|
Specific conditions were recorded using the International
Classification of Disease (ICD-10). Initially, the ICD codes were
clustered into categories intended to be consistent with those used
in the Canadian NPHS (10). Additional categories (for example, brain
injury) were derived from the ICD-10 codes for further exploratory
analysis. Associations were explored by calculating prevalence ratios,
with the prevalence of major depression in subjects with a long-term
condition comprising the numerator of this ratio and subjects without
any long-term condition comprising the denominator.
To identify and adjust for confounding or modifying
effects of additional variables, the sample was stratified on sex,
age, social support, income, employment, and recent life stressors.
The age and income stratification levels were decided upon post
hoc by identifying levels associated with substantial changes in
the prevalence of major depression. The age groups were defined
as under 45 years and 45 years or over. Subjects were categorized
as having deficits in social support if they reported they had no
one to confide in, no one to count on in a crisis situation, no
one to rely upon for advice when making important decisions, or
no one to make them feel loved and cared for. Major life stressors
included any of the following events (occurring to oneself or someone
close to oneself): physical attack, unwanted pregnancy, major financial
crisis, failing at school or training, demotion or pay cut, increased
arguments with partner, going on welfare, or a child moving back
home.
Sampling weights were used to account for both the method of random
digit dialing employed (11), the number of telephones in the household,
the number of household residents, and (post hoc) an overrepresentation
of women in the study sample. Tabular analyses used the CSAMPLE
program (12).
Results
A total of 2542 subjects were interviewed. There
were 958 (38% unweighted) subjects who reported 1 or more long-term
medical conditions. Of these, 201 (21% unweighted) had the CIDI-SFMD
depressive syndrome, compared with 234 (15% unweighted) without
long-term medical conditions. Having 1 or more condition was associated
with an increased prevalence of major depression (prevalence ratio
[PR] = 1.44, CI, 1.17 to 1.76). Because major depression becomes
less common with advancing age, and many medical conditions are
more common with advancing age, the crude association might have
been diminished as a result of confounding by age. Hence, the results
were stratified on age. Additional stratification on sex, major
life stressors, social supports, and employment was carried out.
As shown in Table
1, the association between depression and long-term medical
illness remained evident after stratification, except that it appeared
possibly weakened in the subjects aged 45 years and over.
To explore these associations further, specific groups of illnesses
were examined separately. Table
2 outlines the number of subjects with each type of medical
condition and the prevalence ratios and the CIs for each class of
disorders. The conditions that associated most strongly with depression
were asthma, back problems, gastrointestinal ulcers and other gastrointestinal
disorders, genitourinary conditions, musculoskeletal conditions,
vascular disorders, chronic fatigue syndrome and fibromyalgia, and
with the exception of diabetes, endocrine or metabolic disorders.
Consistent with previous studies (5–7), hypertension was not
associated with an increased prevalence of depression. This association
was stratified by age, because it could be theorized that the psychosocial
impact of hypertension might be greater in young subjects. This
stratification failed to uncover definite differences, but there
was a nonsignificant elevation in prevalence among subjects under
age 45 years (PR = 1.52; CI, 0.73 to 3.20), compared with those
over 45 years (PR = 1.10; CI, 0.58 to 2.08).
Discussion
These results replicate previous reports of an association
between long-term medical conditions and an increased prevalence
of major depression. Consistent with previous research (1–2),
we found that hypertension was not associated with major depression
in the general population sample. Hypertension differs from most
other medical conditions in that it is often asymptomatic and often
has little impact on day-to-day life.
The relation between depression and various long-term
medical conditions is potentially important. Depression is associated
with an elevated risk of major cardiac events following an episode
of unstable angina (13) and is associated with elevated mortality
in various other populations, including the general population (14).
|
|
Long-term medical conditions may be an important risk factor for
depressive disorders. Although the current investigation cannot
argue strongly for an etiologic association because of its cross-sectional
study design, these results do replicate and confirm that long-term
medical conditions, as a group, are associated with major depression
in the general population. We will need prospective studies to clarify
the etiologic forces underlying these associations. Whooley recently
reviewed mechanisms that may underlie the association between depression
and medical illness (15). Depression may increase the risk of medical
illness, and medical illness may increase the risk of depression.
Existing studies are cross-sectional in nature. Prevalence is influenced
both by incidence and duration of illness episodes; for this reason,
cross-sectional associations could emerge as a result of medical
illness having a negative prognostic impact on depressive disorders,
or visa versa. Hence, decreased treatment compliance among depression
patients could contribute to the association.
Other than its reliance on a cross-sectional study design, the major
limitation of this study was its use of a brief predictor for major
depression (rather than a detailed diagnostic interview) and its
reliance upon self-report data about long-term medical conditions.
Presumably, some errors in classifying these variables were made.
If these errors were independent of each other, the expected direction
of bias would be toward the null. As such, bias of this sort could
not account for the associations observed in this study. However,
if the measurement errors were differential (that is, if misclassification
of medical conditions depended upon mood status), the association
may have been overestimated. This possibility cannot be entirely
excluded because people with major depression may be expected to
view their health in a more negative way. Because false positives
could theoretically occur more often in subjects with depression,
this could lead to overestimation of the association. Nevertheless,
the self-report question, as mentioned above, asked about diagnoses
made by a “health professional” as a means of protection
against this sort of bias.
Acknowledgement
Dr Patten is funded by a Population Health Investigator Award from
the Alberta Heritage Foundation for Medical Research. This project
also received support from the Calgary Health Region.
References
1. Ahto M, Isoaho R, Puolijoki H, Laippala P, Romo
M, Kivela SL. Coronary heart disease and depression in the elderly
– a population-based study. Fam Pract 1997;14:436–45.
2. Andersen G, Vestergaard K, Riis JO, Lauritzen
L. Incidence of post-stroke depression during the first year in
a large unselected stroke population determined using a valid standardized
rating scale. Acta Psychiatr Scand 1994;90:190–5.
3. Ciacci C, Iavarone A, Mazzacca G, De Rosa
A. Depressive symptoms in adult coeliac disease. Scand J Gastroenterol
1998;33: 247–50.
4. Mendez MF, Doss RC, Taylor JL, Salguero
P. Depression in epilepsy. Relationship to seizures and anticonvulsant
therapy. J Nerv Ment Dis 1993;181:444–7.
5. Wells KB, Golding JM, Burnam MA. Psychiatric
disorder in a sample of the general population with and without
chronic medical conditions. Am J Psychiatry 1988;145:976–81.
6. Moldin SO, Scheftner WA, Rice JP, Nelson
E, Knesevich MA, Akiskal H. Association between major depressive
disorder and physical illness. Psychol Med 1993;23:755–61.
7. Patten SB. Long-term medical conditions
and major depression in the Canadian population. Can J Psychiatry
1999;44:151–7.
8. Patten SB. Major Depression Prevalence
in Calgary. Can J Psychiatry 2000;45:923–6.
9. Kessler RC, Andrews G, Mroczek D, Ustun
B, Wittchen HU. The World Health Organization composite International
Diagnostic Short-Form (CIDI-SF). International Journal of Methods
in Psychiatric Research 1998;7:171–85.
10. Statistics Canada. National Population
Health Survey. 1994-95. Public Use Microdata Files. Ottawa: Minister
of Industry, 1995.
11. Potthoff RF. Telephone sampling in epidemiological
research: to reap the benefits, avoid the pitfalls. Am J Epidemiol
139:967–78.
12. Kalsbeek W, Frerichs R. CSAMPLE. Analyzing
data from complex survey samples. Epi Info Manual. Atlanta: Centers
for Disease Control and Prevention. 1997;157–84.
13. Lesperance F, Frasure-Smith N, Juneau
M, Theroux P. Depression and 1-year prognosis in unstable angina.
Arch Intern Med 2000;160:1354–60.
14. Wulsin LR, Vaillant GE, Wells VE. A systematic
review of the mortality of depression. J Psychosom Med 1999:61;6–17.
15. Whooley MA. Depression and medical illness
(Letter). Ann Epidemiol 1999;9:281–2.
-----------------------------------------------------
Manuscript received January 2001, revised September 2001, and accepted
January 2002.
1Psychiatry Resident, University of Alberta, Department of Psychiatry,
Edmonton, Alberta.
2Associate Professor, Department of Community Health Sciences, University
of Calgary, Calgary, Alberta.
Address for correspondence: Dr Patten, Associate Professor,
Department of Community Health Sciences, University of Calgary,
3330 Hospital Drive Northwest, Calgary, AB T2N 4N1
e-mail: patten@ucalgary.ca
|
