Background Risk factors measured in middle age do not fully explain the occurrence of future dementia risk. This raises the possibility that pre-adult risk factors may be implicated in its aetiology. Height is a proxy for early life circumstances (early life illness, adversity, nutritional privation, or psychosocial stress). Lower physical stature is associated with elevated cardiovascular disease (CVD) rates. Since CVD and dementia have a shared aetiology, we examined the little-explored association between height and later-life dementia. This individual participant meta-analysis is the largest study to date to examine this relationship.

Methods We pooled raw data from 18 country-wide, independent, prospective cohort studies comprising the Health Survey for England (N = 15) and the Scottish Health Survey (N = 3). Between 1994 and 2008, a total of 181,800 people aged 16 years or older participated in baseline surveys. Height was measured to the nearest millimetre with shoes removed with the study member stretching to their maximum height. Covariate data, including health behaviours, weight, self-reported general health, and socioeconomic status were quantified using standard protocols. Cause-specific mortality was ascertained through linkage to national mortality registers

Results A mean follow-up of 10 years gave rise to 426 dementia deaths in men and 667 in women. There was a dose-response association between shorter stature and risk of dementia death that was stronger for men (age- and sex-adjusted HR per standard deviation decrease in height 1.26 [95% CI 1.14, 1.40]) than women (1.10 [1.14, 1.19]). Incorporating measures of adulthood socioeconomic status, health, and health behaviours into the multivariable model had a negligible effect on the strength of these associations

Conclusion Early life circumstances, as indexed by adult height, may influence later dementia risk. These findings suggest that improved early life conditions may decrease dementia risk.