Abstract
Men have a higher risk than women for most cancers affecting both sexes. Since taller stature is associated with increased cancer risk and men are, on average, taller than women, we investigated to what extent adult body height, as a proxy for stem cell number, explains the elevated cancer risk in men. This population-based cohort study linked adult height information from Swedish conscripts, mothers, and passports to the National Cancer and Cause of Death Registers (1960-2011). We used mediation survival analysis to estimate the proportion of the association between male sex and site-specific cancer risk mediated by adult height, our main outcome. Statistical significance was assessed using two-sided tests with a .05 significance level. Among 6,156,659 adults, we observed 285,778 non-sex-specific cancer cases. Male sex was significantly associated with cancer risk at 33 of 39 sites, and greater height with increased risk at 27 of 39 sites. Height mediated 0.5%-100% of the excess male cancer risk, with the highest proportions for salivary gland cancer, colon cancer, melanoma, and acute myeloid leukemia. The effects of height and its mediated effect were most consistent for malignancies with weak or no established environmental risk factors. These findings indicate that a substantial proportion of the excess cancer risk in men may be explained by height. This highlights the role of stochastic biological processes linked to height, as well as genetic and early-life determinants of height, in contributing to sex differences in cancer risk, beyond influences of adult lifestyle and environmental exposures.