Sexual orientation, body size, and shape

Using MIDUS data, I lumped together homosexuals and bisexuals, giving me 53 gay men and 45 lesbians. I then calculated waist-to-hip ratios for everyone:

Mean waist-to-hip ratio

Straight men .97
Gay men .95
Straight women .85
Lesbians .88

Compared to gay guys, straight men have more tubular (masculine) bodies, while among women, lesbians have the more straight-up-and-down shape. The differences are statistically significant. The guy gap is only two-tenths of a standard deviation (SD), but the female gap is four-tenths of an SD--a moderate difference. 

Next, I calculated the mean Body Mass Index (BMI) for each group:

Mean Body Mass Index

Straight men 28.35
Gay men 27.21
Straight women 27.53
Lesbians 28.27

Straight men and lesbians have similar numbers and are heavier than their gender counterparts. Gay men and straight women also have similar BMIs (gay men are actually thinner). The differences, however, are not statistically significant.

Finally, let's look at height:

Mean height (inches)

Straight men 70.2
Gay men 69.6
Straight women 64.3 
Lesbians 64.8

For the men, straight guys are 6/10 of an inch taller. Lesbians are half of an inch taller than hetero females, but neither of the differences are significant.

Overall, the numbers suggest that gay men are more feminine than straight men, while lesbians are more masculine than heterosexual women.  

Ethnic differences in waist-to-hip ratio

Women participating in the Midlife Development in the United States Study (2004-06) were measured for waist-to-hip ratio (WHR). A ratio of 0.7 is considered to be ideal for health and attractiveness. Here are means by race/ethnic group (N = 1,891).  

Mean waist-to-hip ratio

Mexican .88
Amerindian .88
French .87
Black .86
Scottish .86
Swedish .86
English .85
Irish .85
Polish .85
Russian .85
Norwegian .84
German .84
Asian .84
Jewish .84
Italian .83

The differences seem small, but the gap between Italian and Mexican Americans is sixth-tenths of a standard deviation. The groups are roughly 1 1/2 to 2 standard deviations above the ideal--they're all fat--but keep in mind that this is a sample of middle-aged women, with most of the sample between the ages of 40 and 60.