Marital sex frequency and midcycle female testosterone.

The purpose of the study was to attempt to replicate a finding of Persky et al. (1978) that midcycle peak values of testosterone (T) in women predicted differences in frequency of intercourse among married couples. Luteinizing hormone (LH), total testosterone (TT), and free testosterone (FT) values from 10 to 14 daily midcycle blood samples donated by 43 volunteering wives were analyzed against sexual activity patterns reported by the couples over a longer period of time. All couples were contracepting by means other than exogenous hormones or the rhythm method. Each morning through three menstrual cycles husbands and wives recorded independently and on separate forms answers to a series of questions concerning sexual activity in the previous 24 hr. Wives also recorded basal body temperatures (BBT). We designated midcycle values of TT and FT according to several definitions of midcycle. Total testosterone levels at the day of the BBT nadir and the day before the nadir correlated significantly with average intercourse frequency. Correlations with FT were statistically significant regardless of which midcycle measure was used; the day before the BBT nadir gave the highest correlation, 0.618, p = 0.01. Mean testosterone (TT or FT) values were not significantly related. We conclude that female midcycle total testosterone or free testosterone is indexing some unobserved event that affects the frequency of intercourse of couples. We speculate that this event affects the motivation of females, which influences the set point of the compromise frequency characteristic of couples.