• P Hasselback, K I Lee, Y Mao, R Nichol, and D T Wigle.
• Department of Epidemiology, University of Western Ontario, London.
• Can J Psychiatry. 1991 Nov 1; 36 (9): 655-9.

AbstractThe correlates of suicide rates were determined by conducting a multivariate study of sociodemographic indicators and suicide rates of 261 Canadian census divisions. Twenty-one sociodemographic variables were entered into a stepwise multiple linear regression to develop a model for suicide rates. The important variables were mortality rate for all causes of death, the age of the population, average family income, population density, proportion with no religious affiliation, proportion of Francophones, unemployment, immigration, proportion of Native people, a regional effect for British Columbia and the north, and growth by mobility, explaining 62% of the observed variation. This spatial ecologic analysis highlights the importance of cultural differences in explaining the variation of suicide rates. The analysis supports the previously found negative relationship between income and suicide while contrasting from previous studies in determining a inverse relationship with unemployment and an inverse relationship with the age distribution.

13 keywords...

hide…