Spousal Choice by Height in an Urban Middle-Class Japanese Population

Dear Editor,

In Western populations, the similarity between spouses is well documented for a variety of anthropometric variables such as height and body mass index. However, relatively little information is available on assortative mating in height in other ethnic populations. A study on a small, exclusively agricultural community on Kyushu island (Furusho 1961) was quoted in English literature as representative of the Japanese population. The study showed that spousal correlation coefficient was +0.072 ± 0.057 in 431 couples, and based on this finding, it was concluded that Japanese do not practice homogamy for body size (Spuhler 1982).

Nevertheless, the current agricultural population comprises <3% of the entire population of Japan, and the above study does not represent the Japanese population at large. Further, that study was conducted in the postwar period of 1959, when the farming community suffered from a shortage of male due to compulsory draft and war deaths in preceding decades, making assortative mating for anthropometric variables more or less unrealistic in this population.

Another study addressed spousal correlation for height in Japanese urban populatons in the 1930s (Yoshida 1941). The study subjects were 2,410 couples who were parents of school children and residents in Tokyo and other urban communities. The spousal correlation for height was 0.220 ± 0.019, indicating the presence of positive assortative mating in urban Japanese populations.

We present additional data to indicate a stable, weak spouse correlation for height for urban Japanese population. The study subjects were 421 Japanese couples who gave birth to their children at a company-owned general hospital in Tokyo in 1984–1993. Their heights were measured one month after delivery by the same stadiometer in the clinic by trained pediatric nurses. Of the subjects, either male spouse or female spouse was an employee of the company that runs the hospital.

Demographic characteristics of the study subjects are given in Table 1. The mean (SD) height was 170.5 (5.8) cm for male spouses and 157.3 (5.0) cm for female spouses, each being comparable to the mean height of the 2009 national survey for 17.5-year-olds: 170.8 cm for boys and 157.9 cm for girls. We also compared these mean heights with those of the 1976 national survey for 17.5-year-olds for boys and the 1978 national survey for 17.5-year-olds for girls (168.8 cm for males, 156.6 cm for females). The results indicate that the mean height of our study subjects was slightly taller than that of the 1976–1978 national survey data for 17.5-year-olds of both sexes.

Table 1.

Demographic Characteristics of the Study Subjects

Table 2.

Spousal Correlations in Height

Table 2 shows the correlation coefficient between spousal heights for the 421 couples, along with those of previous studies on Japanese couples. The correlations in the two studies on urban populations are in agreement with each other, but there is no spousal correlation for height in the 431 couples from the agricultural population.

Table 3 shows different levels of sexual dimorphism in stature in the 421 couples studied. The couples were divided into six levels by magnitude of male-female height ratio based on the scheme of Pawlowski (2003). There were 394 male-taller couples (93.6%) and 3 male-shorter couples (0.7%).

The results of the present study provide additional evidence for stable, positive assortative mating for height among urban Japanese population, as measured in the 1930s and in 1984–1993. We also identified that a male-taller norm exists in contemporary urban Japanese population. Collectively, both studies do not support the argument that Japanese do not practice homogamy for body size (Spuhler 1982).

A Swedish population-based growth study (n = 2,402) and the British National Child Development Study (n = 6,192) gave nearly identical results in the spousal correlation coefficient, 0.27 for Swedish population and 0.28 for British population, suggesting a common phenotypic...