M O Agbayewa1, S A Marion, S Wiggins. 1. Department of Psychiatry, University of British Columbia, Vancouver. agbayewa@unixg.ubc.ca
Abstract
OBJECTIVE: The effects of socioeconomic factors on suicide rates in the general population are widely documented. Few of these reports have specifically studied the effects of socioeconomic variables on suicide rates in the elderly population. Elderly persons have the highest suicide rates of any age-group. This group is different from the rest of the population insofar as suicide is concerned. For example, since most elderly persons are no longer in the labour force, it would be expected that they would be affected differently by economic factors such as unemployment. We report the findings of an ecological study of old-age suicide in British Columbia over an 11-year period. METHODS: We obtained information on all suicide counts (International Classification of Diseases [ICD-9] codes E590-959) recorded in the 21 health units of British Columbia over the 11-year period from October 1, 1981, to September 30, 1991, from the Division of Vital Statistics of the Province of British Columbia and Statistics Canada. Social, economic, and demographic information for the health units was obtained from census data and included the number of persons per household, proportion of the population that lived in 1-person households, immigration and migration rates for each region, proportion of the population with less than grade 9 education, proportion with less than grade 12 certification, marital status rates, unemployment rates by gender, average household income, average census family income, and labour-force participation rate by gender. We calculated overall and gender-specific suicide rates for elderly persons (65 years and older) and younger populations. Using Poisson regression analyses, we determined the cross-sectional and longitudinal relative risks associated with the socioeconomic variables for the units, and we also examined trends in suicide rates. RESULTS: There were 4630 suicides in the 11-year period. The mean suicide rate (per 100,000 population) for those over age 9 years was 18.6 (between health unit SD 5.2, 95% confidence interval [CI] = 17.0-20.2). The elderly have a higher suicide rate in every region. The male suicide rates (mean = 26.9, SD 6.4, 95% CI = 24.0-30.0) are higher than female rates (mean = 7.5, SD 1.7, 95% CI = 6.8-8.3) in every region. The factors influencing suicides were different for elderly males and elderly females. In all analyses, suicide rates in elderly females remained essentially stable across age-groups and units and over the years. Elderly male suicide rates varied across units and age-groups and over the years. CONCLUSIONS: Suicide rates are highest in males over age 74 years. There are regional differences in elderly suicide rates and the factors that influence them. Longitudinal and cross-sectional risk factors differ, and there are gender differences in the risk factors. For both elderly males and females, suicide rates appear to be influenced by social factors in the population as a whole, not just in the elderly population. Male and female employment patterns are associated with elderly male suicide rates, even though the latter are not in the labour force. For suicide in elderly women the important factors are population education, income, and migration levels.
OBJECTIVE: The effects of socioeconomic factors on suicide rates in the general population are widely documented. Few of these reports have specifically studied the effects of socioeconomic variables on suicide rates in the elderly population. Elderly persons have the highest suicide rates of any age-group. This group is different from the rest of the population insofar as suicide is concerned. For example, since most elderly persons are no longer in the labour force, it would be expected that they would be affected differently by economic factors such as unemployment. We report the findings of an ecological study of old-age suicide in British Columbia over an 11-year period. METHODS: We obtained information on all suicide counts (International Classification of Diseases [ICD-9] codes E590-959) recorded in the 21 health units of British Columbia over the 11-year period from October 1, 1981, to September 30, 1991, from the Division of Vital Statistics of the Province of British Columbia and Statistics Canada. Social, economic, and demographic information for the health units was obtained from census data and included the number of persons per household, proportion of the population that lived in 1-person households, immigration and migration rates for each region, proportion of the population with less than grade 9 education, proportion with less than grade 12 certification, marital status rates, unemployment rates by gender, average household income, average census family income, and labour-force participation rate by gender. We calculated overall and gender-specific suicide rates for elderly persons (65 years and older) and younger populations. Using Poisson regression analyses, we determined the cross-sectional and longitudinal relative risks associated with the socioeconomic variables for the units, and we also examined trends in suicide rates. RESULTS: There were 4630 suicides in the 11-year period. The mean suicide rate (per 100,000 population) for those over age 9 years was 18.6 (between health unit SD 5.2, 95% confidence interval [CI] = 17.0-20.2). The elderly have a higher suicide rate in every region. The male suicide rates (mean = 26.9, SD 6.4, 95% CI = 24.0-30.0) are higher than female rates (mean = 7.5, SD 1.7, 95% CI = 6.8-8.3) in every region. The factors influencing suicides were different for elderly males and elderly females. In all analyses, suicide rates in elderly females remained essentially stable across age-groups and units and over the years. Elderly male suicide rates varied across units and age-groups and over the years. CONCLUSIONS: Suicide rates are highest in males over age 74 years. There are regional differences in elderly suicide rates and the factors that influence them. Longitudinal and cross-sectional risk factors differ, and there are gender differences in the risk factors. For both elderly males and females, suicide rates appear to be influenced by social factors in the population as a whole, not just in the elderly population. Male and female employment patterns are associated with elderly male suicide rates, even though the latter are not in the labour force. For suicide in elderly women the important factors are population education, income, and migration levels.