a1 Department of Psychiatry, University of Manitoba, Winnipeg, Manitoba, Canada
a2 Department of Psychology, University of Manitoba, Winnipeg, Manitoba, Canada
a3 Manitoba Centre for Health Policy, Winnipeg, Manitoba, Canada
a4 Department of Community Health Sciences, University of Manitoba, Winnipeg, Manitoba, Canada
Background The association between physical disorders and suicide remains unclear. The aim of this study was to examine the relationship between physical disorders and suicide after accounting for the effects of mental disorders.
Method Individuals who died by suicide (n = 2100) between 1996 and 2009 were matched 3:1 by balancing score to general population controls (n = 6300). Multivariate conditional logistic regression compared the two groups across physician-diagnosed physical disorders [asthma, chronic obstructive pulmonary disease (COPD), ischemic heart disease, hypertension, diabetes, cancer, multiple sclerosis and inflammatory bowel disease], adjusting for mental disorders and co-morbidity. Secondary analyses examined the risk of suicide according to time since first diagnosis of each physical disorder (1–90, 91–364, ≥ 365 days). Similar analyses also compared individuals with suicide attempts (n = 8641) to matched controls (n = 25 923).
Results Cancer was associated with increased risk of suicide [adjusted odds ratio (AOR) 1.40, 95% confidence interval (CI) 1.03–1.91, p < 0.05] even after adjusting for all mental disorders. The risk of suicide with cancer was particularly high in the first 90 days after initial diagnosis (AOR 4.10, 95% CI 1.71–9.82, p < 0.01) and decreased to non-significance after 1 year. Women with respiratory diseases had elevated risk of suicide whereas men did not. COPD, hypertension and diabetes were each associated with increased odds of suicide attempts in adjusted models (AORs ranged from 1.20 to 1.73).
Conclusions People diagnosed with cancer are at increased risk of suicide, especially in the 3 months following initial diagnosis. Increased support and psychiatric involvement should be considered for the first year after cancer diagnosis.
(Received July 17 2013)
(Revised June 06 2014)
(Accepted June 14 2014)