Context: Seasonal variation in daylight regulates reproduction in animals living at higher latitude, but the influence of season on the sex hormones in humans remains unclear.

Objective, Design and Participants: A cross-sectional population-based study in Tromsø, Norway (70 degree North) included 1,651 women and 1,540 men aged 25-84 yr. Circulating total estradiol (and calculated free levels), follicle stimulating hormone (FSH), and dehydroepiandrosterone sulfate (DHEAS) were measured between September 1994 and September 1995, and provided a unique opportunity to study effects of extreme seasonal variations in the daylight on hormone levels in an artic population.

Main Outcome measure: Circulating total and free estradiol, FSH and DHEAS.

Results: Total and free estradiol showed differences between monthly means, with peak in June in postmenopausal women (P < 0.001), and in May in men (P = 0.002 and P < 0.001) by analysis of covariance. By cosinor analysis, a seasonal variation in total and free estradiol was evident in women (P = 0.02 and P = 0.03) and in men (P = 0.004 and P = 0.001), but only 0.2-0.9% of the variation in total and free estradiol was explained by season. FSH and DHEAS showed no obvious seasonal variation in either sex.

Conclusions: Seasonal variations should be considered while designing studies and interpreting results of estradiol measurements to avoid bias in comparative studies.