Pii: s1090-5138(00)00063-5

Evolution and Human Behavior 22 (2001) 61±69 Second to fourth digit ratio and male ability in sport: implications for sexual selection in humans aPopulation and Evolutionary Biology Research Unit, School of Biological Sciences, University of Liverpool, bFootball Research Unit, Department of History, University of Liverpool, Liverpool L69 3BX, UK Received 14 July 2000; accepted 16 October 2000 Fetal and adult testosterone may be important in establishing and maintaining sex-dependent abilities associated with male physical competitiveness. There is evidence that the ratio of the length of the 2nd and 4th digits (2D:4D) is a negative correlate of prenatal and adult testosterone. We use ability in sports, and particularly ability in football, as a proxy for male physical competitiveness. Compared to males with high 2D:4D ratio, men with low ratio reported higher attainment in a range of sports and had higher mental rotation scores (a measure of visual±spatial ability). Professional football players had lower 2D:4D ratios than controls. Football players in 1st team squads had lower 2D:4D than reserves or youth team players. Men who had represented their country had lower ratios than those who had not, and there was a significant (one-tailed) negative association between 2D:4D and number of international appearances after the effect of country was removed. We suggest that prenatal and adult testosterone promotes the development and maintenance of traits which are useful in sports and athletics disciplines and in male:male fighting. D 2001 Elsevier Science Inc. All rights reserved.
Keywords: 2nd:4th digit ratio; Sports ability; Sexual selection Darwin (1871) recognised two forms of sexual selection: (a) mate choice, i.e., intersexual selection and (b) competition within a sex, usually male, for resources which facilitate access to mates, i.e., intrasexual selection. Ability in many sports and athletic disciplines may act as a proxy for ability in male±male fighting because they demand good visual±spatial judgement, speed, endurance and strength. Football (``soccer' ) is one such sport. Striking E-mail address: jtmann@liverpool.ac.uk (J.T. Manning).
1090-5138/01/$ ± see front matter D 2001 Elsevier Science Inc. All rights reserved.
J.T. Manning, R.P. Taylor / Evolution and Human Behavior 22 (2001) 61±69 the ball with foot, head and fist (in the case of goalkeepers) and directing it towards fellow players or the goal requires spatial judgement. Playing competitive football for 90 min is a sustained demonstration of cardiovascular efficiency and a test of speed and sometimes strength. Testosterone, particularly prenatal testosterone, has many extragenital influences (Bardin & Catterall, 1981; McEwen, 1981). Androgens are critical for sexual differentiation of the nervous system and the development of abilities such as spatial judgement (Geschwind & Galaburda, 1985; Kimura, 1996, 1999), and in males the formation of an efficient cardiovascular system (Manning & Bundred, 2000). There is evidence that the ratio between the length of the 2nd and 4th digits (2D:4D) is a correlate of prenatal and adult testosterone concentration because (a) it is lower in males than females (Baker, 1888; Manning, Scutt, Wilson, & Lewis-Jones, 1998; Phelps, 1952); (b) the ratio does not appear to change at puberty and is probably fixed prenatally, perhaps by the end of the first trimester (Garn, Burdi, Babler, & Stinson, 1975; Manning et al., 1998); (c) the waist:hip ratio (a positive correlate of testosterone) of mothers is negatively associated with the 2D:4D ratio of their children (Manning, Trivers, Singh, & Thornhill, 1999); and (d) 2D:4D has been reported to be negatively associated with testosterone in men (Manning et al., 1998).
Low 2D:4D ratios are associated with autism (Manning, Baron-Cohen, Wheelwright, & Sanders, 2001), left-hand preference (Manning, Trivers, Thornhill, & Singh, 2000), male homosexuality (Robinson & Manning, 2000) and male membership of a symphony orchestra (Sluming & Manning, 2000). There is evidence that all these traits are correlated with high prenatal testosterone (Geschwind & Galaburda, 1985; but see also Bryden, McManus, & Bulman-Fleming, 1994). The purpose of this work was to examine the relationship between 2D:4D ratio and sports ability, with particular reference to football.
The participants were 128 men recruited from libraries and sports centres. Many were university students. Participants who reported injuries to their 2nd and/or 4th digits were excluded from the study. The length of the right and left 2nd and 4th digits was measured from the crease proximal to the palm to the tip of the digit. Vernier callipers, measuring to 0.01 mm, were used. The 2nd and 4th digits from 50 hands and 50 subjects were Subjects ranked their competitive level in the sport they most regularly practiced on a scale from 1 to 10, with 10 representing international participation (Table 1). This scale has been used to rank middle distance runners and it was shown to correlate with their best times for The subjects were 125 men recruited from libraries, adult education classes and sports centres. The participants were from a variety of educational and socio-economic back- J.T. Manning, R.P. Taylor / Evolution and Human Behavior 22 (2001) 61±69 Participants were asked to choose their most suitable rank for their favorite sporting activity, starting from rank 10 You think you could represent your Country You think you could compete at National level You think you could compete at County level You think you could compete at an organised level grounds. Subjects who had injuries to the 2nd and/or 4th digits were excluded from the study.
As with Study I, the length of the 2nd and 4th digits was measured directly from the hands using Vernier callipers measuring to 0.01 mm. The digits were measured twice in 50 hands The subjects were given a Vandenberg Mental Rotation Test (MRT). This test is known to reveal strong differences in visual±spatial scores between the sexes (Vandenberg & Kuse, 1978). The MRT consisted of 20 items, each composed of a criterion figure, two correct alternatives and two distractors. Correct alternatives were identical to the criterion but were shown in a rotated position. Two credits were given for a line with two correct choices and none if one or both are incorrect. This eliminates the need to apply a correction for guessing.
This study was concerned with 2D:4D ratio in football players. The sample consisted of men who were currently playing or had played in English professional football (n = 304), together with controls (n = 533). Subjects who reported injuries to the 2nd and/or 4th digit were excluded from the study. Photocopies were made of the ventral surface of the right and left hands of all participants. Digit length was measured from the basal crease of the finger proximal to the palm to the tip of the finger. Measurements were made with Vernier callipers measuring to 0.01 mm. For 30 hands from 30 subjects, we measured 2nd and 4th digits length from the digits and from the photocopies. For 100 hands from 100 subjects, we measured the 2nd and 4th digits twice from photocopies.
Professional football in England is organised into four competitions or leagues. The Premier Division is the elite competition with Divisions I, II and III in descending hierarchical order. The football players in our sample were (a) 1st team, reserve and youth (minimum age 17 years) players from clubs playing in the 1998±1999 season of the English Premier Division (Liverpool and Coventry, n = 48) and Divisions I (Sunderland and Tranmere, n = 62), II (Preston and Oldham, n = 77), III (Cambridge and Rochdale, n = 67) J.T. Manning, R.P. Taylor / Evolution and Human Behavior 22 (2001) 61±69 of the English Football League; (b) coaches from these clubs (n = 21), many of whom had played for elite clubs and some of whom had represented their country; (c) internationals from England, Scotland, Wales, Northern Ireland, Eire and Argentina (n = 29) who had played their football in the period 1936±1999 and who had represented or were currently representing many of the foremost clubs in England (O. Ardilles, J. Armfield, J. Barnes, J.
Charles, K. Dalglish, T. Francis, T. Finney, P. Gascoigne, J. Giles, J. Haynes, G. Hoddle, N.
Hunter, G. Hurst, P. Jennings, C. Jones, N. Lofthouse, P. McGrath, D. McKay, F. McLintock, W. Mannion, S. Matthews, A Mullery, T. Paine, B. Robson, J. Royle, I. Rush, L. Shackleton, A. Shearer, P. Shilton). The hands of the internationals were photocopied at the Centenary Celebrations marking the end of the 100th English League Championship. In total, there were 304 players in the sample. The controls had not played professional football and were recruited from colleges, universities, libraries and sports centres. Data from the controls had The repeatability of the 2D:4D ratio was high (r1=.93) and the differences between subjects greater than the measurement error (repeated measures ANOVA, F = 29.96, P = .0001). We therefore concluded that our measurements reflected real between-subject The participants practiced seven sports (running 45% of the sample, football 14%, martial arts 10%, rugby 8%, tennis or squash 8%, swimming 7% and hockey 5%), while there were three subjects (2%) who practiced no sport. Means and standard deviations for the sample were: mean 2D:4D ratio of right and left hand 0.98 ‹ 0.03, sporting rank 5.33 ‹ 2.01, sporting experience 5.33 ‹ 4.44 years and age 22.35 ‹ 4.91 years. Mean 2D:4D ratio was very close to Fig. 1. The relationship between the mean 2nd:4th digit ratio of the right and left hands of 128 male participants and their reported level of sporting attainment. Subjects with low 2D:4D ratio report higher sporting attainment than subjects with high 2D:4D ratio.
J.T. Manning, R.P. Taylor / Evolution and Human Behavior 22 (2001) 61±69 the reported population norm for Liverpool men, and as expected, there was no relationship between 2D:4D and age (b = 0.0001, F = 0.05, P = .82; Manning et al., 1998). There was a significant negative relationship between mean 2D:4D of the right and left hands and sporting rank, i.e., men with low 2D:4D ratio reported higher rank (b = À 22.73, F = 10.99, P = .001; Fig. 1). Removing the two most extreme ratios (0.88 and 1.06) made little difference to the relationship (b = À 21.48, F = 7.11, P = .007). A simultaneous multiple regression analysis with rank as the dependent variable and independent variables mean 2D:4D, age, sports experience and sporting discipline (dummy coded running = 1, football = 2, martial arts = 3, rugby = 4, tennis or squash = 5, swimming = 6, hockey = 7, no sport = 8) showed that 2D:4D remained significantly related to rank when the effect of the other variables was removed (2D:4D b = À 21.84, t = 3.53, P = .006; age b = 0.08, t = 2.32, P = .02; experience b = 0.06, t = 1.47, P = .14; sport b = À 0.25, t = 3.25, P = .002).
The 2D:4D ratio showed high repeatability (r1=.91) and the between-subject differences in 2D:4D were greater than the measurement error ( F = 27.59, P = .0001). We concluded that our measurements of 2D:4D reflected real differences between subjects. Means and standard deviations for mean of left and right hand 2D:4D, age and MRT scores were as follows: 0.99 ‹ 0.03 ratio, 25.36 ‹ 8.24 years, 28.16 ‹ 7.02 MRT. The mean 2D:4D ratio was slightly higher than that reported for Liverpool men (0.98; Manning et al., 1998).
We found that men with low 2D:4D ratio tended to score higher on the MRT than men with high 2D:4D (b = À 83.87, F = 20.17, P = .0001; Fig. 2). Younger subjects also scored higher MRT scores than older men (b = À 0.25, F = 11.59, P = .0009). There was no relationship between age and 2D:4D (b = À 0.0003, F = 0.72, P = .40). A simultaneous multiple regression with dependent variable MRT score and mean 2D:4D and age as independent variables showed that 2D:4D and age were independent negative predictors of MRT (2D:4D b = À 91.27, t = 5.49, P = .0001; age b = À 0.39, t = 5.84, P = .0001).
Fig. 2. The relationship between the mean 2nd:4th digit ratio of the right and left hands of 125 male participants and their MRT scores (a measure of their visual±spatial ability). Subjects with low 2D:4D ratio had higher MRT scores than subjects with high 2D:4D.
J.T. Manning, R.P. Taylor / Evolution and Human Behavior 22 (2001) 61±69 The repeatabilities of the 2D:4D ratios were high and showed significantly higher between-subject differences compared to measurement error (repeated measures ANOVA, measurements of hand and photocopy, n = 30, r1=.91, F = 17.47, P = .0001; measurements of photocopies, n = 100, r1=.99, F = 27.37, P = .0001). We concluded that our measurements of 2D:4D ratio represented real between-individual differences.
The mean age was 29.83 ‹ 13.36 years for the football players and 28.56 ‹ 10.08 years for the controls. There was no evidence for an association between mean 2D:4D of the right and left hands and age (complete sample b = 0.0001, F = 0.58, P = .37). Mean 2D:4D was lower in football players compared to controls (players 0.95 ‹ 0.03, controls 0.98 ‹ 0.04, t = 14.52, P = .0001). The control mean was close to that previously reported for Liverpool men (Manning et al., 1998). There was an overall increase in 2D:4D in the sequence: Interna- tionals, Coaches, Premier Club players, Divisions I, II, III Club players, Controls (b = 0.009, F = 163.48, P = .0001; Fig. 3). A significant positive relationship remained within the players' sample when the controls were removed (b = 0.002, F = 4.58, P = .03).
Within clubs, the regular 1st team players and reserves are mainly drawn from the annual recruitment of youth players. We asked the club coaches to identify 1st team, reserve and youth players. The 1st team players had lower mean 2D:4D ratios than the reserves and the youth players had intermediate ratios (1st team, n = 154, 0.94 ‹ 0.02; reserves, n = 38, 0.96 ‹ 0.03, youth squad, n = 65, 0.95 ‹ 0.03, ANOVA F = 5.58, P = .005). A simultaneous multiple regression analysis with dependent variable mean 2D:4D and independent variables ``Division'' (dummy coded Premier = 0, Divisions I = 1, II = 2, III = 3) and 1st Team/Reserves (dummy coded 1 and 2) showed that 1st Team players had lower 2D:4D ratios than reserves independent of the Division they were playing in (1st Team/Reserves, b = 0.01, t = 2.69, P = .008; Division, b = 0.002, t = 1.48, P = .14).
In the total sample of football players, there were 37 players who had played for England, Scotland, Wales, Northern Ireland or Eire. There were 227 players born in these ``Home'' Fig. 3. The sample means and standard errors of 2nd:4th digit ratios (mean of right and left hands) of professional football players and controls. There is an overall increase in the sequence: Internationals, Coaches (many of whom were internationals), Premier League Players, Divisions I, II and III players and Controls (b = 0.009, F = 163.48, P = .0001). The distribution of 2D:4D suggests a relationship between football ability and low 2D:4D.
J.T. Manning, R.P. Taylor / Evolution and Human Behavior 22 (2001) 61±69 Nations who had not yet played for their Country. The Internationals had lower mean 2D:4D ratio (0.94 ‹ 0.02) than the non-internationals (0.95 ‹ 0.03, t = 2.47, P = .01). Of the 37 ``Home'' Nation international players, there were 31 who were over 40 years and likely to have finished their career. The mean number of international appearances for this group was 44.81 ‹ 33.70. Within this sample, there was a non-significant negative association between mean 2D:4D and (base 10 log) number of international appearances (b = À 6.24, F = 2.40, P = .07; this is a one-tailed test because the predicted relationship is negative). A simultaneous multiple regression with (base 10 log) number of international appearances as the dependent variable and mean 2D:4D and Country (dummy coded England = 1, Scotland = 2, Wales = 3, Northern Ireland = 4, Eire = 5) showed a significant negative relationship between 2D:4D and number of international appearances (b = À 6.78, t = 1.75, P = .045, one-tailed) and a non- significant relationship with Country (b = À 0.15, t = 1.89, P = .07, two-tailed).
We have found that low 2D:4D ratio is associated with a high level of attainment across a number of sports and also with high mental rotation scores. Professional football players had lower 2D:4D ratios than controls; 1st team players had lower ratios than youth team members or reserves; international players had lower 2D:4D ratios than those who have not yet represented their Country; and in a one-tailed test, 2D:4D ratio was negatively related to number of international appearances after the affect of Country was removed. We suggest that low 2D:4D ratio in men is a correlate for high ability in many sports, including football.
The relationship between low 2D:4D ratio and ability in sports may arise in two ways.
First, it has been suggested that high prenatal testosterone promotes the growth of the right hemisphere and facilitates visual±spatial ability (Geschwind & Galaburda, 1985). In support of this, we have found evidence that low 2D:4D is related to high mental rotation scores. This result appears to be a widespread one because a similar negative association between 2D:4D ratio and MRT has been found in samples of men (but not women) from London, Hungary and Merseyside (Sanders, Bereczkei, Csatho and Manning, manuscript). Secondly, prenatal conditions are likely to be powerful antecedents of developmental stability in the vascular system and, in later life, freedom from coronary heart disease (Barker, 1992). There is evidence in the male fetus that first trimester exposure to exogenous estradiol and progesterone may lead to cardiovascular anomalies which include ventricular septal defect, atrial septal defect, pulmonary stenosis, patent ductus arteriosis and transposition of the great vessels (Heinonen, Slone, Monson, Hook, & Sahpiro, 1977; Levy, Cohen, & Fraser, 1973; Lorber, Cassidy, & Engel, 1979). It is probably not coincidental that estradiol/progesterone treatment in the first trimester is also associated with digit anomalies, e.g., polydactyly, digit reduction and fingerisation of thumbs in boys (Levy et al., 1973; Lorber et al., 1979).
Testosterone is also an important prophylactic against malfunction of the vascular system in men (Rosano, 2000). It reduces the effects of exercise-induced myocardial ischemia in men with coronary heart disease (Rosano, Leonard, & Pagnotta, 1999); men with coronary artery disease have lower testosterone levels than controls (English et al., 2000) and survivors of heart attack have lower testosterone and higher estradiol levels than controls (Aksut, Aksut, J.T. Manning, R.P. Taylor / Evolution and Human Behavior 22 (2001) 61±69 Karamehmetoglu, & Oram, 1986; Phillips, Pinkernell, & Jing, 1994; Sewardson, Vylithin- gum, Jailal, & Desai, 1990; Swartz & Young, 1987). It has been suggested that the formation of the cardiovascular system is sensitive to testosterone and estradiol and that low 2D:4D ratio is a marker for high prenatal testosterone and an efficient vascular system in males We have drawn a parallel between abilities that are useful in intrasexual competition and those that are associated with success in sport. Football players and other sportsmen enjoy high status and often substantial earnings. We do not have any data relating to numbers of sexual partners of successful sportsmen. However, it is to be expected that high status and wealth are attributes that are of interest to women in their choice of sexual partner.
Low 2D:4D ratios in men have now been found associated with ability in sports, high sperm counts and testosterone concentrations (Manning et al., 1998), and a possible male display trait, i.e., musical ability (Sluming & Manning, 2000). We therefore suggest that 2D:4D ratio is a negative correlate of male competitiveness, fertility and attractiveness.
For permission to measure, we thank the Football Clubs and The Football League. Thanks also to our participants who generously gave their time.
Aksut, S. V., Aksut, G., Karamehmetoglu, A., & Oram, E. (1986). The determination of serum estradiol, testos- terone and progesterone in acute myocardial infarction. Japanese Heart Journal, 27, 825±837.
Baker, F. (1888). Anthropological notes on the human hand. American Anthropologist, 1, 51±76.
Bardin, C. W., & Catterall, C. F. (1981). Testosterone: a major determinant of extragenital sexual dimorphism.
Barker, D. J. B. (1992). Fetal and infant origins of adult disease. London: BMJ Publishing Group.
Bryden, M. P., McManus, I. C., & Bulman-Fleming, M. B. (1994). Evaluating the empirical support for the Geschwind±Behan±Galaburda model of cerebral lateralisation. Brain and Cognition, 26, 103±167.
Darwin, C. (1871). The descent of man, and selection in relation to sex. London: Raven Press.
English, K. M., Mandour, O., Steeds, R. P., Diver, M. J., Jones, T. H., & Channer, K. S. (2000). Men with coronary artery disease have lower levels of androgens than men with normal coronary angiograms. European Garn, S. M., Burdi, A. R., Babler, W. J., & Stinson, S. (1975). Early prenatal attainment of adult metacarpal± phalangeal rankings and proportions. American Journal of Physical Anthropology, 43, 327±332.
Geschwind, N., & Galaburda, A. M. (1985). Cerebral lateralisation. Biological mechanisms, association and pathology: a hypothesis and a program for research. Archives of Neurology, 42, 428±654.
Heinonen, O. P., Slone, D., Monson, R. R., Hook, E. B., & Shapiro, S. (1977). Cardiovascular birth defects and antenatal exposure to female hormones. New England Journal of Medicine, 296, 67±70.
Kimura, D. (1996). Sex, sexual orientation and sex hormones influence human cognitive function. Current Opinion in Neurobiology, 6, 259±263.
Kimura, D. (1999). Sex and cognition. Cambridge: MIT Press.
Levy, E. P, Cohen, A., & Fraser, F. C. (1973). Hormone treatment during pregnancy and congenital heart defects.
J.T. Manning, R.P. Taylor / Evolution and Human Behavior 22 (2001) 61±69 Lorber, C. A., Cassidy, S. B., & Engel, E. (1979). Is there an embryo±fetal exogenous sex steroid exposure syndrome (EFESSES)? Fertility and Sterility, 31, 21±24.
Manning, J. T., Baron-Cohen, S., Wheelwright, S., & Sanders, G. (2001). The 2nd to 4th digit ratio and autism.
Developmental medicine and child neurology (in press).
Manning, J. T., & Bundred, P. E. (2000). The ratio of 2nd to 4th digit length: a new predictor of disease predisposition? Medical Hypotheses, 54, 855±857.
Manning, J. T., & Pickup, L. J. (1998). Symmetry and performance in middle-distance runners. International Journal of Sports Medicine, 19, 205±209.
Manning, J. T., Scutt, D., Wilson, J., & Lewis-Jones, D. I. (1998). The ratio of 2nd to 4th digit length: a predictor of sperm numbers and levels of testosterone, LH and oestrogen. Human Reproduction, 13, 3000±3004.
Manning, J. T., Trivers, R. L., Singh, D., & Thornhill, R. (1999). The mystery of female beauty. Nature, 399, Manning, J. T., Trivers, R. L., Thornhill, R., & Singh, D. (2000). The 2nd:4th digit ratio and asymmetry of hand performance in Jamaican children. Laterality, 5, 121±132.
McEwen, B. S. (1981). Neural gonadal steroid actions. Science, 211, 1303±1311.
Phelps, V. R. (1952). Relative finger length as a sex-influenced trait. American Journal Human Genetics, Phillips, G. B., Pinkernell, B. H., & Jing, T. Y. (1994). The association of hypotesteronemia with coronary artery disease in men. Arteriosclerosis and Thrombosis, 14, 701±706.
Robinson, S. J., & Manning, J. T. (2000). The ratio of 2nd to 4th digit length and male homosexuality. Evolution Rosano, G. M. (2000). Androgens and coronary heart disease. A sex-specific effect of sex hormones? European Rosano, G. M., Leonardo, F., & Pagnotta, P. (1999). Acute anti-ischaemic effect of testosterone in men with coronary artery disease. Circulation, 99, 1666±1670.
Sewardson, M., Vylithingum, S., Jailal, I., & Desai, R. K. (1990). Abnormalities in sex hormones are a risk factor for premature manifestation of coronary artery disease in South African Indian men. Atherosclerosis, Sluming, V. A., & Manning, J. T. (2000). Second to fourth digit ratio in elite musicians: evidence for musical ability as an honest signal of male fitness. Evolution and Human Behavior, 21, 1±9.
Swartz, C. M., & Young, M. A. (1987). Low serum testosterone and myocardial infarction in geriatric male patients. Journal American Gerontological Society, 35, 39±44.
Vandenberg, S. G., & Kuse, A. R. (1978). Mental rotations, a group test of three dimensional spatial visualisation.
Perceptual and Motor Skills, 47, 599±604.

Source: http://www.kafyeda.co.il/vault/documents/Ratio_08.pdf