Changes in alcohol drinking and subsequent sickness absence.

AIMS: The aim was to examine whether changes in alcohol drinking are associated with sickness absence.
METHODS: Repeated postal questionnaires on alcohol drinking were conducted among employees of the City of Helsinki in 2000-2 and 2007 to assess changes in drinking habits between these two time points. Data on the number of self-certified and medically confirmed sickness absences were derived from the employer's register. Sickness absences were followed from 2007 until the end of 2010 among employees participating in both questionnaire surveys. The study includes 3252 female and 682 male employees 40-60 years old at baseline. Poisson regression was used in the data analysis and population attributable fractions (PAFs) were calculated.
RESULTS: Alcohol drinking was associated especially with self-certified sickness absence. Rate ratios (RRs) and 95% confidence intervals (CIs) for increasing weekly average drinking were 1.38, 1.18-1.62 among women and 1.58, 1.18-2.12 among men. Also stable problem drinking (for women 1.39, 1.26-1.54, for men 1.44, 1.10-1.87) and among women stable heavy drinking (1.53, 1.20-1.94) increased self-certified sickness absence. There were associations between alcohol drinking and medically confirmed sickness absence but these were mainly explained by health and health behaviours. Also, a decrease in weekly average drinking was associated with sickness absence among women whereas among men former problem drinking increased sickness absence. According to the PAF values, problem drinking had a stronger contribution to sickness absence than weekly average drinking.
CONCLUSIONS: Alcohol drinking is particularly associated with self-certified sickness absence. Reducing adverse drinking habits is likely to prevent sickness absence.