The dietary exposome: a brief history of diet, longevity, and age-related health in rodents

Much of the research investigating the impact of dietary restriction (DR) on lifespan and healthspan in mammals has been undertaken in laboratory rats and mice. However, many model and non-model organisms have also been studied [1], including Rhesus macaques Macaca mulatta [2,3], Labrador retrievers Canis lupus familiaris, [4] and the Mouse lemur Microcebus danfossi [5]. In this review, I will highlight some of the early and notable studies that have laid the foundation for using DR as an intervention strategy (Figure 1) to better understand the fundamental mechanisms underlying ageing in mammals. Throughout, I have adopted The Jackson Laboratory's approximate life history stages for C57BL/6J mice (https://www.jax.org/news-and-insights/jax-blog/2017/november/when-are-mice-considered-old#), where a mature adult mouse is between 3 and 6 months (12–24 weeks), a middle-aged mouse is between 10 and 14 months (40–56 weeks), and an old mouse is between 18 and 24 months (72–96 weeks) of age (Figure 2). It should be noted that these approximations vary depending on several factors including sex and strain, with significant strain-specific variation in lifespan seen in rodents under both ad libitum (AL) feeding and DR feeding [1,6–11]. Additionally, I will briefly discuss how genetic background may influence the extent to which DR modulates ageing and explore the potential challenges we face in translating these findings to humans.

Caloric restriction

The recognition that temperance of the appetite can elicit health benefits has been acknowledged for hundreds of years (see [12]). Philip Paracelsus, the Swiss alchemist and physician who lived between 1493 and 1541, famously stated ‘Fasting is the greatest remedy–the physician within’. The earliest and still most widely used dietary restriction paradigm is a reduction in total calories in the absence of malnutrition, termed caloric restriction (CR). Probably the earliest example demonstrating the benefits of CR on ageing in mammals was provided by Osborne and colleagues and published in Science in 1917 [13]. Their interest was initially piqued by observational data showing that while around 70% of individual ‘stock’ rats in growth retardation studies died before reaching 24 months of age, ∼30% lived beyond 24 months under laboratory conditions. This observation led them to formally examine the impact of growth retardation from 1 month of age on fertility and lifespan in female rats. Growth restriction delayed age-associated reproductive decline, and maximum lifespan for all individuals exceeded 24 months, an age when 70% of stock animals had already died [13]. The authors concluded that ‘a preliminary stunting period lengthened the total span of their life’. However, it is important to note that only 4 individuals were studied, the period of restriction ranged from 5 months to 16 months, and no control group was included. Additionally, the cause of death for all animals was attributed to lung disease, and it remains somewhat unclear precisely how growth restriction was achieved, as it may have involved a period of malnutrition.

Several years later, Clive McCay and colleagues began to formally investigate the effect of CR on lifespan. They initially reported that Brook trout Salvelinus fontinalis maintained on a low protein diet with associated very low growth rates lived longer than trout fed high protein diets with typical growth rates [14]. This led the authors to suggest that ‘something was consumed in growth that is essential for the maintenance of life’. McCay and colleagues then shifted their focus to investigating the impact of nutrition on lifespan in rats [15]. In a large-scale study undertaken in both male and female white rats, individuals were divided at weaning into three groups; I) AL controls, II) Food intake restricted from weaning, or III) Food intake restricted from 2 weeks post-weaning. The food, and hence growth, restricted groups were switched back to AL feeding at either 766 days of age or 911 days of age. The diet composition was reported in the original paper and did not differ between groups, except that the restricted rats were periodically supplemented with cod liver oil, yeast and occasionally fresh beef liver. In males, both CR groups showed an increase in average lifespan relative to AL animals (483, 820, and 894 days for groups I, II, and III, respectively). Interestingly, however, no CR effect on longevity was observed in females (801, 775, and 826 days for groups I, II, and III, respectively), marking the first reported instance of sex-specific effects of CR on lifespan in rodents. Although the survival data in the present study were not statistically analysed, a later statistical evaluation confirmed the original observations of CR-induced longevity [16]. Male rats on the AL diet were much shorter lived than AL females, and while males were heavier than females, the percentage of body weight lost under CR was similar between sexes [15]. Tumour incidence at time of death was similarly reduced in male, but not female, rats under CR. This study also provides the first evidence of inter-individual differences in the response to CR on lifespan, and importantly, it sheds light on phenotypic differences, specifically body weight, between rats maintained at different institutions (Yale vs. Cornell). Following these early studies, many research groups began investigating the impact of CR on lifespan and healthspan in rodents. Key researchers include Weindruch and Walford (e.g. [17,18]), Merry and Holehan (e.g. [19,20]), and Masoro and Yu [21,22].

CR has consistently been shown to elicit broad-spectrum health benefits in rodents, including reducing the onset and incidence of both spontaneous [8,17] and experimentally induced tumours [23–25]. Interestingly, CR initiated in mid-life in male C57BL/6 mice increased the incidence of plasma cell neoplasms – despite increasing lifespan – relative to AL controls [26]. CR also protect against many age-related pathologies, including insulin resistance [8,27], sarcopenia [28,29], and cognitive decline [30,31] in rodents, although its impact on the immune system is somewhat less clear [32–35]. CR protects against ischaemia/reperfusion injury [36], and decreases pathogenesis while increasing survival in multiple disease models, including models of Parkinson's disease [37], Alzheimer's disease [38], viral myocarditis [39], and chronic kidney disease [40].

One clear obstacle to the translational potential of CR in humans is that chronic reduction in food intake over years or decades is unlikely to be a popular, or indeed achievable life choice for most people [41]. As a result, there has been significant interest, particularly over the last couple of decades, in determining whether manipulating specific macro- or micronutrients levels within the diet (dietary restriction, DR) can recapitulate the benefits of a CR diet without the need for chronic reductions in food intake. Indeed, the debate over whether it is the reduction in calories or specific dietary components driving the beneficial effects of DR has been a subject of contention for many years (see [12,42–45]). Modulating dietary macro- or micronutrient levels is challenging because the manipulated diet should be isocaloric with the control diet. Reducing or removing one component required replacing it with another, even if the replacement is metabolically inert [46].

Slonaker was likely the first investigator to empirically demonstrate that varying dietary protein affects average lifespan through a remarkable set of studies focussing not only on lifespan but also on energy expenditure, activity, growth, and reproduction in male and female albino rats [47–51]. However, and as discussed elsewhere [46], the lowest protein diets used by Slonaker also contained less fat but more carbohydrate than his ‘control’ diets. Notwithstanding, Slonaker's work was instrumental in highlighting the importance that dietary macronutrients in affecting longevity in rodents. He also provided practical advice on cage design and animal husbandry, suggesting for example that ‘where a large number of rats are being reared it is advisable to procure the refuse from a restaurant or hotel. Table scraps give a fairly balanced diet’ [52]. Ross [53,54] later undertook many studies in rats using semisynthetic, isocaloric diets that varied in their casein and sucrose content. He showed that lifespan was extended in rats given unrestricted access to a low-protein/ high-carbohydrate diet, compared with diets relatively higher in protein or lower in carbohydrate. Interestingly, food intake in the low-protein/high-carbohydrate group was lower than in the other diets, making it difficult to disentangle the effects of reduced calorie intake from altered dietary composition. Indeed, when the same diets were restricted, no difference in lifespan was observed between groups, although the levels of restriction differed significantly (ranging from 10% CR to 63% CR) [53]. Several other groups examined the effects of protein restriction on lifespan around this time, but these effects were modest, especially when compared with the effects of CR itself [55–59]. These findings led to the consensus that lifespan extension through DR was driven by reductions in calories per se rather than by reductions in protein (for further discussions and debate, see [43,46,60]).

However, this consensus has been challenged by numerous studies showing that restricting macronutrients or micronutrients can significantly impact lifespan in rodents [55,61–64]. Large-scale studies by Simpson and colleagues have employed a Geometric Framework approach [65,66] to investigate interactions between dietary components and food intake, lifespan, and metabolic health in mice [44,67–69]. For example, this group systematically examined the influence of 25 diets, fed AL over the entire lifetime, that varied in dietary protein, lipid, and carbohydrate content, and differed in terms of caloric density via the addition of indigestible cellulose (low- 8, medium- 13 or high 17kJ/g energy diet) on lifespan and healthspan in C57BL/6 mice [67]. They found that the greatest median survival in male and female mice was achieved in mice fed a low protein/high carbohydrate (LPHC) diet, but total caloric intake did not influence this outcome. Importantly, the effect of dietary protein was independent of total calorie consumption, and as the ratio between protein and carbohydrate increased, median lifespan decreased. Notably, the diet that promoted longevity also induced positive cardiometabolic effects, although the diets that maximized lifespan were not the same as those that maximized reproductive function [69].

A much shorter term (8 week) study also demonstrated that LPHC diets replicated many of the metabolic benefits of CR, even when provided ad libitum, but no additive effects on metabolic outcomes were observed when the LPHC diet was restricted [68]. More recently this group reported that carbohydrate quality (e.g. resistant starch v fructose and glucose) in these LPHC diets also significantly affects metabolic health in mice [70]. Using a comprehensive meta-analytical approach from both invertebrates and vertebrates studies, Nakagawa and colleagues [1] reported that the amount of protein consumed was more important to life extension under DR than the amount of calories consumed.

Lamming and colleagues at the University of Wisconsin-Madison have published many well-designed and important studies examining the impact of diets containing single or combined reductions in specific amino acids (AAs), and reduced total protein on metabolic health and lifespan in mice. For example, they showed that male C57BL/6J mice fed a diet with reduced branch-chained amino acids (BCAAs) leucine, isoleucine, and valine from 9 weeks of age had improved glucose tolerance and β-cell function relative to chow fed control mice, similar to outcomes in male mice fed a low-AA diet (7% of protein) [71]. Critically, mice in both groups were lighter and leaner than chow fed mice over the 10-week study duration. However, the impact on metabolism was not comparable: mice on a low-AA diet had higher FGF-21 plasma levels and increased Fgf21 mRNA expression levels in liver and muscle, increased plasma adiponectin levels and higher night-time energy expenditure. These parameters were unaffected in BCAA-restricted mice, although, like low AA-fed mice they were hyperphagic [71].

It was later reported that a reduced BCAA-diet leads to fat mass loss, greater insulin sensitivity, and improved glucose tolerance in mice fed a high-fat, high-sugar diet [72]. In contrast with the earlier findings, BCAA-restriction on this obesogenic diet did increase FGF-21 levels and energy expenditure, suggesting that some metabolic effects of BCAA restriction are diet (context)-dependent. Furthermore, this group showed that a reduced BCAA diet increased lifespan in progeroid Lamin A/C deficient mice (Lmna^−/−), although only in females, not males, and in a mouse model of Hutchinson–Gilford progeria syndrome (HGPS; Lmna^G609G/G609G) [73]. In the same paper, the authors then went on to examine the potential geroprotective effects of a BCAA-restricted diet in male and female C57BL/6N mice fed the restricted diet from 16 months of age. Again, the BCAA-restricted diet conferred benefits on glucose metabolism and reduced fat mass and it significantly slowed the onset of frailty during ageing in both males and females. However, no significant effect of on lifespan was detected in either sex when the restricted BCAA-diet was initiated from mid-life [73]. Interestingly, when the BCAA-restricted diet was fed from weaning onwards, male median lifespan increased by 32% relative to male controls, although no effect on lifespan was seen in females, and the diet's positive impact on frailty was only observed in males after 25 months of age. While this diet reduced mTOR signalling in the liver and muscle of male mice, no reduction in mTOR signalling was observed in females. Additionally, a diet containing high levels of BCAAs induced hyperphagia, obesity, and altered appetite signalling within the brain, ultimately shortening lifespan in male and female C57BL/6J mice [74]. These effects appear to be driven by an AA imbalance promoting hyperphagia rather than any intrinsic toxicity of the high BCAA diet per se.

Restriction of specific amino acids within the diet has also been shown to positively impact on lifespan of rodents. Restricting the essential amino acid methionine in the diet (Meth-R: from 0.86% to 0.17% in an otherwise isocaloric diet) significantly increased both median and maximum lifespan of male Fischer 344 rats by ∼40% [75,76]. Preliminary reports also suggest that Meth-R also extends median lifespan across other inbred (Brown Norway) and outbred (Sprague Dawley, Wistar Hannover) rats strains [77]. Meth-R significantly extended lifespan in female CB6F₁ hybrid mice exposed to Meth-R from 6 weeks of age [78]. However, significant early-life mortality was observed in the Met-R group, primarily due to an increased incidence of rectal prolapse. This issue led to the amount of methionine being increased in the diet over the first 6 months of the study to mitigate these effects. Meth-R also slowed immune and lens ageing in these mice and increased resistance to acetaminophen-induced liver damage [78]. However, in contrast with some [79] but not all long-lived mouse models [80], Meth-R does not increase in vitro cellular stress resistance [81]. Exposure to a Meth-R diet from 12 months of age also significantly increased lifespan in male CB6F₁ mice [82].

Similar to CR, Meth-R confers many metabolic benefits in rodents during ageing including lowering visceral adiposity, inducing mitochondrial biogenesis and increasing metabolic rate [83–87]. Meth-R also increases UCP1 levels and induces browning within white adipose tissue, possibly thought increasing serum FGF21 levels. Interestingly, this effect may be sex-specific [85]. Meth-R lowers of serum lipids, insulin, IGF-1, and glucose levels [83,84,87,88]. However, in contrast with CR, the effects of Meth-R on lifespan are not associated with a reduction in food intake, as rodents on Meth-R are hyperphagic [75,76,85]. Meth-R has also been shown to delay pathology and extend lifespan in progeroid mice [89], and protect against renal injury in mice [90].

Several studies have also investigated the impact of restricting other single AAs on rodent lifespan. In female Long-Evans rats, reducing dietary tryptophan (restricted diets contained only 40% [T40] or 30% [T30] of the tryptophan contained in a control diet) increased late-life survival [91]. However, similar to Meth-R in female mice, restricting dietary tryptophan increased early-life mortality (100% of controls survived to year 1, 79% of the T40 group survived, but only 58% of T30 females survived to year 1). The restricted groups were much lighter than control rats [91] and while the initial study suffered from a small sample size and was highly observational, this group had previously reported that tryptophan restricted female rats were reproductively active for longer, their coat condition was preserved, and tumour onset was delayed during ageing [92]. Similarly, in male Swiss albino mice restricting dietary tryptophan modestly increase median and maximum lifespan, though neither parameter was statistically analysed [93]. Short-term (8 weeks) leucine restriction improved several metabolic parameters in male C57BL/6J mice, although these effects were less pronounced compared with Meth-R and, unlike Meth-R, leucine restriction did not reduce hepatic lipogenic gene expression [88]. The impact of leucine restriction on metabolism appears complex, as 10 weeks of leucine restriction in male C57BL/6J mice increased epididymal and dermal white adipose tissue, although these animals were not hyperphagic, with improved glucose tolerance observed after only 3 weeks on the leucine restricted diet [71].

Regarding the restriction of other AAs, it has been reported that reducing dietary levels of histidine, phenylalanine or threonine slows weight gain and adiposity in male C57BL/6J mice over a 3-month period, unlike lysine, methionine or tryptophan restricted diets [94]. The histidine restricted diet also increased insulin sensitivity and reversed both hepatic steatosis and obesity in high fat diet-induced male mice, independently of FGF-21. Improvement in glucose tolerance and a reduction in adiposity was seen in male, but not female, C57BL/6N mice when fed a histidine-restricted diet from 16 months of age, although no effect on frailty or lifespan was observed [94]. In contrast, restricting dietary isoleucine in HET3 mice improved metabolic health, increased food intake and energy expenditure, reduced frailty in both male and female mice, and increased lifespan in both sexes, although to a much greater extent in males [95]. It has also been shown that both male and female C57BL/6J and DBA/2 mice exposed to a high-fat, high-sucrose Western diet in combination with an isoleucine restricted diet are protected against much of the metabolic dysfunction induced by Western diet [96].

Ketogenic diets (KD), which are high in fat and low in carbohydrates, have become a popular weight-loss intervention, and are widely used to help prevent seizures in humans [97,98]). KD have been shown to reduce ataxia, lessen seizures, prevent weight loss, reverse nephropathy in mouse models of type I and type II diabetes mellitus, and increase lifespan in mouse models of neurological disease [99–101]. It was reported that feeding male C57BL/6 mice KD from 12 months of age increased median survival (14% increase relative to mice fed control diet) and preserved motor coordination, muscle mass, and memory during ageing, although glucose tolerance appeared impaired [102]. A second study published around the same time examined lifespan and healthspan in male C57BL/6N mice given access to a KD on a cyclical basis- KD for one week, control diet for one week- from 12 months of age onwards. The KD reduced mortality in middle age, but tended to increase mortality later in life, resulting in a non-significant effect on overall lifespan [103], although the KD diet did preserve memory during ageing. In contrast, the lifespan of C57BL/6J mice fed a KD from 8 weeks of age was not different from chow-fed control animals [104]. While the KD improved glucose tolerance and insulin sensitivity relative to control mice, it also induced hepatic steatosis and inflammation [104]. Recently, it was reported that a KD increased cellular senescence in both C57BL/6J mice and humans [105]. Interestingly, the effects on cellular senescence were induced in mice exposed to a KD every-other-day. Consequently, it is probably still too early to be confident in the effect of KD on lifespan, as outcomes appear to be highly dependent on how and when the diet is administered [106].

Mice subjected to CR regimes typically experience extended periods of fasting, because they consume their rations within 2–3 h [12]. As a result, these animals spend the vast proportion of their day in a fasted state, particularly under feeding protocols that provide food every second or third day. Probably the earliest evidence that intermittent fasting (1 day of fasting every 2, 3, or 4 days) could extend lifespan was provided by Carlson and Hoelzel in 1946 [107]. They reported that rats fasted for 1 day out of every 3 showed the greatest lifespan increase across both males and females. No change in growth rate was reported although mammary tumour development decreased with the amount of fasting. Several studies have employed chronic every-other-day (EOD) feeding regimes in mice, where animals are provided AL access to food one day and no food the next. EOD feeding has been shown to extend lifespan in male Wistar rats [108,109] and male C57BL/6J, A/J and F1 hybrid B6AF1/J mice [110,111]. However, the effect on lifespan in mice appears to be highly dependent on the age of initiation of EOD feeding and genetic background. Indeed, EOD feeding reduced both mean and maximum lifespan in A/J mice when initiated at 10 months of age [110]. In contrast, male Wistar rats switched from AL to EOD feeding at 12 months of age or switched from EOD to AL feeding at 12 months of age lived significantly longer than rats fed an AL diet throughout life. However, neither of the switched groups lived as long as rats maintained on EOD feeding throughout their entire lives [109]. While EOD has been shown to protect hippocampal neurones against excitotoxic injury in C57BL/6 mice [112], protect male Sprague-Dawley rats against ischaemic injury to the heart [113], ameliorate development of neoplastic disease in male C57BL/6J mice, and slow frailty progression in C57BL/6 male, but not female, mice [114], the impact on other age-related phenotypes appears minimal [111].

Longo and colleagues reported that exposing female C57BL/6 retired breeders to a fasting mimicking diet (low calorie/low protein for 8 days per month, with AL access to chow diet for the rest of the month) from 16 months of age significantly extended lifespan and healthspan [115]. Time-restricted feeding protocols where animals are only provided with food for a short period each day, increasing the fasting time, have also been shown to extend rodent lifespan [116,117]. Even under conditions of overnutrition, male C57BL/6 mice exposed a time-restricted feeding protocol had a similar caloric intake to AL high fat fed mice but were protected against metabolic dysfunction and pathology [118]. These time-restricted diets could stabilize and even reverse pathology in mice with preexisting obesity and/or type 2 diabetes [119]. More recently, it was reported that the positive effects of CR on lifespan in male C57BL/6J mice were mediated via energy imbalance rather than by a reduction in absolute energy intake or protein intake [45].

Surprisingly, fewer studies have investigated the impact of fat restriction on longevity. Iwasaki and colleagues [120] examined the effects of fat restriction, without restricting overall caloric intake, in male Fischer-344 rats by limiting corn-oil in the diet from 6 weeks of age. No differences in medium or maximum lifespan were observed between AL rats and the fat-restricted group. However, the incidence of chronic nephropathy, cardiomyopathy, and gastric hyperkeratosis was reduced in the fat restricted group relative to AL rats, while the incidence of lymphoma and leukaemia was higher. It should be noted that the reduction in corn oil in the diet (10g/100g of diet in control diet vs. 6g/100g of diet in fat restricted group) was compensated by a 4g increase (per 100g of diet) in the complex carbohydrate Dextrin [120]. The Yu and Masoro group later provided further evidence of the lack of effect of dietary fat restriction on lifespan [121].

Additionally, it has been reported that dietary lipid composition can influence longevity under DR. C57BL/6J mice subjected to 40% CR relative to control mice (5% CR with soybean oil), but exposed to diets differing in lipid composition – 1) soybean oil, 2) fish oil or 3) lard – all lived significantly longer than the 5% CR group [122]. Intriguingly, the lard-fed 40% CR mice lived significantly longer than the other two 40% CR groups. While no difference in neoplasm incidence was observed between groups, the incidence of great vessel mineralization was much higher in the 40% CR groups, irrespective of dietary fat source [122]. The lard-fed mice also exhibited preserved renal and skeletal muscle structure and function during ageing compared with the other groups [123,124]. Interestingly, 7 weeks of fat restriction (∼50% less dietary soybean oil) in male Wistar rats had no effect on hepatic reactive oxygen species (ROS) production or oxidative stress, unlike protein restriction or CR [125]. Similarly, while average telomere length ratio (ATLR) was greater in long-lived mice fed a LPHC diet, no relationship was found between ATLR and dietary fat intake [126].

As discussed elsewhere in detail [9,127], most mice used in biomedical research are derived from the highly inbred C57BL/6 strain originally bred at the Jackson Laboratories in the 1920s and 30s. It has been estimated that around 90% of all biomedical research using mice employs the original C57BL/6J strain or its sub-strains. It is now evident that the impact of CR/DR on lifespan and health in rodents is dependent on many factors, including genetic background, age-of-onset of diet, level of restriction, method of restriction, and sex [6,8,9,11,96,128–134]. Humans are genetically heterogeneous and DR has typically been studied in inbred (male) rodents. Expanding the genetic diversity of rodent models and studying both males and females is crucial [9,132,135,136]. This is beginning to happen, with dietary manipulations showing significant effects on lifespan and health depending on sex and genetic background [8,9,95,127,137].

As mentioned earlier in this review, lifelong caloric restriction is unlikely to be a life choice that many people are willing or able to make, despite the well-publicized benefits on health and lifespan in multiple model and non-model organisms, and the well-established benefits on metabolic health and age-related outcomes in humans [138–141]. Consequently, diets or drugs that can mimic the benefits of DR without requiring significant changes in food intake, dietary habits and/or behaviour are more likely to succeed [41,60,142], i.e. have your cake and eat it. Several candidate drugs, such as rapamycin, trametinib, and metformin, show geroprotective properties in model organisms [41,143–145] and thus have significant potential as DR mimetics for translation.

What is clear is that the field has come a long way since the pioneering work of Osbourne, McCay and Slonaker (Table 1), but there is still much to learn mechanistically about how these different diets modulate late-life health and longevity. We need to better understand how translatable and safe these diets are before they can be recommended as health interventions [146–148]. It is unlikely that one level of restriction or a specific dietary manipulation will work for everyone. As first proposed by Margaret Ohlson in the 1950s [149], precision nutritional approaches may be needed to maximise health benefits for individuals.

Intermittent fasting protocols have received significant attention over the past decade because they are seen as more realistic intervention strategy, although interestingly, adherence, weight loss, or cardioprotection do not appear to outperform CR [150]. However, the data from human studies on intermittent fasting are mostly very encouraging [151,152], although as expected the effects on health outcomes appear complex [153]. Despite over 100 years passing since Osbourne's initial observations [13], much remains to be understood from both the pre-clinical and clinical perspectives about how exactly these dietary interventions work, and how they interact with other factors such as age, sex, genetics, deprivation, and other interventions known to improve, or impair, overall health [142]. Still, there are certainly reasons to be optimistic.

The author declares that there are no competing interests associated with the manuscript.

This work was funded through the Biotechnology and Biological Sciences Research Council (BBSRC) [grant number BB/S014330/1].

Colin Selman: Writing—original draft, Writing—review & editing.

I am grateful to Will Mair and Matt Piper for initial discussions around this topic and apologise to the multiple authors whose work was not cited in this review due to space limitations.

AA

amino acid

AL

ad libitum

ATLR

average telomere length ratio

BCAA

branch-chained amino acid

CR

caloric restriction

DR

dietary restriction

Meth-R

methionine restriction

EOD

every-other-day

FGF-21

fibroblast growth factor 21

HGPS

Hutchinson–Gilford progeria syndrome

IGF-1

insulin-like growth factor 1

KD

ketogenic diets

Lmna−/−

lamin A/C deficient

LPHC

low protein/high carbohydrate

ROS

reactive oxygen species

UCP1

uncoupling protein 1