| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the Department of Psychiatry (N.S., A.M.), McGill University, Douglas Hospital Research Centre, Montreal Canada; Departments of Psychiatry and Community Health Sciences (J.W.), Faculty of Medicine, University of Calgary, Calgary, Canada; Centre de Recherche Fernand Seguin (A.L.), University of Montreal, Hôpital Louis-H. Lafontaine, Montreal, Canada.
Address correspondence and reprint requests to Norbert Schmitz, Douglas Hospital Research Centre, McGill University, 6875 LaSalle Boulevard, Montreal, Quebec, H4H 1R3, Canada. E-mail: norbert.schmitz{at}mcgill.ca
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Methods: We used data from the Canadian Community and Health Survey Cycle 2.1 (n = 46,262), a nationally representative survey conducted in 2003 by Statistics Canada. Depression, chronic conditions, and functional disability were assessed by personal/telephone interview.
Results: Prevalence of functional disability was higher in subjects with chronic conditions and comorbid major depression (46.3%) than in individuals with either chronic conditions (20.9%) or major depression (27.8%) alone. With no chronic conditions and no major depression as reference and after adjusting for relevant covariates, the odds ratio of functional disability was 2.49 (95% confidence interval (CI), 1.91–3.26) for major depression, 2.12 (95% CI, 1.93–2.32) for chronic conditions, and 6.34 (95% CI, 5.35–7.51) for chronic conditions and comorbid major depression.
Conclusions: The results suggest that there is a joint effect of depression and chronic conditions on functional disability. Research and social policies should focus on the treatment of depression in chronic conditions.
Key Words: depression • chronic conditions • disability • community survey
Abbreviations: CCHS = Canadian Community and Health Survey; CIDI-SFMD = Composite International Diagnostic Interview-Short Form for Major Depression; SE = standard error; OR = odds ratio; CI = confidence interval.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Somatic diseases and major depression co-occur at high levels in epidemiological and clinical settings. Prevalence rates of both conditions range from 3% to 6% in community settings (2) to 5% to 10% in primary care (4) and 8% to 15% in medical/surgical inpatients (5,6). Some evidence indicates that comorbid depression interacts with physical illness to amplify the disability normally associated with many medical conditions (7–9). For example, studies have assessed the impact of depression in patients with diabetes in terms of their functional ability or quality of life. According to data from the 1999 National Health Interview Survey in the United States, subjects with diabetes and major depression had higher functional disability compared with individuals with either diabetes or major depression alone (10). Similar associations have been reported for other chronic conditions. Goethe et al. (11) found that depression had a major impact on functional capacity in patients with asthma. Carels (12) reported that depressive symptoms had a greater impact on quality of life in patients with congestive heart failure than severity of cardiac dysfunction. Several studies provided evidence that depression is an independent risk factor of disability/reduced health-related quality of life in patients with rheumatic diseases (13), ulcers (14), migraine (15), back pain (16), and hypertension (17).
Using data from the recent Canadian Community and Health Survey (CCHS), the objectives of the present study were to evaluate the association between chronic conditions, depression, and functional disability in a community sample and to examine the joint effect of chronic conditions and depression on functional disability. Due to the large sample size of the survey, we were able to evaluate the associations for a broad class of chronic conditions. We hypothesized that individuals with chronic conditions and comorbid depression would have higher disability compared with individuals with chronic conditions or depression alone, even after controlling for potentially confounding factors.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
12 years. The survey participants were selected using multiple-staged, stratified random sampling procedures. Data were collected from 134,072 individuals by trained Statistics Canada interviewers using a computer-assisted personal or telephone interviewing method. The response rate was 80.7%. The study was approved by an Advisory Committee consisting of representatives from health regions, all provincial and territories ministries of health, and Health Canada. The CCHS Cycle 2.1 questionnaire consisted of a main module, which was asked of all respondents, and several submodules. The main module included, among others, questions regarding sociodemographic characteristics, chronic conditions, general health, height/weight, physical activities, smoking, and disabilities. The optional modules included questionnaires that were chosen by regional representatives from a fixed list according to local needs and priorities (e.g., depression, distress, suicidal thoughts, and medication use). The depression module was administered in five provinces (Newfoundland and Labrador, Prince Edward Island, New Brunswick, Ontario, and Alberta) and the three territories (Northwest Territories, Nunavut, and Yukon). Overall, 52,108 subjects were assessed for depression. In the present study, we included only subjects who were assessed for depression and who were between 15 and 79 years at the time of interview, resulting in a sample size of 46,673 subjects.
Assessment
Major depression in the past 12 months was evaluated using the Composite International Diagnostic Interview-Short Form for Major Depression (CIDI-SFMD) (19). The CIDI-SFMD is a brief version of the major depression section from the CIDI, a fully structured diagnostic interview (20). It includes 19 questions covering the symptoms of major depression that are critical to the Diagnostic and Statistical Manual of Mental Disorders, 4th Edition (DSM-IV) diagnostic definition. Validation studies have reported that 75% to 90% of people endorsing 5 symptoms on the CIDI-SFMD have experienced an episode of major depression during the preceding year (19). Therefore, a criterion of at least five symptoms was used to identify subjects having had an episode of major depression in this study.
The survey included 30 questions asking subjects if they suffered from various chronic conditions diagnosed by a health professional at the time of the interview. In the present analyses, we focused on the most prevalent chronic conditions: diabetes, asthma, high blood pressure, heart disease, stomach or intestinal ulcers, arthritis/rheumatism, migraine headaches, and back problems. We focused our analyses on chronic conditions in general (presence of at least one chronic condition; presence of one chronic condition; presence of 2 chronic conditions) as well as on the individual chronic conditions.
The CCHS also collected information about functional disability in the 2 weeks before the interview. Specifically, the participants were asked: "During the past 14 days, did you stay in bed at all because of illness or injuries, including any nights spent as a patient in a hospital?" and "During the past 14 days, were there any days that you cut down on things because of illness or injuries?" Subjects who responded "yes" to either question were asked about the number of disability days (days that respondents had to stay in bed for all or most of the day or did cut down on things for all or most of the day) during the 2 weeks before the survey. In our analyses, the presence of at least one disability day was considered to indicate the presence of functional disability. Disability days were categorized into five categories (none; 1 or 2; 3 to 5; 6 to 13; and 14 disability days) due to the skewed distribution.
The sociodemographic factors being studied are sex, age, education, marital status, employment, and race and urban-rural status. Urban residence was defined as residence in a metropolitan area.
All subjects voluntarily participated in the study. Informed consent was obtained by Statistics Canada. The data were released for public use in 2005 (18).
Statistical Analysis
Sociodemographic characteristics and the prevalence of functional disability were compared across four categories: no chronic condition and no depression, chronic condition only, depression only, and chronic condition and depression. Ordinal logistic regression (21) was used to model the associations between depression/chronic medical conditions and disability days. Ordinal logistic regression is an extension of logistic regression that deals with ordered outcomes—in this case, the five disability days categories (none; 1 or 2; 3 to 5; 6 to 13; and 14 disability days). The ordinal model specifies a log-linear relation for the odds of being in one category as compared with being in a lower category (1 disability days versus no disability days; 3 disability days versus <3 disability days; 6 disability days versus <6 disability days; 14 disability days versus <14 disability days) and calculates a single odds ratio (OR) for the association between a predictor variable and each combination of higher versus lower outcome categories. This approach assumes proportional odds for any dichotomy of the five levels of disability days status. The assumption of proportional odds was tested and found to be valid for the described models.
To evaluate the interaction of depression and chronic medical conditions, we recoded the two risk factors into three dummy variables: a) depression without chronic medical conditions, b) chronic medical conditions without depression, and c) depression with chronic medical conditions. These dummy variables were entered simultaneously into the ordinal logistic regression models to determine if the ORs increased with the presence of both depression and chronic medical conditions. In all regression analyses, we controlled for sex, age, marital status, race, education, employment, and urban-rural status. For the eight individual chronic conditions, we controlled additionally for the presence of other chronic conditions.
To provide additional insight into the joint relationship of depression and chronic condition with functional disability, we also tested an additive OR model using a certain synergy index (22–24). When two causes of an outcome act independently of each other, their effects are additive; that is, the risk for subjects suffering from both conditions equals the sum of the risks in the groups suffering from only one condition. When the risk in the group with both conditions is even greater than that expected under the additivity model, this is called an interaction of the risk factors and it suggests that the two causes might be influencing each other (acting synergistically). For the logistic regression models, the synergy index (SI) was computed using the following formula: SI = ORcd – ORc – ORd + 1, where ORcd is the OR for depression and chronic condition, ORc is the OR for chronic condition without depression, and ORd is the OR for depression without chronic condition. SI > 0 implies a greater than additive interaction effect (synergistic effect), whereas SI = 0 implies no interaction (additive effect), and SI < 0 implies a less than additive effect (negative interaction). Due to the complex sampling design, 95% confidence interval (CI) for the SI was estimated using a bootstrap approach. We generated 2000 bootstrap samples with replacement for each logistic regression model. SI values were calculated for each resampling, the 2.5 and 97.5 percentiles of the empirical distributions identified the 95% CI.
All data were analyzed by using Stata statistical software, release 8.1, which includes commands for the analysis of complex survey data. Logit models were fitted using the svyologit procedure in Stata 8.1, which takes into account the sample’s clustering and stratification. All estimates were weighted by the master weight provided by Statistics Canada. A survey weight was given to each person included in the sample. This weight corresponds to the number of persons in the entire population that are represented by the respondent. Weights were created that adjusted for different probabilities of selection (e.g., geographical adjustment), nonresponse (e.g., household nonresponse and personal nonresponse), and several other factors (e.g., seasonal effects). A relative weight was created for the present analyses by dividing the weight by the mean of the weight.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Of 46,673 participants aged 15 to 79 years examined in our sample, 46,262 participants had complete data on all study variables and provided the basis of this report.
The population prevalence of depression in the past 12 months was 6.0%, although 4.0% of the subjects suffered from depression and at least one chronic condition and 2.0% of the subjects suffered from depression but did not report any chronic condition. Less than half (45.4%) of the subjects did not suffer from depression but reported at least one chronic condition; 48.5% of the subjects did not suffer from either chronic condition or depression.
Less than one fifth (17.5%, standard error of the mean = 0.3) of the subjects reported disability days in the previous 2 weeks (1 or 2 days: 7.4%; 3–5 days: 4.6%; 6–13 days: 2.9%; 14 days: 2.6%).
"Back problems" was the most prevalent (20.6%) chronic condition, followed by arthritis/rheumatism (16.8%), high blood pressure (14.1%), migraine (11.3%), asthma (8.2%), diabetes (4.5%), heart disease (4.3%), and ulcers (3.0%).
The demographic characteristics of subjects suffering from chronic condition and/or depression are shown in Table 1. Sociodemographic characteristics varied across groups. There were more women than men in those with asthma, arthritis, and migraine; the proportion of males was higher than those with diabetes and heart diseases. With two exceptions (asthma and migraine), subjects with chronic condition were more often older than those without chronic conditions. The highest education was observed among those with back problems, whereas the lowest education was observed among those with diabetes.
|
The eight chronic conditions accounted for >72% of all disability days (chronic conditions without depression: 58.3%). The presence of depression substantially increased the number of disability days: 4.0% of the subjects with chronic conditions suffered from comorbid depression, but these subjects reported 13.7% of all disability days.
Table 2 shows the mean number of disability days and adjusted ORs of functional disability by disease category. Ordinal logistic regression suggests that depression had more than an additive effect on disability in subjects suffering from chronic condition. Depression increased the OR for functional disability from 2.1 in chronic conditions only to 6.3 in chronic conditions with comorbid depression. The SI was calculated to be 2.9 (95% CI, 1.65–4.13), indicating a strong positive interaction.
|
An even stronger negative synergistic effect was observed in subjects suffering from 2 chronic conditions. A higher proportion of subjects with depression and comorbid chronic conditions (54%) reported having functional disability in contrast with 28% of those with depression only, 27% of those with chronic conditions only, and 12% of subjects with neither depression nor chronic conditions (SI = 4.6; 95% CI, 2.84–6.54). The negative effect of depression on functional disability was smaller but still present in subjects suffering from a single chronic condition (SI = 0.84; 95% CI, 0.31–1.54).
When comparing individual chronic conditions, a strong effect of depression on functional disability was found for all conditions. The strongest effect was found for heart diseases, suggesting a six-fold increase of functional disability in the presence of depression (Table 3). The lowest effects were found in asthma and migraine.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Our results are compatible with the findings by others; namely, depression is associated with a high burden of disease because >17% of all disability days in the present study were associated with depression (25). The present study adds to this literature by showing that the odds of functional disability were significantly higher in individuals with chronic conditions and comorbid depression than in individuals with either depression or chronic conditions alone. These results are consistent with the findings by Kessler et al. (26) who found that comorbid mental-physical disorders were more impairing than either mental disorders only or physical disorders only, and with Patten’s findings (8) that depression comorbid with at least one chronic condition had a greater than additive effect on disability. However, the study by Kessler et al. included those who were aged 
54 years (26). Recently, Stein et al. found that the presence of both major depression and chronic medical conditions was associated with significantly greater likelihood of healthcare use, increased functional disability and work absence compared with the presence of a chronic physical illness without comorbid major depression (27). However, they found an additive rather than a synergistic effect of comorbid depression on disability. In contrast to our analysis, Stein et al. used a dichotomized measure of disability and controlled for a broad range of physical health conditions (n = 20 conditions) (27). These methodological differences might partially explain the inconsistency between the results of the above-mentioned study and ours.
One possible limitation of the present study might be the use of a self-reported functional disability measure. Disability is a complex medical issue that has many dimensions—social, psychological, and medical. However, the number of disability days, as used in the present study, seems to be a global measure closely associated with health, although part of this association may be spuriously generated by recall and reporting bias. Several studies have shown a good agreement between self-reported work loss days (subject was unable to go to work) and employers’ absence records (28,29).
Another limitation might be the assessment of chronic conditions. Our results are based on self-reports without clinical confirmation of the physical disorders. This may result in an underestimation of chronic conditions. Prior research suggests that self-report of common chronic conditions showed moderate-to-strong agreement with medical records data (30). An additional limitation is that our assessment of chronic conditions does not capture severity or acuteness of symptoms. It is possible that condition-specific severity may moderate the association between depression and disability for some chronic conditions (31). Furthermore, we do not have information on the course of depression symptoms and depression treatments received. The quality of treatment may affect the associations between depression, chronic conditions, and disability.
There are several possible explanations for the finding that depression increases systematically the risk of disability. Some chronic conditions have stressful consequences as a result of the generic disease characteristics. For example, some conditions are accompanied by pain and disabilities, whereas others are troublesome because they are life threatening or have an unpredictable course that may be considered stressful in itself (32). Depressive mood may be a psychological reaction to primary or secondary consequences of the medical disease, resulting in greater risk for subsequent disability (33,34). Furthermore, the depression and physical decline link may be due to a third factor (such as organic brain dysfunctions) that is related to both conditions (34).
The mechanisms by which major depression and chronic conditions interact with each other are not fully understood. Biological, psychological, and social explanations and confounders should be considered. First, we must recognize that the cross-sectional design does not allow for causal inferences nor for the unfolding of disability for disorders that are chronic. On the biological side, depression has been recognized as a multidimensional disorder affecting brain and body. It has been associated with alterations in endocrine, cardiovascular, and immune system (35–37). The increased disability in those suffering from depression and chronic conditions might be caused by psychological mechanisms, too. It is possible that depression may have effects on perceived disability, but not on true disability. In addition to this measurement bias, it is possible that psychosocial characteristics cause the negative effect: depression is thought to interfere with physical recovery by impeding treatment seeking, adherence to treatment, and adoptions of healthy lifestyles (33), hence increasing disability. Therefore, depression and/or persistent somatic symptoms of depression (e.g., fatigue, sleeplessness) may enhance susceptibility to chronic conditions and result in an amplification of symptoms and corresponding subjective reactions to these symptoms (38–40). Social determinants of disability exist in addition to health and mental health determinants, and these determinants are also those of chronic disorders and depression with higher prevalence in socioeconomically disadvantaged populations. Comorbidities may actually capture greater social disadvantages that also determine disability. The availability and quality of emotional and instrumental social support, with their psychological and social determinants, may also mediate or confound the effect of the comorbidities, including depression, on disability.
Our results suggest that depression and most of the major chronic physical conditions are associated with greater disability, and if depression itself is considered a chronic disorder, the accumulation of chronic disorders is associated with greater functional disability. The risk of functional disability increased in subjects with both chronic conditions and depression: nearly half of the subjects with chronic conditions and comorbid depression reported disability days in the previous 2 weeks. The fraction of overall disability in the population associated with these main chronic disorders represent 72% of all disability days and the comorbidity is associated with 14% of all disability days. From a public health perspective, they represent a legitimate target for improved treatment. First, depression is becoming in countries like Canada the first cause of incapacity and it largely goes untreated; in general, public health strategies to increase the treatment of depression are appropriate (41). Functional disability also predicts additional functional decline, increased use of health services, increased healthcare costs, and increased risk of death (42). However, there is increasing evidence that treatment of depression may reduce functional disability because several studies found significant improvements in functional outcome in chronic conditions due to the treatment of comorbid depression (43,44).
Our results suggest that there is a joint effect of depression and chronic conditions on disability. Attracting attention to depression in the presence of other chronic conditions would offer a window of opportunity to treat depression, improve the management of the other chronic condition, and reduce disability. The negative effect of depression on disability suggests that research and social policies should focus on the treatment of depression in chronic conditions.
This analysis is based on Statistics Canada’s Canadian Community Health Survey, Cycle 2.1 (2003), Public Use Microdata file, which contains anonymized data collected. All computations on these microdata were prepared by Norbert Schmitz, JianLi Wang, Ashok Malla, and Alain Lesage. We accept responsibility for the use and interpretation of these data.
|
NOTES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
This research was supported, in part, by a FRSQ Chercheur-Boursier fellowship (N.S.) and CIHR Grant MOP-79464 (N.S., J.W., A.M., A.L.) from the Canadian Institute for Health Research.
DOI:10.1097/PSY.0b013e31804259e0
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  D. Levinson and Y. Lerner Employment Among Persons With Past and Current Mood and Anxiety Disorders in the Israel National Health Survey Psychiatr Serv, May 1, 2009; 60(5): 655 - 662. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
