Strength and Anthropometric Measures in Identical and Fraternal Twins: No Evidence of Masculinization of Females with Male Co-Twins

In several species of litter-bearing mammals, intrauterine positioning predicts postnatal characteristics. ¹ Female animals gestated between two male fetuses are exposed to increased levels of androgens; this has been shown to result in masculinization of the female fetus with regard to a broad range of morphological and behavioral postnatal characteristics. ^2–5

Human twins can be identical (MZ), fraternal of the same sex (DZss), or fraternal of the opposite sex (DZos). Pregnancy with DZos is the human analogy to the intrauterine position phenomenon in litter-bearing animals. The female DZos fetus shares the intrauterine environment with a male fetus and is potentially exposed to higher androgen levels produced by the male co-twin. This exposure has been offered as an explanation for the higher degree of masculinization of female DZos observed in some studies. ^6–8

In the present twin study, we focused on strength and body composition measures, phenotypes that are known to depend on intrauterine positioning in animals, ^1,2,4 and that are also important predictors of human morbidity and disability. ^9,10 We used data from a large newly established panel of middle-aged identical and fraternal twins identified in a population-based registry.

Subjects and Methods

We identified pairs of twins born between 1931 and 1952 through the Danish Twin Registry. ¹¹ The twins had responded to a brief mailed questionnaire (response rate 77%) in 1997 and had declared their willingness to participate in future studies (90% of responders). This questionnaire included items on similarity of the twins, based on which we assigned zygosity, a method found to result in misclassification rates of less than 5%. ¹² Within each of the 22 birth cohorts, we randomly identified 40 pairs of each zygosity (MZ, DSss, and DZos). Half of the retrieved MZ and DZss twin pairs were female. Owing to insufficient numbers of MZ pairs in the birth cohorts of 1933, 1934, and 1936, we retrieved an additional 11 MZ pairs from the birth cohorts of 1931 and 1935. A total of 2,640 pairs were identified. Shortly before the survey we obtained demographic information on these twins from a continuously updated nation-wide population register, the Danish Civil Registration System. Ninety-one twins (1.7%) had died or emigrated in the 2-year period preceding the study, leaving 5,189 twins eligible for participation in the study.

Interviews were conducted by a total of 100 interviewers from the Danish National Institute of Social Research, an approach previously used in twin research projects. ¹³ The survey, which lasted on average 11/2 hours, comprised a questionnaire, tests of cognitive and physical functioning, and sampling of DNA.

All interviewers received a detailed training program by a physician and were closely monitored in the 6-month period (October 1998 through March 1999) during which the interviews were completed. The twins were considered not possible to contact only when at least three unsuccessful attempts to contact them at their residence were made. To avoid interviewer bias, which would inflate twin similarity, twins from a pair were never interviewed by the same interviewer. A pilot study testing all procedures resulted in minor changes only.

We compared participants and non-participants with regard to age, gender, urban residency, and marital status using data from the Civil Registration System. Height in centimeters and weight in kilograms were self-reported and were used to calculate the body mass index (weight/height² kg/m²). With a tape measure, the interviewers took two measurements of the subjects’ waist circumference between the lowest rib margin and the iliac crest and recorded the values to the closest 0.1 cm. The average of the two measures was used in this analysis. For grip strength, we identified the maximum value among the recorded measurements (three for each hand for 97.8% of the sample).

Results

Eighty-five percent of male and 81.3% of female eligible twins participated (Figure 1). Compared with non-participants, participants were slightly younger, more frequently married, and lived less frequently in urban areas (Table 1). These differences remained when the sample was stratified by zygosity (results not shown). Pair-wise participation was greater for MZ pairs (77%) than DZ pairs (72%).

After exclusion of twins with missing values on variables of interest, the sample was reduced to 4,223 twins (97.9%). The mean age of these subjects was 56.9 years (SD = 6.3) and varied 0.3 years or less in sex and zygosity strata (Table 2). Mean values and standard deviations for grip strength and body composition were virtually identical for MZ, DZss, and DZos in both males and females. In particular, the mean values of female DZos for all measures differed by less than 1% from the corresponding means for other female twins (MZ and DZss), and the differences were not consistently in one direction.

Discussion

Animal studies provide strong support for masculinization of female fetuses placed between males in utero. Nevertheless, the results of this study are not consistent with the hypothesis of masculinization of human female DZos. In this large sample of twins, members of DZos pairs were strikingly similar to MZ and DZss of the same sex with respect to an objective measure of strength and anthropometric measures.

Previous studies of the potential masculinization of human female DZos have yielded conflicting results with respect to physical and physiologic variables. Boklage reported an intermediate gender phenotype with regard to craniofacial growth for DZos twins. ¹⁴ McFadden reported a masculinized pattern of spontaneous otoacoustic emissions in female DZos compared with female DZss and female non-twins. ⁶ Studies of birth weight have yielded conflicting results. ^15–17 In a sample of Australian female twins, height and some items related to reproductive functions showed minor effects consistent with a masculinization of female DZos. ¹⁸ In a large Danish study, fecundability measured by waiting time to first ever pregnancy did not differ by zygosity or sex of co-twin. ¹⁹

Studies of behavioral traits, such as sensation seeking ⁷ and attitudes, ⁸ seem to show more consistent results in favor of a more masculine phenotype in DZos. However, the effects of rearing (female DZos growing up with a brother) and intrauterine exposures cannot be adequately discerned in these studies. Physical and physiological measures may in fact also be confounded by the effects of psychosocial rearing in DZos twins, an effect we would expect to be less pronounced in a Danish setting, which is characterized by minimal class variability and sexism. If such confounding is present, however, we would expect it to bias female DZos phenotypes toward a greater similarity with male phenotypes. Clearly this is not the case in the present study.

The participation rate in the present study was high. Non-participation was not highly selected, but showed patterns recognized from other studies: a tendency for lower participation of fraternal twins, persons who were not married, and urban area dwellers. Interestingly, males were more likely to participate in this as well as in previous Danish twin studies, ^13,20 a pattern opposite to what we expected from the literature. ²¹

In conclusion, the data of this large study of twins using both objective measurements and self-reported data provide no evidence of a more masculinized phenotype in female DZos.