4. General Discussion [...]
Our review pertaines to 32 relationships between putative indicators of prenatal androgenization (left hand 2D:4D, right hand 2D:4D, Dr-l ) and measures of circulating levels of sex hormones (total T, bioavailable T, E2, LH, FSH, and progesterone). Several significant relationships, involving T, FSH, and LH, were obtained for men. However, these relationships depended largely or entirely on one highly atypical sample of infertile men (Manning et al., 2004). Without the results from this sample, we obtained data on 29 relationships. Only one of them, pertaining to right hand 2D:4D and FSH in men, was statistically significant. For a number of reasons, this finding does not challenge the usefulness of 2D:4D as an indicator of prenatal androgenization. First, given the large number of associations examined, this finding may well have arisen from chance. Second, although statistically significant, the relationship was very small (r = .12). Third, and most importantly, considerations regarding the effects of sex hormones on behavior and cognition focus on T in males and on T, estrogen, and progesterone in females (e.g., Fitch and Denenberg, 1998; Keefe, 2002). We are not aware of results or considerations that suggest an effect of current FSH levels on men’s behavior or cognition. For the theoretically more important hormones, T, estrogen, and progesterone, the present evidence does not suggest an association with 2D:4D or Dr-l in the normal population. A word of caution is in order regarding the relationship between 2D:4D and circulating T levels in men. Recently, van den Bergh and Dewitte (2006) found that 2D:4D predicts T spikes triggered by sexual cues in men. Thus, researchers should ensure that obtained relationships between 2D:4D and actual behavior do not reflect effects of a transient T spike (e.g. caused by the presence of an attractive experimenter) on that behavior. 14 Our review was entirely based on published studies. The results of published studies may not be representative for the results of all studies that have investigated the particular question of interest. This is because statistically significant studies as compared to statistically non significant studies have a higher chance to get published. The resulting underrepresentation of non significant results is a potential threat to the validity of meta-analyses (e.g., Begg, 1994). This is not true in our case because unpublished, non significant results would strengthen but not weaken our conclusions. However, one should keep in mind that the absence of evidence is not the same as the evidence of absence. Those of our results that did not pertain to T were often based on only one sample, and the respective confidence intervals were often large. Thus, we cannot rule out that future studies may reveal moderate associations between 2D:4D or Dr-l and levels of circulating sex hormones in the normal population. However, the present evidence suggests that neither 2D:4D nor Dr-l are associated with adults’ sex hormone levels in healthy adults. This supports the validity of 2D:4D as a means to study the effects of prenatal androgenization on human behavior and cognition
www.livescience.com
www.elitedaily.com
www.researchgate.net
