No sexual dimorphism in human prenatal metacarpal ratios

No sexual dimorphism in human prenatal metacarpal ratios Stefan Van Dongen a, , Frietson Galis Sofie Delen a , Jessica Bots

b,c ,

Clara Ten Broek a , Kristiina Heikinheimo d,e , Liliane C.D. Wijnaendts

b,1 ,

a

Evolutionary Ecology Group, Department of Biology, University of Antwerp, Antwerp, Belgium b Department of P athology, VU University Medical Centre, Amsterdam, The Netherlands c NCB Naturalis, Leiden, The Netherlands d Department of Oral and Maxillofacial Surgery, Institute of Dentistry, University of Turku, Turku, Finland e Department of Oral Diagnostic Sciences, Institute of Dentistry, University of Eastern Finland, Kuopio, Finland a

abst ract Background: Ratios of digit lengths are studied intensively as markers of prenatal sex hormone levels. Aim: Study sexual dimorphism in ratios of metacarpals, which received less attention. Methods: We studied six metacarpal ratios in deceased human fetuses of ages 10 to 42 weeks. Results and conclusion: We found no indication of a sexual dimorphism at this early stage of development.

1. Introduction In humans, the second to fourth digit ratio (2D:4D) — among other digit ratios — is smaller in males than in females [8]. T his sexual dimorphism established early in life [2,3,7], is driven by exposure to prenatal sex hormones [6,10], and to some extent influenced by homeobox genes [5,19]. Nevertheless, it is influenced by at least 20 genes that control development of many organ systems including the limb skeleton and the brain [9,20]. Because prenatal environmental conditions — including hormone exposure — can have significant effects on individual morphology, behavior and performance as an adult, the study of variation in digit ratios in humans and primates has received much interest (e.g., [1,8]). While finger ratios have been studied intensively, those of metacarpals have received much less attention. Nevertheless, similar sexual dimorphism has been observed in chimpanzees, gorillas and baboons, as well as in humans, although the dimorphism appeared to be more strongly affected by metacarpal 5 compared to metacarpal 4 [4,11–13,15]. T o date, however, no study has been conducted during early devel- opment, which would lead to important insights into the developmental stage where the dimorphism emerges (e.g. [3]). Here, we investigate sexual dimorphism in six metacarpal ratios (2M:3M; 2M:4M; 2M:5M; 3M:4M; 3M:5M and 4M:5M) in over 600 deceased and electively aborted human fetuses from two populations, to test if sexual dimorphism already emerges during early fetal development. 2. Materials and methods 2.1. Study populations 2.1.1. Population 1 Between 1964 and 1974, human fetuses from elective abortions were collected by scientists at the Institute of Dentistry, University of T urku (Finland) to study craniofacial development. For ethical reasons all background information on the mothers has been discarded, leaving only data on the gestation age, sex and size for the majority of the fetuses. Complete fetuses were preserved by the ‘clearing and single staining’ procedure and kept on glycerin. T his technique allows studying skeletal traits by making eviscerated specimens semi-transparent with potassium hydroxide and staining ossified parts with alizarin red. All fetuses were externally evaluated for skeletal malformations (including craniofacial and limb defects) by a pediatric pathologist (LW) and no major abnormalities were reported. Bots et al. [21] did report the pres- ence of a high frequency of cervical ribs and rudimentary 12th ribs. Measurements of the metacarpals were obtained for 182 individuals (95 females and 83 males), aged 10 to 20 weeks of gestation (mean 15.2 weeks ± 2.1). Each fetus' left and right hand was photographed independently and repeatedly with a Canon 300D digital SLR camera. Digital pictures were scaled using a transparent ruler (resolution of 0.1 cm) and analyzed in Image J 1.42q [22]. Metacarpals II to V were

measured from the center of the proximal end to the center of the distal end of the bones (Fig. 1). Digits could not be measured for most fetuses because fingers had become rigid such that it was impossible to straighten them for accurate measurement. All measurements were re- peated twice independently for each picture to evaluate repeatability (see below). 2.1.2. Population 2 Since 1980, deceased human fetuses presented for autopsy at the VU Medical Centre in Amsterdam (T he Netherlands) have been routinely radiographed ventrally and laterally with hands taped in as straight a position as possible (23 mA, 70–90 kV, 4–12 s, Agfa Gevaert D7DW Structurix films). T he ventral radiographs of 1195 fetuses that died between 1990 and 2009 were used for this study. T o allow digital image analysis, we digitalized these radiographs using a standard set- up with a Canon 300D digital SLR camera positioned in a fixed distance from a glass plate with flash underneath. In total, pictures of 655 male and 540 female fetuses (mean, 28 weeks ± 11 weeks; range, 14–42 weeks) were examined. For the left and right hand of each fetus, the six metacarpal ratios were determined in Image J 1.46. Meta- carpals II to V were measured from the center of the proximal end to the center of the distal end of the bones (Fig. 1). Only pictures were included where the hands were in a flat position, leading to measurements for at least one ratio in 264 male and 271 female fetuses. In addition to the metacarpals, digits 2 and 4 were measured (see [3] for details). Measurements were repeated twice independently for each picture (if avail- able) to evaluate repeatability. Data on ethnicity were not available for individual fetuses in either sample, but the patient populations of both samples were predominantly Caucasian. Gestational ages were provid- ed by the gynecologist, based on the time since the first day of the last menstruation (post menstruation). 2.2. Statistical analysis We first investigated the repeatability of our measurements using a mixed ANOVA model with individual and the interaction between pic- ture and individual (i.e., a nested picture effect within individual) as random effects. Repeatabilities ranged between 0.82 and 0.95 and did not show marked differences between populations. T hus, the meta- carpal ratios were estimated with high accuracy. T o investigate sexual dimorphism in each study population, meta- carpal ratios (2M:3M; 2M:4M; 2M:5M; 3M:4M; 3M:5M and 4M:5M), and digit ratio 2D:4D for population 2, averaged across the left and right hand, were treated as dependent variables in linear models with sex as fixed factor and age as continuous covariate. T he two-way inter- action between sex and age was also tested (i.e., an ANCOVA model with different slopes). T o investigate covariation in metacarpal and digit ratios, we estimated and tested correlations between the digit and metacarpal ratios after correcting for possible age and gender ef- fects, modeled in the ANCOVAs. Analyses performed separately for left and right hand measurements gave very similar results (data not shown). All analyses were performed in R (version 2.10.1) using the function lm for the ANCOVA models and lmer in the package lme4 for the mixed ANOVA. Descriptive statistics are reported as mean ± SE, and effect sizes of the sexual dimorphism is provided as Cohen's D. For comparison with other studies, sexual dimorphism in 2D:4D and 2M:4M is also provided for both sides separately. 3. Results T able 1 provides an overview of the ANCOVA model results. No sex- ual dimorphism was detected in either population, and in none of the six metacarpal ratios (T able 1). Although not significantly so for each test, the metacarpal ratios appeared to decrease with age in population 1 (i.e., between gestation weeks 10–20), and to increase slightly in pop- ulation 2 (mainly data after an age of 20 weeks of gestation) (see sign of correlation coefficient in T able 1). None of the interactions with sex was significant, indicating that the changes in metacarpal ratio with age were comparable between male and female fetuses (T able 1). Effect sizes were small ranging between − 0.14 and 0.14 with an average of 0.037. Estimated associations and raw data for the M2:M4 ratio in males and females in both populations are provided in Fig. 2 (results for the other ratios were very similar). Because population 2 consisted of de- ceased fetuses, many of which showed severe congenital abnormalities, we also plotted the data for the fetuses of population 2 which did not show any major abnormalities after medical examination (red symbols in Fig. 2). T hese observations did not appear to behave differently (Fig. 2). For the 2D:4D digit ratio in population 2, in contrast, the sexual di- morphism was statistically significant (F 1,257 = 10.8, p = 0.001), with 2D:4D being smaller for males (mean = 0.916, SE = 0.002, N = 140) than for females (mean = 0.928, SE = 0.003, N = 120). T he 2D:4D decreased with age (slope = − 0.0009, F 1,257 = 18.4, p b 0.0001), and this decrease was similar for male and female fetuses (i.e., no signif- icant interaction between sex and age, F 1,256 = 0.01, p = 0.93). T he 2D:4D did not correlate with any of the metacarpal ratios after

correcting for age and sex differences (if present) with correlation coef- ficients raging between −0.16 and 0.04 (all p N 0.3). Average 2M:4M and 2D:4D for males and females on both sides appeared to show very similar differences (population 1: 2M:4M right: males 1.161 (0.062) — females 1.172 (0.061) Cohen's D = 0.18; 2M:4M left: males 1.156 (0.059) — females 1.161 (0.063) Cohen's D = 0.08; population 2: 2M:4M right: males 1.149 (0.068) — females 1.140 (0.081) Cohen's D = −0.12; 2M:4M left: males 1.139 (0.065) — females 1.148 (0.074) Cohen's D = 0.12; 2D:4D right: males 0.917 (0.035) — females 0.924 (0.037) Cohen's D = 0.19; 2D:4D left: males 0.916 (0.033) — females 0.929 (0.038) Cohen's D = 0.37). 4. Discussion dimorphism in the prenatal digit ratio (2D:4D) emerges during early fetal development [2,3,7,8], even though the extent of the sexual dimor- phism was smaller in fetuses compared to adults [3]. T his finding is in agreement with the hypothesis that early levels of sexual hormones have a lasting influence on 2D:4D ratios. T he lack of a significant sexual dimorphism in metacarpal ratios does not seem to be due to a lack of statistical power as we did detect a significant dimorphism in 2D:4D in population 2 in this study with a similar sample size. In addition, our sample sizes were larger than those in McFadden and Bracht [11–13]. Robertson et al. [15] performed a much larger study of sexual dimorphism in digit and metacarpal ratios (N N 3000), showing that the sexual dimorphism in 2D:4D (males: 0.908 (0.022); females: 0.922 (0.021); Cohen's D = 0.65) was much more pronounced than that in 2M:4M (males: 1.152 (0.031); females: 1.157 (0.033); Cohen's D = 0.16). In addition, Haugen et al. [4] also reported stronger sexual dimorphism in 2D:4D (males: 0.910 (0.024); females: 0.926 (0.026); Cohen's D = 0.64) than that in 2M:4M (males: 1.168 (0.032); females: 1.173 (0.042); Cohen's D = 0.13) in over 1000 adults. Clearly, our study did not have sufficient statistical power to detect such slight differences in metacarpal ratios. Nevertheless, the average effect size we found for the metacarpal ratios was 0.037, which does appear somewhat lower than those reported by Robertson et al. [15] and Haugen et al. [4] study- ing adults. Also important to note is that Robertson et al. [15] and Haugen et al. [4]only studied metacarpals 2 and 4, and thus could not confirm that other ratios (especially those relative to metacarpal 5) showed larger dimorphism as suggested by McFadden and Bracht [11–13]. T his larger contribution of metacarpal 5, however, is not supported by our study. Other evidence that the sexual dimorphism in ratios of digit and metacarpal lengths differs during the fetal develop- ment, are the opposite associations with age (where 2D:4D decreased with age and at least some metacarpal ratios increased with age in population 2) and the lack of any correlation between 2D:4D and the metacarpal ratios. Finally, one could argue that our sample was very heterogeneous (lack of good ethnicity data, different ages) and that therefore, sexual dimorphisms would become obscured. However, stan- dard deviations of digit and metacarpal ratios in our study were remark- ably similar to those reported by Robertson et al. [15] and Haugen et al. [4]. T here is thus no reason to suspect the higher heterogeneity in our sample, at least with respect to the ratios studied. We therefore con- clude that sexual dimorphism in metacarpal ratios in our sample were likely to be smaller than those reported in adults. Due to the smaller sample sizes in this study, however, it is impossible to rule out the pres- ence of very small sexual dimorphisms. About the reason for the absence of dimorphisms of similar sizes as in adults in our sample of fetuses we can only speculate. Possibly, differential timing of development between metacarpals and phalanges plays a role. T he patterning of the limbs proceeds along a proximo-distal axis, such that more distal elements are laid out later than more proximal ones (reviewed in [16]). Although the differences are small, at least in mice it has been shown that ossification of metacarpals starts before that of phalanges [14]. T hus, the pattern formation of the metacarpal bones might start earlier as well, possibly leading to a lack of differential exposure to androgens in males and females. However, this cannot ex- plain the presence of sexual dimorphism in metacarpal ratio in adults [13,15], suggesting that the adult dimorphism most likely arises after birth. T he fact that also for 2D:4D digit ratio there are indications that it is modified during postnatal growth [3,17,18], at least it does not preclude that similar modifications occur in metacarpals. Further study is clearly needed and given the discrepancy between patterns observed in fetuses (this study) and adults [13,15] special attention should go to the metacarpal ratios in children and adolescents.

Acknowledgments T his research was financially supported by the Grant for a Short Sci- entific Mission for the COST 23B Action (Oral Facial Development and Regeneration) of the European Science Foundation to FG and by the Fund for Scientific Research (FWO) — Flanders (postdoctoral grant to JB and research program G.0027.07 to SVD). Ethical approval (approval number 648 ⁄ 32 ⁄ 300 ⁄ 05) to study the human collection of embryos and fetuses owned by the University of T urku was obtained from the National Supervisory Authority for Welfare and Health of Finland (VALVIRA). References [1] Fink B, Thanzami V, Seydel H, Manning JT. Digit ratio and hand-grip strength in German and Mizos men: cross-cultural evidence for an organizing effect of prenatal testosterone on strength. Am J Hum Biol 2006;18:776–82. [2] Garn SM, Burdi AR, Babler WJ, Stinson S. Early prenatal attainment of adult metacarpal–phalangeal rankings and proportions. Am J P hys Anthropol 1975;43:327–32. [3] Galis F, Ten Broek CMA, Van Dongen S, Wijnaendts LCD. Sexual dimorphism in the prenatal digit ratio (2D:4D). Arch Sex Behav 2010;39:57–62. [4] Haugen IK, Niu J, Aliabadi P , Felson DT, Englund M. The associations between finger length pattern, osteoarthritis, and knee injury: data from the Framingham Commu- nity Cohort. Arthritis Rheum 2011;63:2284–8. [5] Kondo T, Zakany J, Innis JW, Duboule D. Of fingers, toes and penises. Nature 1997;390:29. [6] Lutchmaya S, Baron-Cohen S, Raggat P , Knickmeyer R, Manning JT. 2nd to 4th digit ratio, fetal testosterone and estradiol. Early Hum Dev 2004;77:23–8. [7] Malas MA, Dogan S, Evcil EH, Desdicioglu K. Fetal development of the hand, digits and digits ratio (2D:4D). Early Hum Dev 2006;82:469–75. [8] Manning JT, Scutt D, Wilson J, Lewis-Jones DI. The ratio of 2nd to 4th digit length, a predictor of sperm numbers and concentrations of testosterone, luteinizing hormone and oestrogen. Hum Reprod 1998;13:3000–4. [9] Manning JT. Resolving the role of prenatal sex steroids in the development of digit ratio. P roc Natl Acad Sci U S A 2011;108:16143–4. [10] Manning JT, Kilduff LP , Trivers R. Digit ratio (2D:4D) in Klinefelter's syndrome. Andrology 2013;1:94–9. [11] McFadden D, Bracht M. The relative lengths and weights of metacarpals and meta- tarsals in baboons (P apio hamadryas). Horm Behav 2003;43:347–55. [12] McFadden D, Bracht M. Sex differences in the relative lengths of metacarpals and metatarsals in gorillas and chimpanzees. Horm Behav 2005;47:99–111. [13] McFadden D, Bracht M. Sex and race differences in the relative lengths of meta- carpals and metatarsals in human skeletons. Early Hum Dev 2009;85(2): 117–24. [14] P atton JT, Kaufman MH. The timing of ossification of the limb bones, and growth rates of various long bones of the fore and hind limbs of the prenatal and early postnatal laboratory mouse. J Anat 1995;186:175–85. [15] Robertson J, Zhang W, Liu JJ, Muir KR, Maciewicz RA, Doherty M. Radiographic assessment of the index to ring finger ratio (2D:4D) in adults. J Anat 2008; 212:42–8. [16] Tamura K, Yonei-Tamura S, Yano T, Yokoyama H, Ide H. The autopod, its formation during limb development. Dev Growth Differ 2008;50:S177–87. [17] Trivers R, Manning J, Jacobson A. A longitudinal study of digit ratio (2D:4D) and other finger ratios in Jamaican children. Horm Behav 2006;49:150–6. [18] Van Dongen S. Second to fourth digit ratio in relation to age, BMI and life history in a population of young adults: a set of unexpected results. J Negat Results Ecol Evol Biol 2009;6:1–7. [19] Zakany J, Duboule D. The role of Hox genes during vertebrate limb development. Genet Dev 2007;17(4):359–66. [20] Zheng Z, Cohn MJ. Developmental basis of sexually dimorphic digit ratios. P NAS 2011;108:16289–94. [21] Jessica Bots, Liliane CD, Wijnaendts, Sofie Delen, Stefan van Dongen, Kristiina Heikinheimo, Frietson Galis. Analysis of cervical ribs in a series of human fetuses. J Anat 2011;219:403–9 [issn 0021-8782]. [22] Abramoff MD, Magalhaes P J, Ram SJ. Image processing with ImageJ. Biophotonics Int 2004;11(7):36–42.