Abstract
Der Heide et al., 2013) and mental health (Howard et al., across socioeconomic groups, this might affect the extent to
Background: Although it is known that health literacy (HL) plays an explanatory role in educational inequali- 2006). which HL explains socioeconomic inequalities in health in ties in health, it is unknown whether this role varies across age groups.
Objective: The purpose of this study Although various studies have investigated the mediatdifferent age groups. was to investigate whether the mediating role of HL in educational inequalities in four health outcomes varing role of HL in socioeconomic inequalities in health (Stor- Third, over the last decades, individuals are increasingly ies across age groups: age 46 to 58 years, age 59 to 71 years, and age 72 to 84 years.
Methods: We used data macq et al., 2019), none have addressed whether this role expected to manage their own health (Du & Yuan, 2010; from the Dutch Doetinchem Cohort Study, which included 3,448 participants. We included years of education varies across age groups. However, there are various reasons Geboers, 2017). Much more than in the past, care systems are as predictor, chronic illness prevalence and incidence, mental and self-perceived health as outcomes, and for why this may be the case. First, evidence shows that relacharacterized by shared decision-making and the expecta- HL, based on self-report, as mediator. We used multiple-group mediation models to compare indirect effects tive socioeconomic inequalities in health are largest between tion that patients play a proactive role in their own health and across age groups. Key
Results: In the complete sample without age stratification, HL partly mediated the ages 50 and 60 years and decrease thereafter (Dupre, 2007; health care (Bodenheimer et al., 2002; Epstein & Street, 2011). effect of education on all health outcomes except for incidence of chronic diseases. These indirect effect esti- Huisman et al., 2004). Explanations for this fact include vari- Citizens are nowadays more intensively informed about premates were larger for subjective (self-perceived health, proportion mediated [PM] = 37%, and mental health, ation in the distribution of risk factors across birth cohorts vention and health behaviors and expected to use this infor- PM = 37%) than for objective health outcomes (prevalence of chronic disease, PM = 17%). For the prevalence (i.e., smoking and alcohol misuse) and selective survivormation to monitor and optimize their health (Mackenbach, of chronic disease, the indirect effect estimate was significantly larger among individuals age 46 to 58 years ship (i.e., lower educated individuals are more likely to die at 2012). Nevertheless, current younger-age groups may be compared to individuals age 59 to 71 years and for incidence of chronic disease also compared to individuals younger ages than higher educated individuals, thereby remore intensively exposed and receptive to this development a