Edinburgh Napier University

  Women's biology has been used as a powerful justification for social inequality (Fausto-Sterling, 1992). However, evolutionary arguments suggesting females are unlikely to be equipped biologically for status and dominance remain controversial and unresolved (c.f. Blaffer-Hrdy, 1999; Campbell, 2002). Hormone-competition studies have been one means of assessing the relationship between biology and social dominance. However, despite their relative importance there is a scarcity of studies involving female participants. In the absence of reliable information concerning Testosterone (T) dynamics, those few studies already conducted have adopted the sampling protocols of male-focused studies. Where researchers have taken care to account for the dynamic nature of T levels in the design of studies involving male participants they have tended to control for the circadian component of variation only, by collecting samples at approximately the same time of day and subsequently disregarding episodic fluctuation. In addition, bio-behavioural studies exploring the relationship between T and competition in males have tended to utilise a limited number of samples in order to determine T levels at baseline, pre and post-competition. Any deviation from the baseline is assumed to be a function of the competitive event, or the participant's interpretation of the event. Comparatively little accurate research is available about detailed daily patterns of T in females; particularly the biologically active 'free' component, as measured in saliva. Given that females may also experience episodic fluctuation in levels of T, this limited or single time-point sampling approach for bio-behavioural studies may be inappropriate. Consequently, we conducted two studies which sought to provide a comprehensive picture of female salivary free-T dynamics over one day and two non-consecutive days. Fig. 1 illustrates that, within the group mean, female participants demonstrate a clear circadian rhythm. A within-subjects repeated measures ANOVA, utilising the Greenhouse-Geisser correction procedure, revealed a significant main effect, e.g. from 9am to 11pm, F(3.5, 146) =3.190, p<0.05. A subsequent pair-wise Sidak comparison revealed the differences to occur between 9am-9pm and 9am-11pm. It is perhaps of interest to note that whilst ANOVA did not reach significance between 9am and 7pm this actually represents a mean 17% difference. Fig. 3 illustrates that the circadian profile across two non-consecutive days appears only marginally similar. A within subject two-way repeated measures ANOVA with 2 (days) X 8 (time) revealed a main effect for time F(4, 216) = 10.57, p<0.05. Day X time interaction effects were non-significant. Employing a Spearman rho correlation, reliability of mean T at 8 points across the two non-consecutive days ranged from r=0.1 through r=0.48. Perhaps more importantly, throughout the course of the day T concentrations were highly variable with episodic fluctuation of individual data points exceeding 80% of 9am levels. Fig. 2 provides an illustrative sample of individual T levels from 6 subjects. These results suggest that, as well as circadian activity which is variable across non-consecutive days, females also experience episodic fluctuation at levels which call into question the use of single time-point measurements of T in bio-behavioural studies. Moreover, there is little evidence that T is any less labile in the afternoon compared with the morning. We argue that circadian variation and episodic fluctuation really do confound interpretation of measurements from one-off salivary T samples. Single measurements appear to be of little value-representing not much more than the level at the time they were collected.