Latitudinal gradient in 2D:4D.

The potential dependence of second-to-fourth digit ratio (2D:4D) on latitude has been discussed in recent papers in this journal (Hurd & van Anders, 2007; Loehlin, McFadden, Medland, & Martin, 2006, 2007). Such an association was first suggested by Manning (2002), who proposed that 2D:4D was highest at intermediate latitudes and lowest at the equator and at high latitudes. In their Fig. 1, Hurd and van Anders (2007) plotted the combined data of Loehlin et al. (2006) and Manning (2002). This figure shows a clear positive latitudinal trend in 2D:4D, except that data from Finland seem an outlier from this otherwise rather linear trend. Thus, the suggestion of low 2D:4D at high latitudes depends critically on these data from Finns. Here, we provide new, larger data on the 2D:4D of Finnish women and show that Finns have a much higher mean 2D:4D than those reported by Manning (2002) and Hurd and van Anders (2007). The mean 2D:4D of Finnish women reported by Manning (2002), based on measurements from 54 women, was approximately 0.94. Note that Hurd and van Anders (2007) seem to have used even a smaller sample (but higher mean 2D:4D of 0.95) of only 17 women from Finland, reported in Manning et al. (2000). We collected data on the right and the left hand 2D:4D from a geographically diverse sample of 287 Finnish women born 1946–1958, who participated in national screening program for cervical cancer in Turku, South-West Finland (Helle & Lilley, in press). These women are an ethnically homogeneous group, as all nonnative Finns were excluded from the analysis and none of those included were born in the Sami parishes of Northern Finland. The mean 2D:4D for these women was 0.975, which is significantly higher compared to the data reported by Manning (2002) (0.94 vs. 0.975, t(286) = 15.1, p \ .0001). When replacing data from Manning (2002) with our estimate into Hurd and van Anders’s (2007) Fig. 1, Finland does not anymore seem to be an outlier (Fig. 1). Nevertheless, regressing women’s 2D:4D on latitude still shows a tendency for a convex association, i.e., that 2D:4D first increases towards high latitudes but then levels off or even begins to decrease again towards very high latitudes (Fig. 1). When the data from Finns are excluded, the analysis does not anymore give statistical support for a curvilinear association between latitude and women’s 2D:4D (p = .22 for quadratic association). Instead, this association seems linear, b(95% CIs) = 0.00092 (0.00012, 0.00171), F(1, 8) = 7.06, p = .029). This suggests that data from Finland are still responsible for the curvilinear effect of latitude on 2D:4D. More data on northern latitudes are thus needed to confirm whether 2D:4D ceases to increase or begins to decline towards the extreme high latitudes in the northern hemisphere. In order to gain more insight to this problem, we investigated also within-population geographical variation of women’s 2D:4D in Finland. While our sample was collected from women that lived in the same city at the time of the measurements, there was large geographic variation in the location of their town of birth (latitude: from 60 10 to 66 50 N (corresponding to ca 740 km); longitude: from 21 20 to 29 30 E). We took advantage of this variation and examined whether either latitude or longitude of the location of birth was related to 2D:4D, using a general linear mixed model (Littell, Milliken, Stroup, & Wolfinger, 2006; see Table 1). Laterality and its two-way interactions with latitude and longitude were included into the model to examine whether these gradients differed between the right and the left hand 2D:4D. S. Helle (&) T. Laaksonen Section of Ecology, Department of Biology, University of Turku, 20014 Turku, Finland e-mail: [email protected]