The contribution of sleep to social inequalities in cardiovascular disorders: a multi-cohort study

Abstract

Aims: Sleep disturbances exhibit a strong social patterning, and inadequate sleep has been associated with adverse health outcomes, including cardiovascular disorders (CVD). However, the contribution of sleep to socioeconomic inequalities in CVD is unclear. This study pools data from eight European cohorts to investigate the role of sleep duration in the association between life-course socioeconomic status (SES) and CVD.

Methods and results: We used cross-sectional data from eight European cohorts, totalling 111 205 participants. Life-course SES was assessed using father's and adult occupational position. Self-reported sleep duration was categorized into recommended (6-8.5 h/night), long (>8.5 h/night), and short (<6 h/night). We examined two cardiovascular outcomes: coronary heart disease (CHD) and stroke. Main analyses were conducted using pooled data and examined the association between life-course SES and CVD, and the contribution of sleep duration to this gradient using counterfactual mediation. Low father's occupational position was associated with an increased risk of CHD (men: OR = 1.19, 95% CI [1.04; 1.37]; women: OR = 1.25, 95% CI [1.02; 1.54]), with marginal decrease of the gradient after accounting for adult occupational position (men: OR = 1.17, 95% CI [1.02; 1.35]; women: OR = 1.22, 95% CI [0.99; 1.52]), and no mediating effect by short sleep duration. Low adult occupational position was associated with an increased risk of CHD in both men and women (men: OR = 1.48, 95% CI [1.14; 1.92]; women: OR = 1.53, 95% CI [1.04; 2.21]). Short sleep duration meaningfully contributed to the association between adult occupational position and CHD in men, with 13.4% mediation. Stroke did not exhibit a social patterning with any of the variables examined.

Conclusion: This study suggests that inadequate sleep accounts to a meaningful proportion of the association between adult occupational position and CHD, at least in men. With sleep increasingly being considered an important cardiovascular risk factor in its own terms, our study additionally points to its potential role in social inequalities in cardiovascular disease.

Keywords: Cardiovascular disorders; Life-course; Mediation; Sleep duration; Socioeconomic status.

Figure 1

Directed acyclic graphs representing the counterfactual mediation model for the association between SES indicators and cardiovascular outcomes, mediated by sleep duration. (A) NDE, Natural direct effect: Effect of the predictor (SES) on the main outcome (CVD), through pathways which do not involve the mediator (sleep duration). (B) NIE: Natural indirect effect: Effect of the predictor (SES) on the main outcome (CVD), through pathways which involve the mediator (sleep duration). (C) Confounding effects by covariates. COV, covariates (age, cohort, study period, health behaviours, flexible working hours); SES, Adult/Father’s occupational position; M, mediator—sleep duration; CVD, cardiovascular disorders; MTE, marginal total effect of the predictor (SES) on the main outcome (CVD): NDE + NIE (not represented). This figure was realized with MS Office-Excel.

Figure 2

Counterfactual mediation estimates for the association between SES indicators and cardiovascular disorders, mediated by short sleep duration (<6 h/night), using pooled cohort data. (A) Association between father’s occupational position and CVD, adjusted for age, cohort, study period, flexible working hours and health behaviours. (B) Association between father’s occupational position and CVD, adjusted for age, adult occupational position, cohort, study period, flexible working hours and health behaviours. (C) Association between adult occupational position and CVD, adjusted for age, cohort, study period, flexible working hours and health behaviours. Sample size (A, B, C): Men: N=36 987 CHD, N=36 759 stroke; Women: N=72 863 CHD, N=72 819 stroke. CHD, coronary heart disease; MTE, marginal total effect (OR 95% CI); NDE: Natural direct effect (OR 95% CI); NIE, natural indirect effect (OR 95% CI); PM, proportion of the association between occupational position and cardiovascular disorders which is mediated by short sleep duration (*, significant mediation; lower and upper arrow indicate that CIs extend beyond the limits of the graph). This figure was realized with MSOffice-Excel.