Limitations
The World Mental Health version of the Composite International Diagnostic Interview (CIDI), which was used in the Canadian Community Health Survey (CCHS): Mental Health and Well-being, has yet to be validated. Therefore, the extent to which clinical assessments by health care professionals would agree with assessments based on CCHS data is not known.
In this study, the association between depression and work impairment was based on self-reported data rather than objective measures of work impairment. The degree of bias stemming from recall error or from the impact of depression on respondents’ perceptions of their own work impairment is not known.
Some variables used in cross-sectional analysis were not included in the longitudinal National Population Health Survey (NPHS) (alcohol and drug dependence, anxiety disorder, self-perceived work stress, coping, co-worker support, supervisor support) or were defined slightly differently (depression in previous year, at least one disability day in past two weeks, chronic conditions, low emotional social support, occupation). Consequently, the cross-sectional and longitudinal models are similar but not identical.
Because NPHS interviews are conducted every two years, work impairment subsequent to depression reported at the baseline interview pertains to the situation two years later. If depression-associated work impairment occurred within this two-year interval, it would not be captured in the survey. Therefore, longitudinal associations between depression and subsequent work impairment may be underestimated.
As a result of a skip-pattern error, no information was collected on the pregnancy status of 2,093 employed women aged 25 to 49 at the time of their CCHS interview. Therefore, those who were pregnant and whose weight exceeded their non-pregnant weight may have been placed in an incorrect BMI category. However, the impact of this oversight on the prevalence and odds ratios reported in this paper is probably negligible.
Smoking, a potential confounder in the relationship between depression and work impairment, was not available in the 2002 CCHS, and so could not be accounted for in the cross-sectional multivariate analysis. However, it was included in the longitudinal analysis using NPHS data.
|