Home  |  Archive  |  AJA @ Nature  |  Online Submission  |  News & Events  |  Subscribe  |  APFA  |  Society  |  Links  |  Contact Us  |  中文版

Second to fourth digit ratio (2D:4D) and testosterone in men

J. T. Manning1, S. Wood2, E. Vang2, J. Walton3, P. E. Bundred4, C. van Heyningen5, D. I. Lewis-Jones2,6
Department of Psychology, University of Central Lancashire, Preston PR1 2HE, UK
2Reproductive Medicine Unit , Liverpool Women's Hospital, Crown Street, Liverpool L8 7SS, UK
3Department of Medical Imaging, The University of Liverpool, PO Box 147, Liverpool L69 3BX,UK
4Department of Primary Care, The University of Liverpool, Liverpool L69 3GX,UK
5University Hospital Aintree, Liverpool L9 7AL, UK
6Department of Obstetrics & Gynaecology, University of Liverpool, Liverpool Women's Hospital, Crown Street, Liverpool L8 7SS, UK

Asian J Androl 2004 Sep; 6: 211-215

Keywords: 2nd to 4th digit ratio; 2D:4D; testosterone

Aim: To investigate the relationship between 2D:4D and testosterone in men attending an infertility clinic and men drawn from the general population. Methods: Data on 2D:4D and testosterone from two samples were collected: (1) 43 men attending an infertility clinic, and (2) 51 men drawn from the general population without regard to fertility. Results: In sample (1) there were negative associations between 2D:4D and testicular function, and men with lower 2D:4D in their right compared to left hand had higher testosterone levels than men with higher 2D:4D in their right compared to left hand. Sample (2) showed no significant associations between 2D:4D or side differences in 2D:4D and testosterone. Conclusion: Adult levels of testosterone may be related to aspects of 2D:4D in samples which contain men with compromised testicular function, but not in men from normative samples. Associations between 2D:4D and fertility-associated traits probably arise from early organisational effects of testosterone rather than from activational effects of current testosterone.

1 Introduction

The formation of the urinogenital system and the fingers is influenced by two groups of Homeobox genes known as Hoxa and Hoxd [1, 2]. These and similar observations led to the suggestion that aspects of gonad function, e.g. testosterone and sperm production, may be reflected in patterns of finger morphology [3, 4]. One obvious finger trait which may relate to gonad function is the ratio of the length of the 2nd ('index' finger) and 4th ('ring' finger) digits (2D:4D). The 2D:4D ratio shows sex differences such that males have lower mean values than females [3, 5], and the sex difference is stable across ethnic and racial groups [4]. Within samples of men drawn from infertility clinics high female-like 2D:4D has been shown to be associated with germ cell failure or lowered sperm counts [3], and unsuccessful surgical sperm retrieval [6].

An early origin of the sexual dimorphism in 2D:4D is indicated by reports that (i) adult ratios of finger lengths are established in utero by about week 14 [7] (ii) children as young as 2 years show the sex difference and there seems little change in 2D:4D at puberty [3] (iii) children with severely compromised growth in utero (birth weight <1 500 g) show correlations between 2D:4D and patterns of dermatoglyphic ridges which are fixed by the fourth month of fetal life [4] and (iv) 2D:4D in adult males is related to measures of their size at birth, such that men with low 2D:4D had below average placental weight and above average neonatal crown-heel length [8].

It has been suggested that prenatal sex steroids may be responsible for the sexual dimorphism in 2D:4D [3]. In support of this it has been found that (i) some traits that are seen more frequently in males are also associated with low values of 2D:4D e.g. left hand preference [9], autism [10], good visuo-spatial perception [11] and poor verbal fluency [4] (ii) the ratio of the circumference of the waist to hips (WHR) is sexually dimorphic such that women with low WHR have high oestrogen and low testosterone. The WHR of mothers is negatively related to the 2D:4D ratio of their children [12] (iii) the WHR of women is negatively related to their 2D:4D ratio [13] (iv) children with congenital adrenal hyperplasia, a trait associated with high prenatal androgen, have lower 2D:4D than healthy controls [14, 15] and (v) sensitivity to testosterone is determined by the number of CAG triplets found at one end of the androgen receptor gene. Normal variation in triplet number varies between 11 and 30 CAG repeats. Individuals with low numbers of CAG repeats tend to have low 2D:4D [16]. In many of these studies it is the 2D:4D ratio of the right hand which has turned out to be the strongest predictor of the index trait [4]. This may arise because sexually dimorphic traits often show the most intense expression of the 'male form' on the right side of the body [17]. The trait Dr-l (2D:4D right hand - 2D:4D left hand) has been shown to be related to testosterone as indicated by the structure of the androgen receptor gene [16]. That is men with low 2D:4D in their right hand relative to their left hand 2D:4D (negative values of Dr-l) are more sensitive to testosterone than men with positive values of Dr-l.

Therefore there is accumulating evidence that 2D:4D ratio is related to prenatal testosterone. There are less data to suggest that 2D:4D is associated with adult testosterone concentrations. The question is an important one because it affects our interpretation of 2D:4D associations. That is, if a trait such as sperm concentration per ejaculate is related to 2D:4D does this mean that it is influenced by early organisational effects of androgen or by both organisational and adult activational effects of the hormone? A relationship between 2D:4D and adult testosterone is suggested by studies of WHR. The 2D:4D ratio shows little evidence of change at puberty but the WHR shows increased sexual dimorphism at puberty. Despite this 2D:4D is negatively related to WHR in men and women [4]. WHR is positively associated with testosterone [18], so this is indirect evidence that 2D:4D is a correlate of adult androgen. However, to date there is only one study which has found a relationship between 2D:4D and adult testosterone. Manning et al [3] have reported a negative association between right hand 2D:4D and testosterone in 58 males attending an infertility clinic. The problems of extending that finding to males in the general population are obvious. Some men attending an infertility clinic will have compromised testicular function, others will not. The variance in testicular function in such a sample may be very great. A significant negative association between 2D:4D and testosterone may therefore arise because a substantial proportion of infertile males have high 2D:4D and low testosterone. Such an association may not be apparent in a normative sample in which infertile men are likely to be present in small numbers. We therefore investigated relationships between 2D:4D and testosterone in two samples. We measured 2D:4D and testosterone in (1) male participants attending an infertility clinic for elective surgical sperm retrieval and IVF and (2) a sample of males recruited from the general population without regard to fertility.

2 Materials and methods

2.1 Sample (1)

This consisted of 44 males undergoing elective surgical sperm retrieval with subsequent cryopreservation of spermatozoa. The characteristics of this sample are reported elsewhere [6]. In brief all patients were diagnosed as azoospermic following two separate semen analyses. The participants were azoospermic due to a history of non-obstructive azoospermia, previous vasectomy, or congenital absence of the vas deferens. The non-obstructive group were likely to have severely compromised testicular function. The remainder were likely to have normal testes. Non-obstructive azoospermia was diagnosed from testicular histology, low testicular volume and elevated levels of FSH. Blood samples (5 mL) were taken at the same time of day. Serum levels of total testosterone, LH and FSH were measured using radioimmunoassay (DPC Coat-a-Count, Euro/DPC Ltd, Gwynedd, UK: inter-laboratory coefficient of variation 8 to 12 %, UK-NEQAS). Testis volume was estimated by comparing testis size with a series of ovoids of known volume [19]. There was no information available regarding use of medication in this sample.

Digit lengths were measured from the ventral crease proximal to the palm to the tip of the finger, using vernier callipers recording to 0.01 mm. Second measurements were made in order that repeatabilities of the 2D:4D ratio could be calculated. All the measurements were made by one observer with right and left hands measured first and this procedure repeated after a period of at least 5 minutes blind to first measurements. One participant was excluded from the study due to a previous amputation of the index finger. The protocol was agreed by the local ethical committee, and signed informed consent was obtained from all patients.

2.2 Sample (2)

The second sample consisted of a normative sample of males who were contacted via their family doctor. The relationship between 2D:4D and testosterone has not been previously investigated in a sample of males from the general population. We were therefore unable to perform a power calculation. There were 52 males in the sample. The participants were informed that the study concerned measurements of finger length and of testosterone from the blood. One participant reported taking medication (lipid-lowering drugs). He was excluded from the sample. All others reported good health. Blood samples (5 mL) were taken between 10.00 am and 4.00 pm. Serum levels of total testosterone were measured using the DPC Coat-a-Count immunoassay (Euro/DPC Ltd, Gwynedd, UK: inter-laboratory coefficient of variation 8 to 12 %, UK-NEQAS). Digit lengths and repeated measurements were made as in sample (1). The methodology was agreed by the local ethical committee, and signed informed consent was obtained from all participants.

3 Results

3.1 Repeatability of 2D:4D Measurements

We used Model II single factor ANOVA tests to calculate the intra-class correlation coefficient or repeatability (r1) of our measurements (r1= groups MS - error MS/groups MS + error MS; where MS=mean squares) [20]. Repeated measures ANOVA tests were used to calculate the ratio (F) between groups MS (i.e. differences between individuals) and error MS (i.e. measurement error). The intra-class coefficient was high for both studies, and the variance of our between-subject measures was significantly greater than the variance of our repeated measures i.e. our measurement error (Study 1 left hand r1 = 0.982, F = 104.38, P = 0.0001; right hand r1 = 0.991, F = 135.80, P = 0.0001; Study 2 left hand r1= 0.997, F = 605.77, P = 0.0001, right hand r1 =
0.999, F = 1419.05, P = 0.0001). We concluded that our measurements of 2D:4D showed real differences between subjects which were much greater than our measurement error.

3.2 Statistics

Descriptive statistics (means and SD's) of the samples were as follows:

Sample (1) age 39.007.57 years; 2D:4D right hand 0.970.04, left hand 2D:4D 0.970.03; testosterone 16.996.66 nmol/L; LH 6.073.85 U/L; FSH 12.29 1.77 U/L; mean volume of right and left testis 19.615.83 mL.

Sample (2) age 56.484.80 years; 2D:4D right hand 0.970.03, left hand 2D:4D 0.960.03; testosterone 16.705.99 nmol/L.

Therefore testosterone levels were similar in Samples (1) and (2) but participants in Sample (2) were older than Sample (1). The samples showed similar means for 2D:4D.

3.3 Digit Ratios and Testosterone

For sample (2) we found no difference in testosterone concentrations across the period of sampling from 10.00 to 4.00 (simple linear regression b = 0.26, F = 0.29, P = 0.59). There were also no significant differences in testosterone with age in either sample (Sample 1 b = -0.15, F = 1.21, P = 0.27; Sample 2 b = 0.01, F = 0.004, P = 0.95).

The relationships between 2D:4D, Dr-l and testosterone are given in Table 1. In sample (1) these were non-significant for right 2D:4D (P = 0.54) and left 2D:4D (P = 0.14). Dr-l showed a significant negative relationship with testosterone (P = 0.004, Figure 1). We also found that participants with negative values of Dr-l had higher testosterone concentrations than participants with positive values of Dr-l (negative Dr-l mean testosterone 19.077.02 nmol/L, postive Dr-l mean testosterone 13.494.24 nmol/L, t = 2.88, P = 0.006). For Sample (2) we found no significant associations between testosterone and right 2D:4D (P = 0.54), 2D:4D left (P = 0.45), or Dr-l (P = 0.10). The later correlation was the strongest in Sample (2) and in the same direction as in Sample (1). However, participants with negative Dr-l did not have significantly higher testosterone than participants with positive Dr-l (- Dr-l mean testosterone 17.556.84 nmol/L, + Dr-l mean testosterone 15.895.06 nmol/L, t = 0.99, P = 0.33).

Table 1. Relationships (simple linear regression analysis) between testosterone and finger ratios in (1) 43 azoospermic men and (2) 51 normal controls.


Right hand 2D:4D

Left hand 2D:4D

Dr-l (right-left)







Azoospermic men







Normative sample







Figure 1. Relationship between total serum testosterone and Dr-l (right hand 2D:4D - left hand 2D:4D) in 43 azoospermic men with azoospermia. Participants with lower right hand 2D:4D than left hand ratio have high testosterone concentrations. The equation for the line is y = -107.88x + 16.84, r2 =0.18.

3.4 Digit Ratios and Testicular Function

Table 2 shows relationships between 2D:4D, Dr-l and measures of gonadotrophins and testicular function i.e. LH, FSH and mean testis volume in Sample (1). The 2D:4D ratio was positively associated with LH concentrations (right P = 0.001; left P = 0.0001). Significant positive associations were also seen for FSH (right P = 0.01, left P = 0.03). Mean testis volume was significantly negatively related to right 2D:4D (P = 0.02) but was just above significance for left 2D:4D (P = 0.07). Dr-l was not significantly related to LH, FSH or testis volume.

Table 2. Relationships (simple linear regression analyses) between measures of gonadotrophin and testicular function and finger ratios in 43 azoospermic men.


Right hand 2D:4D

Left hand 2D:4D

Dr-l (right-left)





















Testis Volume







4 Discussion

The length of the 2nd and 4th digit (2D:4D) is related to prenatal testosterone. Men with low right hand 2D:4D and low right hand 2D:4D in relation to left may have experienced high in utero testosterone. The association between 2D:4D and adult testosterone is less clear. One report, from a sample of men attending an infertility clinic, showed a negative correlation between 2D:4D and testosterone. However, the relationship may have been strongly influenced by subjects with compromised testicular function.

We have found evidence that digit ratio is related to testosterone in a sample of men attending an infertility clinic. Participants with lower right 2D:4D than left 2D:4D (negative Dr-l) had higher testosterone than men with higher right than left ratio (positive Dr-l). Also LH was low in subjects with low 2D:4D. Testosterone exerts negative feedback on LH [21]. The associations between low values of 2D:4D and high testosterone and low LH indicate that gonadal production of testosterone is related to 2D:4D in men attending infertility clinics. 2D:4D was also predictive of correlates of germ cell function. Damage to the seminiferous epithelium leads to azoosper-mia, a rise in FSH and a reduction in testicular volume [19]. In this sample men with low 2D:4D had low FSH and large testes. We found no evidence for significant associations between right and left 2D:4D and testosterone in our normative sample. As with the infertility clinic sample a negative relationship was found between Dr-l and testosterone but this was not significant.

These data and those of Manning et al [3] suggest that adult testosterone concentrations may be negatively related to 2D:4D and Dr-l. However, the associations are only significant in samples from infertility clinics i.e. where high numbers of azoospermic men are present. The relationships are weaker when we consider a normative sample. This suggests that 2D:4D or Dr-l are not strong correlates of adult concentrations of testosterone in men who are sampled without regard to fertility. Of course we cannot completely exclude a weak association between 2D:4D and testosterone in normative samples. Indeed the association between 2D:4D and WHR supports such a relationship. However, the impression remains that 2D:4D is a better indicator of prenatal testosterone than of adult testosterone.


[1] Mortlock DP, Innis JW. Mutation of HOXA13 in hand-foot-genital syndrome. Nat Genet 1997; 15: 179-80.
[2] Kondo T, Zakany J, Innis JW, Duboule D. Of fingers, toes and penises. Nature 1997; 390: 29.
[3] 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, luteinising hormone and oestrogen. Hum Rep 1998; 13: 3000-4.
[4] Manning JT. Digit Ratio: a Pointer to Fertility, Behavior and Health. New Jersey: Rutgers University Press; 2002. p 24-40.
[5] Phelps VR. Relative index finger length as a sex-influenced trait in man. Am J Hum Genet 1952; 4: 72-89.
[6] Wood S, Vang E, Manning J, Walton J, Troup S, Kingsland C, et al. The ratio of second to fourth digit length in azoospermic males undergoing surgical sperm retrieval: predictive value for sperm retrieval and on wubsequent fertilization and pregnancy rates in IVF/ICSI cycles. J Androl 2003; 24: 871-7.
[7] Garn SM, Burdi AR, Babler WJ, Stinson S. Early prenatal attainment of adult metacarpal-phalangeal rankings and proportions. Am J Phys Anthropol 1975; 43: 327-32.
[8] Ronalds G, Phillips DI, Godfrey KM, Manning JT. Fetal growth and the ratio of second to fourth digit length. Early Hum Dev 2002; 68: 21-6.
[9] Manning JT, Barley L, Walton J, Lewis-Jones DI, Trivers RL, Singh D, et al. The 2nd:4th digit ratio, sexual dimorphism, population differences, and reproductive success. evidence for sexually antagonistic genes? Evol Hum Behav 2000; 21: 163-83.
[10] Manning JT, Baron-Cohen S, Wheelwright S, Sanders G. The 2nd to 4th digit ratio and autism. Dev Med Child Neurol 2001; 43: 160-4.
[11] Manning, JT, Taylor RP. Second to fourth digit ratio and male ability in sport: implications for sexual selection in humans. Evol Hum Behav 2001; 22: 61-9.
[12] Manning JT, Trivers RL, Singh D, Thornhill R. The mystery of female beauty. Nature 1999; 399: 214-5.
[13] Manning JT, Trivers RL, Thornhill R, Singh D. The 2nd:4th digit ratio and asymmetry of hand performance in Jamaican children. Laterality 2000; 5: 121-32.
[14] Okten A, Kalyoncu M, Yaris N. The ratio of second- and fourth-digit lengths and congenital adrenal hyperplasia due to 21-hydroxylase deficiency. Early Hum Dev 2002; 70: 47-54.
[15] Brown WM, Hines M, Fane BA, Breedlove SM. Masculinized finger length patterns in human males and females with congenital adrenal hyperplasia. Horm Behav 2002; 42: 380-6.
[16] Manning JT, Bundred PE, Newton DJ, Flanagan BF. The 2nd to 4th digit ratio and variation in the androgen receptor gene. Evol Hum Behav 2003; 24: 399-405.
[17] Tanner JM. Foetus into Man: Physical Growth from Conception to Maturity. Princeton: Harvard University Press; 1990. p27.
[18] Singh D. Adaptive significance of female physical attractiveness: role of waist-to-hip ratio. J Pers Soc Psychol 1993; 65: 293-307.
[19] Pryor JP. Azoospermia. In: Hargreave TB, editor. Male Infertility. Berlin: Springer-Verlag; 1983. p212-27.
[20] Zar JH. Biostatistical Analysis. London: Prentice-Hall; 1984. p323.
Wu F. Endocrinology of male infertility and fertility. In: Hargreave TB, editor. Male Infertility. Berlin: Springer-Verlag, 1983. p75-86.

Correspondence to: Dr. J. T. Manning, Department of Psychology, University of Central Lancashire, Preston PR1 2HE, UK.
Tel: +44-1772-894 467, Fax: +44-1772-892 925
Email: jtmanning@uclan.ac.uk
Received 2003-12-24    Accepted 2004-05-08