Background Hazardous alcohol consumption and unemployment are both associated with high mortality in Russia. Unemployment adversely affects mental and physical health. The relationship between alcohol and employment is complex as alcohol consumption may lead to unemployment but unemployment may also increase hazardous consumption. Very few longitudinal studies have investigated the effects of alcohol use on employment status. The objectives of this study were to investigate alcohol intake, drinking behaviour and drinking patterns as predictors of employment status in a longitudinal study of working-age men in Izhevsk, Russia.

Methods Participants were 1217 men aged 25–54 resident in Izhevsk, Russia in regular paid employment at baseline (2003–6) who were followed up and re-interviewed (2008–9). Alcohol use was measured by total volume of ethanol consumed from beverage alcohol, drinking behaviour (abstainers, beverage alcohol drinkers with no problem drinking behaviour, beverage alcohol drinkers with problem drinking behaviour and non-beverage alcohol drinkers (e.g. eau de cologne)), and three aspects of drinking pattern (drinking spirits without food, drinking alone and drinking before noon). Problem drinking was defined as twice weekly or more frequency of hangover, excessive drunkenness, sleeping in clothes because of drunkenness or one or more episodes of zapoi in the past year (defined as continuous drunkenness of two or more days during which participants were withdrawn from normal life). Logistic regression was used to investigate whether alcohol use at baseline predicted employment status at follow up after adjusting for socio-demographic variables.

Results Total volume of ethanol from beverage alcohol at baseline did not predict whether men were still in regular paid employment at follow-up (P value for linear trend=0.21). Problem drinkers (adjusted odds ratio 2.82 95% CI 1.23, 6.46) and non-beverage alcohol drinkers (adjusted odds ratio 2.42 (95% CI 1.18, 4.93) were more likely not to be in regular employment at follow-up than non-problem beverage alcohol drinkers. Among drinkers drinking spirits without food (adjusted odds ratio 2.13 95% CI 1.28, 3.54), drinking alone (adjusted odds ratio 1.65 95% CI 1.09, 2.50) and drinking before noon (adjusted odds ratio 2.49 95% CI 1.65, 3.78) all predicted employment status at follow-up.

Conclusion Drinking behaviour and drinking patterns predicted future employment status in a sample of working-age men in Izhevsk Russia. However the volume of alcohol consumed per year did not, suggesting how alcohol is consumed and whether it leads to problem behaviour is more important than the overall amount consumed when considering the impact of alcohol on employment.