LAUSR

For approximately a half-century, epidemiologists have pursued the links between diet and cancer risk, using case–control and cohort studies complemented by findings of experimental and animal studies. The resulting literature is voluminous and numerous associations of cancer risk with foods and methods of food preparation and with microand macronutrients have been found, leading to recommendations for prevention. Given persistent uncertainties about diet and cancer, however, research continues, based in well-established and documented cohorts and in case–control studies of specific cancer sites. This issue of the International Journal of Cancer includes two articles that indicate a further complexity in the association between diet and cancer risk: the timing of meals in relation to sleep. Kogevinas and colleagues completed population-based case–control studies of breast and prostate cancer in Spain. Diet was assessed with a food frequency questionnaire and questionnaires were also used to evaluate the timing of meals in relation to sleep and general characteristics of sleep. The findings were parallel for breast cancer and prostate cancer: a lengthier interval between supper and nocturnal sleep was associated with reduced risk for cancer. Srour and colleagues report comparable findings from the NutriNetSante0 cohort study in France, which included 428 incident breast cancers and 179 incident prostate cancers. Risk was increased for those eating supper later; for breast cancer the hazard ratio (HR) was 1.48 (95% confidence interval (CI) 1.02–2.17) for those having supper after 9:30 p.m. versus earlier, while the HR for prostate cancer was 2.20 (95% CI: 1.28–3.78). The findings of the two studies are similar, but based on different designs and quite distinct study populations. Bias cannot readily be offered as an explanation. The sleep-related exposure variables are subject to misclassification, likely random rather than differential, and potential confounding has been addressed, even though relevant confounders cannot be readily proposed. The findings are novel and raise questions as to how risk of cancer could be affected by the timing of meals in relation to sleep. In explaining their findings, the authors of both papers turn to circadian rhythms and the timing of meals in relation to the sleep–wake cycle. Disruption of circadian rhythm by shift-work has already been classified as a possible carcinogen by the International Agency for Research on Cancer (IARC) with the relevant epidemiological evidence largely related to breast cancer. A recent meta-analysis, based on 28 studies of breast cancer, found an overall relative risk (RR) of 1.14 [95% confidence interval (CI) 1.08–1.21] for circadian disruption. The link between timing of meal consumption and cancer risk could also be mediated through the effects of circadian rhythm on glucose and lipid metabolism. Alterations in the circadian rhythm, with accompanying disturbances in energy intake, have been associated with adverse effects on glucose metabolism leading to increased predisposition for insulin resistance, glucose intolerance, and type 2 diabetes. Variability in feeding patterns, particularly if desynchronized with the intrinsic circadian clock, can disturb energy homeostasis, impair glucose homeostasis and heighten the risk of metabolic sequelae. Consequently, it is not surprising that time-restricted feeding with consumption of more energy throughout the day is preferable to evening consumption, particularly for metabolic health. Thus, delayed meal timing and close proximity of the evening meal to the sleep period could provide the metabolic milieu for increased risk for development of cancer. In fact, the greater propensity for cancer in people with type 2 diabetes has been well known for decades. Carcinomas of the breast, pancreas, liver, colon, and endometrium are more prevalent in people with type 2 diabetes than people without. Data accumulated over a 40-year period in a Swedish registry show that the incidence of hepatic, pancreatic, esophageal, colorectal, cervical, endometrial, ovarian, and renal carcinoma is higher in people with type 2 diabetes than in the general population. Although the association with specific cancer types has not been consistent across available studies, the higher risk of cancer in type 2 diabetes has been documented in cohort studies from Korea, Israel, Japan and the United States. Even in the absence of overt type 2 diabetes, disorders of glucose metabolism, such as hyperglycemia and insulin resistance, have been reported to increase the risk for cancer. Elevated glucose levels have been linked to a higher risk of death from gastric, colorectal, hepatic, and pancreatic cancer and the associated risk increases linearly with glucose concentrations. Hyperglycemia can promote neoplastic transformation and/or increase the risk of progression of existing cancer, possibly through the formation of intracellular glycation endproducts which alter signaling pathways and lead to DOI: 10.1002/ijc.31838 Correspondence to: Naresh M. Punjabi, Johns Hopkins University School of Medicine, Baltimore, MD, E-mail: [email protected] International Journal of Cancer IJC