Toward chronobiologic pattern discrimination of the risk of developing breast cancer and other diseases

With a view of the prevention as well as treatment of cancer and other diseases, it is important to quantify health positively and on an individual basis by chronobiologic methods. These include the assessment of the characteristics of certain circadian, circannual and other endocrine rhythms and trends, for the recognition of risk (prior to the occurrence of a given disease) by an alteration of the same rhythm characteristics and even by a time-specified single sample. With this aim in mind, a small number of selected (rather than randomly picked) women of 3 age groups was extensively sampled for 12 plasma hormones around the clock and the calendar, in 2 geographic locations. Such data revealed correlations of the familial risk of developing breast cancer with the circannual amplitudes of circulating prolactin and TSH. The risk of several other conditions was also correlated with hormonal rhythm characteristics; for example, the risk of developing diseases associated with a high blood pressure was correlated with the circannual amplitude of plasma aldosterone. The mapping of circannual characteristics, however, is time-consuming and costly and may not be warranted as a first step, for example when a physician is not in a position to wait for a year to make a diagnosis. With the possibility in mind that sampling requirements may be reduced to one or at most two samples, a chronobiologic pattern discrimination analysis was undertaken on the original data from young adults. The results are presented to indicate the method and to suggest the singling-out of certain variables for further testing on a larger, properly stratified and randomized sample, rather than as definitive results. Different classifiers and different corresponding reference values from variables that undergo circadian and circannual rhythms may perhaps withstand the test (and, with Vergil, the tooth) of time. If so, reference values that are time-specified may well prove to be a sine qua non in the assessment of certain neuroendocrine aspects of developing certain diseases, including breast cancer.