Bergmann's rule and the thrifty genotype

One of Roberts' key contributions was his work demonstrating the applicability of several ecological rules to human populations (Roberts [1953] Am. J. Phys. Anthropol. 11:533-558; [1978] Climate and Human Variability, 2nd ed., Menlo Park, CA: Cummings). His finding that average body weight systematically covaries with mean annual temperature was widely taken as confirmation of Bergmann's rule for humans. More recently his findings on weight and temperature have been extended and confirmed (Ruff [1994] Yrbk. Phys. Anthropol. 37:65-407; Katzmarzyk and Leonard [1995] Hum Biol Council Program Abstracts 132) although the strength of the association may be decreasing when considering more recent surveys (Katzmarzyk and Leonard [1995]). Roberts noted in 1953 that Oceanic populations may be somewhat of an exception to Bergmann's rule, and we propose that Neel's ([1962] Am. J. Hum. Genet. 14:353-362) thrifty genotype model may account for some of the deviation from predicted weights among these populations. We provide an updated version of the thrifty genotype model, suggesting that selection for energetic efficiency may have occurred for some Oceanic populations during the voyaging to and settlement of their island homes. Under conditions of modernization the thrifty genotype may be manifesting as high rates of obesity and NIDDM among Polynesians and Micronesians. First, using measurements of adult male weight from 19 Oceanic populations, we demonstrate the extreme nature of their deviation from predicted weight based on Roberts' regression of weight on mean annual temperature. Next, we regress the deviations from predicted weight on NIDDM prevalence for these 19 populations, producing a highly significant regression (R2 = 0.46; P < 0.001), consistent with expectations if the thrifty genotype is responsible for the high weights.