3The contemporaneous association of socioeconomic status (SES) with health is well-established, whereas much 4 less is known about the health-related effects of social mobility (i.e., movements across different SES). This 5 study investigates the impact of SES in childhood and adulthood on health satisfaction across the life course.
6Using data from the German Socio-Economic Panel (SOEP) and education as a central marker of SES, we test 7 whether parental education (i.e., childhood SES) affects adult health satisfaction, directly and/or indirectly 8 through own educational attainment (i.e., adult SES) as a mediating variable. Moreover, we apply diagonal 9 reference models to disentangle the independent effect of intergenerational educational mobility. Our findings 10 show that parental education has both direct and indirect effects. Yet, the relative weight of parents' education as 11 a predictor of health satisfaction is found to depend on when in the life course health satisfaction is measured: 12 parental education shows an increasing relevance as a predictor of health satisfaction at higher ages. On top of 13 (additive) effects of parental and own education, we find significant mobility effects in earlier adulthood: upward 14 educational mobility is conducive to health satisfaction and the reverse for downward educational mobility.
15
Introduction
16A higher socioeconomic status (SES) is conducive to health and healthy ageing; the better educated, those in 17 higher occupational positions, and those who can dispose of greater financial resources tend to enjoy better 18 health at all ages. Whereas this positive relationship between current SES and health is well-established (Eikemo 19 et al., 2008; Mackenbach et al., 2008; Marmot and Wilkinson, 1999; Sanderson and Scherbov, 2014), much less 20 is known about the health effect of socioeconomic trajectories across the life course. Most available research on 21 the impact of childhood SES suggests that childhood conditions matter a great deal for adult health. Children of 22 parents at a low SES are at a higher risk of an earlier onset and a faster progression of functional health problems 23 in later life (Agahi et al., 2014) and of developing and dying from adult cardiovascular disease (Blackmore and 24 Ozanne, 2015; Galobardes et al., 2006; Poulton et al., 2002). The mortality risk from some cancers is also higher among those of lower SES in childhood (Galobardes et al., 2008). Based on such observations, Hayward and 1 Gorman (2004) put forward the notion of a 'long arm of childhood', i.e., of an enduring influence of early life 2 conditions on adult morbidity and mortality. While the statistical association between childhood SES and adult 3 health is undisputed, a crucial question is whether childhood conditions have a direct and lasting effect on adult 4 health, or whether the impact of childhood conditions is only indirect, insofar as SES in childhood has a defining 5 impact on later life-course trajectories and health exposures. Prior research shows m...