Fasting and Longevity: What Research Shows About Fasting and Lifespan

A team around Di Francesco studied 960 genetically diverse mice across their entire lives. They compared 20 and 40 percent calorie reduction as well as intermittent fasting with one and with two fasting days per week.

Both strategies extended lifespan, and the more pronounced the restriction, the stronger the effect. At the same time, the animals' genetics had a greater influence on lifespan than the diet. 40 percent calorie reduction extended life the most, but led to loss of lean mass and to changes in the immune repertoire that might make animals more susceptible to infection.

The sentence from this work that matters most to me: the metabolic improvements, meaning less body fat and lower fasting glucose, were not associated with the longer lifespan. Looking healthier and living longer are evidently not the same thing.

A consortium around Harrison gave genetically heterogeneous mice the mTOR inhibitor rapamycin in their feed. Not from youth, but only from day 600 of life, which in humans corresponds roughly to middle to later adulthood.

Measured at the age by which 90 percent of the animals had died, the treated females lived 14 percent and the males 9 percent longer. The effect appeared at three independent test centres.

For you this means: the nutrient sensor mTOR is demonstrably linked with lifespan in mammals. It does not mean that fasting does the same thing as a drug that inhibits mTOR in a targeted way.

A group around Guevara-Aguirre followed these people for 22 years. Because of the defective growth hormone receptor, their IGF-1 is also permanently low, the messenger through which growth hormone takes effect in the tissues.

Within this group, one single non-fatal cancer occurred during that time and not a single case of diabetes. In the comparison group of relatives, the frequencies were 17 percent for cancer and 5 percent for diabetes. Fasting insulin was 1.4 versus 4.4 micro-units per millilitre, the insulin resistance index HOMA-IR 0.34 versus 0.96.

In cell experiments using the serum of these people, there were fewer DNA breaks and lower activity of the RAS, PKA and TOR signalling pathways. Whether these people live longer overall as a result is not answered by the paper. What it does show: the insulin and IGF-1 pathway is closely tied to disease burden in humans too.

A group around Weir studied in the nematode Caenorhabditis elegans how dietary restriction and the energy sensor AMPK extend lifespan. The result: through maintenance of mitochondrial network homeostasis and through their cooperation with the peroxisomes in fatty acid burning.

When mitochondrial fusion or division was specifically blocked, the lifespan-extending effect of dietary restriction disappeared. For the lifespan extension mediated by intermittent fasting, active remodelling of the network was even required.

That is a very clean mechanism. It comes from a worm with around a thousand cells. In humans it has not been shown in this form.

A team around Shimazu showed that beta-hydroxybutyrate is an endogenous inhibitor of certain histone deacetylases, that is, of enzymes that co-determine how strongly a gene is read.

Fasting and calorie restriction increased histone acetylation in tissue in mice. Among the genes read more strongly were FOXO3A and MT2, which belong to the defence against oxidative stress. Mice given beta-hydroxybutyrate were considerably better protected against oxidative stress.

For you this means: the switch into fat burning is not only a change of fuel. It may also be a switching process in the cell nucleus. This has been shown in mice and cells.

A review by Lee and colleagues sums up the state of things soberly. In model systems, sirtuins delay cellular senescence, among other things through slower telomere loss, better genome stability and promoted DNA repair.

At the same time it states explicitly that the lifespan-extending effect remains contested and appears to depend on which sirtuin is present in which amount and in which tissue. The activator resveratrol was studied intensively without a robust longevity result in humans coming out of it.

For you this means: sirtuins are a real biological system. They are not evidence that a supplement extends your life.

A group around Jamshed had eleven adults with overweight eat, in a crossover design and for four days, either between 8 am and 2 pm or between 8 am and 8 pm. Glucose course, hormones and gene expression in blood were measured.

The early eating window lowered mean 24-hour blood glucose by 4 milligrams per decilitre and glucose swings by 12 milligrams per decilitre. In the morning, ketones rose along with the expression of a stress response and ageing gene and of an autophagy gene. In the evening, expression of a central nutrient-sensitive growth gene was elevated.

That is a genuine signal that eating breaks change something at exactly these switches in humans. It is eleven people over four days and it is gene expression data from blood cells, not a measured clean-up performance in tissue.

A consortium around Ravussin randomised 218 healthy, non-obese adults between 21 and 51 years of age to two years with 25 percent calorie reduction or to free eating. 25 percent was planned, 11.7 percent on average was actually achieved.

The fasting group maintained a weight reduction of around 10 percent. The thyroid hormone T3 fell more than in the control group after 12 and 24 months, the inflammatory marker TNF-alpha after 24 months. Cardiometabolic risk factors improved more clearly, without quality of life suffering.

This is the most robust human study that exists on this topic. It measures risk markers over two years. It does not measure lifespan.

A group around Waziry later examined blood samples from the same trial with DNA methylation clocks, that is, with laboratory methods that estimate biological age.

The result: calorie restriction slowed the pace of ageing on the DunedinPACE measure. The clocks PhenoAge and GrimAge, by contrast, did not change significantly. The authors describe the effect sizes explicitly as small and point out that a conclusive test would only be possible through long-term studies with real disease and mortality endpoints.

For you this means: there is a first, cautious signal that the pace of ageing can be influenced in humans. It is not more than a signal yet.

In the University of Wisconsin study, a roughly 30 percent calorie reduction from young adulthood onwards significantly improved age-related and all-cause survival in rhesus monkeys. In the parallel study by the National Institute on Aging there was no survival difference.

A joint analysis of both datasets by a team around Mattison worked out the differences: different age at start, different feed composition, and control animals in the NIA study that, compared with a large primate database, were effectively already being fed restrictively themselves.

The conclusion of both groups: the health benefits of calorie restriction remain in primates, but timing and composition have a say.

A team around Trepanowski compared three groups over twelve months: alternate-day fasting, daily calorie reduction and a control group without intervention.

Weight loss was practically the same in both intervention groups. For blood pressure, heart rate, triglycerides, fasting glucose, fasting insulin, insulin resistance, CRP and homocysteine, no significant differences were found between the intervention groups. LDL cholesterol after twelve months was even 11.5 milligrams per decilitre higher in the fasting group than under daily calorie reduction.

Behaviour was also striking: on fasting days more was eaten than planned, on feast days less. The body regulates against it.

A working group around Liu randomised 139 people with obesity over twelve months to calorie reduction with an eating window from 8 am to 4 pm or to calorie reduction alone.

The time window group lost 8.0 kilograms, the comparison group 6.3 kilograms. The difference of 1.8 kilograms was not statistically significant. For waist circumference, body fat, lean mass, blood pressure and metabolic risk factors there was likewise no advantage of the time window.

In a further study around Lowe with 116 people over twelve weeks, a 16:8 window without a calorie target also brought no significant weight advantage. Striking there was a greater loss of appendicular lean mass in the fasting group.

A group around Wilkinson followed 19 people with metabolic syndrome who had previously been eating across more than 14 hours a day. Most were already taking statins or antihypertensives.

Over twelve weeks they ate within a self-chosen ten-hour window. Weight, blood pressure and atherogenic blood lipids improved, in addition to ongoing standard therapy.

This study had no control group. But it shows where the lever is probably largest: not in someone who already eats in an orderly way, but in someone whose eating window has so far covered almost the whole day.

A group around Acosta-Rodríguez noticed that mice under classic calorie restriction devour their feed in about two hours and then fast for 22 hours. The researchers therefore separated three factors cleanly: calorie amount, length of the fasting phase and time of day of feeding.

30 percent calorie reduction alone extended lifespan by 10 percent. When a daily fasting phase was added and feeding fell in the animals' active night phase, it was 35 percent. And this was independent of body weight.

For you this means: it may depend less on the calorie amount than on whether there is a genuine night rest for metabolism at all and whether eating fits the internal clock.

A team around Mitchell compared free feeding, 30 percent calorie restriction and a single daily meal without calorie reduction in mice. Two different feed compositions were tested in addition.

Both the calorie restriction and the single meal extended lifespan, and did so regardless of how the feed was composed. The animals on a single meal fasted for long phases of their own accord and had less disease burden.

This too is a mouse finding. But it fits very well with the idea that the length of the eating break may be a factor in its own right and not just a side effect of fewer calories.

A team around Romashkan analysed the adverse events of the two-year calorie restriction. Overall the intervention was safe and well tolerated, there were no deaths.

Three participants did, however, have to be taken out of the intervention for safety reasons. Within the fasting group, participants of normal weight more frequently had complaints of the nervous system, the musculoskeletal system and the reproductive system than those with overweight. After 12 and after 24 months, bone density at the lumbar spine, hip and femoral neck was below that of the control group.

The authors explicitly recommend watching for bone loss and anaemia. That is not an argument against fasting. It is an argument for supervision.

A group around Maegawa showed that the age-related shift of DNA methylation patterns, known as epigenetic drift, occurs across species in mice, rhesus monkeys and humans, and that its pace is related to the life expectancy of the species.

Rhesus monkeys aged 22 to 30 years who had received 30 percent fewer calories since the age of 7 to 14 years showed an attenuated drift. Their methylation age in blood appeared around seven years younger than their actual age. In mice under 40 percent restriction the effect was even more pronounced.

This fits with the cautious signal from the CALERIE follow-up analysis in humans. It does not replace it.