (35.8)
No
1395 (63.6)
Town
419 (19.1)
1110
(50.7)
Village
660 (30.2)
Excellent/good
Formal education
Primary school/high school only
2189
1312
(60.7)
0
Self-rated health
2161
28
Fair/poor/very poor
620 (28.7)
BMI < 30
1681 (77.5)
Longer vocational/academic education
229 (10.6)
BMI > 30
489 (22.5)
1751
No
Gambled on slot mashines at least once a month
438
447 (25.5)
1304
(74.5)
Gambled on bingo at least once a month
Yes
2181
8
2172
17
2181
8
2170
19
2189
0
1599
590
1833
356
Health behavior
Gambled on the lotteries at least once a month
No
3
794 (36.4)
Obesity
Gambling behavior during past year
2186
1387 (63.6)
Short vocational education (less than 3 years)
Yes
Missing
data
Mental symptoms during past two weeks
Age groups
Nuuk
N
1888
301
459 (24.3)
1429
(75.7)
1679
510
Daily smoker
793 (36.2)
Yes
1369 (63.8)
No
793 (36.2)
Harmful alcohol use past year (CAGE-C)
Yes
521 (32.6)
No
1078 (67.4)
Frequent use of marijuana past year (smoked at least once a
month)
Yes
No
Gambled on cards and dice at least once a
month
Yes
No
68 (4.1)
1611
(95.9)
1731
458
210 (12.1)
1521
(87.9)
Lie/bet screen
Non-problem gambler/ never gamble
1542
No/never gamble
No/ never gamble
No
1583 (86.4)
Sedentary time during past 7 days
3 hours or less daily
1082 (50.8)
More than 3 hours daily
1047 (49.2)
465
254(14.7)
1470
(85.3)
1500
113 (7.5)
1387
(92.4)
Yes
617
No
1572
Use of health services during past 3 months
1724
Selfrated problem gambling
Yes, past year or previously (lifetime)
647
196 (12.7)
1346
(87.3)
Spent too much time or money on gambling
Yes, past year or previously (lifetime)
250 (13.6)
Unhealthy diet
Problem gambling
Lifetime problem gambler
Yes
Yes
841 (43.9)
No
1073 (56.1)
2129
60
2189
0
1914
275
689
Sociodemographic variation applied to the response rates for questions regarding gambling behavior, problem gambling and the measures of health
behavior. For the questions regarding gambling behavior, harmful alcohol use, frequent use of marijuana and having spent too much time or money
on gambling, the odds ratio (OR) for answering was higher among men compared to women and among inhabitants in the capital compared to
inhabitants from towns and villages. The OR decreased with age and increased with level of education. Following the same pattern, response to selfrated problem gambling was associated with gender, age group and formal education but not place of residence and response to the lie/bet screen
was only associated with gender and age group. Response to use of health services was associated with place of residence and formal education, but
with a decrease in ORs for having answered with increase in level of education.
Table 2. Prevalence and odds ratio (OR) for different measures of health status and health behavior among frequent gamblers (at least once a month
during past year) on the lotteries, bingo, cards/dice and slot maschines. Past year problem gamblers (lie/bet screen) were excluded in the analyses.
The general prevalence was weighted for regional, gender and age differences in the sample. Binary logistic regressions were performmed for men
and women seperately and both unadjusted and adjusted for age group, formal education and place of residence. Odds ratios are shown with 95%
confidence intervals. Greenland Inuit 2006-2010.
HEALTH STATUS
Monthly gambling on….
The lotteries (N=1751)
Yes (men=263; women=184)
No (men=564; women=740)
p-value
OR, Men [C.I. 95%]
Unadjusted
Adjusted
OR, Women [C.I. 95%]
Unadjusted
Adjusted
Bingo (N=1888)
Yes (men=153; women=306)
No (men=695; women=734)
p-value
OR, Men [C.I. 95%]
Unadjusted
Adjusted
OR, Women [C.I. 95%]
Unadjusted
Adjusted
Cards/dice (N=1731)
HEALTH BEHAVIOR
Anxiety
Depression
Longstanding
illness
27.6
30.3
0.212
20.0
24.4
0.045
33.2
32.9
0.902
1.04
[0.73-1.48]
1.10
[0.75-1.60]
0.95
[0.65-1.39]
1.01
[0.67-1.52]
Harmful
Sedentary time
alcohol use
during past 7
(past year) Frequent use days (> 3 hours
CAGE-C of marijuana
daily)
Poor selfrated
health
Obesity (BMI
>30)
Daily smoker
33.3
34.2
0.711
24.7
22.8
0.387
55.9
60.9
0.048
31.0
31.6
0.818
15.4
17.4
0.308
0.86
1.01
[0.63-1.19] [0.73-1.39]
0.93
1.05
[0.65-1.32] [0.74-1.48]
1.50
[1.04-2.16] *
1.19
[0.81-1.76]
1.07
[0.79-1.46]
1.19
[0.85-1.66]
0.67
[0.47-0.94] *
0.68
[0.47-0.99] *
1.41
[0.95-2.10] †
1.13
[0.74-1.73]
0.83
1.33
0.90
[0.58-1.20]
[0.94-1.86] [0.63-1.27]
1.03
1.37
1.06
[0.70-1.51] [0.95-1.98] † [0.73-1.53]
1.03
[0.71-1.50]
1.06
[0.71-1.58]
0.82
[0.59-1.15]
0.97
[0.68-1.39]
0.89
[0.59-1.39]
0.94
[0.61-1.44]
35.1
34.2
0.727
23.5
23.2
0.884
70.3
56.7
>0.0001
21.8
33.7
>0.0001
15.7
16.1
0.864
1.22
1.81
[0.81-1.83] [1.20-2.72] *
1.26
1.90
[0.82-1.93] [1.24-2.92] *
0.97
1.24
[0.67-1.42] [0.85-1.80]
0.77
1.12
[0.52-1.53] [0.75-1.64]
1.04
[0.66-1.64]
0.94
[0.58-1.52]
1.57
[1.07-2.31] *
1.69
[1.13-2.55] *
0.82
[0.54-1.24]
0.73
[0.47-1.12]
1.04
[0.79-1.37]
1.00
[0.74-1.34]
1.21
0.98
[0.91-1.61] [0.75-1.32]
0.99
0.81
[0.73-1.34] [0.60-1.10]
0.89
2.27
[0.65-1.22] [1.67-3.10] ***
0.71
2.21
[0.51-0.99] * [1.59-3.08] ***
0.94
[0.66-1.34]
0.91
[0.62-1.33]
1.33
[0.95-1.86] †
1.62
[1.13-2.30] **
32.6
29.4
0.217
29.9.6
22.4
0.002
1.08
[0.81-1.45]
0.96
[0.71-1.31]
30.8
34.0
0.227
Unhealthy
diet
Use of health
services during
past 3 months
58.8
54.6
0.107
27.2
29.3
0.376
41.0
40.6
0.898
1.12
[0.83-1.52]
0.88
[0.64-1.23]
1.06
[0.76-1.46]
1.16
[0.81-1.66]
1.14
[0.82-1.60]
0.97
[0.67-1.40]
0.73
1.86
[0.40-1.32] [1.32-2.63] ***
0.59
1.53
[0.31-1.11]
[1.05-2.21] *
0.90
[0.62-1.31]
0.98
[0.65-1.47]
1.22
[0.86-1.73]
1.17
[0.80-1.69]
47.7
56.4
0.002
26.3
29.0
0.275
44.2
40.1
0.166
1.11
[0.68-1.83]
1.18
[0.70-2.01]
0.64
[0.45-0.92] *
0.72
[0.49-1.05] †
1.24
[0.85-1.81]
1.59
[1.06-2.41] *
1.26
[0.85-1.87]
1.12
[0.74-1.69]
1.02
[0.62-1.67]
1.29
[0.75-2.23]
0.79
[0.60-1.05]
1.06
[0.79-1.44]
1.11
[0.82-1.50] 0.89 [0.67-1.20]
1.31
0.95
[0.94-1.82]
[0.69-1.30]
Yes (men=106; women=104)
No (men=701; women=820)
p-value
OR, Men [C.I. 95%]
Unadjusted
Adjusted
OR, Women [C.I. 95%]
Unadjusted
Adjusted
Slot maschines (N=1679)
Yes (men=44; women=24)
No (men=742; women=869)
p-value
OR, Men [C.I. 95%]
Unadjusted
Adjusted
OR, Women [C.I. 95%]
Unadjusted
Adjusted
38.8
29.4
35.4
22.3
27.9
32.6
33.8
34.0
17.9
23.2
79.7
58.0
36.1
30.9
25.4
15.9
53.5
55.8
33.3
27.4
36.6
42.2
0.018
<0.0001
0.246
0.964
0.146
<0.0001
0.239
0.005
0.699
0.122
0.229
1.48
1.62
[0.92-2.36] [0.96-2.65] †
1.39
1.61
[0.86-2.27] [0.97-2.66] †
0.87
0.98
[0.55-1.38] [0.62-1.56]
0.93
0.99
[0.57-1.52] [0.62-1.59]
1.13
[0.66-1.94]
1.36
[0.77-2.39]
1.89
[1.17-3.05] **
1.50
[0.91-2.46]
1.42
[0.87-2.33]
1.35
[0.82-2.25]
1.63
[0.94-2.80] †
1.61
[0.91-2.84]
0.60
[0.39-0.94] *
0.67
[0.42-1.06] †
1.46
[0.95-2.24]
1.32
[0.84-2.08]
0.87
[0.53-1.42]
0.99
[0.59-1.65]
1.17
1.70
[0.76-1.80] [1.10-2.63] *
1.00
1.41
[0.64-1.55]
[0.90-2.20]
1.10
1.19
[0.71-1.72] [0.77-1.85]
1.06
1.04
[0.67-1.69] [0.66-1.63]
0.71
3.70
[0.43-1.20] [2.06-6.66] ***
0.66
3.23
[0.39-1.13] [1.79-5.86] ***
0.98
[0.57-1.66]
0.84
[0.49-1.46]
1.50
[0.78-2.90]
1.57
[0.77-3.18]
1.30
[0.84-2.00]
1.74
[1.10-2.75] *
1.31
[0.84-2.03]
1.32
[0.84-2.10]
0.84
[0.54-1.31]
0.84
[0.53-1.33]
21.1
30.2
0.102
14.3
23.3
0.079
0.77
[0.33-1.79]
0.77
[0.27-1.90]
0.93
[0.40-2.15]
0.71
[0.27-1.90]
39.4
33.9
0.334
24.3
23.2
0.826
65.7
59.6
0.304
37.9
31.3
0.291
19.7
16.7
0.529
52.4
56.2
0.388
45.7
27.6
0.001
53.2
41.3
0.062
0.96
[0.41-2.23]
1.03
[0.44-2.43]
0.90
1.08
[0.43-1.87] [0.53-2.20]
0.98
1.41
[0.45-2.14] [0.63-3.15]
1.43
[0.66-3.13]
1.41
[0.60-2.73]
1.46
[0.71-3.01]
1.37
[0.65-2.87]
1.27
[0.59-2.74]
1.27
[0.57-2.82]
1.87
[0.85-4.12]
1.51
[0.67-3.42]
1.40
[0.70-2.78]
1.26
[0.61-2.58]
1.41
[0.70-2.83]
1.29
[0.62-2.70]
1.50
[0.73-3.10]
1.69
[0.79-3.63]
0.62
[0.20-1.90]
0.77
[0.25-2.41]
1.15
1.65
[0.45-2.94] [0.66-4.09]
0.96
1.67
[0.36-2.53] [0.63-4.44]
1.69
[0.66-4.34]
1.67
[0.63-4.44]
1.65
[0.59-4.62]
2.32
[0.81-6.64]
0.38
[0.09-1.71]
0.38
[0.08-1.73]
1.03
[0.23-4.58]
0.91
[0.20-4.21]
0.88
[0.36-2.20]
0.62
[0.24-1.63]
1.58
[0.61-4.06]
2.19
[0.82-5.83]
0.95
[0.35-2.55]
0.81
[0.30-2.22]
*** p < 0.001; ** p < 0.01; * p < 0.05; † < 0.10
28.8
32.9
0.483
Table 3. Prevalence and odds ratio (OR) for different measures of health status and health behavior among lifetime problem gamblers (different
measures applied). The general prevalence was weighted for regional, gender and age differences in the sample. Binary logistic regressions were
performmed for men and women seperately and both unadjusted and adjusted for age group, formal education and place of residence. Odds ratios
are shown with 95% confidence intervals.
HEALTH STATUS
HEALTH BEHAVIOR
Anxiety
Depression
Longstanding
illness
34.0
31.9
0.547
30.7
24.3
0.048
37.3
32.1
0.146
1.36
1.72
[0.88-2.10] [1.11-2.66] *
1.08
[0.71-1.63]
1.44
1.68
[0.92-2.25] [1.08-2.64] *
1.23
[0.79-1.92]
1.64
[1.10-2.46]
*
1.64
[1.09-2.48]
*
Unadjusted
1.61
1.86
[1.01-2.56] * [1.16-2.97] *
1.29
[0.80-2.07]
1.21
[0.76-1.93]
0.82
[0.48-1.41]
Adjusted
1.62
1.76
[1.01-2.61] * [1.08-2.86] *
1.32
[0.80-2.16)
1.16
[0.71-1.88]
0.77
[0.44-1.33]
2.33
[1.32-4.10]
**
2.26
[1.26-4.04]
**
36.5
31.9
0.155
40.5
34.0
0.049
15.4
24.3
0.002
66.5
59.1
0.028
0.96
[0.63-1.47]
1.61
[1.08-2.40]
*
1.08
[0.70-1.67]
1.62
[1.08-2.43]
Lifetime problem gambling
Lie/bet screen (N=1542)
Problem gamblers (men=116;
women=80)
Non-problem gamblers/ non-gamblers
p-value
OR, Men [C.I. 95%]
Unadjusted
Adjusted
OR, Women [C.I. 95%]
Too much time and money… (N=1724)
Yes (men=117; women=137)
No (men=694; women=776)
p-value
OR, Men [C.I. 95%]
40.9
29.1
>0.0001
Unadjusted
1.96
[1.29-2.97] **
Adjusted
1.97
[1.29-3.02] **
33.5
22.0
>0.0001
2.02
[1.31-3.12]
**
2.05
[1.32-3.19]
**
Poor selfrated
health
43.7
34.8
0.013
Obesity
(BMI >30) Daily smoker
20.0
23.6
0.260
73.3
59.9
>0.0001
1.09
[0.66-1.79]
1.35
[0.89-2.06]
1.16
[0.69-1.95]
1.29
[0.83-2.01]
Harmful
alcohol use
(past year) Frequent use
CAGE-C of marijuana
45.8
29.5
>0.0001
37.6
14.6
>0.0001
2.24
[1.44-3.48]
1.74
*** [1.07-2.83] *
2.25
[1.44-3.54]
1.66
*** [1.00-2.76] *
Sedentary
time during
past 7 days
(> 3 hours
daily)
Unhealthy
diet
Use of
health
services
during past
3 months
60.5
53.9
0.076
43.7
34.8
0.013
43.8
42.0
0.664
1.06
1.66
[0.71-1.58] [1.10-2.50] *
1.45
[0.93-2.27]
†
1.11
1.47
[0.74-1.68] [0.95-2.27] †
1.46
[0.92-2.32]
1.35
[0.84-2.16]
1.46
[0.90-2.37]
1.10
[0.67-1.81]
1.46
[0.89-2.38]
1.53
[0.92-2.53]
1.11
[0.67-1.85]
1.46
[0.82-2.60]
2.75
[1.44-5.25]
**
2.81
[1.42-5.56]
**
39.9
28.9
0.002
24.4
15.9
0.002
51.7
55.7
0.254
38.0
27.5
0.001
51.7
38.9
0.001
0.65
[0.37-1.13]
1.35
1.83
1.77
[0.89-2.06] [1.19-2.83] ** [1.09-2.87] *
0.79
[0.53-1.18]
1.73
[1.15-2.60]
**
0.63
[0.35-1.11]
1.26
1.86
1.60
[0.82-1.94] [1.19-2.89] ** [0.98-2.64] †
0.79
1.62
[0.52-1.18] [1.06-2.47] *
1.72
[1.12-2.65]
*
1.76
[1.12-2.74]
*
1.58
[0.90-2.77]
OR, Women [C.I. 95%]
0.64
[0.41-1.00]
1.52
2.00
* [1.02-2.27] * [1.32-3.03] **
0.55
[0.35-0.88]
1.38
1.87
*
[0.92-2.08] [1.22-2.87] **
Unadjusted
1.24
[0.86-1.79]
1.35
[0.98-1.96]
1.36
[0.93-1.98]
1.08
[0.74-1.57]
Adjusted
1.11 [0.761.61]
1.13
[0.76-1.67]
1.24
[0.84-1.83]
1.01
[0.68-1.49]
41.0
31.4
35.6
24.4
48.5
32.2
44.6
35.0
18.8
23.8
69.3
60.4
0.046
0.012
0.001
0.054
0.252
1.07
[0.60-1.92]
1.25
[0.71-2.21]
1.23
[0.66-2.27]
Self-rated problem gambling (N=1500)
Yes (men=54; women=59)
No (men=663; women=724)
p-value
OR, Men [C.I. 95%]
Unadjusted
Adjusted
OR, Women [C.I. 95%]
1.69
[0.94-3.04] †
1.76
[0.95-3.25] †
Unadjusted
1.66
[0.97-2.82] †
Adjusted
1.69
[0.98-2.92] †
2.44
[1.37-4.35]
**
2.54
[1.39-4.63]
**
2.45
[1.44-4.19]
**
2.44
[1.40-4.24]
**
2.49
[1.45-4.25]
**
2.21
[1.27-3.84]
**
1.35
[0.76-2.41]
0.96
[0.66-1.39]
1.34
[0.90-1.97]
0.98
[0.66-1.45]
1.49
[0.81-2.76]
1.18
[0.80-1.76]
1.37
[0.91-2.07]
0.99
[0.66-1.48]
38.4
30.7
24.5
17.0
62.6
53.8
34.7
28.5
53.8
41.3
0.077
0.140
0.065
0.088
0.187
0.030
1.09
[0.54-2.16]
1.15
[0.65-2.06]
2.00
[1.08-3.72] *
1.73
[0.89-3.36]
1.27
[0.72-2.24]
1.44
[0.81-2.57]
1.50
[0.81-2.78]
1.24
[0.69-2.24]
1.05
[0.51-2.13]
1.26
[0.68-2.33]
2.13
[1.12-4.06] *
1.47
[0.73-2.96]
1.06
[0.59-1.91]
1.39
[0.75-2.56]
1.58
[0.83-3.01]
1.50
[0.88-2.56]
0.75
2.04
[0.40-1.42] [1.08-3.85] *
1.10
[0.59-2.06]
1.01
[0.39-2.64]
1.49
[0.81-2.75]
1.49
[0.86-2.59]
0.64
1.99
[0.33-1.22] [1.04-3.81] *
1.00
[0.53-1.91]
1.15
[0.43-3.10]
1.35
1.43
[0.79-2.33]
[0.82-2.49]
1.70
[0.96-3.00]
1.71
† [0.96-3.05] †
1.51
[0.81-2.81]
*** p < 0.001; ** p < 0.01; * p < 0.05; † < 0.10. Analyses of the association between the included measures of lifetime problem gambling showed the
measures were significantly associated with each other (p<0.0001, results not tabulated). The OR for having spent too much time and money on
gambling, if you were a lifetime problem gambler according to the lie/bet screen, was 9 (C.I. 95%: 6.3-12.7) and the OR for self-rated problem
gambling was 17 (C.I. 95%: 10.7-25.9). If you had spent too much time and money on gambling, the OR for self-rated problem gambling was 21 (C.I.
95%: 13.0-33.2). However, a kappa<0.4 for all two-way combinations, indicated a poor agreement between the included measures of lifetime
problem gambling.
ISSN: 1601-7765
ISBN: 978-87-7899-263-5
ISBN: 978-87-7899-264-2