Effects of competition outcome on testosterone concentrations in humans: An updated meta-analysis

A contribution to a special issue on Hormones and Human Competition. Since Archer's (2006) influential meta-analysis, there has been a major increase in the number of studies investigating the effect of competition outcome on testosterone reactivity patterns in humans. Despite this increased research output, there remains debate as to whether competition outcome modulates testosterone concentrations. The present paper examines this question using a meta-analytic approach including papers published over the last 35years. Moreover, it provides the first meta-analytic estimate of the effect of competition outcome on testosterone concentrations in women. Results from a meta-analysis involving 60 effect sizes and >2500 participants indicated that winners of a competition demonstrated a larger increase in testosterone concentrations relative to losers (D=0.20)-an effect that was highly heterogeneous. This 'winner-loser' effect was most robust in studies conducted outside the lab (e.g., in sport venues) (D=0.43); for studies conducted in the lab, the effect of competition outcome on testosterone reactivity patterns was relatively weak (D=0.08), and only found in studies of men (D=0.15; in women: D=-0.04). Further, the 'winner-loser' effect was stronger among studies in which pre-competition testosterone was sampled earlier than (D=0.38, after trim and fill correction) rather than within (D=0.09) 10min of the start of the competition. Therefore, these results also provide important insight regarding study design and methodology, and will be a valuable resource for researchers conducting subsequent studies on the 'winner loser' effect.