Calorie restriction is the best studied of all interventions shown to slow aging and extend life in short-lived laboratory species. In humans it produces significant health gains, somewhat greater than any established medical technology can provide to essentially healthy individuals, at least until the broader advent of senolytic drugs. Unfortunately, it does not extend life by any great degree in long-lived species such as our own. The response to calorie restriction serves to increase evolutionary fitness during periods of famine, increasing the odds of individuals surviving to reproduce once food is plentiful again. Seasonal famines are of a given length, long relative to a mouse life span, short relative to a human life span, so only the mouse evolves to live 40% longer in calorie restricted conditions.
The mechanisms by which calorie restriction produces benefits broadly overlap with those of fasting, and in recent years some research groups have made inroads in finding the 80/20 point of calorie intake in humans at which a low calorie intake produces most of the benefits of a zero calorie intake. Calorie restriction upregulates the operation of cellular maintenance processes such as autophagy and the unfolded protein response, which leads to better cell and tissue function over the long term. It also produces sweeping changes in the operation of cellular metabolism, but autophagy appears to be the critical mechanism that mediates effects on long term health and longevity.
Some of the effects of calorie restriction on metabolism are similar enough to aspects of diabetes for the state to be called pseudo-diabetes, or beneficial diabetes. Relatedly, mTOR inhibitors are used to mimic some of the effects of calorie restriction, and the first generation of such inhibitors have undesirable side-effects that are somewhat diabetes-like. In today's open access paper, the author argues that the research community too readily categorized the side-effects of mTOR inhibitor rapamycin as entirely harmful, those mediated by inhibition of the mTORC2 protein complex, and in fact much of it may be pseudo-diabetes and thus of benefit. I'm not sure that I entirely agree, but this is an interesting position, given that a strong focus in the present clinical development of mTOR inhibitor drugs is to find a way to avoid these specific effects by focusing on inhibition of only the mTORC1 protein complex rather than all activities of mTOR.
In 2012, a paper entitled "Rapamycin-induced insulin resistance is mediated by mTORC2 loss and uncoupled from longevity" turned everything upside down. Data were misinterpreted to indicate rapamycin causes diabetes. Because the paper was published in a high profile journal, basic researchers believe that rapamycin is harmful and causes diabetes, which prompted calls for development of rapamycin-like drugs without rapamycin effects. In fact, however, this paper does not show that rapamycin causes type 2 diabetes; it shows that prolonged treatment with rapamycin causes glucose intolerance and insulin resistance in mice, which is in agreement with earlier studies. Furthermore, the study showed that these metabolic alterations were associated with increased longevity, indicating better health.
In humans, diagnosis of diabetes depends on the arbitrary choice of a threshold for fasting blood glucose: it was 140 mg/dl before 1997 and 126 mg/dl after 1997. But what is the arbitrary diagnostic threshold in mice? It is not defined. Is the slight increase in fasting glucose sufficient for a "diagnosis" of diabetes in mice? Does such hyperglycemia decrease life span or cause nephropathy? It does not. As we will discuss next, similar glucose intolerance and insulin resistance can also be caused by prolonged fasting and extreme very low calorie diet (VLCD). Prolonged fasting and starvation cause a condition well known in the past but unknown to modern researchers: starvation pseudo-diabetes.
During starvation or prolonged fasting, glucose utilization by nonbrain tissues is inhibited in order to feed the brain. Prolonged fasting is characterized by low insulin levels, gluconeogenesis, lipolysis, ketogenesis, and ketosis (ketone bodies in the blood), glucose intolerance, and hepatic resistance to insulin. When a starved animal is fed with glucose, it cannot utilize the glucose (glucose intolerance), leading to transient glycosuria (glucose in the urine) and polyuria (high urine volume).
Given that rapamycin is a starvation- or CR-mimetic, its metabolic effects can be viewed as "starvation-mimicking side effects." I advance the hypothesis that in animals (and humans) rapamycin can cause a reversible and benevolent condition, identical to starvation pseudo-diabetes. If so, this condition may, in theory, prevent development of genuine type 2 diabetes and its complications. For example, rapamycin prevents diabetic nephropathy. The results of recent studies are consistent with the idea that rapamycin-induced metabolic alterations are reversible and beneficial in nature. Hyperglycemia may be a marker of beneficial processes, given that rapamycin ameliorates nephropathy, despite elevating blood glucose levels in a mouse model of type 2 diabetes.
The notion of benevolent insulin resistance also resolves the insulin resistance paradox; that is, insulin resistance is associated with both decreased or increased life span. Insulin resistance due to activation of mTOR shortens life span, whereas insulin resistance due to inhibition of mTOR increases life span. Simply stated, insulin resistance associated with TOR overactivation is bad, but insulin resistance associated with inactive TOR is good. This is the mTOR-centric view on glucose metabolism. Detrimental metabolic alterations should have detrimental consequences, such as diabetic complications, but there is no evidence that rapamycin-induced glucose intolerance is detrimental. On the contrary, rapamycin improves nephropathy in diabetic mice, despite increasing blood glucose levels.