Relationship between the 2D:4D and prenatal testosterone, adult level testosterone, and testosterone change: Meta‐analysis of 54 studies

Abstract

The ratio between the hands' second to the fourth finger (2D:4D) is commonly hypothesized to result from prenatal testosterone. The 2D:4D has also been hypothesized to relate to adult‐level testosterone and, more recently, to the testosterone response to a challenging situation. In the present work, we tested these core assumptions. Drawing from, in total, 54 studies and 8077 participants, we investigated whether the 2D:4D is related to adult level testosterone (44 studies), testosterone change (6 studies), and prenatal testosterone (10 studies). We found no evidence of the relationship between the above testosterone types and digit ratios. Furthermore, there was no relationship between testosterone and the right and left 2D:4D, male and female 2D:4D, and the 2D:4D and testosterone measurement (i.e., measured in blood or saliva). However, we found some evidence that prenatal testosterone measured in amniotic fluid (but not cord blood) might be related to the digit ratios—further studies are necessary to validate this observation. In summary, considering the current state of knowledge, any conclusions drawn from the assumption of the digit ratios as the proxy for testosterone (prenatal, adult level, or testosterone change under a challenging situation) warrant great caution.

The association between 2D:4D digit ratio and sex-typed play in children with and without siblings

“…A more general limitation could be the usefulness of 2D:4D as a marker for prenatal testosterone itself. Evidence in humans for direct associations between 2D:4D and prenatal testosterone is scarce and heterogeneous 55 , 70 – 72 , although, some evidence exists and 2D:4D is a commonly used surrogate to investigate associations of prenatal testosterone with other variables of interest 73 . As effect sizes between 2D:4D and sources of prenatal testosterone have to be generally considered as small 73 , there seems to be another source of variance that has not yet been identified.…”

“…Evidence in humans for direct associations between 2D:4D and prenatal testosterone is scarce and heterogeneous 55 , 70 – 72 , although, some evidence exists and 2D:4D is a commonly used surrogate to investigate associations of prenatal testosterone with other variables of interest 73 . As effect sizes between 2D:4D and sources of prenatal testosterone have to be generally considered as small 73 , there seems to be another source of variance that has not yet been identified. Also Hönekopp and Watson 2 noted the significance of identifying these other sources that may affect 2D:4D in one of the first meta analyses on 2D:4D and to control for them in future studies.…”

The fallacy of global comparisons based onper capitameasures

“…Based on our experiences, we think that there are three major reasons why ratios are still used for normalization/standardization: (i) ratios are indeed very simple to calculate and are thus intuitively appealing, which gives the impression that they are simple to interpret, which is not at all true; (ii) a tradition in a given field that is difficult to abandon, as researchers tend to use a technique they learned during their training. This may include the use of the second-to fourth-digit ratio in biological anthropology (for recent meta-analysis, see [31]), which is still very common, although it was previously criticized [12,32]; and (iii) a lack of proper mathematical knowledge about functions. The problems with ratios are often not included in practical statistical/mathematical courses.…”

“…a tradition in a given field that is difficult to abandon, as researchers tend to use a technique they learned during their training. This may include the use of the second- to fourth-digit ratio in biological anthropology (for recent meta-analysis, see [ 31 ]), which is still very common, although it was previously criticized [ 12 , 32 ]; and…”