Weight gain in relation to cancer risk

Abstract

INTRODUCTION: Obesity has been recognized as an increasing problem in the western countries. Obesity, determined as a single measurement of weight (and height) at a certain moment, has already been associated with an increased risk for several cancers. Nevertheless, weight change seems to be a more appropriate measure for adiposity in humans. AIM: Systematically review the effect of weight gain (throughout life) on the risk of several types of cancers. MATERIAL AND METHODS: We selected 58 studies, which investigated the relation of weight gain with breast cancer or/and ovarian cancer or/and endometrial cancer or/and prostate cancer or/and colorectal cancer. Furthermore, of all the studies we evaluated the study design, the characteristics of the study populations, the method of exposure assessment, specific cancer outcomes, confounding factors and effect modifiers. RESULTS: The studies discussed differed with respect to study design, sample size and confounding factors (within a certain cancer outcome). The majority of studies assessed weight gain information retrospectively (by recall), and were entailed by numerous biases. Only a few studies assessed weight gain prospectively. However, this studies only investigated the effect of short-term weight gain. Increases of risk for post-menopausal cancers in relation to weight gain were reported in many studies, however only in non-HRT users. Studies which also stratified on menopausal status, did not find this effect in pre-menopausal women. For the other cancers discussed, no consistent increases or decreases in risk were found. CONCLUSION: Weight gain seems to increase the risk for post-menopausal breast cancer. In contrast, weight gain does not seem to change the risk for pre-menopausal breast cancer. Ovarian cancer, endometrial cancer, prostate cancer and colon-rectal cancer were not studied as extensively as breast cancer in relation to weight gain. Before drawing any conclusions about these type of cancer, more large, prospective studies are needed.