# Prospective Effects of Self-Rated Health on Dementia Risk in Two Twin Studies of Aging

## TL;DR

This study finds that genetic factors largely explain the link between poor self-rated health and increased dementia risk, with a small independent effect observed in one sample.

## Contribution

The study uses twin data to disentangle genetic and environmental influences on the subjective health-dementia risk association.

## Key findings

- Genetic and confounding factors explain most of the effect of self-rated health on dementia risk.
- A small correlation between self-rated health and dementia risk remained after adjusting for genetics in the Danish sample.
- Subjective health ratings may still have predictive value for dementia risk despite genetic influences.

## Abstract

Subjective health ratings are associated with dementia risk such that those who rate their health more poorly have increased risk for dementia. The genetic and environmental mechanisms underlying this association are unclear, as prior research cannot rule out whether the association is due to genetic confounds. The current study addresses this gap in two samples of twins, one from Sweden (N = 548) and one from Denmark (N = 4,373). Using genetically-informed, bivariate regression models, we assessed whether additive genetic effects explained the association between subjective health and dementia risk as indexed by a latent variable proxy measure. Age at intake, sex, education, depressive symptomatology, and follow-up time between subjective health and dementia risk assessments were included as covariates. Results indicate that genetic variance and other sources of confounding accounted for the majority of the effect of subjective health ratings on dementia risk. After adjusting for genetic confounding and other covariates, a small correlation was observed between subjective health and latent dementia risk in the Danish sample (rE = − .09, p < .05). The results provide further support for the genetic association between subjective health and dementia risk, and also suggest that subjective ratings of health measures may be useful for predicting dementia risk.

The online version contains supplementary material available at 10.1007/s10519-024-10182-1.