Socioeconomic influences on late-life health and mortality : exploring genetic and environmental interplay

Abstract

The objective of this thesis was to increase the understanding of socioeconomic differences in health and mortality in old age - in a genetically informative setting. Data from the Swedish Twin Registry (STR), different statistical methods, and family-based designs were applied to investigate socioeconomic circumstances over the life-course and how these affect cognitive function, frailty, and mortality in late life.

In Study I, we studied the effect of rearing social class on late-life cognitive ability. An association between rearing social class and cognitive ability at age 65 was observed, but there was no effect on cognitive change. After controlling for familial influences, the association between rearing social class and late-life cognitive ability no longer remained and could instead be attributed to genetic influences.

In Study II, we used a classical twin design to investigate if childhood and attained socioeconomic indicators moderates the effects of genes and environment on late-life cognitive abilities. Cognitive ability was measured via four cognitive tests and a general ability score. Estimates of intercepts from growth models centered at age 75 and two linear slopes (before and after age 75) were utilized in the moderation models. The results from the moderator models for the four socioeconomic indicators showed similar patterns for the intercept. For cognitive change, moderation differed depending on cognitive test and socioeconomic indicator.

In Study III, we investigated mortality inequalities by comparing preventable and non-preventable mortality using a survival model. Familial confounding was analyzed using a co-twin control method. We observed a social gradient for mortality for the adult socioeconomic measures, which was stronger for preventable mortality than for non-preventable mortality. Adjustments for familial confounding did not change the observed associations between the attained socioeconomic indicators and mortality. However, the associations between rearing social class and mortality did not remain in the co-twin control analyses of the reared apart twins.

In Study IV, we explored the influence of attained socioeconomic indicators on frailty and mortality in men and women. Additional co-twin control analyses indicated familial confounding. Frailty was operationalized as the Frailty Index. There were robust sex differences in frailty. Socioeconomic influences on frailty were stronger for women than for men. In the co-twin control analyses, the effect remained the same for men, but for women the within-pair effect was strongly attenuated. No differences could be observed dependent on zygosity. The socioeconomic gradient in the relationship between frailty and mortality was stronger in men, but was not influenced by familial factors.