© International Epidemiological Association 2000
International Journal of Epidemiology 2000;29:470–477
Printed in Great Britain
Country of birth, acculturation status and
abdominal obesity in a national sample of
Mexican–American women and men
Jan Sundquista,b and Marilyn Winklebya
Background Few studies have examined the influence of country of birth and acculturation
status on indicators of obesity using national samples of Mexican–American women
and men.
Methods
We analysed data for 1387 Mexican–American women and 1404 Mexican–
American men, ages 25–64, from the third National Health and Nutrition Examination Survey (1988–1994). We examined whether waist circumference and
abdominal obesity varied by country of birth and acculturation status (primary
language spoken), and whether among those with abdominal obesity, number of
associated cardiovascular disease (CVD) risk factors varied by country of birth and
acculturation status.
Results
Both country of birth and, to a lesser degree, acculturation status were significantly associated with waist circumference and abdominal obesity. Mexican-born
women and men had the smallest waist circumference (90.4 cm, 94.0 cm
respectively), US-born English-speaking women and men had intermediate waist
circumference (93.6 cm, 97.3 cm), and US-born Spanish-speaking women and
men had the largest waist circumference (96.9 cm, 97.7 cm), after accounting for
age, education, per cent of energy from dietary fat, leisure-time physical activity,
and smoking. All women had high prevalences of abdominal obesity, particularly
US-born Spanish-speaking women (68.7%). In addition, US-born Spanish-speaking
women with abdominal obesity were significantly more likely than their counterparts to have one or more of the following CVD risk factors: high serum insulin,
non-insulin dependent diabetes, high blood lipids, and/or hypertension.
Conclusions These findings illustrate the heterogeneity of the Mexican–American population
and suggest that country of birth and lack of acculturation to the majority
culture, as well as secondary lifestyle changes, may explain the significant clinical
differences observed in abdominal obesity within Mexican–American population
subgroups.
Keywords
Mexican–American, migration, acculturation, waist circumference, abdominal
obesity
Accepted
15 October 1999
Obesity has reached epidemic proportions in industrialized
societies and is becoming increasingly prevalent in the United
States. Data from the National Health and Nutrition Examination Survey (NHANES) indicates that the prevalence of obesity
a Stanford Center for Research in Disease Prevention, Stanford University
School of Medicine, USA.
b Karolinska Institutet, Family Medicine Stockholm, Novum, 14157
Huddinge, Sweden.
Reprint requests to: Professor Jan Sundquist, Karolinska Institutet, Family
Medicine Stockholm, Novum, 14157 Huddinge, Sweden. E-mail:
[email protected]
(body mass index [BMI] >30.0) increased from 14.5% in 1976–
1980 to 22.5% in 1988–1994.1 This increase was observed in
both women and men, and in black, Mexican–American, and
white ethnic groups.1 High prevalence of obesity, as well as
increasing prevalence, have been documented by other studies
conducted in the US and internationally.2–5 These trends are important given the association between obesity and adverse
health outcomes, including all-cause and cardiovascular disease
(CVD) mortality.6,7 Furthermore, there is evidence that abdominal obesity is associated with CVD risk factors, as indicated by
high serum insulin, non-insulin dependent diabetes (NIDDM),
high blood lipids, and hypertension.8–10
470
ABDOMINAL OBESITY IN MEXICAN–AMERICAN ADULTS
There are large differences in the prevalence of obesity across
ethnic groups,1,11 with the problem being particularly important in the Mexican–American population,2–17 which it is increasing rapidly. Between 1980 and 1990 the Hispanic population, of
whom 63% are Mexican–American, increased by 45% compared with 7% for the overall population, with projected
increases in population size from 31.4 million people in 2000 to
96.5 million people in 2050.18,19
There are also differences in the prevalence of obesity within
the Mexican–American population, indicating a heterogeneity
that is important to examine further. For example, the San
Antonio Heart Study showed that Mexican–American women
living in more affluent suburbs had a mean BMI of 24.5 while
those living in the most socioeconomically disadvantaged
neighbourhoods had a mean BMI of 28.3.13 Corresponding
BMI levels for Mexican–American men in these areas were 26.9
and 27.8, respectively.
There are additional variations in obesity among populations
that have migrated to new countries, suggesting that factors
associated with the migration process (e.g. change in diet, exercise, and stress) may adversely influence body weight. Prospective studies have shown significant BMI increases following
immigration to a new country. A prospective study of 654 adult
Tokelauans who migrated to New Zealand because of a natural
disaster showed that their BMI levels increased from 24.1 to
28.7 between 1968 and 1982 compared with non-migrants
whose BMI levels increased from 24.8 to 26.1.20 Moreover, an
8-year prospective study from Sweden showed a significant net
increase in BMI levels among men who immigrated from South
Europe to Sweden compared to a reference group of men who
had not immigrated (BMI levels for immigrant men increased
from 25.5 to 27.0 compared to 24.6 to 25.0 for non-immigrant
men, after adjustment for age, leisure-time physical activity,
smoking, education, and health status).5
In addition to the possible adverse effects of migration on
body weight, several studies have suggested that the degree to
which people acculturate to the majority culture following
migration may influence weight. A cross-sectional analysis of
Mexican–American women and men from the Hispanic Health
and Nutrition Examination Survey (HHANES) showed that
preference for speaking Spanish (indicating a lower level of
acculturation) was associated with a significantly higher BMI in
women but not in men.21 Hazuda et al.22,23 examined the
multidimensional aspects of acculturation and its association
with body weight by constructing the following three acculturation scales: functional integration with mainstream American
culture, worth placed on conserving Mexican culture, and
attitudes toward traditional family structure. The scales showed
different associations with body weight, with the first scale (low
functional integration) showing the strongest association, but
only for women and not for men.23 In contrast, the well-known
study on acculturation and weight, which involved a cohort of
4653 men of Japanese ancestry living in Hawaii, found that
men who were most traditional had significantly lower levels of
BMI (mean of 23) than those who were more acculturated
(mean of 26).24
In this analysis we used data from the Third National Health
and Nutrition Examination Survey (NHANES III), a large
national sample of women and men living in the US. We
examined whether abdominal obesity differed among three
471
groups of Mexican–American women and men who reflect
different stages of migration and acculturation: those born in
Mexico, those born in the US whose primary language was
Spanish, and those born in the US whose primary language was
English. We also examined whether among those with
abdominal obesity, risk of associated CVD risk factors (indicated
by high serum insulin, NIDDM, high blood lipids, and hypertension)25–27 varied according to the three migration/
acculturation groups.
We chose waist circumference as our measure of abdominal
obesity for the following reasons. First, waist circumference is
more strongly related to the metabolically active visceral adipose
tissue and therefore more biologically interpretable than the
more widely used measure of abdominal obesity, the waist to
hip ratio (WHR).28,29 Second, several studies have concluded
that waist circumference alone is more closely associated with
the amount of abdominal, visceral obesity than the WHR, and
recommend that waist circumference should be used to estimate risk of CVD28,30 and NIDDM.31 Third, a 7-year prospective study revealed that waist circumference was the best
predictor of NIDDM compared to BMI, WHR, triceps and subscapular skinfolds in Mexican–American women and men.32
Methods
The NHANES III study, conducted between 1988 and 1994 by
the National Center for Health Statistics, was designed to collect
information to assess the health status of the US civilian noninstitutionalized population aged >2 months.33 The NHANES
III sample design is similar to that of the previous NHANES,
which used a stratified multistage probability design.34,35 It
included an oversampling of both the Mexican–American and
black American populations so that the sample could produce
statistically reliable health estimates for the two largest ethnic
minority groups in the US.
The NHANES III data were collected via standardized questionnaires administered by bilingual interviewers and examiners at participants’ homes, and laboratory tests conducted at
NHANES mobile examination centres. Of the 40 600 people
invited to participate, 86% completed the home questionnaire,
and 78% completed both the medical examination and the
home questionnaire (n = 33 994).
The sample for our analyses includes 1387 Mexican–
American women and 1404 Mexican–American men, 25–64
years of age, who completed both the home questionnaire and
medical examination. We used age 25 as our lower age cutpoint because educational attainment (a covariate in our multivariate regression models) is often not completed before this
age. We used age 64 as our upper age cut-point because of lifestyle changes associated with retirement and ageing after 65.36
We excluded data for pregnant women (n = 69) and surveys
coded as unreliable (n = 3). Missing data were as follows for the
outcome variables; waist (4.0%), BMI (0.4%), serum insulin
(3.6%), NIDDM (0.2%), high-density lipoprotein cholesterol
(4.3%), serum triglycerides (3.7%), and hypertension (4.4%).
Definition of variables
Migration and acculturation status were indicated by country
of birth and primary language spoken at home. Race-ethnic
groups were based upon combinations of the race and ethnicity
472
INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
reported by the participants. Respondents who chose Mexican
or Mexican–American ethnicity are included in this analysis.
Waist circumference was measured with the person in the
erect position, with an inelastic tape at the level of the natural
waist (the most narrow part of the torso).37 Abdominal obesity
was defined as a waist circumference of >102 cm for men and
>88 cm for women, these cut-points representing a substantially increased risk of future adverse health outcomes and CVD
risk factors.10,38 As a measure of general obesity, a risk factor
associated with CVD, we used BMI.
The CVD risk factors were defined as follows. Non-insulin
dependent diabetes mellitus was defined as having an >8-hour
fast and plasma glucose levels >126 mg/dl and/or a medical
history of diabetes (other than during pregnancy) with an age
of onset .25 years.39 Non-high-density lipoprotein cholesterol
(non-HDL-C) was calculated as the difference between total cholesterol and high-density lipoprotein cholesterol. Non-HDL-C,
which may be a better indicator of atherogenic lipoprotein particles than indirectly estimated low-density lipoprotein cholesterol,40 does not require fasting blood samples and therefore
allowed the use of the entire NHANES III sample. Non-HDL-C
(mmol/l), insulin (mU/ml), and triglycerides (mmol/l) were
measured from serum specimens. All blood specimens were
obtained by a venipuncture. High insulin levels were defined as
>mean + 2 standard deviations (SD) for the entire NHANES III
sample. High non-HDL-C cholesterol and high triglyceride
levels were defined as >4.0 mmol/l and >2.26 mmol/l, respectively.41 Blood pressure was measured on the right arm by
health examiners or physicians while the participant was seated
during the medical examination; the mean of the second and
third of three blood pressure readings was used. Hypertension
was defined as systolic blood pressure >140 mmHg and/or
diastolic blood pressure >90 mmHg and/or current use of
antihypertensive medications.42
Data analysis
Primary analyses, using linear models, were carried out in
SUDAAN, to adjust for the complex sample design of NHANES
III.43 All analyses incorporated sampling weights that adjusted
for unequal probabilities of selection. Sample weights were also
adjusted for non-respondents’ characteristics. Analyses were
run separately for women and men. The primary outcome variable was waist circumference; the predictor variables were
country of birth and acculturation status (three groups—born in
Mexico and used as the reference group, born in the US and
Spanish-speaking, born in the US and English-speaking). The
following potential confounding factors were included as
covariates in the linear models: age (in years, centred at the
sample mean to aid in the interpretation of the regression
coefficients), years of education (continuous and centred), per
cent of energy from dietary fat (based on a single 24-hour
dietary recall and administered by trained dietary interviewers);
no leisure-time physical activity (based on whether individuals
had engaged in any leisure time physical activity in the past
month, including exercises, sports, or physically active hobbies—
those who reported no leisure-time activities were considered
physically inactive); and current cigarette smoker (determined by
self-report and based on whether individuals had smoked ù100
cigarettes during their lifetime and whether they were current
smokers). We validated self-reported smoking by examining a
biochemical measure, serum cotinine, and found that underreporting did not differ significantly between the three groups
for either women or men. We included all first order interactions between predictor variables in the linear models.
Results
Individuals born in the US who were Spanish-speaking were
older, less likely to live in urban areas, and more likely to live
in the South (predominantly Texas), than their counterparts
(Table 1). Mexican–Americans born in the US who were Englishspeaking had substantially higher levels of education, lower
levels of poverty, and higher employment rates than the other
two groups. Although all groups reported moderately high
percentages of energy from total fat, those born in Mexico
reported the lowest percentages. Women born in Mexico
reported the lowest levels of leisure-time activity and were the
least likely to be current cigarette smokers.
Both country of birth and, to a lesser degree, acculturation
status (indicated by primary language spoken) were significantly associated with waist circumference in the linear model
(Table 2). After adjustment for age, education, per cent of energy
from dietary fat, leisure-time physical activity, and smoking,
waist circumference was significantly larger in US-born women
and men who spoke English or Spanish than Mexican-born
women and men, indicating an influence of country of birth.
There was also a significant difference in waist circumference
between Spanish- and English-speaking US-born women
(P-value = 0.04), indicating an influence of acculturation
(evident when examining further contrasts in the linear model).
There was no significant difference between Spanish- and
English-speaking US-born men.
The magnitude of the differences in waist circumference by
country of birth was large; for example the beta coefficient of
6.5 for the comparison of US-born Spanish-speaking women to
Mexican-born women indicates that the former group had, on
average, a waist circumference that was 6.5 cm larger than the
Mexican-born women having comparable age, education, smoking, leisure-time physical activity and energy from fat. There
were two modestly significant interactions for men, but when
examined in more depth, neither were clinically meaningful.
Table 3 presents the unadjusted and adjusted means for waist
circumference and BMI by country of birth and acculturation
status. After adjustment for age, education, dietary fat, leisuretime physical activity, and smoking, Mexican-born women and
men had the smallest waist circumference (90.4 cm, 94.0 cm
respectively), US-born English-speaking women and men had
intermediate waist circumference (93.6 cm, 97.3 cm), and USborn Spanish-speaking women and men had the largest waist
circumference (96.9 cm, 97.7 cm). The corresponding adjusted
levels for white women and men (presented for reference only)
were 87.9 cm for women and 96.8 cm for men.
Using the recommended cut-points for abdominal obesity
(>88 cm waist circumference for women and >102 cm for
men)10,38 we found a gradient effect (Figure 1): US-born
Spanish-speaking women and men had the highest levels of
abdominal obesity (68.7% and 39.5% obese, respectively),
US-born English-speaking women and men had intermediate
levels (58.6% and 31.8% obese), and Mexican-born women
and men had the lowest levels (55.6% and 21.4% obese). The
ABDOMINAL OBESITY IN MEXICAN–AMERICAN ADULTS
473
Table 1 Sample sizes, weighted sociodemographic and lifestyle characteristics by country of birth and acculturation status, Mexican–American
women and men, ages 25–64, NHANES III, 1988–1994a
Women
Born in
Mexico
Estimated US populationb
Men
Born in US
Born in US
English-speaking
Spanish-speaking
Born in
Mexico
1 300 000
1 057 300
416 000
1 600 300
1 138 300
342 000
626
502
259
696
475
233
38.6
39.6
44.4
36.5
38.4
45.6
5.1
3.7
3.8
5.0
3.5
4.1
34.4
100
100
37.3
100
100
Sample sizec
English-speaking Spanish-speaking
Sociodemographic factors
Age (mean)
Family size (mean number of persons)
Portion of life lived in the US (%)
Speaking English at home (%)
Living in an urban area (%)
Years of education (mean)
7.2
100
0
7.2
100
0
67.7
57.2
35.9
69.4
59.6
35.9
7.0
11.9
8.9
7.3
12.2
8.7
Living below the poverty level (%)
57.8
24.4
45.8
48.0
17.1
44.6
Employed (%)
44.4
70.1
48.9
86.1
89.0
66.8
Region of residence
% West (predominantly California,
Arizona, and New Mexico)
64.2
49.2
27.8
61.5
53.5
27.0
% South (predominantly Texas)
27.2
40.8
67.2
27.1
37.7
66.1
8.6
10.0
5.0
11.4
8.8
6.9
Calories from total fat (%)
31.0
35.0
34.3
30.9
33.4
33.4
No leisure-time physical activity (%)
56.3
27.1
36.0
37.1
16.8
33.5
Current cigarette smoker (%)
11.7
18.2
20.4
33.0
27.3
33.3
% Northeast/Midwest
Lifestyle factors
a Source: National Center for Health Statistics, NHANES III, 1988–1994. Means and percentages were calculated with normalized sample weights.
b Projected estimates based on weighted percentages from NHANES III for the defined sample.
c Number who participated in both the home questionnaire and medical examination, unweighted.
Table 2 Regression model coefficients and P-values for the predictor variables and waist circumference, Mexican–American women and men,
ages 25–64, NHANES III, 1988–1994a
Waist circumference (cm)
Women
Men
b
±SE
P
b
±SE
P
Born in US, English-speakingb
3.0
Born in US, Spanish-speaking
6.5
1.3
0.02
4.7
1.0
,0.001
1.4
,0.001
4.7
1.2
0.4
,0.001
0.08
,0.001
0.4
0.05
,0.001
Main effects
Country of birth/acculturation status
Born in Mexico
Age
Education
–0.5
0.1
,0.001
0.03
0.09
0.78
Energy from dietary fat
0.03
0.05
0.62
0.09
0.04
0.03
3.2
1.0
,0.01
0.2
1.08
0.84
–1.1
1.2
0.35
–1.5
0.7
0.04
Age ´ Spanish-speaking
–0.2
0.1
0.02
Education ´ English-speaking
–0.6
0.2
0.03
No leisure-time physical activity
Current cigarette smoker
Interactions
a Source: National Center for Health Statistics, NHANES III, 1988–1994. Linear models incorporated sampling weights. Predictor variables were country of
birth/acculturation status, age, education, percent of energy from dietary fat, leisure-time physical activity, smoking, and all first-order interactions between
predictor variables. Regression coefficients are unstandardized.
b Further contrasts in the linear model showed a significant difference (P
English-speaking men.
, 0.04) for Spanish- versus English-speaking women, but not for Spanish- versus
474
INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
Table 3 Waist circumference and body mass index (BMI) by country of birth and acculturation status, Mexican–American women and men, ages
25–64, NHANES III, 1988–1994a
Born in US,
English-speaking
Born in Mexico
Born in US,
Spanish-speaking
White, Non-Hispanicc
Unadjusted
Adjustedb
SD
Unadjusted
Adjusted
SD
Unadjusted
Adjusted
SD
Waist (cm)
91.1
90.4
13.1
91.9
93.6
15.0
96.9
96.9
14.9
87.9
87.9
15.3
BMI (kg/m2)
28.1
27.8
5.5
28.5
29.1
6.8
30.2
29.7
6.6
26.3
26.3
6.4
Waist (cm)
93.3
94.0
9.9
97.5
97.3
13.0
100.6
97.7
13.2
96.7
96.8
12.7
BMI (kg/m2)
26.7
26.9
4.0
28.2
28.2
5.0
28.7
28.1
5.3
27.0
27.0
5.0
Unadjusted Adjusted
SD
Women
Men
a Source: National Center for Health Statistics, NHANES III, 1988–1994. Means and percentages were calculated with normalized sample weights.
b Adjusted for age, education, per cent of energy from dietary fat, leisure-time physical activity, smoking, and all first order interactions.
c Data for white, non-Hispanic women and men are presented for reference.
more likely to have one or more CVD risk factors than their
counterparts. Among men, there were no significant differences.
Discussion
Figure 1 Prevalence of abdominal obesity by country of birth and
acculturation status, adjusted for age and years of education. Waist
circumference >88 cm for women and >102 cm for men with 95% CI
corresponding percentages for white women and men were
45.4% and 31.4%, respectively. It is important to note the substantially higher levels of abdominal obesity in women than
men for all subgroups.
High insulin levels, NIDDM, elevated cholesterol and triglycerides, and hypertension are among the main risk factors associated with CVD.44 In Table 4 we present the percentages with
high serum insulin (>mean + 2 SD), NIDDM, high non-HDL
cholesterol (>4.0 mmol/l), high triglycerides (>2.26 mmol/l),
and hypertension among those with abdominal obesity.10,38
Although there were some inconsistencies among the individual risk factors, women born in the US who were Spanishspeaking and who had abdominal obesity were significantly
In this analysis of a large national sample of Mexican–American
women and men living in the US, we found important and significant clinical differences in waist circumference and abdominal obesity among three population groups, suggesting that
country of birth (for both women and men) and acculturation
status (for women only) are associated with the pathogenesis of
obesity. We found that Mexican-born women and men had the
smallest waist circumference, English-speaking US-born women
and men had intermediate waist circumference, and Spanishspeaking US-born women and men had the largest waist circumference, when age, education, per cent of energy from dietary
fat, leisure-time physical activity, and smoking were taken into
account. The prevalence of abdominal obesity paralleled these
findings, and showed that 68.7% of Spanish-speaking US-born
women and 39.5% of Spanish-speaking US-born men had
abdominal obesity, indicating substantial risk of CVD.
It is well-known that obesity is strongly associated with
behavioural, cultural, and societal factors.5,45–48 This association has been established, in part, by migration studies that
have shown increasing levels of obesity among populations that
migrate, as well as those that have low levels of acculturation to
the majority culture.5,20,49 Mechanisms proposed as possible
explanations are lifestyle changes,5,50 dietary changes,51,52 lack
of educational or occupational opportunities,47,48,53 and/or other
structural inequalities.50
Our findings of an association between country of birth and
abdominal obesity (waist circumference) are consistent with
findings from the Hispanic HANES that showed that BMI was
larger for second and third generation Mexican–Americans than
for first generation Mexican–Americans.21 In addition, Hispanic
HANES found that Mexican–American women with a preference for the English language had significantly lower levels of
BMI than those with a preference for the Spanish language, a
finding which agrees with the present study. The San Antonio
Heart Study also demonstrated that Mexican–American women
with lower acculturation levels (low functional integration) had
higher levels of central obesity (subscapular-to-triceps skinfoldthickness ratio) than women with higher acculturation levels.23
ABDOMINAL OBESITY IN MEXICAN–AMERICAN ADULTS
475
Table 4 Cardiovascular disease risk (CVD) factors among Mexican–American women and men with abdominal obesity, ages 25–64, NHANES III,
1988–1994a
Womenb
Menb
Born in
Mexico
Born
in US,
Englishspeaking
Born
in US,
Spanishspeaking
White,
NonHispanicc
Born in
Mexico
Born
in US,
Englishspeaking
Sample size with abdominal obesity
348
294
178
919
149
% with high serum insulin
5.2
4.1
4.5
4.1
7.4
13.1
12.3
15.1
9.5
% with high non-high-density
lipoprotein cholesterol
4.0
5.1
6.7
% with high triglycerides
5.5
6.1
6.7
% with hypertension
17.8*
18.7*
% with >1 of the above CVD risk factors
30.8*
31.0*
% with non-insulin-dependent diabetes
Born
in US,
Spanishspeaking
White,
NonHispanicc
151
92
546
9.3
10.9
5.3
13.0
20.9
18.4
15.5
5.1
5.4
4.0
0.0
1.6
5.7
8.1
6.6
4.4
6.8
35.4
32.1
27.5*
39.7
44.6
40.1
46.6
39.2
45.6
55.0
54.3
48.4
a Source: National Center for Health Statistics, NHANES III, 1988–1994. Abdominal obesity defined as >88 cm for women, and >102 cm for men.
b If there was a significant difference among the three Mexican–American groups, then pairwise x2 tests were conducted to identify which groups were
different. An asterisk indicates a significant (P , 0.01) difference for the indicated group compared to the group with the highest value.
c Data for white, non-Hispanic women and men are presented for reference and not included in the pairwise x2 tests.
Underlying mechanisms
It is possible that the biological pathway between migration,
acculturation, and abdominal obesity may be mediated by a
long-term stress reaction.54–56 Stress can result in either an
acute fight or flight reaction, the first leading to control and the
second leading to a long-term defeat reaction, characterized by
depression or helplessness.57,58 Abdominal obesity has been
shown to be associated with the long-term defeat reaction, in
which the hypothalamic-pituitary-adrenal axis is activated with
an increase of cortisol levels, which changes fat metabolism and
can result in a central distribution of fat.26,59–61 Migration,
whether forced by poverty or persecution, leading to settlement
in a different social, political and cultural context, might result
in a long-term defeat reaction regardless of previous health. In
addition to migration stress, not speaking the language of the
majority culture may indicate a separation from the majority
culture and create long-term acculturative stress.
It is unlikely that the differences in abdominal obesity that we
found between the three migration/acculturation groups can be
explained solely by genetic differences. Although it is estimated
that genetic factors may explain from 30% to 70% of cases of
obesity in the US, the rapid increase in the prevalence of obesity
over the last two decades is likely due to environmental factors
since the gene pool has been stable over the same time interval.62 Furthermore, in a population of Mexican–American
adults from the San Antonio Heart Study, it was estimated that
genes accounted for only 15% to 30% of the phenotypic variation in measures of obesity.63 Although genes may play an
important role in the metabolism of adipose tissue, the large
differences in weight-related outcomes that we observed within
a single ethnic group of Mexican–Americans suggests that the
current epidemic of obesity is largely environmental, either
directly or via an interaction with genetic influences.
Our findings that a high proportion of women and men have
abdominal obesity suggest that many Mexican–Americans may
be at risk for additional CVD risk factors and metabolic complications.64 Björntorp26 suggested that the primary factors in the
‘syndrome of visceral obesity’ or the metabolic syndrome, are
overeating, physical inactivity, smoking, and stress. These may
act via neuroendocrine pathways to induce insulin resistance,
increased serum insulin and abdominal obesity, which can lead
to diseases such as diabetes. The Metabolic Syndrome could be
called a ‘Civilization Syndrome’ because of its strong relationship with an urban sedentary lifestyle and a positive energy
balance (overeating and lack of leisure-time physical activity).10
We suggest that the severe stress that can be associated with
lack of acculturation to the majority culture and secondary lifestyle changes may induce an increase in weight, consistent with
the findings in the present study.
Strengths and limitations
The NHANES III study is one of the most comprehensive
national surveys to date. Extensive and complete data are available from both the home survey and medical examination, including standardized measures of waist circumference. As noted
previously, response rates were high and there was minimal
missing data. Unlike many surveys, NHANES III represents a
sample of the US population and therefore results are generalizable to the population of Mexican–Americans living in the US. It
also included an oversampling of Mexican–Americans to ensure
that large numbers would be available for analyses. In addition,
NHANES III assessed BMI as well as waist circumference, which
may be a stronger predictor of obesity complications than BMI.65
Despite these strengths, there are several limitations to the
NHANES III survey. First, it is based on a cross-sectional survey
design that does not allow one to draw inferences about causal
pathways. Second, there were limitations to several of the
variables we used, including our outcome measure, waist circumference. Although exact methods exist for measuring the
visceral adipose tissue accumulation, such as magnetic resonance imaging and computerized tomography, NHANES III did
not use these more exact methods. However, studies of
anthropometric correlates have found that waist circumference
is the preferred clinical and epidemiological method to use
because of its association with visceral adiposity, low cost and
ease of measurement.10,28,38
476
INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
In addition, NHANES III lacks measures of physical activity at
work, which may influence the gender differences that we
found for leisure-time physical inactivity (significant for women
but not for men, Table 2). Finally, our measure of acculturation,
indicated by primary language spoken at home, is a static factor
that does not measure core beliefs and practices in relation to
specific CVD risk factors or medical conditions, and does not
consider the social context of behaviour.66,67
In summary, we found substantial variation in waist circumference and abdominal obesity among three subgroups of
Mexican–Americans living in the US, distinguished by country
of birth and acculturation status. All women, but in particular
US-born Spanish-speaking women (68.7%), had high prevalence of abdominal obesity. The variation we found in abdominal
obesity illustrates the heterogeneity of the Mexican–American
population and suggests that both country of birth and
acculturation status influence indicators of obesity. To address
these variations, social and economic interventions are needed
as well as lifestyle and public health policy interventions, such
as those that influence food consumption patterns and stimulate physical activity.66,67 As with all interventions, there is a
need for primary prevention efforts that begin early in life that
are linguistically and culturally appropriate.
7 Lindsted KD, Singh PN. Body mass and 26-year risk of mortality
among women who never smoked: findings from the Adventist
Mortality Study. Am J Epidemiol 1997;146:1–11.
8 Pi-Sunyer
FX. Medical hazards of obesity. Ann Intern Med
1993;119:655–60.
9 Björntorp P. Abdominal obesity and the metabolic syndrome. Ann Med
1992;24:465–68.
10 Björntorp P. Obesity. Lancet 1997;350:423–26.
11 Deurenberg P, Yap M, van Staveren WA. Body mass index and percent
body fat: a meta analysis among different ethnic groups. Int J Obes
Relat Metab Disord 1998;22:1164–71.
12 Winkleby MA, Fortmann SP, Rockhill B. Health-related risk factors in
a sample of Hispanics and whites matched on sociodemographic
characteristics: the Stanford Five-City Project. Am J Epidemiol
1993;137:1365–75.
13 Stern MP, Rosenthal M, Haffner SM, Hazuda HP, Franco LJ. Sex
difference in the effects of sociocultural status on diabetes and
cardiovascular risk factors in Mexican Americans: the San Antonio
Heart Study. Am J Epidemiol 1984;120:834–51.
14 Haffner SM, Stern MP, Hazuda HP, Pugh JA, Patterson JK, Malina R.
Upper body and centralized adiposity in Mexican–Americans and
non-Hispanic whites: relationship to body mass index and other
behavioral and demographic variables. Int J Obesity 1986;10:493–502.
15 Winkleby MA, Gardner CD, Taylor CB. The influence of gender and
socioeconomic factors on Hispanic/white differences in body mass
index. Prev Med 1996;25:203–11.
Acknowledgements
This research was supported by an Established Investigatorship
Award and National Grant-in-Aid from the American Heart
Association and a US Public Health Service Grant from the
National Heart, Lung, and Blood Institute to Dr Winkleby, and
by grants from the Swedish Medical Research Council
(K97–21P-11333–01A), Swedish Heart Lung Foundation,
Swedish Institute, Swedish Society of Medicine, and from the
Henning and Johan Throne Holst’s Foundation for the Promotion of Scientific Research to Dr Sundquist. The authors
thank Drs David Ahn, Lori Beth Dixon, Christopher Gardner,
Michaela Kiernan, Helena C Kraemer, and Ms Pam Rief for
their insightful comments on an earlier draft, and Ms Alana
Koehler for her technical assistance.
References
1 Flegal KM, Carroll MD, Kuczmarski RJ, Johnson CL. Overweight and
obesity in the United States: prevalence and trends, 1960–1994. Int J
Obes Relat Metab Disord 1998;22:39–47.
2 Kuczmarski RJ, Flegal KM, Cambell SM, Johnson CL. Increasing
16 Winkleby MA, Kraemer HC, Ahn DK, Varady AN. Ethnic and
socioeconomic differences in cardiovascular disease risk factors:
findings for women from the Third National Health and Nutrition
Examination Survey, 1988–1994. JAMA 1998;280:356–62.
17 Winkleby MA, Robinson TN, Sundquist J, Kraemer HC. Ethnic
variation in cardiovascular risk factors among children and young
adults: findings from the Third National Health and Nutrition
Examination, 1988–1994. JAMA 1999;281:1006–13.
18 Day JC. Population Projections of the United States by Age, Sex, Race, and
Hispanic Origin. US Bureau of the Census, Current Population Reports.
Washington, DC: US Government Printing Office, 1996, Rep. No.
P25–1130.
19 Schick FL, Schick R. Statistical Handbook on US Hispanics. Phoenix, AZ:
Oryx Press, 1991.
20 Salmond CE, Prior IAM, Wessen AF. Blood pressure patterns and
migration: a 14-year cohort study of adult Tokelauans. Am J Epidemiol
1989;130:37–52.
21 Khan LK, Sobal J, Martorell R. Acculturation, socioeconomic status,
and obesity in Mexican Americans, Cuban Americans, and Puerto
Ricans. Int J Obesity 1997;21:91–96.
22 Hazuda HP, Haffner SM, Stern MP, Eifler CW. Effects of acculturation and
socioeconomic status on obesity and diabetes in Mexican Americans:
The San Antonio Heart Study. Am J Epidemiol 1988;128:1289–301.
prevalence of overweight among US adults: the National Health and
Nutrition Examination Surveys, 1960 to 1991. JAMA 1994;272:
205–11.
23 Hazuda HP, Mitchell BD, Haffner SM, Stern MP. Obesity in Mexican–
3 Prentice AM, Jebb SA. Obesity in Britain: gluttony or sloth? Br Med J
24 Reed D, McGee D, Cohen J, Yano K, Syme SL, Feinleib M.
1995;311:437–39.
4 Seidell JC, Verschuren WM, Kromhout D. Prevalence and trends of
obesity in The Netherlands 1987–1991. Int J Obes Relat Metab Disord
1995;19:924–27.
5 Sundquist J, Johansson S-E. The influence of socioeconomic status,
ethnicity and lifestyle on body mass index in a longitudinal study. Int
J Epidemiol 1998;27:57–63.
6 Stevens J, Cai J, Pamuk ER, Williamson DF, Thun MJ, Wood JL. The
effect of age on the association between body-mass index and
mortality. N Engl J Med 1998;338:1–7.
American subgroups: findings from the San Antonio Heart Study.
Am J Clin Nutr 1991;53:1529S–34S.
Acculturation and coronary heart disease among Japanese men in
Hawaii. Am J Epidemiol 1982;115:894–905.
25 Lundgren H, Bengtsson C, Blohme G, Lapidus L, Sjostrom L.
Adiposity and adipose tissue distribution in relation to incidence of
diabetes in women: results from a prospective population study in
Gothenburg, Sweden. Int J Obes 1989;13:413–23.
26 Björntorp P. Visceral obesity: a ‘Civilization Syndrome’. Obesity Res
1993;1:206–21.
27 Björntorp P. Regional obesity and NIDDM. Adv Exp Med Biol
1993;334:279–85.
ABDOMINAL OBESITY IN MEXICAN–AMERICAN ADULTS
477
28 Pouliot MC, Despres JP, Lemieux S et al. Waist circumference and
47 Winkleby MA. Accelerating cardiovascular risk factor change in
abdominal sagittal diameter: best simple anthropometric indexes of
abdominal visceral adipose tissue accumulation and related
cardiovascular risk in men and women. Am J Cardiol 1994;73:460–68.
ethnic minority and low socioeconomic groups. Ann Epidemiol
1997;7:S96–103.
29 Taylor RW, Keil D, Gold EJ, Williams SM, Goulding A. Body mass
index, waist girth, and waist-to-hip ratio as indexes of total and
regional adiposity in women: evaluation using receiver operating
characteristic curves. Am J Clin Nutr 1998;67:44–49.
30 Han TS, van Leer EM, Seidell JC, Lean MEJ. Waist circumference
action levels in the identification of cardiovascular risk factors:
prevalence study in a random sample. Br Med J 1995;311:1041–45.
31 Seidell JC, Han TS, Feskens EJ, Lean ME. Narrow hips and broad
waist circumferences independently contribute to increased risk of
non-insulin-dependent diabetes mellitus. J Intern Med 1997;242:
401–06.
32 Wei M, Gaskill SP, Haffner SM, Stern MP. Waist circumference as the
best predictor of noninsulin dependent diabetes mellitus (NIDDM)
compared to body mass index, waist/hip ratio and other
anthropometric measurements in Mexican Americans—a 7-year
prospective study. Obes Res 1997;5:16–23.
33 National Center for Health Statistics. Plan and operation of the Third
National Health and Nutrition Examination Survey, 1988–1994, series
1: programs and collection procedures. Vital Health Stat
1994;32:1–407.
34 National Center for Health Statistics. Plan and operation of the Health
and Nutrition Examination Survey, United States, 1971–1973. Vital
Health Stat 1977;1:10b.
35 McDowell A, Engel A, Massey JT, Maurer K. National Center for
Health Statistics: plan and operation of the second National Health
and Nutrition Examination Survey. 1976–1980. Vital Health Stat
1981;1:15.
36 House JS, Kessler RC, Herzog AR. Age, socioeconomic status, and
health. Milbank Q 1990;68:383–411.
37 Callaway CW, Chumlea WC, Bouchard C et al. Circumferences. In:
Lohman TG, Roche AF, Martorell R (eds). Anthropometric Standardization Reference Manual. Champaign, IL: Human Kinetics Books, 1988.
38 Lean MEJ, Han TS, Morrison CE. Waist circumference indicates the
need for weight management. Br Med J 1995;311:158–61.
48 Sundquist J, Winkleby MA. Cardiovascular risk factors in Mexican
American adults: a transcultural analysis of NHANES III, 1988–1994.
Am J Public Health 1999;89:723–30.
49 Tomson Y, Aberg H. Risk factors for cardiovascular disease—a com-
parison between Swedes and immigrants. Scand J Prim Health Care
1994;12:147–54.
50 Williams R, Bhopal R, Hunt K. Coronary risk in a British Punjabi
population: comparative profile of non-biochemical factors. Int J
Epidemiol 1994;23:28–37.
51 Cardoso MA, Hamada GS, de Souza JM, Tsugane S, Tokudome S.
Dietary patterns in Japanese migrants to southeastern Brazil and their
descendants. J Epidemiol 1997;7:198–204.
52 Hjern A, Kocturk-Runefors T, Jeppson O. Food habits and infant
feeding in newly resettled refugee families from Chile and the Middle
East. Scand J Prim Health Care 1990;8:145–50.
53 Sundquist J. Living conditions and health. A population-based study
of labour migrants and Latin American refugees in Sweden and those
who were repatriated. Scand J Prim Health Care 1995;13:128–34.
54 Sundquist J, Eng M, Johansson S-E. Body mass index and distribution
of body fat in female Bosnian refugees. A study in primary health
care. Public Health 1999;113:89–94.
55 Rosmond R, Lapidus L, Björntorp P. The influence of occupational and
social factors on obesity and body fat distribution in middle-aged men.
Int J Obes Relat Metab Disord 1996;20:599–607.
56 Rosmond R, Björntorp P. Psychiatric ill-health of women and its
relationship to obesity and body fat distribution. Obesity Res
1998;6:338–45.
57 Frankenhaeuser M, von Wright MR, Collins A, von Wright J, Sedvall G,
Swahn CG. Sex differences in psychoneuroendocrine reactions to
examination stress. Psychosom Med 1978;40:334–43.
58 Henry JP, Grim CE. Psychosocial mechanisms of primary hyper-
tension. J Hypertens 1990;8:783–93.
59 Laatikainen TJ. Corticotropin-releasing hormone and opioid peptides
in reproduction and stress. Ann Med 1991;23:489–96.
39 Expert Committee on the Diagnosis and Classification of Diabetes
60 Björntorp P. The regulation of adipose tissue distribution in humans.
Mellitus. Report of the expert committee on the diagnosis and
classification of diabetes mellitus. Diabetes Care 1997;20:1183–97.
61 Rosmond R, Dallman MF, Björntorp P. Stress-related cortisol secretion
40 Havel RJ, Rapaport E. Management of primary hyperlipidemia. N Engl
J Med 1995;332:1491–98.
41 The Long-term Intervention with Pravastatin in Ischaemic Disease
(LIPID) Study Group. Prevention of cardiovascular events and death
with pravastatin in patients with coronary heart disease and a broad
range of initial cholesterol levels. N Engl J Med 1998;339:1349–57.
Int J Obes Relat Metab Disord 1996;20:291–302.
in men: relationships with abdominal obesity and endocrine, metabolic and hemodynamic abnormalities. J Clin Endocrinol Metab
1998;83:1853–59.
62 Eckel RH. Obesity and heart disease: a statement for healthcare pro-
fessionals from the Nutrition Committee, American Heart Association.
Circulation 1997;96:3248–50.
42 The Joint National Committee on Detection, Evaluation and
63 Mitchell BD, Kammerer CM, Blangero J et al. Genetic and environ-
Treatment of High Blood Pressure. The fifth report of the Joint
National Committee on Detection, Evaluation and Treatment of High
Blood Pressure (JNC V). Arch Intern Med 1993;153:154–83.
mental contributions to cardiovascular risk factors in Mexican
Americans: the San Antonio Family Heart Study. Circulation
1996;94:2159–70.
43 Shah BV, Barnwell BG, Hunt PN, Nileen P, LaVange LM. SUDAAN
64 Lean MEJ, Han TS, Seidell JC. Impairment of health and quality of life
User’s Manual, Release 5.50. Research Triangle Park, NC: Research
Triangle Institute, 1991.
65 Bray GA. Obesity: a time bomb to be defused. Lancet 1998;352:160–61.
44 Despres JP, Moorjani S, Lupien PJ, Tremblay A, Nadeau A, Bouchard
66 Molina C, Zambrana RE, Aguirre-Molina M. The influence of culture,
C. Regional distribution of body fat, plasma lipoproteins, and cardiovascular disease. Arteriosclerosis 1990;10:497–511.
class, and environment on health care. In: Molina C, Aguirre-Molina
M (eds). Latino Health in the US: A Growing Challenge. Washington, DC:
American Public Health Association, 1994.
45 Kaplan GA, Keil JE. Socioeconomic factors and cardiovascular
disease: a review of the literature. Circulation 1993;88:1973–98.
46 Anderson NB. Behavioral and sociocultural perspectives on ethnicity
and health. Health Psychol 1995;14:589–91.
in people with large waist circumference. Lancet 1998;351:853–56.
67 Marin G, Amaro H, Eisenberg C, Opava-Stitzer S. The development of
a relevant and comprehensive research agenda to improve Hispanic
health. Public Health Rep 1993;108:546–50.