Abstract
Breast tumor incidence is compared in C3CF1 mice, receiving at one of 6 different circadian stages, daily injections of melatonin or of a vehicle or no treatment, the latter in a slightly cooler environment. Results are summarized when approximately 33% of all animals have a breast tumor. Vehicle treatment increases tumor incidence. This environment-handling-vehicle effect is reduced by injecting melatonin, as indicated by life table analysis. A circadian-stage-dependence of the melatonin effect is suggested by the statistical significance of a 12-h component fitted to the tumor incidence data. The administration time of melatonin is an indispensable complement to dosing, notably when effects upon tumorigenesis are involved.