A multiple-group approach to the study of
12-month psychiatric comorbidity across
the adult lifespan:
the EU-WMH project.
Ron McDowell*, Assumpta Ryan*,
Siobhan O’Neill**, Brendan Bunting**
*Institute of Nursing and Health Research, Coleraine
**Psychology Research Institute, Derry
Overview
• Introduction
• The EU-WMH project
• Methodology
• Results
• Discussion
• Conclusions
Introduction
• Studies of mental illness using standardised diagnostic instruments
often report very low levels of clinical disorders in later life
(McDowell et al. 2013)
• Scholars are divided as to whether this finding is realistic (Ernst and
Angst 1995) or not (Snowdon 2001)
• Psychiatric comorbidity is also a feature of mental illness (Hettema
2008)
• Again the literature is divided as to whether this is real or artifactual
(First 2002)
• Purpose of the analysis is to examine the psychiatric comorbidity of
common mood and anxiety disorders in Europe across the adult
lifespan. Where are the older adults?
The EU-WMH project
• 10 European countries participated in the World Mental Health
(WMH) surveys (n=37,289):
–
–
–
–
Belgium
Bulgaria
France
Germany
-Italy
-The Netherlands
-Northern Ireland
-Portugal
-Romania
-Spain
• Nationally representative multistage probability samples of adults
18+ years assessed using the Composite International Diagnostic
Instrument (CIDI) (Kessler and Üstün 2004)
• Common 12-month DSM-IV mood and anxiety disorders (without
hierarchy) assessed in all 10 countries:
–
–
–
–
major depressive episode (MDE)
dysthymia (DYS)
generalised anxiety disorder (GAD)
agoraphobia with/without panic (AGO)
- panic disorder (PD)
- social phobia (SO)
- specific phobia (SP)
• Results compared with a wider range of disorders assessed in
Bulgaria, N. Ireland, Portugal and Romania
Methodology
• Multiple-group latent class analysis (Lazarfeld and Henry 1968)
used to group the 12-month disorders according to patterns
suggested by the data
• Measurement invariance assessed at the item-level for the
groups/countries and age, and at the scale-level for the other
covariates (gender, martial status and urbanicity).
• Constraints on predictors of class membership and direct covariate
effects assessed using e.g. Modification Indices, Expected
Parameter Changes, statistical power (Saris et al. 2009)
• Model fit statistics included BIC, SSA-BIC
• Models fitted in Mplus (Muthén and Muthén 1998-2013)
Results
• 4-class solution most appropriate for the common 12-month mood and
anxiety disorders (BIC 46759.60, SSA-BIC 46554.04, Entropy 0.88)
• One class varied slightly in interpretation between countries due to
measurement non-invariance associated with MDE, SP
• Highly depressed class present in countries with the highest (N.
Ireland, Portugal) and lowest (Bulgaria, Romania) 12-month disorders
Results
• Comorbid class at least twice as large in N.Ireland as elsewhere
.8
.6
.4
.2
0
Item probabilities
1
N. Ireland
MDE
DYS
GAD
AGO
Disorder
PDS
SO
Class 1: Minimally affected (83.0%)
Class 2: Moderately depressed (10.9%)
Class 3: Highly depressed (1.8%)
Class 4: Comorbid (4.3%)
SP
Results
Age
Minimally
affected
Moderate
depression
Highly
depressed
Comorbid
25
100
23.34
(14.02,38.85)
2.12
(1.16,3.86)
4.09
(1.10,15.17)
45
100
17.26
(10.74,27.73)
4.20
(2.46,7.19)
5.25
(2.30,11.93)
65
100
9.58
(5.99,15.34)
2.24
(1.49,3.36)
3.03
(1.52,6.03)
80
100
5.12
(0.29,9.01)
0.59
(0.29,1.18)
1.19
(0.45,3.12)
-4
-6
-8
Log odds
-2
• Odds of belonging to the moderately depressed (relative to
minimally affected class) declines with age in most Western
European countries. Highly depressed and comorbid classes peak
mid-life.
Log odds of Class 2 (moderately depressed)
e.g. Unadjusted Odds (N. Ireland)
v Class 1(minimally affected) membership
20
40
60
Age (years)
N. Ireland
Italy
Spain
Belgium
Romania
80
France
Netherlands
Germany
Portugal
Bulgaria
100
Results
-4
Minimally
affected
Moderate
depression
Highly
depressed
Comorbid
25
100
3.02
(0.99,9.15)
0.22
(0.07,0.64)
0.04
(0.01,0.61)
45
100
7.45
(4.55,12.18)
0.69
(0.35,1.34)
0.16
(0.01,2.19
65
100
11.44
(7.29,17.96)
1.52
(0.91,2.54)
0.06
(0.003,1.57)
80
100
11.56
(6.00,22.28)
2.21
(0.88,5.57)
0.01
(0.00,0.55)
-6
Age
-8
-10
Log odds
-2
• Odds of belonging to the moderately depressed or highly
depressed classes (relative to minimally affected class) increases
with age in Eastern European countries. Comorbid class peaks
mid-life.
Log odds of Class 3 (highly depressed)
v Class 1(minimally affected) membership
e.g. Unadjusted Odds (Bulgaria)
20
40
60
Age (years)
N. Ireland
Italy
Spain
Belgium
Romania
80
France
Netherlands
Germany
Portugal
Bulgaria
100
• The shape of these relationships is unchanged after adjustment for
sociodemographic variables.
Results
• Probability of 12-month dysthymia increases with age across
all classes and all countries (β=0.25 (SE 0.06), p<0.001)
• Probability of 12-month social phobia decreases with age
across all classes and countries (β=-0.23 (SE 0.06), p<0.001)
• Makes a substantive difference in the highly depressed and
comorbid classes
N. Ireland latent class profile, 25 years
.6
.4
0
0
.2
.2
.4
.6
Item probabilities
.8
.8
1
1
N. Ireland latent class profile, 80 years
MDE
DYS
GAD
AGO
Disorder
PDS
SO
SP
MDE
DYS
GAD
AGO
Disorder
PDS
SO
SP
Class 1: minimally affected (77.1%)
Class 2: moderately depressed (17.6%)
Class 1: minimally affected (93.5%)
Class 2: moderately depressed (4.7%)
Class 3: highly depressed (16.8%)
Class 4: comorbid (3.6%)
Class 3: highly depressed (5.5%)
Class 4: comorbid (1.2%)
Results
• These effects are unchanged after adjustment for
sociodemographic variables.
• Effects associated with gender, marital status and urbanicity on
class membership are homogenous across classes and
countries e.g. odds ratios for gender (relative to minimally
affected class)
* Bulgaria, N. Ireland, Portugal and Romania only
• Measurement non-invariance for gender at the scale level:
probability of 12-month specific phobia higher in females than
males across all classes and countries (β=0.95 (SE 0.10),
p<0.001)
• When a wider range of disorders is included
–
–
–
–
–
intermittent explosive disorder (IED)
bipolar disorders (BI)
posttraumatic stress disorder (PTSD)
alcohol abuse/dependence (AL)
adult separation anxiety (ASA)
mental illness among older adults is reported primarily in terms
of mood and phobic disorders.
.2
.4
.6
.8
1
N. Ireland latent class profile, 80 years
0
Item probabilities
Results
MDE
DYS
GAD
AGO
PDS
SO
SP
Disorder
PTS
BI
AL
ASA
Class 1: minimally affected (87.3)
Class 2: cothymic (11.3%)
Class 3: highly depressed (0.7%)
Class 4: comorbid (0.7%)
IED
Discussion
• Patterns of comorbidity don’t correspond exactly to the DSM
metastructure or other commonly suggested patterns (Krueger
1999) and vary slightly with the number of countries and
disorders studied. High levels of comorbidity in N. Ireland
worrying.
• Older adults increasingly unlikely in the West to experience
depression/anxiety in later life. How realistic is this?
– In N. Ireland, among adults 25 years and over with mental illness
who were known to their Health Care Trust (excluding those with
dementia) 10.2% were 75 years and over (DHSSPSNI 2008)
– Lieden 85+ study: 15.4% of participants with clinically significant
depression (Stek et al. 2004)
• The CIDI hasn’t been specifically validated among older
community-dwelling adults and not at all in Eastern countries.
Discussion
• Although the number of adults in the highly depressed class
declines with age in NI/Portugal (and increases with age in
Romania/Bulgaria), within this class the probability of ‘double
depression’ (MDE/DYS) increases with age.
• Double depression has a worse course trajectory than major
depression alone (Rhebergen et al. 2009): patients recover
more slowly, relapse more quickly, have more intense
experience of suicidality (Keller and Shapiro 1982 ,Klein et al.
2006, Holm-Denoma et al. 2006).
• Decreasing probability of social phobia with age- common in
the literature. Is this likely to be realistic? Social isolation
(common among older adults) a risk factor for mental illness
(Chou et al. 2011).
Discussion
– Situations giving rise to social phobia decline with age or become
less distressing (Mohammadi et al. 2006, Cairney et al. 2007)
– Older adults may not find their fears ‘unreasonable’: the construct
may be invalid among older adults (Karlsson et al. 2010).
• Mental illness in later life is dominated by mood and phobic
disorders: how realistic is this?
– 15% of older adults are self-confessed worriers (Wisocki 1994)
– Common fears include personal health, loved ones, falls, crime
and victimisation (Kogan and Edelstein 2004, OFMDFM 2009,
Blay and Marinho 2012)
– Need for more research specifically about anxiety in later life
(Wittchen et al. 2011)
– Incidence of alcohol-use disorders relatively high among older
adults and this is not always reflected in research (Caputo et al.
2012), as is the burden of bipolar disorder among adults in
general (Karam 2013)
Conclusions
•
•
•
•
Limitations:
Cross-sectional data: confounding of temporal effects
The disorders assessed vary slightly between countries
Limited sociodemographic variables: lack of comparable data
across countries, computationally intensive
• Analyses just consider illness at the 12-month disorder level
• Are clinical diagnoses appropriate in later-life?
• Conclusions:
• Continued doubts on suitability for CIDI to assess mental
illness among older adults: needs proper validation
• Prevalence of common mental disorders likely to be
underestimated among older adults in a wide variety of
countries
• Reasons for differences among younger adults between East
and West not clear (Layard 2005).
References
•
•
•
•
•
•
•
•
•
•
•
Blay, S.L. and Marinho, V. 2012. Anxiety disorders in old age. Current Opinion in Psychiatry, 25 (6), 462467.
Cairney, J., McCabe, L., Veldhuizen, S., Corna, L.M., Streiner, D. and Herrmann, N. 2007. Epidemiology
of social phobia in later life. American Journal of Geriatric Psychiatry, 15 (3), 224-233.
Caputo, F., Vignoli, T., Leggio, L., Addolorato, G., Zoli, G. and Bernardi, M. 2012. Alcohol use disorders in
the elderly: a brief overview from epidemiology to treatment options. Experimental gerontology, 47 (6),
411-416.
Chou, K.L., Liang, K. and Sareen, J. 2011. The association between social isolation and DSM-IV mood,
anxiety, and substance use disorders: wave 2 of the National Epidemiologic Survey on Alcohol and
Related Conditions. Journal of Clinical Psychiatry, 72 (11), 1468-1476.
DHSSPSNI. 2008. Adult Community Statistics: 1st April 2007-31st March 2008. Belfast: Dept. of Health,
Social Services and Public Safety.
Ernst, C. and Angst, J. 1995. Depression in old age: Is there a real decrease in prevalence? A review.
European archives of psychiatry and clinical neuroscience, 245 (6), 272-287.
First, M.B. 2002. DSM-IV and psychiatric epidemiology. 333-342.
Hettema, J.M. 2008. The nosologic relationship between generalized anxiety disorder and major
depression. Depression and anxiety, 25 (4), 300-316.
Holm-Denoma, J.M., Berlim, M.T., Fleck, M.P. and Joiner, T.E.,Jr. 2006. Double depression in adult
psychiatric outpatients in Brazil: distinct from major depression?. Psychiatry research, 144 (2-3), 191-196.
Karam, K. 2013. Under-reporting Bipolar Disorder in Large-Scale Epidemiologic Studies. [online]. Harvard
Medical School, World Mental Health Survey Initiative. Available from:
http://www.hcp.med.harvard.edu/wmh/ [Accessed 15 July 2013].
Karlsson, B., Sigstrom, R., Waern, M., Ostling, S., Gustafson, D. and Skoog, I. 2010. The prognosis and
incidence of social phobia in an elderly population. A 5-year follow-up. Acta Psychiatrica Scandinavica,
122 (1), 4-10.
References
•
•
•
•
•
•
•
•
•
•
•
Keller, M.B. and Shapiro, R.W. 1982. "Double depression": superimposition of acute depressive episodes
on chronic depressive disorders. American Journal of Psychiatry, 139 (4), 438-442. .
Kessler, R.C. and Üstün, T.B. 2004. The World Mental Health (WMH) Survey Initiative Version of the
World Health Organization (WHO) Composite International Diagnostic Interview (CIDI). International
Journal of Methods in Psychiatric Research, 13 (2), 93-121.
Klein, D.N., Shankman, S.A. and Rose, S. 2006. Ten-year prospective follow-up study of the naturalistic
course of dysthymic disorder and double depression. American Journal of Psychiatry, 163 (5), 872-880.
Kogan, J.N. and Edelstein, B.A. 2004. Modification and psychometric examination of a self-report
measure of fear in older adults. Journal of anxiety disorders, 18 (3), 397-409.
Krueger, R.F. 1999. The structure of common mental disorders. Archives of General Psychiatry, 56 (10),
921-926.
Layard, R. 2005. Happiness: Lessons from a new science. New York, NY, US: Penguin Books/Penguin
Group.
Lazarfeld, P.F. and Henry, N.W. 1968. Latent Structure Analysis. Boston: Houghton Mifflin.
McDowell, R.D., Ryan, A., Bunting, B.P., O'Neill, S.M., Alonso, J., Bruffaerts, R., de Graff, R., Florescu,
S., Vilagut, G., de Almeida, J.M.C., de Girolamo, G., Haro, J.M., Hinkov, H., Kovess-Masfety, V.,
Matschinger, H. and Tomov, T. 2013. Mood and anxiety disorders across the adult lifespan: a European
perspective. Psychological Medicine, (Available on CJO 2013 doi:10.1017/ S0033291713001116 (epub
ahead of print))
Mohammadi, M.R., Ghanizadeh, A., Mohammadi, M. and Mesgarpour, B. 2006. Prevalence of social
phobia and its comorbidity with psychiatric disorders in Iran. Depression & Anxiety, 23 (7), 405-411.
Muthén, L. K., & Muthén, B. O. 1998-2013. Mplus User's Guide. Sixth Edition. Sixth Edition ed. Los
Angeles, CA: Muthén & Muthén.
OFMDFM. 2009. A profile of older people in Northern Ireland. Belfast: Office of the First Minister and
Deputy First Minister.
References
•
•
•
•
•
•
Rhebergen, D., Beekman, A.T., Graaf, R., Nolen, W.A., Spijker, J., Hoogendijk, W.J. and Penninx, B.W.
2009. The three-year naturalistic course of major depressive disorder, dysthymic disorder and double
depression. Journal of affective disorders, 115 (3), 450-459.
Saris, W.E., Satorra, A. and van der Veld, W.M. 2009. Testing structural equation models or detection of
misspecifications? Structural Equation Modeling, 16 (4), 561-582.
Snowdon, J. 2001. Is depression more prevalent in old age? Australian & New Zealand Journal of
Psychiatry, 35 (6), 782-787.
Stek, M.L., Gussekloo, J., Beekman, A.T., van Tilburg, W. and Westendorp, R.G. 2004. Prevalence,
correlates and recognition of depression in the oldest old: the Leiden 85-plus study. Journal of affective
disorders, 78 (3), 193-200.
Wisocki, P.A. 1994. The experience of worry among the elderly. 247-261.
Wittchen, H.U., Jacobi, F., Rehm, J., Gustavsson, A., Svensson, M., Jonsson, B., Olesen, J., Allgulander,
C., Alonso, J., Faravelli, C., Fratiglioni, L., Jennum, P., Lieb, R., Maercker, A., van Os, J., Preisig, M.,
Salvador-Carulla, L., Simon, R. and Steinhausen, H.C. 2011. The size and burden of mental disorders
and other disorders of the brain in Europe 2010. European Neuropsychopharmacology, 21 (9), 655-679.