Article from cell.com.
REVIEW| VOLUME 185, ISSUE 9, P1455-1470, APRIL 28, 2022
Caloric restriction in rodents
There is a large literature on the beneficial impact of CR on indices of health including the onset and progression of multiple age-related diseases and conditions in rodents (Richardson, 2021). The mechanisms responsible for the effects of CR on longevity and disease involve the nutrient responsive signaling genes discussed above, although additional mechanisms, including the prevention of insulin resistance and metabolic diseases, are also involved. In fact, decreased adiposity is a hallmark of CR in rodents and is associated with changes in fat function, including the secretion of protein and lipid factors associated with metabolic homeostasis (Miller et al., 2017). Not all strains respond equivalently to the same imposed degree of restriction, and more recently sex dimorphism has been a focal point in nutritional studies of aging (Mitchell et al., 2016). Despite a well-established dampening of growth response in rodents on CR, there is evidence that the immune response is actually improved (Palmaet al., 2021), although it will be important to determine whether CR renders organisms sensitive to specific infectious diseases. In rodent studies the pressing question is whether the documented benefits of CR can be harnessed through manipulation of diet composition or timing of feeding, and it seems there may be interactions between the two (see below). Since the literature on CR in rodents has been reviewed extensively over the years, in this review we will focus on the effects of CR in monkeys and humans.
Caloric restriction in monkeys
Nonhuman primates are a highly translational model for biomedical research, and the rhesus monkey Macaca mullata is one of the best characterized for aging studies (et al., 2017). Two prominent studies of the impact of long-term CR on aging and health were conducted over 30 years, one at the Wisconsin National Primate Research Center and the other within the National Institute on Aging intramural program at the National Institutes of Health (Colman et al., 2009, Colman et al., 2014; Mattison et al., 2012). Although initial reports seemed contradictory, subsequent comparisons between the studies resolved that the monkeys that weighed less and ate less lived longer and were healthier until later in life (Mattison et al., 2017).
The hallmarks of CR in monkeys mirror those identified in mouse studies and include lower adiposity, lower fasting glucose and insulin, greater insulin sensitivity, and more favorable lipid profiles (Kemnitz, 2011). Importantly, these same features are also detected in a short-term clinical trial of CR in humans (Most et al., 2017). CR in nonhuman primates is associated with several indices of healthy aging. MRI studies of monkeys on CR indicate delayed brain aging, based on the preservation of age-related loss of gray matter volume (Colman et al., 2009), which was subsequently linked to improved insulin sensitivity (Willette et al., 2012) and preservation of white matter (Bendlin et al., 2011). CR was also effective in delaying sarcopenia, preserving muscle mass, preventing an age-related decline in physical activity, and lowering the metabolic cost of movement that becomes elevated with age (Yamada et al., 2013). There is also evidence that CR preserves the T cell repertoire (Messaoudi et al., 2006), opposing a phenotype of age that is thought to be directly linked to disease vulnerability. In skeletal muscle, CR increases expression of genes involved in energy metabolism and proteostasis, decreases the expression of genes involved of immune and inflammatory pathways, preserves fiber metabolism and cross-sectional area, and delays fibrosis and fat infiltration (Rhoads et al., 2020). At the system level, delayed muscle aging is linked to insulin sensitivity. In hepatic tissue, CR induces gene expression related to oxidative phosphorylation, lipid metabolism and peroxisomal pathways, proteostasis, and RNA processing, while downregulating immune and inflammatory pathways (Rhoads et al., 2018). Serum metabolomics reveal similarities between the rodent and monkey responses to CR (Aon et al., 2020), including enrichment of ketone bodies, fatty acids, and factors associated with fasting including succinate, glutamine, and lactate. These studies suggest that the biology of CR is at least partially conserved from mice to nonhuman primates.
Caloric restriction in humans
The NIH/NIA-sponsored 2-year CALERIE study demonstrated that the systemic hallmarks of CR observed in rodents and monkeys are largely recapitulated in humans (Most et al., 2017). CR induced a loss of total body weight and a reduction in adiposity resulting in fat-free mass being higher as a percent of total body mass in individuals on the CR regimen. CR was associated with greater insulin sensitivity (Ravussin et al., 2015), lower risk scores for cardiovascular disease (Kraus et al., 2019), and improved biomarkers of liver health (Dorling et al., 2021). Analysis of clinical and plasma biomarkers from CALERIE subjects indicates that the pace of aging is delayed (Belsky et al., 2017), which was later corroborated using the methylation clock (Belsky et al., 2020). The similarities between the human, nonhuman primate, and mouse responses to CR argue for strong conservation in the underlying mechanisms by which CR impacts health in mammals, with links to improved longevity consistent between mice and monkeys.
The biology of mammalian CR
Despite species specificity in how aging manifests and in the final determinants of mortality, the underlying cellular biology of CR is conserved. Key processes involved in making the transition to healthier status include autophagy, proteostasis, energy metabolism and the switch to lipid fuel usage, changes in growth signaling including translation and synthetic pathways, and engagement of gene regulatory mechanisms such as RNA processing, all of which have been linked to longevity regulation in short-lived species (Figure 1). The overall picture indicates that longevity is associated with a reduction in the activity of growth pathways and a switch to metabolic patterns associated with fasting responses. These changes are coincident with reduced inflammation without a general impairment of immune function, which could contribute to protection against diseases ranging from cancer, to cardiovascular disease, to Alzheimer’s and autoimmune diseases. Although these outcomes are shared in effective models of longevity enhancement, there are important caveats. Timing of onset of the diet, sex, and existing metabolic and genetic status all influence the efficacy of these diets in producing beneficial effects. The striking influence of genetics on the health and longevity response to CR observed in shorter-lived laboratory animals, many of which are substantially inbred, is unlikely to be so dramatic in outbred populations or in primates and humans, but genetics must be considered as a factor in optimizing dietary interventions. The take-home message from aging and nutrition research is that one size does not fit all, but it is almost certain that specific nutrition patterns can optimize health and longevity.
In this section we bring together evidence from mammalian studies on the beneficial effects of fasting, describing the different models of fasting implementation and their effects on disease risk factors, health, and longevity.
Although there are many different types of intermittent fasting regimens, ranging from restricting eating to a limited number of hours (time-restricted feeding [TRF] in model organism and time-restricted eating [TRE] in humans) to alternate day fasting, to fasting for 2 days a week, here we will focus on the most common form of IF, which in most cases entails 12–23 h of fasting per day (Longo et al., 2021).
TRF in rodents
TRF has gained a lot of public attention due to ease of implementation and the promise of health benefit. In this regimen, the time period in which there is access to food is reduced but composition of the food is not changed. Much of the early work was not focused on aging but on correcting or avoiding metabolic dysfunction associated with obesogenic diets (Chaix et al., 2014). Compared to a high-fat diet (60% fat) fed ad libitum without timing limitations, the same diet limited to 9 h of feeding followed by 15 h of fasting (TRF) activated pathways in liver including those involved in metabolism, proteostasis, RNA processing, and repair and defense pathways (Chaix et al., 2019). On a standard diet, benefits of TRF are also observed, and although there are similarities with CR even at the molecular level, TRF is not as effective as CR in delaying aging (Aon et al., 2020; Mitchell et al., 2019; Velingkaar et al., 2020). The beneficial metabolic effects of TRF (9 h feeding/15 h fasting) in mice fed a “Western” diet (45% fat) ad libitum compared to those fed the same ad libitum diet without imposed fasting periods are observed in young and mature adult mice and also include improved survival in response to lipopolysaccharide (LPS) (Chaix et al., 2021). There are interesting sex-specific effects of TRF benefits, reflecting differences in the innate response to high-fat-diet feeding and adding to the growing evidence across aging studies that females are not the same as males. Thus, the impact of nutrient limitation by CR and the effect of TRF on health indices is observed whether animals are on standard diets or on high-fat diets, although the specifics of the metabolic reprograming are not identical (Diaz-Ruiz et al., 2021). It seems that the context matters, perhaps reflecting the importance of metabolic status, and the best strategy to harness health- and longevity-associated pathways that may be different when starting from a metabolically compromised position.
TRE in humans
There is growing evidence of the beneficial effects of TRE also in humans (Duregon et al., 2021). Most of the clinical trials to date have focused on weight loss or correcting existing metabolic impairment, with studies involving subjects with obesity, metabolic syndrome, or type 2 diabetes (T2D). The regimen used is usually a 8–10 h daily eating window, with duration varying from 4 to 12 weeks, and with some studies imposing TRE only 5 out of 7 days per week. Almost all studies of TRE report weight loss and a reduction in adiposity (Wilkinson et al., 2020) or waist circumference (Schroder et al., 2021) when measured. Several studies report improvements in circulating factors linked to cardiovascular disease (Che et al., 2021; Schroder et al., 2021; Wilkinson et al., 2020), but this is not always the case. Few studies report improvements in glucoregulatory parameters. A notable exception is a study involving healthy weight individuals, where cardiovascular disease indices were unaltered but circulating glucose levels were lowered (Martens et al., 2020). More stringent TREs (6 h) are effective in improving insulin sensitivity but are more challenging to undertake (Sutton et al., 2018). Epidemiology studies are less clear about TRE. Longer daily fasting periods that involve breakfast skipping have been consistently associated with increased mortality, which is particularly high for cardiovascular disease (Rong et al., 2019). 4 weeks of another form of intermittent fasting in which subjects fast every other day (alternate day fasting) was also effective in improving cardiovascular markers, reducing trunk fat, improving the fat-to-lean ratio, and increasing b-hydroxybutyrate, even on non-fasting days (Stekovic et al., 2020).
In summary, TRE appears to have beneficial effects in both rodents and humans, but both compliance issues and side effects point to a 11–12 h daily eating period as ideal at least until additional studies identify TRE lengths that are safe, feasible, and effective.
Periodic fasting and fasting-mimicking diets
In humans, markers or risk factors for aging and age-related disease, including IGF-1, insulin, glucose, insulin resistance, HbA1c, C reactive protein (CRP), hypertension, and high cholesterol, can be affected by dietary composition and by fasting periods. As described in the earlier section, intermittent fasting (IF) regimens—ranging from TRE, to alternate day fasting, to 2 days of fasting per week—require frequent and long-term restrictions, and only specific types are effective and not associated with side effects. Periodic fasting (PF) cycles, which do not require frequent fasting, are emerging as an alternative to IF. The disadvantage is that they require longer periods of fasting lasting 2 days and in most cases 4 sequential days, but the advantage is that they are in the great majority of cases adopted twice a month or less and may begin to be beneficial even if applied only a few times a year, since they have been shown to provide long-term protective effects even months after PF cycles end (Wei et al., 2017). Thus, PF can be adopted at regular intervals such as once a month or can be used analogously to drugs based on the need to treat a condition or disease such as cancer (Longo et al., 2021). Although periods of water-only fasting (i.e., only water is consumed) lasting 3 or more days are feasible, the extreme nature of this intervention underlines both safety and compliance concerns, particularly when involving relatively healthy subjects who are not motivated by the need to treat a disease or condition. In fact, the initial trials focused on the use of water-only fasting in cancer patients proceeded very slowly, since the intervention was difficult for patients and was met with skepticism by oncologists (Longo et al., 2021). For this reason, but also in search for nutritional compositions able to enhance the effects of water-only fasting, fasting-mimicking diets (FMDs) were developed and tested both in animal and clinical studies.
Periodic fasting/FMDs in rodents
FMDs are plant-based low-calorie, low-protein, low-sugar, and high-fat nutritional compositions normally provided to animals or human subjects in a pre-packaged form developed and studied for the purpose of replacing water-only fasting while maintaining and possibly exceeding its effect on key markers of the fasting response including changes in IGF-1, IGFBP1, glucose, and ketone bodies (Longo et al., 2021). They are part of an emerging nutri-technology field focused on applying both specific ingredients and complex food compositions as medicine to accompany or replace pharmacological or biological therapies.
In mice, FMD cycles are protective in both type 1 and type 2 diabetes models, prevent the premature death caused by a high-fat/calorie diet, reduce the symptoms and pathology associated with multiple autoimmune diseases, reduce the incidence and progression of a range of tumors, and extend lifespan (Longo et al., 2021). Notably, in humans the beneficial effects of FMD cycles on disease markers/risk factors including IGF-1 and leptin continue for weeks after the return to a normal diet, consistent with what has been observed in mice (Caffa et al., 2020). Although the specific mechanisms responsible for the protective and rejuvenating effects of FMD cycles are only beginning to be understood, many of the beneficial outcomes induced by CR including reduced adiposity, improved insulin sensitivity, and lowered inflammation are also detected in response to periodic FMD treatments. Specific mechanisms observed during the FMD cycles in mice include the activation of stem cells and developmental-like programs in multiple systems, but the re-feeding phase in which mice are switched from the FMD to a high-protein and calorie nutrition appears to be central for the regenerative effects. Consistent with both the anti-inflammatory and regenerative effects, FMDs cause a reduction in autoimmune cells leading to reduced inflammation, re-myelination, and reduced pathology in mouse models of multiple sclerosis (Longo et al., 2021). In leptin receptor deficient a type 2 diabetes db/db mouse model, FMD lowers insulin resistance, pointing to a fundamental role of fasting periods in recalibrating metabolic integrity. FMD also promotes a gene expression profile in the pancreas similar to that observed during embryonic development, leading to the reversal of the depletion in functional beta cells and reduced insulin production in a type 1 diabetes model (Cheng et al., 2017). FMD cycles applied for 5 days once a month to mice on a high-fat/calorie diet, lower body fat, improve cardiac function, lower cholesterol, and restore lifespan to the levels observed in mice on a standard diet (Mishra et al., 2021). Additional health benefits of FMD cycles lasting 4 days include extended longevity, reduced tumor incidence, and delayed cognitive decline, even when started at middle age (Brandhorst et al., 2015). Furthermore, either CR or FMD cycles protect against cancer in an implanted xenograft model (Pomatto-Watson et al., 2021). Metabolite signaling is integral to fasting biology, and recent studies demonstrated enhanced longevity in mice fed the TCA intermediate alpha ketoglutarate (Asadi Shahmirzadi et al., 2020). Perhaps the stimulation of mitochondrial activity by ketone bodies or TCA cycles intermediates, in addition to increased stress resistance and anti-inflammatory and regenerative effects, could mediate part of the effects of periodic fasting/FMDs.
Periodic fasting/FMDs in humans
In humans, PF and FMDs have been studied both in normal subjects and in disease treatment. A randomized crossover study of 100 patients of which 71 received 3 monthly 5-day FMD cycles showed reduced body weight, trunk, and total body fat; lowered blood pressure; and decreased IGF-1. A post hoc analysis also indicated a reduction in fasting glucose, triglycerides, total and low-density lipoprotein cholesterol, and C-reactive protein in participants with high levels of these risk factors at baseline (Wei et al., 2017). A number of studies have now also investigated the role of FMD in cancer treatment, including a 125-patient randomized study indicating that FMD increases the efficacy of chemotherapy on clinical and pathological responses in women with breast cancer, even if the majority of patients complete only two cycles of the dietary intervention (de Groot et al., 2020). In addition, a 36-patient feasibility study in which FMD combined with hormone therapy to treat breast cancer was found to be safe and reduce markers and risk factors associated with cancer progression without reducing muscle function or mass (Caffa et al., 2020).
Therefore, FMD cycles have been associated with potent anti-inflammatory, metabolic, and regenerative effects in mice and with improvements in disease risk factors or clinical response in multiple clinical studies. Because the beneficial changes caused by FMD cycles can last for months, this dietary intervention has the potential to be effective and should be tested in clinical trials for the prevention and treatment of many diseases when applied for only 3–4 times per year without requiring but while preferring improvements in the daily eating habits.
Macronutrient composition and levels
The role of nutrition on lifespan and age-related disease is widely accepted, yet we are far from a consensus on what type of nutrition affects healthspan. Fortunately, the nutrition response mechanisms affecting health and longevity are quite well conserved in species ranging from simple organisms to rodents to humans, making it possible to take advantage of both basic science and human studies to identify the type and levels of macronutrients and nutrition patterns that will be effective in regulating adiposity and aging in most individuals, although the diet will also need be tailored to account for not only age, sex, and genetics but also lifestyle and the health status of an individual. Soon, multi-omic analysis, eventually aided by artificial intelligence, will allow an even more sophisticated personalization of nutritional therapies, but these approaches will not be discussed here since they are only beginning to be applied effectively. It is now well established that diets that augment central adiposity can cause major increases in insulin resistance and the risk for diabetes, cancer, and neurodegenerative disease in mice and humans (Saltiel and Olefsky, 2017). In the sections below we focus on how macronutrient composition, levels, and source affect biomarkers and risk factors for aging and age-related diseases in rodents and humans.
In rodents and humans, increasing calorie intake above the level required for the required energy expenditure increases lipogenesis, fat storage, and obesity, contributing to major age-related disease (Janssen, 2021). Excess glucose is directed to the synthesis of triglycerides in the liver, which are transported to adipose tissue and muscle by VLDL. Genetics influence the response to dietary interventions, but in general a diet providing high levels of saturated fats and sugars appears to be effective in generating obesity, insulin resistance, high cholesterol, and a shortened lifespan (Mishra et al., 2021; Wali et al., 2020). In mice and rats, the calories from fat intake that promote obesity and insulin resistance is typically in the 40%–60% range, but these high-fat diets in general also contain high levels of sugars (Wali et al., 2020).
Since 1970, daily calorie intake in the United States has increased by 20% or by about 425 Kcal/day (Janssen, 2021), but the increase in total calorie intake is not the only qualitative difference of the Western diet. With high calories comes also elevated sugar, starch, saturated fat, and protein content. Collectively, Western diets result in elevated insulin, hyperglycemia, high IGF-1, and high cholesterol and triglyceride levels; on one hand they activate pro-aging pathways and on the other hand promote insulin resistance and obesity, outcomes linked to a host of age-related diseases (Figure 2). Thus, the combination of these factors appears to contribute to disease and mortality both by accelerating the aging process and by promoting morbidities independently of aging.
Low-carbohydrate and ketogenic diets
In humans, most low-carbohydrate diets limit daily carbohydrate consumption to 50–60 g, with the rest of the calories coming from high levels of fat and moderate to high levels of proteins. 100 years ago, Dr. Wilder at Mayo Clinic described how the previously known benefits of fasting in children with epilepsy could be also achieved by a diet able to produce higher levels of ketone bodies and called it “ketogenic diet” (KD) (Wilder, 1921). This was later described as a diet providing 1 g of protein per Kg of body weight, less than 15 g of carbohydrates/day, and the rest coming from fat. In the 1970s, the ketogenic diet was modified and made popular by Robert Atkins to achieve a higher compliance and weight loss in adults by allowing a much higher level of proteins but maintaining carbohydrate intake low to very low (Weber et al., 2020). However, this popularization of the KD resulted in the adoption of a low-carbohydrate diet that allows more than 15 g/day of carbohydrates but also promotes the consumption of ingredients typical of Western diets.
The ability of very-high-fat-content diets to confer health benefits has prompted widespread interest in the KD. In mice, a KD applied in fully mature adult animals modestly increases lifespan and improves indices of metabolic, physical, and cognitive function (Roberts et al., 2017). A cyclical application of the diet also affords metabolic and cognitive benefits (Newman et al., 2017). In mice KD improves cerebrovascular function (Ma et al., 2018), and in rats it improves cognitive scores together with changes in metabolite transport systems in the prefrontal cortex (Hernandez et al., 2018). Cognitive performance also improves in Alzheimer’s disease models treated with KD (Pawlosky et al., 2020), and treatment with ketone bodies in Alzheimer’s disease mice fed a standard diet improves cognition and lowers plaque burden in a manner that was linked to hippocampal neuronal mitochondrial function (Wu et al., 2020). Within hours, the KD induces autophagy in liver, a step critical for the synthesis of ketone bodies. Mechanistically, the induction of the ketogenic program is linked to key lipid metabolism regulator PPARa and the autophagy-dependent removal of an inhibitor complex (Saito et al., 2019). The link between lipid metabolism and autophagy is likely to contribute to the health benefits of the KD as both are linked to longevity regulation in shorter-lived species. Notably, in many studies the KD involves a low protein intake, so it is possible that some of the reported benefits of KD on longevity and disease may be linked at least in part to lower protein/amino acid intake. It will also be important to know how different animal- and plant-derived fats affect the effect of KD on aging and disease.
Ketogenic and other low-carbohydrates diets have also been studied extensively in humans. In obese humans, a recent meta-analysis suggested that ketogenic/low carbohydrate consumption was no more effective than a balanced diet including low-calorie, low-fat/high-carb, or low-protein/high-carb diets, with equivalent effects on body mass index (BMI), circulating levels of total cholesterol, lipoprotein profiles, and triglycerides (López-Espinoza et al., 2021). Some large epidemiological studies have specifically focused on carbohydrate intake and mortality. One of these studies followed 85,168 women (aged 34–59 years at baseline) and 44,548 men (aged 40–75 years at baseline) without heart disease, cancer, or diabetes, for 26 years and 20 years, respectively. This study showed that a low-carbohydrate diet based on animal food sources was associated with higher all-cause mortality in both men and women, whereas a low-carbohydrate diet with a higher content of plant-based food was associated with lower all-cause and cardiovascular disease mortality rates. Men on an animal products-based low-carb diet also displayed a 66% increased risk of cancer mortality, whereas women on the same diet displayed a 26% increased risk of dying of cancer (Fung et al., 2010).
In a meta-analysis of multiple cohorts involving 432,179 participants, both a low carbohydrate consumption (<40% of energy) and high carbohydrate consumption (>70% of energy) increased mortality risk compared to moderate carbohydrate intake. The risk of overall mortality increased by over 50% in the group consuming less than 20% of energy from carbohydrates compared to that for the group consuming 50%–55% of energy from carbohydrates (Seidelmann et al., 2018). Notably, the low carbohydrate intake necessitates increases in protein and fat intake, raising the possibility that the higher protein and/or fat intake may be more important for mortality than the low carbohydrate consumption. In addition to macronutrient balance, the source of macronutrients was also found to be key. Mortality risk was about 18% higher when animal-derived proteins or fats replaced carbohydrates but 18% lower when plant-based proteins or fats replaced carbohydrates. These epidemiological studies considered groups that were consuming low carbohydrate levels but far from the very low levels (<50 g) allowed in the strict ketogenic diets. Although it is well established that long-term consumption of the very restrictive KD is not feasible for the great majority of the population, these studies are important to understand whether certain plant-based diets providing moderately low carbohydrate levels could represent a more realistic option for the public. They also underline the importance of analyzing the relative macronutrient content instead of focusing on a specific one but also point to very different effects of animal- versus plant-based sources of fats and proteins on health, mortality, and longevity. These results also clarify the importance of combining basic and human studies to begin to identify the age-specific nutrition that can extend healthspan.
Low-protein and amino acids diets
A groundbreaking study in mice compared 25 different diets varying in fat, protein, and carbohydrate (Solon-Biet et al., 2014). Quantitation of survival and health outcomes indicated that the low-protein and high-carbohydrate diets were the most beneficial, although there was sex dimorphism in how diet interacts with mortality risk. In a follow-up study, low-protein diets could recapitulate to some extent the beneficial effects of CR on cognition with evidence of nutrient signaling pathway activation and preservation of neuronal architecture related to connectivity in the hippocampus (Wahl et al., 2018).
In contrast, a very-low-protein diet causes mice to eat less, due in part to altered signaling in the hypothalamus (Wu et al., 2021). These studies indicate that the composition of the diet influences feeding behaviors and may illicit distinct feeding signaling patterns in hypothalamic centers.
The relative importance of specific amino acids in the diet is an active area of investigation. Methionine restriction (MR) increases longevity in mice, and recent studies have identified the potential use of this restriction in combination with standard treatments for cancer (Gao et al., 2019). Notably, methionine levels are very low in legumes and other plant-based protein sources compared to those in animals. A key signaling molecule in the mechanisms of both protein restriction (PR) and MR is the liver-derived signaling peptide FGF21 (Hill et al., 2020). The remodeling of adipose tissue by MR requires activation of FGF21 receptors in the brain, indicating a cross-talk among tissues (Forney et al., 2020). In a high-fat-diet background, MR dampens inflammation locally in tissues and systemically, although the effect on inflammation is independent of FGF21 (Sharma et al., 2019). MR has beneficial effects on cognition even when applied at advanced age, and in this case the metabolic and structural changes induced in the hippocampus are FGF21 dependent (Ren et al., 2021). The increase in circulating levels of branched chain amino acids (BCAAs) in models of metabolic dysfunction has prompted considerable research interest. Glucoregulatory improvement and the anti-inflammatory effects of PR are dependent on low levels of BCAAs, and increasing the levels of dietary BCAAs under standard diet feeding conditions is sufficient to drive overeating and increased adiposity (Solon-Biet et al., 2019). Lifelong restriction of BCAAs improves health and extends lifespan in males but not in female mice (Richardson et al., 2021). Molecular analysis of skeletal muscle from these animals shows the enrichment of pathways involved in peroxisomes, lipid metabolism, and growth signaling in males but not females.
The ability of protein/amino acid restriction to extend rodent longevity is linked to a reduction in the levels of IGF-1, in agreement with the role of pro-growth signaling in blunting longevity in organisms ranging from yeast to mice (Figure 2) (Longo et al., 2021). In humans, CR results in beneficial changes in cardiometabolic risk factors but is not associated with reduced IGF-1 levels unless participants are also protein restricted (Fontana et al., 2008). In both mice and humans, a low-protein diet imposes a reduction in growth factors/signaling both upstream of IGF-1 (GHRH, GH) and downstream of it (mTOR, S6K). With PR diets, lower growth signaling goes hand in hand with lower insulin and improved insulin sensitivity, and although clinical studies more often focus on insulin, it is clear that there is a connection between these pathways.
The role of protein intake in increasing mortality and reducing longevity appears to be also conserved in humans, although this relationship is complex. There is evidence that diet should be tailored to age. Whereas consumption of more than 20% of calories in the form of proteins is associated with a 75% increase in overall mortality risk and 400% increase in the risk of cancer mortality in subjects 65 years old or younger compared to consumption of less than 10% of calories from proteins, these associations are not observed in those 66 and older (Levine et al., 2014). These results are in agreement with those in mice in which, prior to 85 weeks of age, mortality is minimized by a low protein consumption, but as animals aged beyond 85 weeks, a major increase in the protein to carbohydrate ratio is necessary to minimize mortality (Senior et al., 2019).
In subjects younger than 65, IGF-1 levels correlate with level of protein intake but not in subjects 66 and older (Levine et al., 2014). These findings suggest that protein restriction in the elderly may no longer provide protection against overall and cancer mortality in part because it no longer inhibits pro-aging pathways. In light of these results, the low carbohydrate and high mortality correlation described earlier may also be re-interpreted by focusing on the protein content of the different diets. The group on the lowest carbohydrate diet consumed 37.2% energy from carbohydrate versus 60.5% in the highest intake group, but the lowest-carb group also obtained 22.3% of the energy from proteins versus 15% in the high-carb group (Fung et al., 2010), raising the possibility that the increased all-cause, cardiovascular, and cancer mortality observed for the low-carbohydrate group may be also due to the high protein intake. Instead, for the group consuming a vegetable-based low-carbohydrate diet, which was associated with a reduced all-cause and cardiovascular mortality, the protein intake was similar to that of the high-carbohydrate group (18.7% versus 17.5%) (Fung et al., 2010). These studies suggest that animal-derived proteins play an important role in age-related mortality and diseases, underline the importance of the balance of all macronutrients within the diet, and demonstrate that diet efficacy can be age-range specific.
Low-fat and high-fat diets
For decades, low-fat diets have been adopted by the public and recommended by the medical community to combat obesity. Although the consumption of fat has decreased in the United States, obesity has continued to increase pointing to increased total calorie intake and to modern diet composition rather than simply the intake of fat as culprits. In fact, when 7,447 participants at high risk for cardiovascular disease were randomized to a Mediterranean diet supplemented with extra-virgin olive oil, or mixed nuts, or to a control diet with the advice to reduce dietary fat, the risk of major cardiovascular events was about 30% lower in the Mediterranean diet groups supplemented with healthy fats from olive oil or nuts compared to the group recommended a low-fat diet (Estruch et al., 2018). These results are also in agreement with the epidemiological data discussed earlier and showing that a diet high in animal fat and animal protein increases mortality compared to a high-carbohydrate diet, but that a low-carbohydrate diet is beneficial when high in vegetable-based food sources (Fung et al., 2010). The consensus from these studies is that a relatively high-carbohydrate diet is ideal but that the balance of macronutrients is important, and the source of nutrients can determine whether the diet is more or less healthy.
Several studies indicate that pesco-vegetarians but not vegans display reduced risk for overall mortality compared to meat eaters, although a vegan dietary pattern is also associated with reduced risk of cancer, hypertension, and diabetes compared to that for regular meat eaters (Segovia-Siapco and Sabaté, 2019). Notably, the vegan diet has been associated with a 43% increased risk in all fractures and 2.3-fold increase in hip fractures compared to non-vegan diets (Tong et al., 2020). This frailty may be explained in part by deficiencies of certain amino acids. In fact in the EPIC-Oxford study, 16.5% of vegan men and 8.1% of vegan women had a protein intake lower than their requirement, which could be made worse by the reliance on amino acids solely from legumes, which provide very low levels of methionine and other amino acids (Mariotti and Gardner, 2019). In summary, the data are consistent with remarkable benefits of a vegan diet against aging and diseases but also an association of vegan diets with fewer benefits compared to vegetarian or pesco-vegetarian diets, possibly because these diets prevent the frailty associated with vegan diets in the general population.
A multi-pillar approach for nutrition and healthspan
The evidence from the literature to date underscores the need for hypothesis-driven and multi-disciplinary assessment of nutrition and healthspan to identify the complex dietary patterns that promote healthy longevity. Alone, an “epidemiological” comparison of how a low versus a high consumption of an isolated macronutrient and its association with health and mortality may not only fail to identify protective or detrimental nutrition patterns but may lead to misleading interpretations. For example, many epidemiological studies have pointed to the increased mortality risk in subjects with low IGF-1, leading to the conclusion that IGF-1 should be maintained higher, but several studies have pointed to both the lowest and highest IGF-1 levels being associated with higher mortality, pointing to mid-range IGF-1 as consistently linked to low mortality (Burgers et al., 2011). Thus, epidemiology, which is clearly a central pillar in determining the ideal ranges of a nutrient or factor for health and longevity, should be complemented by at least three additional pillars that account for age, sex, and underlying metabolic status and that assess risk factors in addition to biological age: (1) basic research focused on lifespan and healthspan, (2) carefully controlled clinical trials, and (3) studies of individuals and populations with record longevity.
The longevity diet
Based on all of the studies discussed in this review and representing all the pillars of longevity listed above, we can begin to point to a common denominator for healthy longevity. These pillars indicate that the everyday normocaloric longevity diet associated with low or very low side effects and extended lifespan and healthspan is characterized by a mid to high carbohydrate and low but sufficient protein intake that is mostly plant based but includes regular consumption of pesco-vegetarian-derived proteins (Longo, 2018). For example, animal products represented about 1% of the traditional diet of the record longevity Okinawans (Willcox et al., 2007), and occasional meat or animal-product consumption also characterized the populations of the Sardinian and Loma Linda areas with high prevalence of centenarians or high average lifespan (Levine et al., 2014). The benefits of such a diet are supported by evidence from the calorie and protein restriction studies in short-lived species, are in agreement with the epidemiological data described in earlier sections, and are consistent with the evidence from large clinical trials. Thus, the low but sufficient protein diet or a normal protein intake with high legume consumption and therefore relatively low content of methionine and other amino acids contributes to the reduction in the levels/activity of the pro-aging GHR, IGF-1, insulin, and TOR-S6K signaling (Figure 2). However, in over 65 individuals the low-protein diet does not appear to reduce further the circulating IGF-1 already lowered during the aging process and may instead contribute to lean body mass loss and frailty. In the absence of obesity and insulin resistance, the relatively high complex carbohydrate consumption may also contribute to avoiding frailty at all ages but particularly in the elderly, thus providing energy without increasing insulin and activating glucose signaling pathways.
A fat consumption providing about 30% of energy mostly from plant-based and pro-longevity sources is also part of the longevity diet and is again consistent with the basic research, and epidemiological and clinical data, although the traditional Okinawan diet provided a much lower level of fats, confirming that there are variations of the optimal longevity diet that could be equally effective. The high circulating fat content does not appear to have the pro-aging effects of the protein- and sugar-endocrine axes, possibly because fat catabolism, fatty acids, and ketone bodies are at the center of fasting responses. A recent study based on meta-analyses and data form the Global Burden of Disease 2019 study including studies from the United States, China, and Europe provides evidence in support of the longevity diet. A sustained change from the typical Western diet to an optimal diet rich in legumes, whole grains, and nuts with reduced red and processed meats is associated with an increase in life expectancy of 10.7 years in females and 13 years in males if started at age 20, and over 8 years of increased life expectancy when started at age 60 (Fadnes et al., 2022).
An important caveat is that the longevity diet should be designed to avoid malnourishment, particularly in the over-65 population, to prevent frailty and diseases that may result from reduced bone or muscle mass or low blood cell counts. Ideally, the longevity diet would also include a 12–13 h daily fasting period that has been shown to be safe, feasible, and effective in many studies. The periodic use of a FMD in those age 18 to 70 may be key in reversing the insulin resistance generated by a high-calorie diet. In fact, maintaining a BMI lower than 25 and an ideal sex- and age-specific body fat and lean body mass levels and distribution and abdominal circumference should be used as guidelines to establish daily food intake rather than a set calorie level. FMD cycles can also lower IGF-1, blood pressure, total cholesterol, and inflammation, particularly in at risk subjects.
In summary, we propose that the longevity diet would be a valuable complement to standard healthcare and that, taken as a preventative measure, it could aid in avoiding morbidity, sustaining health into advanced age.
This work was supported in part by awards to V.D.L. including the Associazione Italiana per la Ricerca sul Cancro (AIRC) ( IG#17605 and IG#21820 .), the BC161452 grant of the Breast Cancer Research Program (US Department of Defense) and the US National Institute on Aging-National Institutes of Health (NIA–NIH) grants P01 AG055369 ). R.M.A. is supported by NIH -NIA RF1AG057408 , R01AG067330 , R01AG074503 , Veterans Adminstration Merit Award BX003846 , and by Impetus Grants and the Simons Foundation . This work was made possible by support from the William S. Middleton Memorial Veterans Hospial Madison Wisconsin. V.D.L. has equity interest in L-Nutra, a company that commercializes medical food.
Declaration of interests
V.D.L. has equity interest in L-nutra, and is a member of its scientific advisory board. V.D.L. has patents related to fasting mimicking diets, discussed in this review.
See the full article can be found at cell.com.