Educational and income inequalities across diseases in Denmark: a register-based cohort study

Abstract

Background

Educational attainment and income are often, but not always, associated with disease incidence. Existing research typically examines single diseases, resulting in disparate analyses with little comparability. In this study, we aimed to assess educational and income inequalities across diseases in Denmark.

Methods

This register-based study included all Danish residents aged at least 30 years between Jan 1, 2013, and Dec 31, 2022, who were born after 1920, and with known educational attainment and income. We used a disease-wide approach to assess associations between education and income and the incidence of 751 diagnostic codes determined upon hospital admission. We estimated age-standardised incidence rates and incidence rate ratios (IRRs) using Poisson regression, adjusted for birth cohort and stratified by sex. Participants were followed up until time of diagnosis, death, emigration, or until Dec 31, 2022.

Findings

4 541 309 individuals aged 30 years and older were registered as living in Denmark between Jan 1, 2013, and Dec 31, 2022. 121 083 were excluded due to limited or missing information about educational attainment. 4 420 226 individuals were included in the analysis of educational inequalities (2 232 200 [50%] were female and 2 188 026 [50%] were male). 23 708 were excluded due to absence of income information, and 4 396 518 were included in the analysis of income inequalities (2 223 217 (51%) were female and 2 173 301 (49%) were male). Socioeconomic differences in incidence rates were observed across all disease groups; incidence rates of most diseases decreased with higher educational attainment and income. The magnitude of the socioeconomic inequalities varied substantially. Among non-communicable diseases, the strongest positive association with regard to education was observed in chronic obstructive pulmonary disease for female individuals (low education vs high education, IRR 2·7 [95% CI 2·4–3·0]) and schizophrenia for male inividuals (low education vs high education, IRR 4·4 [2·2–8·8]), and the strongest negative association was in melanoma and other skin cancers for females (low education vs high education, IRR 0·7 [0·7–0·8]) and melanoma and skin cancers for males (low education vs high education, 0·7 [0·6–0·8]). With regard to income, for females, the strongest positive association was observed in schizophrenia (quartile 1 [Q1] vs quartile 4 [Q4], IRR 10·1 [6·1–17·2]), whereas the strongest negative association was in melanoma and other skin cancers (Q1 vs Q4, IRR 0·5 [0·5–0·6]). For males, the strongest positive assocation was schizophrenia (Q1 vs Q4, IRR 18·4 [95% CI 8·5–39·9]) and the strongest negative association was also melanoma and other skin cancers (Q1 vs Q4, IRR 0·5 [0·5–0·6]). The most prevalent disease category, other digestive diseases, was also strongly positively associated with education (low education vs high education, IRR 1·6 [95% CI 1·6–1·6] for females; IRR 1·5 [1·4–1·5] for males) and income (Q1 vs Q4, IRR 1·5 [1·5–1·5] for females; IRR 1·3 [1·3–1·4] for males).

Interpretation

Our study provides a detailed representation of the association between two socioeconomic indicators and disease incidence. A broad spectrum of diseases, and not only the most prevalent, show socioeconomic disparities. This finding highlights the need for not only policies that address specific diseases, but also universal policies addressing the root causes of socioeconomic disparities and their health consequences.

Funding

The Novo Nordisk Foundation.