Insulin resistance can lead to β
-cell decompensation and eventually type 2 diabetes [40
]. We propose that the selection pressure for insulin resistance was relaxed first in Europeans when dietary carbohydrate increased 12,000 years ago with the advent of agriculture [15
]. In accordance with this long-term exposure, Europeans have experienced a lower prevalence of diabetes, even when overweight and obese (see Section 6
), compared to other population groups. High carbohydrate intake over the past millennia may have also increased positive selection for genes associated with higher β
-cell mass. In Swedish and Finnish populations, variants in 11 genes (TCF7L2, PPARG, FTO, KCNJ11, NOTCH2, WFS1, CDKAL1, IGF2BP2, SLC30A8, JAZF1,
) are significantly associated with the risk of type 2 diabetes independently of clinical risk factors [41
]. Variants in 8 of these genes were associated with impaired β
-cell function. Populations that adopted agriculture more recently such as the Pima Indians are likely to have a higher prevalence of genes transcribing insulin resistance or impaired β
-cell function than those exposed for thousands of years.
The Carnivore Connection hypothesis has been tested recently in sample populations from the Asian steppes [42
]. Ten candidate genes for insulin resistance, anthropometry, and physiological measures, including HOMA insulin resistance, were compared in traditional herders (pastoralists = high-protein diet) and farmers (agriculturalists = high-carbohydrate diet). While none of the genes tested showed causal mutations with higher frequency in herders, tests of neutrality showed some genes (SLC30A8, LEPR, and KCNQ1) could have been involved in past adaptations to diet. Consistent with the Carnivore Connection hypothesis, the prevalence of insulin resistance was significantly greater in herders compared to farmers, despite no major differences in current diet.
Environmental pressures such as geographic isolation and/or starvation may have led to further increases in the prevalence of insulin resistance gene(s) in certain population groups. Geographic isolation has led to genetic bottlenecks and reduced genetic diversity in the Nauruans and Pima Indians [43
]. Both populations have also been exposed to food shortages and starvation in the recent past. As occurs with low carbohydrate intake, starvation results in the need for increased gluconeogenesis and peripheral insulin resistance [44
]. This may have selected for those with a profound degree of insulin resistance which was inherited by future offspring. Women with polycystic ovarian syndrome are known to be exceptionally insulin resistant and may represent a group that was highly fertile in a low-carbohydrate environment [45
Insulin resistance has previously been proposed to be the mechanism for coping with variable food intake during evolution [46
]. Neel's thrifty gene hypothesis postulates that cycles of feast and famine selected for a “quick insulin trigger” (postprandial hyperinsulinemia) as a mechanism to increase fat stores during food abundance and available during food scarcity [48
]. An alternative hypothesis by Reaven [49
] suggests that muscle insulin resistance was the key to survival during food scarcity because it conserved glucose by minimizing gluconeogenesis and preserving lean body mass.
Both these hypotheses are based on the assumption that there were challenging periods of food scarcity prior to the advent of agriculture. However, this is not supported by the scientific literature [50
]. While hunter gatherers would have been exposed to seasonal and geographical changes in food supply, severe food shortages or starvation were rare and more likely to occur after
the transition to agriculture (preindustrialization). Specific mechanisms for coping with low carbohydrate intake, rather than total dietary energy, probably afforded the greatest reproductive and survival advantages.
Genome-wide scans and other research have been directed towards discovering the gene(s) associated with insulin resistance and type 2 diabetes. However, the complexity of human metabolism means that there are likely to be multiple gene systems involved, including those that influence insulin sensitivity and others that influence β
-cell function [51
]. PC-1 (plasma cell membrane glycoprotein-1), which interferes with insulin receptor tyrosine kinase activity, thereby inhibiting subsequent cellular signalling [52
], has been associated with insulin resistance [53
]. Reduced expression of the “susceptibility” gene CAPN10 (calpain-10) has been linked to decreased glucose uptake in skeletal muscle [54
]. Recently, the ACAD10 (acyl coenzyme A dehydrogenase 10) gene has been associated with fatty-acid-induced insulin resistance [55
] as it may catalyze mitochondrial fatty-acid oxidation [56