14-year diabetes incidence: the role of socio-economic status

Abstract

Background: Diabetes prevalence is associated with low socioeconomic status (SES), but less is known about the relationship between SES and diabetes incidence.

Data and methods: Data from eight cycles of the National Population Health Survey (1994/1995 through 2008/2009) are used. A sample of 5,547 women and 6,786 men aged 18 or older who did not have diabetes in 1994/1995 was followed to determine if household income and educational attainment were associated with increased risk of diagnosis of or death from diabetes by 2008/2009. Three proportional hazards models were applied for income and for education--for men, for women and for both sexes combined. Independent variables were measured at baseline (1994/1995). Diabetes diagnosis was assessed by self-report of diagnosis by a health professional. Diabetes death was based on ICD-10 codes E10-E14.

Results: Among people aged 18 or older in 1994/1995 who were free of diabetes, 7.2% of men and 6.3% of women had developed or died from the disease by 2008/2009. Lower-income women were more likely to develop type 2 diabetes than were those in high-income households. This association was attenuated, but not eliminated, by ethno-cultural background and obesity/overweight. Associations with lower educational attainment in unadjusted models were almost completely mediated by demographic and behavioural variables.

Interpretation: Social gradients in diabetes incidence cannot be explained entirely by demographic and behavioural variables.