Prevalence of obesity in multi-racial vs. mono-racial individuals

Despite the weird sounding title (why "ethnic admixture adults "? ... why not: "ethnically admixed adults" or something like that?) this paper (see below) examines quite an interesting question.
Do mixed-race individuals have a higher prevalence of obesity than single-race individuals?
First off, they mention that there is a high proportion (20%) of individuals who report two or more ethnicities in Hawaii, compared to the rest of the US (2%).
The introduction does not state why they expect to see the difference that they are looking for.
They have a humongous sample size (n=215,000), but the data is based on self-report which may bias the results.

The highest prevalence of overweight was for Hawaiian/Latino men (88%,) and black/Latina women (74.5%). The highest prevalence of obesity occurred in Hawaiian/Latino men (53.7%) and Hawaiian women (39.2%). The prevalence of obesity in men and women with Asian/white, Hawaiian/white, Hawaiian/Asian, Latina/white, and Hawaiian/Asian/white ethnic admixtures was significantly higher (P less than 0.0001) than the average prevalence of the ethnic groups with whom they share a common ethnicity/race.

and the money line:

Across all of the ethnic admixtures, the prevalences for both overweight and obesity were similar to or higher than the average of the prevalence estimates for their shared ethnicities; in contrast, none of the admixture prevalences were less than the average of their component ethnicities.

I can't believe they don't provide very much discussion about some of the potential mechanisms that would underlie their findings. They do find that:

"a high caloric intake (e.g., calories from fat and alcohol) and exercise did modestly decrease this difference" (referring to difference in prevalence between ethnic admixturesand monorace adults)"

There might be social and psychological factors that bi-racials experience more often than mono-racials, and this may negatively influence their health.
Nevertheless, this study is interesting and the finding merits future exploration.

The Prevalence of Obesity in Ethnic Admixture Adults
Cheryl L. Albright, Alana D. Steffen, Lynne R. Wilkens, Brian E. Henderson and Laurence N. Kolonel
Obesity (2008) 16 5, 1138–1143. doi:10.1038/oby.2008.31

Abstract: Objective: To determine whether the prevalence of obesity in ethnic admixture adults varies systematically from the average of the prevalence estimates for the ethnic groups with whom they share a common ethnicity.Methods and Procedures: The sample included 215,000 adults who reported one or more ethnicities, height, weight, and other characteristics through a mailed survey.Results: The highest age-adjusted prevalence of overweight (BMI 25) was in Hawaiian/Latino men (88% ; n = 41) and black/Latina women (74.5% ; n = 79), and highest obesity (BMI 30) rates were in Hawaiian/Latino men (53.7% ; n = 41) and Hawaiian women (39.2% , n = 1,247). The prevalence estimates for most admixed groups were similar to or higher than the average of the prevalences for the ethnic groups with whom they shared common ethnicities. For instance, the prevalence of overweight/obesity in five ethnic admixtures—Asian/white, Hawaiian/white, Hawaiian/Asian, Latina/white, and Hawaiian/Asian/white ethnic admixtures—was significantly higher (P less than 0.0001) than the average of the prevalence estimates for their component ethnic groups.Discussion: The identification of individuals who have a high-risk ethnic admixture is important not only to the personal health and well-being of such individuals, but could also be important to future efforts in order to control the epidemic of obesity in the United States.