Components of the illness flexibility model as explanations of socioeconomic differences in sickness absence

The authors studied the social gradient in sickness absence in relation to components of the illness flexibility model, which highlights conditions affecting whether people attend work when they are ill. The conditions studied were: adjustment latitude, attendance requirements at work, stimulating work, and health. The population sample was part of a panel originating in 1994 when 15,154 inhabitants of Stockholm County were randomly selected to receive a questionnaire covering, among other things, health and psychosocial conditions. New questionnaires were sent to the respondents in 1998 and 2002. This article analyzes the 2002 data, for 1,634 women and 1,063 men. These respondents had reported being employed or on leave of absence. In this sample, a social difference was found in sickness absence of 31 days or more per year. For manual workers, women had an odds ratio (OR) of 2.8 and men an OR of 3.4 for such absence compared with nonmanual workers of both sexes in high socioeconomic positions. All single potential confounders decreased these ORs. Adding all characteristics decreased the OR by 78 percent for women and 67 percent for men. The results indicate that the social gradient in sickness absence is due to differences in health and in working conditions when one is ill.