论文部分内容阅读
Background: Urban-rural differences in suicide rates have been observed but results are inconsistent across studies and tend to differ by sex.Urban-rural differentials for different causes of suicide have rarely been examined, yet the identification of populations with higher rates of cause-specific suicide can help uncover potential causal factors and pathways for prevention, thereby facilitating targeted health promotion activities.The aim of this study was to examine the urban-rural continuum and leading causes of suicide in Canadian men and women.Methods: Municipalities were classified according to an 8-category urban-rural continuum based on a typology that geographically differentiates areas of Canada outside census metropolitan areas (metros) and mid-size cities, according to degree (strong, moderate, weak, none) of urban influence (UI).Four types of urban areas were distinguished: (a)large metros (>1 million residents), (b) medium metros (500,000-999,999 residents), (c) small metros (100,000-499,999 residents), and (d) mid-size cities (10,000-99,999 residents).Four types of rural areas (<10,000 residents) were distinguished, based on the percentage of residents commuting to urban cores: (a) strong UI with >30% commuters, (b) medium UI with 5%-30% commuters, (c) weak UI with <5% commuters, and (d) no UI with 0% commuters.Using the 1991-2001 Canadian Census Mortality Follow-up Study cohort (N=2,735,152), we estimated urban-rural differentials (hazard ratio, HR; 95% confidence interval, CI) in all-cause and cause-specific (poisoning, hanging/strangulation/suffocation, firearm, jumping) suicide with Cox proportional hazards regression, accounting for socioeconomic and demographic characteristics, in men and women separately.Results: For males, the risk of suicide was greater in most areas relative to large metros; whereas for females, elevated risk was present only in the two most remote zones.Associations with the urban-rural continuum varied by cause.Compared with large metros, strong and significantly higher hazards were present only in rural areas with no UI for hanging among males (HR=2.55, 95% Cl 1.26-5.17).Residents of rural areas with weak UI had a higher risk of poisoning suicide, with a stronger association for females (HR=2.38, 95% Cl 1.22-4.64) than males (HR=1.61, 95% Cl 1.01-2.82).An increased risk was also present in small metros for males (HR=1.61,95% Cl 1.04-2.48).Hazards were significantly higher in rural areas with no UI for jumping (HR=3.54, 95% Cl 1.21-10.4) and rural areas with strong (HR=2.02, 95% Cl 1.02-4.02) and weak (HR=2.59, 95% Cl 1.46-4.62) UI for firearm suicides (both sexes combined).Conclusion: Urban-rural differences observed for hanging, poisoning, firearm and jumping suicides were mostly confined to the two most rural zones.Relative to large metros, elevated risks were also observed in small metros for poisoning suicides among males and in rural areas with strong Ul for firearm suicides.Integrating a rural-urban perspective is important for prevention of suicide, as the socio-cultural context of rural areas is likely to differ from urban ones.