Disability associated with multimorbidity including mental health conditions: a cross-sectional analysis of Canadian population-based data

This study will improve our understanding of the association between disability and mental disorders that co-exist with physical chronic conditions. Depression, a common mental disorder, results in more years of life lost to disability than any other illness. Multimorbidity, the co-existence of two or more conditions, is a global health burden linked with a greater risk of death, disability and hospitalization. Mental disorders that co-exist with other physical conditions may be the “perfect storm” in relation to disability. Recent studies show that associations of physical health with disability are stronger when mental disorders are present. This analysis of 50,000 participants from the Canadian Longitudinal Study on Aging, explores how mental disorders co-existing with other chronic conditions shape disability. The study responds to calls for more research identifying common patterns of multimorbidity and adds to the limited literature on multimorbidity in Canadian adults. The results will direct researchers and decision-makers on future research and resource allocation priorities.

The study used Baseline CLSA data, for 51,338 community-dwelling Canadians between 45-85 years of age. The outcome was disability, measured using self-reported limitations on any of 14 Basic or Instrumental Activities of Daily Living (ADL/IADL). Multimorbidity was measured by the number of self-reported chronic conditions, captured using participant responses to the following question about conditions lasting 6+ months: “has a doctor ever told you that you have ___”. Mental health conditions were included in the conditions, and reported mood and anxiety were used to identify the presence of mental health disorders. The prevalence of disability at each level of multimorbidity was compared for those with and without mental health conditions, and stratified analyses were performed to identify potential confounders (e.g., age, sex, income, education, social support, living arrangement). Logistic regression was used to identify and compare the odds ratios (ORs) of disability for people with different levels of multimorbidity. Separate models were run for those whose multimorbidity included mental health conditions versus those that didn’t, and all models were adjusted for confounders and used a reference category of no chronic conditions.

We found that fifty-one percent (51%) of the cohort was female, 58% were between the ages of 45 and 64, 20% reported that they had mood or anxiety disorders, and 10% had disability (i.e., reported the existence of at least one of 14 ADLs/IADLs). Thirty-five percent (35%) reported multimorbidity, with the most prevalent chronic conditions being arthritis, hypertension, diabetes, eye conditions and respiratory disorders. On average, participants reported having 2.2 chronic conditions, with 2.1 conditions reported for those without mental health disorders compared to 2.8 for those with mental health disorders.

The proportion with a mood/anxiety disorder was higher for women (62.9% women versus 37.1 % men), younger ages (66.6% for 45-64 versus 33.4% for 65+), and people with more multimorbidity (70.9% for 3+ chronic conditions vs 29.1% for 0-2). Social restrictions due to health were more likely to be reported by people with a mood/anxiety disorder (14.1% of those with a mood/anxiety disorder versus 5.9% without) and disability was more likely to be reported by people with a mood/anxiety disorder (17.2% of those with a mood/anxiety disorder versus 8.3% without). For each level of multimorbidity, the prevalence of disability increased consistently and was higher for women than men. There was evidence of a differential (higher) impact on disability for those with versus without mental health conditions, particularly at the higher levels of multimorbidity (4 and 5+ chronic conditions) and among women. The prevalence of disability increased with age for both sexes.

Logistic regression results showed that the odds of disability increased with increasing levels of multimorbidity. At all levels of multimorbidity, the point estimate odds of disability (OR) were lower for those without mental health conditions compared to those with mental health conditions. The overlap in confidence intervals (CIs) was minimal at most levels of multimorbidity, and there was no overlap for two multimorbidity levels (2& 5+).

With respect to the age/sex-adjusted odds of disability for the two groups, the ORs for those with no mental health conditions were similar to those in the original model run, but now the ORs for those with mental health conditions were higher at all levels of multimorbidity and there was no overlap of the CIs at any level of multimorbidity.

The analyses of ADLs and IADLs run as separate outcomes showed similar results to those above (composite outcome); however, these results should be interpreted cautiously due to the small number of limitations reported at various multimorbidity levels. The factor analysis showed that mental health conditions clustered with highly symptomatic conditions (as hypothesized), which included respiratory conditions, migraine headaches, and bowel and intestinal disorders.
Study findings suggest that disability is higher for those with mental health conditions compared to those without at all levels of multimorbidity. This highlights the unique, differential impact of mental health in shaping disability, which can result in other negative outcomes, such as more self-management challenges, higher healthcare service use, and mortality. When depressive symptoms were captured in addition to self-reported mental health conditions, the odds of disability were higher. Capturing depressive symptoms is likely to provide a more realistic portrait of the disability facing people with mental health disorders, because these conditions are known to be under-reported. It is also important to note that these results pertain to a relatively healthy Canadian cohort, therefore it may underestimate the impact of mental health on disability that would be seen in diseased populations.