Height and Weight
As noted under Animal Studies, energy restriction powerfully reduces mammary tumor incidence in rodents. This relationship is difficult to evaluate directly in humans, because estimates by adults of their energy intake, especially during childhood, are unlikely to be sufficiently precise, and any analysis would also need to account for physical activity with high accuracy.

However, because children who experience energy deprivation during growth do not attain their full potential height, attained height may be used as a proxy for childhood energy intake. Nevertheless, this is not a specific indicator, because protein restriction and genetic factors also affect stature. In Japan, for instance, a substantial increase in average height occurred during the twentieth century, presumably due to improved nutrition. Among countries, height is positively correlated with breast cancer rates, supporting the hypothesis that childhood and adolescent energy intake may influence breast cancer rates decades later.

Most of the case control and cohort studies of attained height and risk of breast cancer suggest a modest positive association. In a follow-up of 7,259 postmenopausal women in the Netherlands, a more than twofold increase in risk was observed for a 15-cm difference in height. In a follow-up of the National Health and Nutrition Examination Survey I (NHANES I) population, in which women at risk for malnutrition had been oversampled,a similar increase in risk was observed. In the NHANES I study, height was positively associated with later age at menarche (protective against breast cancer) and late age at first birth, low parity, higher socioeconomic status, and alcohol use (risk factors for breast cancer); these findings suggest that height may be confounded by other cancer risk factors. Controlling for these variables in multivariate analyses, however, had little influence on the association between height and breast cancer.

Among women in the Nurses’ Health Study, a significant positive association was seen between height and breast cancer among postmenopausal women but not among premenopausal women. Several large cohort studies have been conducted in Scandinavia, and in all of them significant associations have been observed, ranging from 1.1 (for a 5-cm increment) to 2.0 (for an increment larger than 8 cm). In the studies of Vatten and Kvinnsland, the positive trend between height and risk of breast cancer was most nearly linear in the birth cohort of women (1929 to 1932) who lived through their peripubertal period during World War II, a time in which food was scarce and average attained height was reduced. Collectively, these data provide convincing evidence that attained height is associated with a modest increase of breast cancer.

Age at menarche, an established risk factor for breast cancer, provides a second indirect indicator of energy balance during childhood. Nutritional factors - in particular, energy balance - appear to be the major determinants of age at menarche. In prospective studies among young girls, the major predictors of age at menarche were weight, height, and body fatness.

A marginally significant inverse association between dietary fat and age at menarche was seen in one study, but no relation was observed in the others. The potential for energy balance to influence breast cancer risk through age at menarche is greater than might be appreciated by examining the distribution of this variable in modern Western countries. Although the average age at menarche in these countries is now 12 to 13 years, in rural China the typical age has been 17 to 18 years, similar to that of Western countries 200 years ago. An effect of growth rate on breast cancer risk may begin even before birth, as an inverse relation between birth weight and breast cancer risk has been observed in one study of mainly premenopausal women.

The mechanisms by which age at menarche and attained height are related to risk of breast cancer are probably multiple. Early onset of menstrual cycles exposes the breast to ovarian hormones at a younger age and for a longer duration over a lifetime. Also, in several studies, an early menarche has been associated with higher estrogen levels at later ages.

Height has been suggested to be a surrogate for mammary gland mass, which may be related to higher risk. IGF-I could be a key intermediary, because this hormone is directly involved in regulation of growth during childhood, and plasma levels in premenopausal women have been associated with risk of breast cancer. The temporal nature of this relationship remains unclear, because an effect of IGF-I on breast cancer risk could be mediated entirely during the years of growth and development, with the association between adult levels and breast cancer risk only reflecting the correlation of blood levels over time. Alternatively, the blood IGF-I levels may have a temporally more proximate relation to risk, so that reductions in levels during adulthood would reduce risk of breast cancer. The determinants of blood IGF-I levels in humans are not known and may be both genetic and nutritional. In animals, energy restriction reduces IGF-I levels, and infusion of IGF-I appears to negate the effects of energy restriction on bowel tumors in mice.