Methods Participants One hundred and twenty female members of a fitness club in Oberhausen (Germany) who trained for at least 3 h per week were recruited for the study. As we applied a cross-over design in order to control for sequence effects, these females were randomly assigned to two experimental groups by throwing a die. In the first group (n ¼ 66), physical exercise (PE) was followed by reading a newspaper (RN), while in the second group (n ¼ 54), the sequence was the reverse. Of these 120 females, 14 dropped out prior to the commencement of the experiment due to scheduling problems, a further 6 participants refused to have their photos taken in the standardised catsuit, 12 began the experiment but did not complete the study, 3 had to be excluded from the analyses because their questionnaires contained missing data, and a further 20 participants only finished the first activity (PE or RN). Thus, N ¼ 65 females participated in the whole experiment and were included in the statistical analyses. Of these females, n ¼ 47 were in the first (PE followed by RN) and n ¼ 18 in the second group (RN followed by PE).