Re: Night shift work, light at night, and risk of breast cancer.

Two studies recently published in the Journal showed that disruption of diurnal sleep–wakefulness rhythms, especially through nighttime shift work, was associated with higher subsequent breast cancer incidence (1,2). Both of the articles and the accompanying editorial cited melatonin suppression as a likely mediator via reduced suppression of ovarian estrogen production due to light exposure during the time when melatonin would otherwise be highest. It has also been postulated that melatonin serves as an antioxidant in tumor cells and stimulates cytokine release from activated T cells. The relationship between sleep disruption and breast cancer risk may alternatively be explained by other risk factors for breast cancer that are associated with sleep disruption. For example, our laboratory has recently reported, also in the Journal, that loss of normal diurnal variation in cortisol predicts early mortality in metastatic breast cancer (3). Cortisol normally shows marked diurnal variation, peaking in early morning and declining throughout the day. In our study, patients with flattened salivary cortisol rhythms or aberrant peaks and troughs suffered earlier mortality. The prognostic effect of cortisol rhythm on survival emerged approximately 1 year after cortisol assessment and extended at least 7 years after the assessment (3). A similar association between 24-hour rest–activity rhythms and survival has been noted in patients with colorectal cancer (4). Furthermore, patients at high risk for primary breast cancer show abnormal circadian patterns among an array of hormones including cortisol (5). Similar endocrine disturbances have been linked with advancing age and with sleep loss in young subjects (6). In our study, patient self-reports of sleep disruption were associated with flattening of the diurnal cortisol rhythm (3). There is evidence that the early morning transition from dim to bright light not only suppresses melatonin secretion; it also induces a marked elevation of cortisol levels (7). Thus, sleep disruption coupled with exposure to light at night may not only suppress the melatonin peak but may also interfere with the normal cortisol nadir. Although mere sleep disruption was not associated with elevated breast cancer risk in the Davis et al. study (2), it is likely that shift workers have altered diurnal cortisol patterns engendered through alterations of sleep and nutritional patterns, abnormal photic stimuli, and altered rest–activity cycles. In addition, the stress incurred from frequent transitions between daytime and nighttime shift work should not be overlooked. Indeed, chronic sleep debt has been linked with the disruption of carbohydrate metabolism and thyroid hormone function, elevation of sympathetic nervous system activity, as well as with hypothalamic–pituitary–adrenal axis dysfunction. Because these effects are similar to those seen in normal aging, it is reasonable to propose that sleep debt may increase the incidence or severity of age-related diseases, including cancer, by processes similar to those seen with advancing age. Thus the association of sleep disruption with cancer risk may be mediated by a number of pathways, including, but not limited to, the disruption of melatonin secretion. There are now sufficient data to support the notion that the loss of normal diurnal variation in cortisol is another possible mechanism that could account for the observed association between shift work and breast cancer incidence.