Upper abdominal obesity, insulin resistance and breast cancer risk.

PURPOSE: A majority of prospective studies show breast cancer risk to be higher in obese postmenopausal women with upper abdominal adiposity than in those with overall adiposity. The evidence is more limited and inconsistent in the case of premenopausal women. The review examines evidence that aberrant insulin signalling may be involved in the promotion of mammary carcinogenesis. The aetiology and concomitants of abdominal visceral obesity are examined. MECHANISMS: Clinical and experimental evidence suggests that the higher breast cancer risk associated with greater abdominal visceral obesity may be related to aberrant insulin signalling through the insulin receptor substrate 1 pathway, leading to insulin resistance, hyperinsulinaemia and increased concentrations of endogenous oestrogen and androgen. The putative role of aberrant insulin signalling in the promotion of mammary carcinogenesis may help to explain clinical relationships between breast cancer risk and age at menarche, pregnancies and onset of obesity.
CONCLUSION: Overall adiposity in women adversely affects breast cancer risk mainly by greater exposure of mammary epithelial tissue to endogenous oestrogen. Upper abdominal adiposity appears to involve an additional effect related to the presence of insulin resistance. Aetiological factors in the development of hyperinsulinaemic insulin resistance are still uncertain but may involve aberrant susceptibility genes in adipocyte insulin receptors or in the insulin receptor substrate 1 pathway. Epigenetic factors are also likely to contribute, including high free fatty acid levels and obesity. Dietary fatty acids, particularly polyunsaturated fatty acids, are known to regulate adipocyte differentiation through the nuclear peroxisome proliferator-activated receptor gamma, and may also have a role in insulin resistance. These aetiological factors are likely to be relevant to the high risk of postmenopausal breast cancer in industrialised Western populations.