Population risk and burden of health behavioral-related all-cause, premature, and amenable deaths in Ontario, Canada: Canadian Community Health Survey-linked mortality files

PURPOSE: To examine the association of all-cause and premature mortality with four modifiable lifestyle behaviors and quantify the burden of behavioral-related premature death in Ontario, Canada.

METHODS: We analyzed a cohort of 149,262 adults in the 2000-2010 Canadian Community Health Surveys, linked to vital statistics data to ascertain deaths until December 31, 2015. The strength of the association between behaviors (smoking, body mass index, physical inactivity, and alcohol consumption) and all-cause and premature mortality was estimated using sex-specific Cox proportional hazards models. We estimated the proportion of deaths from causes amenable to the health system by behavior.

RESULTS: After full adjustment, hazard ratios (95% confidence interval) for premature mortality were significantly increased for heavy smokers versus nonsmokers [males: 5.48 (4.55-6.60); females 4.45 (3.49-5.66)]; obese class III versus normal weight [males: 2.47 (1.76-3.48); females: 1.73 (1.29-2.31)]; and physically inactive versus active [males: 1.25 (1.07-1.45); females: 1.70 (1.41-2.04)]. In both sexes, a disproportionate burden of amenable deaths were experienced by heavy smokers, severely obese, physically inactive, and heavy drinkers.

CONCLUSIONS: The findings emphasize the importance of prevention to reduce the prevalence of risk behaviors that contribute to a large burden of premature deaths that are amenable to the health system.