UNEMPLOYMENT, SOCIODEMOGRAPHIC BACKGROUND AND CONSUMPTION OF ALCOHOL BEFORE AND DURING THE ECONOMIC RECESSION OF THE 1990S IN FINLAND

Background Some studies suggest that people's alcohol consumption incr eases during unemployment whereas others suggest the opposite. All stu dies, however, deal with situations marked by relatively low national unemployment rates. We studied alcohol use among individuals in relati on to unemployment, education, marital status and sex during times of both low and high unemployment in Finland. Methods A group of 44 391 r espondents, aged 18-64 years, from nationally representative, consecut ive annual samples of 5000 people from 1982 to 1995 was utilized. Over all response rate for men was 77% and for women 80%. Results Univariat e analyses indicated that unemployment was associated with the amount of reported alcohol use. However, when logistic regression was used to analyse interactions between alcohol consumption, unemployment, educa tion and marital status, the picture changed. During a low unemploymen t period (e.g. 1982-1990), being unemployed was not associated with th e upper consumption level of alcohol use (defined as greater than or e qual to 8 drinks/week for men, greater than or equal to 5 for women); nor was it during a high unemployment period (1991-1995), except among single people. During a high unemployment period poorly educated, sin gle, unemployed men (odds ratio [OR] = 1.6, 95% confidence interval [C I] : 1.1-2.4), showed a significantly higher risk of upper level of al cohol consumption than otherwise similar but employed men (OR = 0.8, 9 % CI: 0.6-1.0). The reference group consisted of highly educated, marr ied, employed men who did not exceed the upper drinking limit. Similar ly, the risk of upper consumption level drinking was significantly hig her among highly educated, unemployed single women (OR = 2.4, 95% CI: 1.388-4.3) than among otherwise similar but employed women (OR = 1.1, 95% CI:1.0-1.386). Conclusion Thus, unemployment was weakly but signif icantly related to the upper consumption level of alcohol use among si ngle people during the recession but not in the preceding period of ec onomic growth.