Good News for “Alice”: Height and Sex Differences in Cancer Risk

Abstract

In Lewis Carroll’s Alice’s Adventures in Wonderland, the eponymous heroine falls down a rabbit hole and finds a flask labeled “Drink Me” (1). Following this instruction, Alice shrinks to a size small enough to fit through the door to the mysterious world of Wonderland (Figure 1). Let us imagine another, less fanciful Alice: a woman of less-thanaverage height, who if she is both sensible and lucky is a neversmoker, drinks alcohol only in moderation, and is able to maintain a healthy weight through diet and exercise. We do not fully understand why, but the good news for our Alice is that as a shorter woman with a healthy lifestyle she has a lower risk of cancer than most of her taller and/or male peers. In this issue of the Journal, Walter and colleagues (2) bring together these two mysteries of cancer epidemiology: a greater incidence among men than women of cancer at shared anatomic sites, and the association of greater height with increased risks for many cancer sites in both sexes. Some of the reasons why Alice and other women have a lower incidence of cancer than men are already understood. Leaving aside cancers that are sex-specific due to anatomical differences between women and men, known environmental risk factors including alcohol intake, smoking, and occupational exposures to carcinogens are likely to contribute to sex differences in cancer risk at several shared sites (3). For example, in most populations the prevalence of smoking has been lower in women than men, and therefore women have had a lower rate of smoking-related cancers such as lung cancer. As smoking patterns of men and women have become more similar in developed countries, the disparity between the sexes in lung cancer risk has largely disappeared (3–5). However, cancer incidence at several other sites is greater among men than women by 50% or more, a finding that is consistent across countries at different stages of economic development but which cannot easily be explained by known risk factors (3). Even less is known about why our Alice’s height should be associated with her risk of cancer. Overall cancer risk increases by 10% to 15% per 10 cm (4 in) of height in both men and women, again consistent across different countries (6). Adult height can be measured accurately in middle age, but it is a marker of developmental processes and exposures that occur in early life and has been linked to a very large number of genetic (7) and environmental (8) factors. It is unknown whether a relatively small number of early-life factors might be conspiring to produce the height–cancer association through multiple mechanisms, or if instead the large number of genetic and environmental determinants of height might influence cancer risk through a single, intermediate mechanism. For example, there has been considerable interest in insulin-like growth factor 1 (IGF-1), a correlate of growth in childhood and of risk for some cancers (9). But taller people may simply have more cells, or it may be that many determinants of growth during normal development (perhaps including IGF-1) also have general effects on tumor growth (8). Given the obvious relationship between height and sex, it is surprising that height and sex differences in cancer risk seem not to have been investigated together before. Walter et al. looked at whether height could statistically account for sex differences in cancer risk in the Vitamins and Lifestyle (VITAL) Study, a cohort of approximately 33 000 women and 32 000 men (2). They found that differences in height might account for a third to a half of the excess risk in men of cancers at shared anatomic sites. As with most previous studies, power was limited for specific cancer sites, but the cancers for which height accounted for a large proportion of the sex differences in risk (kidney cancer, melanoma, and hematological malignancies) have previously been found to be associated with height (6).