The thrifty gene hypothesis
The thrifty genotype hypothesis was proposed by the genetic epidemiologist J.V. Neel in 1962, and has been highly influential since then. As originally proposed, the hypothesis assumed diabetes to be a single gene disorder, characterised by a metabolic abnormality present from birth. This abnormality, proposed to result in faster and more efficient insulin action, would confer a small but useful metabolic advantage upon heterozygote carriers of the gene, but would produce juvenile diabetes - a lethal trait - in homozygotes. The thrifty gene would (by analogy with sickle cell anaemia) exist in balanced polymorphism within the gene pool, conferring a selective advantage upon heterozygotes which balanced the mortality in homozygotes. Despite the obsolete nature of some of these assumptions, the hypothesis has had enduring appeal, potentially helping to explain the epidemic nature of diabetes in newly affluent populations. There are however many serious objections to the hypothesis, not least the failure to identify major thrift genes. Instead we have a thrifty genome which represents the interaction of hundreds of genes capable of influencing adaptations to feast and famine.
Two broad explanations of the origins of type 2 diabetes have been proposed; the thrifty genotype hypothesis and the thrifty phenotype hypothesis. The thrifty genotype hypothesis originated with the population geneticist J.V. Neel (1915-2000). In his formative years Neel worked with Drosophila, sickle cell anemia and the effect of ionizing radiation upon the survivors of nuclear explosions in Japan. Later he moved to Ann Arbor, Michigan, and established the Department of Human Genetics. He also became known for his work examining the genetic and phenotypic characteristics of isolated human populations living under conditions approximating to those of our distant ancestors, but his thrifty gene hypothesis predated this experience, and appears to have been strongly influence by his interaction with Dr Stephen Fajans at Ann Arbor.
At the time Neel wrote, diabetes was thought to be a single gene disorder. One common hypothesis was that homozygosity – two copies of the gene – was responsible for juvenile diabetes, and that heterozygosity was responsible for late onset diabetes. Since diabetes is so common, it was reasonable to suppose that the gene conferred some form of selective advantage, an idea first expressed by Aschner and Post in 1957[^2]. Neel, who appears to have had little previous interest in diabetes, drew upon a number of ideas that were then current in the field. It was commonly assumed that the metabolic defect associated with diabetes was present from birth. This being the case, and given the high prevalence of diabetes, it might be supposed that the abnormality conferred some selective advantage. Neel speculated that the gene might in some way allow for more rapid insulin release resulting in more efficient glucose disposal, allowing small but useful excess amounts of fat to be stored against times of scarcity. Unfortunately for the hypothesis, he shared the contemporary belief in a circulating insulin antagonist, and he went on to suggest that rapid release of insulin would, in times of plenty, provoke secretion of this antagonist, resulting in diabetes[^3].
A gene that enabled nutrients to be stored more efficiently could, as he pointed out, be of great benefit during times of famine. The natural analogy was with sickle cell anaemia, in which heterozygosity conferred a selective advantage in terms of resistance to malaria, whereas homozygosity was lethal. The balance between selective advantage and death would determine the spread of the gene within the gene pool, but would prevent it from taking over completely. This was the concept of balanced polymorphism.
It can readily be seen that the hypothesis rested upon a number of key assumptions, some of which are now obsolete. It was assumed that diabetes was a single gene disorder, and that this disorder was associated with a metabolic adaptation which would prove useful in times of scarcity, but maladaptive and harmful in times of plenty. The concept of circulating insulin antagonists proved to be fantasy. In sum, neither the gene nor the defect appear to exist in the form proposed by Neel, but the concept has had undoubted resonance, and has indeed been described as “arguably one of the most influential hypotheses in genetic epidemiology”.
Subsequent views of the Hypothesis
Despite its limitations, the thrifty gene (or genotype) has marked intuitive appeal. It could for example help to account for the rise of type 2 diabetes in affluent westernised societies early on in the twentieth century, and for its explosive spread to newly affluent societies later in the century. It is reasonable to believe that metabolic adaptations designed to improve survival during times of famine are coded for within the human gene pool, and “obesity genes” such as FTO have indeed been identified.
Like other alluring notions, the thrifty genotype hypothesis does not withstand much scrutiny, for it is both self-evident and uninformative. It is self-evident that evolution has shaped us to be energy efficient; the human population has always expanded to the limits of subsistence, and we are descended from the survivors of semi-starvation and episodic famine over countless generations. It is uninformative because individual gene variants such as FTO which undoubtedly enhance energy efficiency are merely one among the multitudinous pathways involved. Ours is a thrifty genome, fine-tuned by evolution to maximize efficient energy storage in a range of environments, and in differing ways in different populations. How this might translate into obesity and diabetes under conditions of plenty remains unclear.
Neel himself acknowledged some of these limitations in a retrospect written in 1998, two years before his death. He accepted that single gene effects must only play a minor role in genetic susceptibility, and the overwhelming preponderance of lifestyle influences upon the “diseases of civilization”. In particular he emphasised the links between hypertension, hyperlipidaemia and diabetes, and proposed that we should instead think of the “syndromes of altered genetic homeostasis”.