Predictors of female sexual desire: testing the putative effects of negative mood and the menstrual cycle using mixed models analyses.

The aim of this dissertation was to expand and clarify previous findings regarding the effects of negative mood and menstrual cycle on sexual desire. This was done by asking female university students (n=213) to rate their level of depression, anxiety and sexual desire every day for two full menstrual cycles. Multilevel random-effects models were used to estimate average effects for the entire sample and to test for variability across participants in the associations between negative mood and sexual desire, controlling also for potential influences of the menstrual cycle. The effect of the menstrual cycle was modeled as a cosine coefficient of time. Thus, fixed cosine effects of the menstrual cycle and random cosine slopes of the menstrual cycle were estimated. Random intercepts were also calculated to look at between-women differences in mean level of sexual desire. Previous findings showing that some women report decreased sexual desire and others increased sexual desire when depressed or anxious, were confirmed. These effects have been defined by Bancroft and colleagues as a paradoxical effect of negative mood (between-groups). More importantly, for both depression and anxiety, results demonstrated the presence of within-person paradoxical associations, whereby, there are some women for whom both low and high levels of negative mood are associated with the same change (an increase or a decrease) in sexual desire. Thus, paradoxical associations between negative mood and sexual desire are also present at low levels of negative mood. Finally, although 1.6% of the variance in sexual desire was associated to an average effect of the menstrual cycle, the crucial findings of these analyses regard the presence of significant individual differences in this association. In fact, individual differences in the effect of the menstrual cycle on changes in sexual desire explained 7.8% of the variance in sexual desire. Moreover, 21.5% of the variance in sexual desire was explained by between-person differences in their mean level of sexual desire. The discussion underlines the importance of considering individual variability and multifactorial non-linear models when studying sexual desire.