Caloric restriction improves health and survival of rhesus monkeys

Julie A. Mattison; Ricki J. Colman; T. Mark Beasley; David B. Allison; Joseph W. Kemnitz; George S. Roth; Donald K. Ingram; Richard Weindruch; Rafael de Cabo; Rozalyn M. Anderson

doi:10.1038/ncomms14063

Caloric restriction improves health and survival of rhesus monkeys

Mattison, Julie A.; Colman, Ricki J.; Beasley, T. Mark; Allison, David B.; Kemnitz, Joseph W.; Roth, George S.; Ingram, Donald K.; Weindruch, Richard; de Cabo, Rafael; Anderson, Rozalyn M. 2017-01-17 00:00:00 ARTICLE Received 31 May 2016 | Accepted 24 Nov 2016 | Published 17 Jan 2017 DOI: 10.1038/ncomms14063 OPEN Caloric restriction improves health and survival of rhesus monkeys 1, 2, 3,4, 3 2,5 Julie A. Mattison *, Ricki J. Colman *, T. Mark Beasley *, David B. Allison , Joseph W. Kemnitz , 6 7 8,9 1, 8,9, George S. Roth , Donald K. Ingram , Richard Weindruch , Rafael de Cabo ** & Rozalyn M. Anderson ** Caloric restriction (CR) without malnutrition extends lifespan and delays the onset of age-related disorders in most species but its impact in nonhuman primates has been controversial. In the late 1980s two parallel studies were initiated to determine the effect of CR in rhesus monkeys. The University of Wisconsin study reported a signiﬁcant positive impact of CR on survival, but the National Institute on Aging study detected no signiﬁcant survival effect. Here we present a direct comparison of longitudinal data from both studies including survival, bodyweight, food intake, fasting glucose levels and age-related morbidity. We describe differences in study design that could contribute to differences in outcomes, and we report species speciﬁcity in the impact of CR in terms of optimal onset and diet. Taken together these data conﬁrm that health beneﬁts of CR are conserved in monkeys and suggest that CR mechanisms are likely translatable to human health. 1 2 Translational Gerontology Branch, National Institute on Aging, Baltimore, Maryland 21224, USA. Wisconsin National Primate Research Center, University of 3 4 Wisconsin-Madison, Madison, Wisconsin 53715, USA. Department of Biostatistics, University of Alabama, Birmingham, Alabama 35294, USA. Geriatric Research Education and Clinical Center, Birmingham/Atlanta Veterans Administration Hospital, Birmingham, Alabama 35233, USA. Department of Cell and 6 7 Regenerative Biology, University of Wisconsin-Madison, Madison, Wisconsin 53792, USA. GeroScience, Pylesville, Maryland 21323, USA. Pennington Biomedical Research Center, Baton Rouge, Louisiana 70808, USA. Department of Medicine, University of Wisconsin, Madison, Wisconsin 53792, USA. Geriatric Research Education and Clinical Center, William S. Middleton Memorial Veterans Hospital, Madison, Wisconsin 53705, USA. * These authors contributed equally to this work. ** These authors jointly supervised this work. Correspondence and requests for materials should be addressed to R.d.C. (email: [email protected]) or to R.M.A. (email: [email protected]). NATURE COMMUNICATIONS | 8:14063 | DOI: 10.1038/ncomms14063 | www.nature.com/naturecommunications 1 ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14063 he rhesus monkey (Macaca mulatta) is an excellent model CR however, and analysis was based on comparing 109 ad libitum for human ageing. The rhesus monkey genome shares fed males and females from colony records at that facility, 1,2 TB93% sequence identity with the human genome , and including insulin resistant and diabetic animals, to only eight numerous aspects of their anatomy, physiology, neurology, male CR monkeys. Two other studies focused speciﬁcally on the endocrinology and immunology directly parallel those of impact of CR in healthy male and female rhesus monkeys: one at 3,4 humans . Rhesus monkeys have a lifespan measured in the National Institute on Aging (NIA) involving 121 monkeys; decades, and develop, mature and age in similar ways to and the other at the University of Wisconsin Madison (UW) with humans. In terms of ageing this includes greying and thinning 76 monkeys. The same University of Alabama at Birmingham of hair, redistribution of body fat, loss of skin tone, loss of vigour based statistical team was engaged for analysis of data from 4–7 and loss of muscle tone . With age there are increases in clinical both studies. The UW study has reported beneﬁcial effects of CR, manifestations of diseases and disorders that also increase in including signiﬁcant improvements in health and age-related 19 20 prevalence with advancing age in humans, including diabetes, survival , and all-cause survival . In contrast, the NIA study neoplasia, sarcopenia, bone loss, altered immune function and reported no signiﬁcant impact of CR on survival, although 3,4,8 21 cognitive decline . Like humans, nonhuman primates are improvements in health were close to statistical signiﬁcance . genetically heterogeneous so that phenotypes of ageing are The basis for the contrasting outcomes from these two parallel non-uniformly manifested among individual animals. Variance studies has not been established. Analysis of limited published can be somewhat offset through comprehensive life-long physical bodyweight data indicated that the controls were not equivalent and medical health records, where age-related changes are viewed between the two studies , pointing to fundamental differences in from the perspective of each animal individually. Nonhuman study design and implementation. Therefore, to more fully assess primates exhibit feeding patterns and sleeping behaviour similar possible explanations for the discrepant ﬁndings between the two to those of humans, and a key advantage of nonhuman primate studies, we have conducted a comprehensive assessment of studies over human studies is that all variables including the longitudinal data from both sites highlighting differences that environment and dietary intake can be fully deﬁned. In short, may have contributed to the dissimilar outcomes. nonhuman primates are vital models for translating basic research into clinical application. A clear understanding of the biology of ageing, as opposed to Results the biology of individual age-related diseases, could be the critical Intrinsic differences in study design. Most of the early rodent turning point for novel approaches in preventative strategies to CR studies involved very young onset life-long CR initiated post- facilitate healthy human ageing. Caloric restriction (CR) offers a weaning, usually in inbred genetic backgrounds. In the 1980s it powerful paradigm to uncover the cellular and molecular basis for became clear that adult onset CR (12-month-old mice) was also the age-related increase in overall disease vulnerability that is effective in delaying ageing and extending lifespan in rodents, shared by all mammalian species. CR extends median and albeit to a lesser extent than the young onset model . Many maximum lifespan in most strains of laboratory rodents and also rodent CR studies opt to feed control animals ad libitum amounts 9–12 delays the onset of age-related diseases and disorders . of food while others provide less than ad libitum amounts arguing Lifespan is also extended by CR in most short-lived species, that this strategy avoids the confounding effects of obesity including the unicellular yeast, nematodes and invertebrates. and reduces variability in food intake among individuals. There has been rapid progress in identifying potential With the launch of the NIA rhesus monkey study in 1987, the 13–17 mechanisms of CR utilizing these models . These implementation of CR was such that the control monkeys were short-lived species are well suited for the investigation of the not free-fed. Food allotments were determined in accordance with underlying mechanisms of CR due to the relative ease in their data published by the National Research Council to provide genetic manipulation, extensive genetic and developmental approximate ad libitum intake based on their age and bodyweight characterization, low cost, and signiﬁcantly reduced timeframe for the maturing control monkeys without overfeeding . Rations for completion of longevity studies. A key question underpinning were increased to maintain growth and development until full this body of work is whether the biology of CR, and its ability to stature was attained. CR monkeys received 30% less food than delay ageing and the onset of disease, applies to humans and height-, age- and sex-matched control monkeys. The intervention human health. was initiated as young-onset and old-onset groups of males, and To date three independent studies of rhesus monkeys (Macaca young, adult, and old-onset groups of females (Table 1). mulatta) have tackled the question of translatability of CR to Launched in 1989, the UW study initiated the CR diet in adult primate species. The University of Maryland rhesus monkey animals only, after full stature was achieved (B8 years of age for study was the ﬁrst to report a positive association of CR with rhesus monkeys) . Food was provided at levels approximating survival with a 2.6-fold increased risk of death in control animals ad libitum to control animals. To accommodate heterogeneity in compared to restricted . The primary focus of the study was not the feeding behaviours within the cohort, the ad libitum reference Table 1 | Study design. Site Group Control N CR N Age of onset Genetic origin NIA Male Juvenile 10 10 1–2 years Indian/Chinese Adolescent 12 10 3–5 years Indian/Chinese Old 10 10 16–23 years Indian/Chinese Female Juvenile 9 9 1–3 years Indian Adult 15 11 6–14 years Indian/Chinese Old 8 7 16–21 years Indian UW Male 23 23 7–14 years Indian Female 15 15 9–15 years Indian 2 NATURE COMMUNICATIONS | 8:14063 | DOI: 10.1038/ncomms14063 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14063 ARTICLE for each individual was established using baseline food intake deprived overnight. While there were considerable differences in measured over 3–6 months, and CR was implemented on a study design as outlined above, it should be noted that animal per-individual basis. The rationale for these design features at housing and routine animal care were equivalent at NIA and UW UW was to implement a study as it might have been conducted in primate facilities. This included identical housing conditions, humans. temperature and humidity range, light cycles, and the use of tap The source of the monkeys in each cohort and the population water, which was continuously available. Both studies included type represented is also a point of difference for the two studies. animal monitoring several times per day, and a designated The UW monkeys were born and raised at the Wisconsin veterinary staff that inspected the animals routinely and provided National Primate Research Center and were all of Indian origin. outstanding care as needed. The NIA monkeys were sourced from several locations and included monkeys of both Indian and Chinese origin. Chinese male rhesus monkeys are generally heavier and longer than their Impact of CR on survival. The initial goal of both NIA and UW Indian counter parts with the reverse being the case for females, studies was to determine the impact of CR on the health of rhesus and Chinese rhesus monkeys are also thought to exhibit greater monkeys, as it was not a foregone conclusion that CR would be sexual dimorphism . Monkeys of different origin are sufﬁciently an appropriate intervention in long-lived species. The investiga- genetically different that they can be distinguished using a panel tion of the impact of CR on longevity was not considered a of single nucleotide polymorphisms . Apart from population primary outcome at either study location. Even though 121 differences, rhesus monkeys share a similar degree of inter- monkeys were enrolled in the NIA study, the differences in age of individual genetic variation as humans . In this way, the onset (from 1 to 23 years) precluded the animals from being contribution of population type to differences in outcomes of grouped together for data analyses. Although the age range for the two studies as opposed to the contribution of individual time of onset is smaller for the UW study (ages 7–15), with only genetic heterogeneity is difﬁcult to ascertain. 38 outbred genetically distinct monkeys per group (including The diet compositions were another important difference both sexes), it seemed unlikely that the study would have the between the two studies. First, the source of diet components was statistical power required to test CR’s effect on longevity. While different. A naturally sourced diet was employed at the NIA neither study reports longevity data, both studies have yielded facility to ensure that micronutrients such as phytochemicals and survival data. For rhesus monkeys in captivity, the previously trace minerals were provided, acknowledging that there was reported median survival was B26 years of age, 10% survival was potential for seasonal variation. In contrast, a semi-puriﬁed diet B35 years of age and maximal survival was B40 years of age . was employed at UW to ensure that intake could be fully deﬁned Mortality curves were generated separately for UW and NIA and consistent throughout the course of the study. Second, (Fig. 1). Survival estimates for monkeys at both sites were although diets at both locations had a similar caloric density, the calculated based on data captured up to July 2015 using the three relative macronutrient composition of the diets was not most common statistical methods: Kaplan-Meier product-limit equivalent (Table 2). Compared to the UW diet, the NIA diet method; Cox proportional hazard regression and parametric was lower in fat, higher in protein and higher in ﬁbre. Finally, the survival analysis assuming a Weibull distribution (Table 3). nutrient content of the diets was also different. At both locations Because the Weibull distribution is a special case of the diets containedB60% carbohydrates by weight, but sucrose generalized extreme value distribution, it can accommodate comprised less than 7% of total carbohydrates at NIA and 45% estimation of the upper quantiles of a survival distribution and of total carbohydrates at UW. Diets at both locations were replete maximal lifespan, especially when there are censored data due to for vitamins that were provided at or above the recommended animals that remain alive . daily allowance. In the UW adult-onset study, the estimated survival of UW Feeding practices also differed between studies. At NIA, the control animals was close to that of the average recorded for monkeys were fed two meals at B6:30 and 13:00 each day. Any monkeys in captivity (B26 years of age). Considering both males food remaining after the morning meal was removed after about and females together, a statistically signiﬁcant effect of CR in 3 h, and a low calorie treat was provided, typically in the form of a increasing survival was observed (Cox regression P¼ 0.017; small piece of fruit. The afternoon meal was not removed so that Supplementary Table 1). The hazard ratio (HR) of 1.865 (95% monkeys had access to food at night. At UW, all monkeys were conﬁdence interval (CI): 1.119–3.108) indicated that at any time- fed in the morning atB8:00 and any remaining food was point the control monkeys had almost twice the rate of death removed at B16:00 when a treat of fresh fruit or vegetable, which when compared to CR animals. The effect of sex on the response was quickly and completely eaten, was provided. Food allotment to CR was not statistically signiﬁcant. Kaplan-Meier analysis for control animals was adjusted to ensure that there was always showed that median survival estimates were greater for CR some uneaten food to be removed at the end of the day. In this animals for both males and females (Table 3). In the NIA study way UW animals were ad libitum fed during the day but food large differences in ages of monkeys at time of recruitment to the Table 2 | Diet composition at each location. Diet component NIA UW Control/CR Nutrient sources Control CR Nutrient sources (% by weight) (% by weight) (% by weight) Protein 17.3 soybean and ﬁsh meal 13.13 13.13 lactalbumin Carbohydrate 56.9 wheat, corn, sucrose (6.8%) 60.92 58.31 corn, sucrose (45%), dextrin Fat 5.0 soy, corn, ﬁsh oils 10.6 10.6 corn oil Fibre 6.5–9.0 cellulose 5.0 5.0 cellulose Vitamins B140% RDA 100% RDA B130% RDA Calories kcal g 3.9 3.9 3.8 NATURE COMMUNICATIONS | 8:14063 | DOI: 10.1038/ncomms14063 | www.nature.com/naturecommunications 3 ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14063 100 100 Control: 2 CR: 5 Control: 1 CR: 5 Average Average 60 60 onset onset UW females UW males 10 20 30 40 10 20 30 40 Control: 10 CR: 7 Control: 9 CR: 7 Average Average onset onset NIA J/A females NIA J/A males 0 0 10 20 30 40 10 20 30 40 Control: 0 CR: 1 Control: 1 CR: 0 Average Average onset 60 onset 40 40 NIA old females NIA old males 0 0 10 20 30 40 10 20 30 40 Age in years Control Age in years CR Figure 1 | Mortality curves for monkeys at UW and at NIA. These curves depict data for male and female monkeys on the UW study and on the NIA study. Animals are grouped by age where male J/A include juvenile and adolescent onset animals, female J/A include juvenile and adult onset animals, and old include the advanced age onset animals. Inset boxes indicate animals still alive, dashed line marks 50% mortality. Statistics related to this ﬁgure are provided in Supplementary Information, Supplementary Table 1. Table 3 | Survival estimates. Groups KM KM KM KM Weibull Weibull Median IQR Mean SE Median IQR All cause survival estimates Males UW Control 26.11 6.82 25.28 1.03 25.75 5.57 CR 28.32 6.19 26.86 1.23 27.63 8.04 NIA J/A Control 28.78 NE 26.00 1.50 29.22 14.31 CR 26.31 13.06 23.71 1.91 25.86 17.05 NIA old Control 34.78 9.16 34.19 1.89 34.80 7.23 CR 37.10 10.42 34.82 1.95 35.88 8.71 Females UW Control 23.86 7.92 23.56 1.29 24.57 8.62 CR 29.68 15.60 25.78 2.18 27.49 14.37 NIA J/A Control 25.67 12.65 24.58 1.71 25.81 14.21 CR 22.63 NE 19.79 1.20 23.53 15.25 NIA old Control 27.19 8.98 28.57 1.59 29.71 8.91 CR 27.87 11.88 26.13 2.65 26.49 8.80 KM, Kaplan-Meier; IQR, interquartile range. study (Table 1) prompted a separation of data from the early and Although Cox proportional hazard regression indicated that the late onset groups. Here and throughout this report, NIA male differences in survival between J/A control and CR were not juveniles and adolescents (J/A) were grouped and female juveniles statistically signiﬁcant (Supplementary Table 1), CR monkeys and adults were grouped (J/A). The Kaplan-Meier median reached 80% mortality before the controls for both sexes. With estimated survival was not different between NIA control and 38% of the NIA J/A cohort still alive, the survival curves are CR animals for the J/A onset groups of males or females (Fig. 1). incomplete and the impact on survival remains to be determined; 4 NATURE COMMUNICATIONS | 8:14063 | DOI: 10.1038/ncomms14063 | www.nature.com/naturecommunications Percent survival Percent survival Percent survival NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14063 ARTICLE however, the early mortality suggests that for some individuals complications due to endometriosis. A further contributing factor implementation of CR in the very young may confer a survival relates to the policy on treatment of clinical conditions. At UW risk. For old-onset CR, Kaplan-Meier estimated survival was not the policy to treat clinical conditions was implemented from the different between control and CR groups for either males or outset. At NIA, although acute pain and suffering were always females (Table 3), but survival estimates were higher than those of treated, chronic medical conditions, including endometriosis, J/A monkeys and UW controls. For both males and females, were monitored but not medically treated. A policy change was survival estimates for the NIA old-onset cohort were comparable implemented in 2010 due to the high incidence of endometriosis. to or exceeded those for UW CR. The power to assess the impact of CR on survival for NIA J/A Although there were slight discrepancies in the estimated females has been compromised somewhat by this one condition. median survival between the non-parametric Kaplan-Meier and parametric Weibull estimation methods, the survival comparisons between study sites using either analysis were consistent. Biometric and food intake measures from both studies. For A certain degree of sexual dimorphism was observed in survival over a quarter of a century during these studies, bodyweight, body outcomes where incidence of early death appeared to be greater composition and food intake were measured for all 197 monkeys. for females. This observation might be explained in part by Bodyweight was determined in fasted and anesthetized monkeys endometriosis, which is the proliferation of endometrial tissue 2–4 times per year during routine procedures. Longitudinal data outside of the uterus. Endometriosis can occur at relative for all monkeys were averaged by age of the animal (Fig. 2a). As is high incidence in monkeys in captivity (B25%), and risk is the case for humans, monkeys often experience cachexia or 31,32 considerably greater for nulliparous females . Incidence of end-of-life rapid weight loss. To avoid confounding effects of endometriosis was equivalent for control and CR groups. For the weight change that is not related to food intake or diet, data from J/A cohorts in the NIA study, 12 of the 44 females died of the last year of life for each monkey were excluded. To facilitate complications due to endometriosis, and of these the juvenile comparisons among the cohorts, data were grouped into three age onset females were conﬁrmed nulliparous. Females recruited to categories representing young adult (11–13 years of age), late the UW study, in contrast, had at least one but no more than mid-age (18–20 years of age) and advanced age (25–27 years of three healthy infants , and only 2 of 30 females died of age) (Supplementary Tables 2 and 3). 7 7 12 13 12 12 7 9 13 11 9 9 12 11 11 8 7 17 17 11 14 13 12 11 18 18 11 10 19 11 4 4 11 4 11 8 7 11 10 11 8 7 7 7 19 14 12 7 7 5 4 9 5 8 5 4 3 3 3 15 4 16 6 17 21 7 4 3 19 18 18 19 3 19 3 5 18 17 5 5 NIA J/A females UW females NIA old females 15 20 25 20 25 30 10 15 20 25 Age in years Age in years Age in years 16 16 15 21 21 19 17 19 18 14 14 20 14 17 18 15 19 9 19 19 17 11 19 18 19 19 19 18 12 19 12 20 7 17 7 18 20 19 8 12 20 6 7 9 15 12 14 19 6 8 8 15 15 13 12 15 13 10 8 18 19 20 17 9 10 8 10 15 14 10 21 21 9 18 17 9 8 18 10 6 9 15 8 17 8 8 8 7 8 8 8 8 18 18 NIA J/A males UW males NIA old males 10 15 20 25 10 15 20 25 20 25 30 Age in years Age in years Age in years Control CR b Males Females 11–13 years 18–20 years 25–27 years 11–13 years 18–20 years 25–27 years 20 20 0 0 –10 –10 –20 –20 –30 –30 UW control NIA J/A control NIA old control UW CR NIA J/A CR NIA old CR Figure 2 | Bodyweight data for monkeys at NIA and UW. (a) Bodyweight (kg) for male and female monkeys at UW and at NIA grouped by age where male J/A include juvenile and adolescent onset animals, female J/A include juvenile and adult onset animals, and old include the advanced age onset animals. Digits shown in white within the boxes are the numbers of individual animals contributing to each data point, data are shown as mean s.e. of the mean. (b) Comparison of bodyweight averages for monkeys from UW and NIA studies with records of the internet Primate Aging Database (iPAD). Average bodyweight for control and CR monkeys at both study locations were determined by age category including adult (11–13 years of age), late mid-age (18–20 years of age) and advanced age (25–27 years of age). Data are expressed as percent deviation from the iPAD average for females and males from each age category. Statistics related to this ﬁgure are provided in Supplementary Information, Supplementary Tables 2 and 3. NATURE COMMUNICATIONS | 8:14063 | DOI: 10.1038/ncomms14063 | www.nature.com/naturecommunications 5 Weight in kg Weight in kg % difference from iPAD average ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14063 Considering ﬁrst the female monkeys, bodyweight for the NIA average (B18% for males; B19% for females), and CR monkeys J/A was not signiﬁcantly different between control and CR had lower bodyweight than the iPAD average (B12% for males; monkeys for any age categories. UW CR females weighed B11% for females) (Fig. 2b). For NIA J/A, control males were the signiﬁcantly less (17–26%) than controls throughout the study same to slightly heavier (5–10%) than the iPAD average and CR period, and UW female controls weighed signiﬁcantly more than weighed less than the iPAD average (B20%), while control and NIA J/A female controls throughout (Supplementary Table 2). CR female monkeys both weighed less than the iPAD average For NIA old-onset females, bodyweight was not signiﬁcantly throughout the study (B10% and B20% respectively). All NIA different between controls and CR, and was signiﬁcantly lower old-onset monkeys weighed less than the iPAD average for both than bodyweight of UW female controls. In summary, for NIA control (B15% for females;B10% for males) and CR (B22% for J/A and old-onset female cohorts, bodyweight for control and CR females; B21% for males) monkeys. In summary, bodyweights of monkeys was not different from each other and all were UW and NIA control monkeys were not equivalent to each other, signiﬁcantly lower than the UW controls. Considering next the and apart from J/A males, were respectively higher and lower of male monkeys, NIA J/A CR males weighed signiﬁcantly less the iPAD average. (19–22%) than their control counterparts throughout the study. To gain insight into differences in the effect of age and diet on The difference between UW control and CR was slightly greater body composition, dual X-ray absorptiometry measures were (24–35%), with CR males weighing signiﬁcantly less than conducted at intervals throughout the course of the two studies controls. The average peak weight for NIA J/A control males (Fig. 3). Since each animal had multiple measures taken over was B15% lower than that of UW control males, but differences time, estimates of the average percent adiposity (fat/bodyweight in bodyweight were signiﬁcant for the young age category only expressed as percent) were adjusted for age (Supplementary (Supplementary Table 3). Bodyweight of the old-onset NIA Fig. 1). Within groups a main effect of age on adiposity was control and CR males were not signiﬁcantly different at either detected for NIA J/A and UW cohorts. A main effect of diet was mid-age or advanced ages, and old-onset NIA male controls detected for NIA J/A males and for both males and females from weighed signiﬁcantly less than UW controls. In summary, NIA the UW study, where CR was associated with signiﬁcantly lower J/A and UW male cohorts showed a clear bodyweight response to adiposity. The NIA J/A control and CR females did not differ CR, but old-onset NIA control and CR males were not different from each other in adiposity and neither of the NIA old-onset from each other and were signiﬁcantly lower than the UW monkey groups had a main effect of CR on adiposity. Combining controls. the data from NIA J/A and UW, a difference in adiposity was The internet Primate Aging Database (iPAD; http://ipad. detected between controls on the two studies for both males and primate.wisc.edu) is a repository of clinical and biometric data females, where NIA monkeys had signiﬁcantly lower percent from healthy, non-experimental, captive nonhuman primates body fat. Control monkeys from NIA J/A were not statistically housed at research facilities across the USA. Using data from over different from UW CR in percent body fat for both sexes. These 1,200 individual rhesus monkeys of Indian origin, mean body- data show an impact of age on adiposity in all three groups and weights were calculated for the above age categories for males reveal that the impact of CR on adiposity was observed for both (11.6, 12.1, 11.5 kg respectively) and females (7.4, 8.4, 7.8 kg groups of UW monkeys and at NIA for J/A males only. respectively). UW control and CR monkeys fell on either side of Food intake was monitored daily at both sites. At UW daily these averages; control monkeys were heavier than the iPAD measures of food intake were used to calculate means. At NIA 40 40 40 3 7 30 24 10 30 30 4 21 3 24 12 24 20 11 5 1917 3 4 2021 4 6 7 5 17 15 22 3 4 3 8 2 141510 21 2 20 20 11 20 15 17 16 21 15 5 3 16 3 12 23 17 5 21 18 12 4 23 22 2 2 3 7 14 17 6 23 23 2 13 2 13 11 10 11 10 10 2 13 11 12 4 9 5 6 6 5 UW females NIA J/A females NIA old females 10 15 20 25 10 15 20 25 30 35 20 25 30 Age in years Age in years Age in years 40 40 40 34 5 30 30 37 35 2921 10 4 30 NIA old males 21 34 15 39 39 4 5 30 10 18 14 12 6 17 13 20 6 19 20 23 4 20 10 20 13 25 6 7 17 35 31 17 13 23 5 18 32 6 36 33 12 8 24 36 11 16 5 6 12 2 4 14 19 37 15 5 5 5 13 4 32 5 13 36 3 6 4 10 13 10 36 10 7 3 15 3 6 12 13 8 2 3 3 3 NIA J/A males UW males 10 15 20 25 10 15 20 25 30 35 25 30 Age in years Age in years Age in years Control CR Figure 3 | Adiposity data for female and male monkeys at NIA and UW. Percent adiposity (fat (g)/total bodyweight (g)) calculated from DXA (dual energy X-ray absorptiometry) measures conducted during the course of the studies for male and female monkeys at UW and at NIA grouped by age where male J/A include juvenile and adolescent onset animals, female J/A include juvenile and adult onset animals, and old include the advanced age onset animals. Digits shown in white within the boxes are the numbers of individual animals contributing to each data point, data are shown as mean s.e. of the mean. 6 NATURE COMMUNICATIONS | 8:14063 | DOI: 10.1038/ncomms14063 | www.nature.com/naturecommunications Adiposity % Adiposity % NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14063 ARTICLE 750 750 750 NIA J/A females 26 24 25 22 22 NIA old females 650 650 17 650 40 16 22 21 19 17 14 35 13 22 18 550 19 550 550 32 22 17 20 22 21 19 21 22 22 4 13 15 22 9 4 29 17 34 10 14 14 11 23 22 18 9 19 4 450 40 12 15 450 450 6 7 38 6 16 9 7 9 7 25 26 7 20 6 25 6 5 UW females 14 21 8 350 350 10 15 20 15 20 25 20 25 30 Age in years Age in years Age in years 900 15 16 900 40 29 34 30 36 28 13 19 38 15 42 37 42 44 36 20 32 17 800 800 34 NIA old males 15 28 36 28 31 15 19 8 8 8 12 16 12 700 700 12 700 16 14 16 14 27 16 14 12 16 13 22 11 13 31 32 33 13 28 29 7 36 10 14 15 13 31 26 15 11 13 600 20 600 13 16 25 600 500 500 500 13 11 5 15 9 16 NIA J/A males UW males 400 400 400 10 15 20 25 10 15 20 25 25 30 Age in years Age in years Age in years Control CR Figure 4 | Food intake data for monkeys at NIA and UW. Food intake (daily values in Kcalories) for male and female monkeys at UW and at NIA grouped by age where male J/A include juvenile and adolescent onset animals, female J/A include juvenile and adult onset animals, and old include the advanced age onset animals. Digits shown in white within the boxes are the numbers of individual animals contributing to each data point, data are shown as mean s.e. of the mean. food intake means were calculated based on measures conducted an intervention that imparts improved health even in the absence during a single week per year as representative of typical intake. of increased longevity, is viewed as a highly favourable and Longitudinal data for all monkeys were averaged by age of the legitimate example of an ageing intervention. With advancing animal (Fig. 4). Data from the last year of life of each monkey age, rhesus monkeys are vulnerable to many of the same were excluded to avoid confounding effects of end-of-life feeding conditions observed in humans. Among the most prevalent are behaviours that usually include loss of appetite. Considering ﬁrst cancer, cardiac disease, and conditions related to immune the females and using the age categories deﬁned above for both dysfunction and inﬂammation, and examples of each were UW and NIA J/A, the controls consumed signiﬁcantly more identiﬁed in monkeys on the ageing and CR studies at both calories than CR at both young and mid-age, but the difference NIA and UW (Supplementary Table 4). persisted only for UW female monkeys at advanced age. For the Fasting glucose measures were common to both studies and the old-onset NIA, caloric intake was not different between control longitudinal data are shown (Fig. 5). In healthy adult rhesus and CR. Among control monkeys, UW females consumed monkeys fasting glucose levels are 64–68 mg dl (refs 18,35). signiﬁcantly more calories than NIA J/A at mid-age and advanced For NIA J/A, fasting glucose levels were equivalent for controls age and more than old-onset at advanced age. Considering next and CR up to B23 years of age, after which the control and CR the males, the NIA J/A controls consumed signiﬁcantly more males, but not females, began to diverge. Both control and CR calories than CR at young and mid-age and the difference females showed an age-related increase in fasting glucose levels between control and CR was signiﬁcant for UW at mid-age only. after B21 years of age. For UW monkeys, the control males had Old-onset males at NIA differed signiﬁcantly in their caloric higher fasting glucose levels than CR from 15 years of age with a intake between control and CR only at advanced age. Among further divergence of the curves after B23 years of age, while a controls, caloric intake was not different for NIA J/A and UW noticeable difference between control and CR females emerged males at any point in the study, but old-onset males consumed after only B21 years. For the NIA old-onset cohorts, fasting signiﬁcantly less than UW males and NIA J/A males at mid-age. glucose was consistently low for the duration of the study period. In summary, signiﬁcant differences in caloric intake were These data point to an age-related increase in fasting glucose identiﬁed between control and CR monkeys for male and female for rhesus monkeys and single out the UW control males as NIA J/A and UW cohorts, but not for old-onset cohorts until being predisposed to elevated circulating glucose in the fasted advanced age and then for males only. Comparing between sites, state. Using multilevel modelling to investigate the relationship caloric intake for NIA female controls of both J/A and old-onset between adiposity and fasting glucose levels a signiﬁcant was lower than that of UW controls, and for males, caloric intake relationship was identiﬁed for UW males only (P¼ 0.005). of NIA J/A and UW controls were not different from each other A signiﬁcant age by diet interaction was also detected (P¼ 0.014), but old-onset NIA controls were lower than both. suggesting that the impact of age on the relationship between adiposity and glucoregulatory parameters is distinct for control and CR monkeys. Impact of CR on incidence of disease. The concept of health- Veterinarians documented body condition and overall health span is a fairly recent development in ageing research, where a of monkeys biannually at both study locations and indicators of distinction is drawn between chronological age and health diseases or disorders identiﬁed. The age at which a monkey was ﬁrst diagnosed with an age-related condition was used status . Traditionally, an increase in both median and maximum lifespan was considered the hallmark of delayed ageing, and to generate morbidity curves (Fig. 6). Age-related conditions improvements in health were deemed to be a necessary and included sarcopenia, osteoporosis, arthritis, diverticulosis, obvious component of longevity. The perspective has shifted cataracts and persistent heart murmurs, in addition to age- somewhat towards greater emphasis on health and morbidity, so related diseases including cancer, diabetes and cardiovascular NATURE COMMUNICATIONS | 8:14063 | DOI: 10.1038/ncomms14063 | www.nature.com/naturecommunications 7 Daily intake in Kcal Daily intake in Kcal ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14063 130 130 110 110 29 21 14 NIA J/A females NIA J/A males 90 90 19 36 26 13 31 27 34 32 37 30 38 24 36 39 20 38 70 31 39 25 70 37 44 39 29 22 50 37 41 44 39 24 48 41 46 32 59 53 36 45 41 30 60 58 41 35 42 27 24 30 38 32 50 47 50 34 32 42 52 42 44 52 42 50 50 10 15 20 25 30 10 15 20 25 30 Age in years Age in years 130 130 110 110 UW males UW females 40 33 42 42 19 14 36 36 25 36 21 14 17 20 22 33 24 17 16 22 24 38 24 22 17 70 32 34 70 22 19 15 25 22 23 18 12 18 26 21 21 25 40 41 39 14 22 39 39 22 22 17 22 14 10 26 50 50 10 15 20 25 30 10 15 20 25 30 Age in years Age in years 130 130 110 110 90 90 NIA old females NIA old males 13 6 70 14 70 14 18 11 15 10 25 24 26 21 10 9 17 17 17 10 18 13 24 7 21 19 21 7 7 17 16 9 25 22 15 12 11 18 10 6 7 50 50 10 15 20 25 30 10 15 20 25 30 Age in years Age in years Control CR Figure 5 | Fasting glucose values for monkeys at NIA and UW. Circulating levels of glucose (mg dl ) are shown for male and female monkeys at UW and at NIA grouped by age where male J/A include juvenile and adolescent onset animals, female J/A include juvenile and adult onset animals, and old include the advanced age onset animals. Digits shown in white within the boxes are the numbers of observations contributing to each data point, data are shown as mean s.e. of the mean. disease. Cox proportional hazard regression modelling indicated clinical care at both locations, although clinical deﬁnitions that age-related conditions occurred at B2.7 times the rate in representing different disease stages were employed at the two 38,39 control animals compared to CR for UW monkeys (HR: 2.665; sites. Similar to humans , insulin resistance occurs in advance 35,40 CI: 1.527–4.653; P¼ 0.0006). In the NIA J/A cohort, age-related of impaired fasting glucose in rhesus monkeys , which occurs conditions occurred at twice the rate in control monkeys before transition to full diabetes. At UW loss of insulin sensitivity compared to CR (HR: 2.091; CI: 1.169–3.641; P¼ 0.0125) was used to diagnose glucoregulatory impairment and was (Supplementary Table 5; Supplementary Fig. 2). The advanced deﬁned as fasting insulin levels (470mUml ) and an insulin age of the old-onset NIA monkeys precluded detection of the ﬁrst sensitivity index (S )o2(E ) as determined by a frequently occurrence of an age-related condition. sampled intravenous glucose tolerance test. At NIA, fasting The incidence of age-related conditions prevalent in glucose (4100 mg dl ), glucosuria and HbA1c (46.5%) human populations, such as cancer, cardiovascular disease and measures were used to deﬁne diabetes. CR animals had lower glucoregulatory dysfunction/diabetes, was determined for incidence of glucoregulatory dysfunction than controls at both monkeys at UW and NIA, with J/A and old-onset combined. UW and NIA study sites. Multilevel modelling was used to Diagnoses were made clinically by veterinary staff upon investigate possible relationships between adiposity and presentation, and disease-related pathology was subsequently morbidity but no signiﬁcant effects were detected for any group conﬁrmed upon necropsy by a board-certiﬁed pathologist. from either study. Similarly, for all NIA groups and for UW Clinically silent pathologies were identiﬁed at necropsy. To males, no relationship between adiposity and mortality was compare studies, cancer and cardiovascular disorders are reported detected. For UW females, adiposity was associated with a modest as incidence upon necropsy. Adenocarcinoma was the leading reduction in risk for death (HR: 0.927; P¼ 0.01) but only after cause of death at both study sites, consistent with previous reports correcting for age and diet. These data suggest that the inﬂuence 36,37 on cancer incidence in rhesus monkeys . The incidence of of adiposity on survival risk is sexually dimorphic and changes adenocarcinoma or other less common neoplasms was lower in with age. CR monkeys at both locations (Fig. 6). The most common diagnosis of cardiovascular disease in live animals was mitral valve dysfunction, while valvular endocardiosis, cardiomyopathy Discussion and myocardial ﬁbrosis were detected at necropsy. Incidence of Data from both study locations suggest that the CR paradigm is cardiovascular disorders was lower in CR monkeys than controls effective in delaying the effects of ageing in nonhuman primates at UW. There was no apparent impact of diet on incidence of but that the age of onset is an important factor in determining the cardiovascular disease for monkeys at NIA; however, incidence extent to which beneﬁcial effects of CR might be induced. In the for both control and CR monkeys was lower than UW controls. UW study, reduced bodyweight, reduced adiposity and reduced Assessment of glucoregulatory function was part of routine food intake of the CR monkeys were associated with improved 8 NATURE COMMUNICATIONS | 8:14063 | DOI: 10.1038/ncomms14063 | www.nature.com/naturecommunications –1 –1 –1 Glucose (mg dl ) Glucose (mg dl ) Glucose (mg dl ) NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14063 ARTICLE survival, with CR monkeys of both sexes surviving longer than Control controls, B28 and B30 years of age for males and females CR respectively, and longer than the median age for monkeys in captivity (B26 years of age). Although an impact of CR on Average survival was not detected within the NIA old-onset cohort, onset comparison to the UW study shows that bodyweight was signiﬁcantly lower in both control and CR groups of males and females than in their UW control counterparts, and was largely equivalent to that of UW CR. All males and females from the NIA UW old-onset groups consumed fewer calories than their counterpart controls from UW, instead both control and CR were closely aligned with food intake values of UW CR. Importantly, the 10 20 30 median survival estimates for old-onset males were very high, Age in years similar to what has been reported previously as the 90th 100 percentile for this species (B35 years of age). Six of the original 20 monkeys have lived beyond 40 years of age, the previous Control 80 maximal lifespan recorded, and one old-onset CR male monkey is CR currently 43 years old, which is a longevity record for this species. Average 60 onset Median survival estimates for old-onset females, B27 and B28 for controls and CR respectively, were also greater than national median lifespan estimates, with one remaining female currently 38 years of age. The clear beneﬁt in survival estimates for 20 NIA monkeys within the old-onset cohort compared to UW controls suggests that food intake can and does inﬂuence survival. The lack of difference between control and CR old-onset monkeys 10 20 30 suggests that a reduction in food intake beyond that of the Age in years controls brings no further advantage. The minimum degree of restriction that confers maximal beneﬁt in rhesus monkeys has b Cancer Ins. resistance not yet been identiﬁed but is an active topic of investigation. 50 50 60 Taken together, data from both UW and NIA studies support the concept that lower food intake in adult or advanced age is 40 40 associated with improved survival in nonhuman primates. 30 The interpretation of the outcomes from NIA J/A cohorts is more complicated. The J/A cohorts reveal a sex-dependence in 20 20 the relationship between bodyweight and food intake that was not observed in the UW cohorts. The NIA J/A CR males consumed 10 10 signiﬁcantly fewer calories than controls for most of the study, and this was reﬂected in a signiﬁcantly lower bodyweight and a 0 0 0 modest reduction in adiposity. Bodyweight of the NIA J/A UW NIA UW control males was in between that of UW controls and CR Cardiovascular disorder Diabetes monkeys even though caloric intake was equivalent between UW 50 15 and NIA control animals, and age-adjusted average adiposity of NIA J/A males was equivalent to that of UW CR males. Median survival estimates for the NIA J/A male controls and CR were not statistically different at B29 and B26 years, respectively, but estimates for the J/A controls were numerically equivalent to UW CR. These data suggest that in the J/A male cohort, the NIA diet 20 20 5 is associated with lower bodyweight, lower adiposity and improved survival in the absence of CR, and that no further 10 10 advantage is gained by lowering food intake or bodyweight below that of controls. A different picture emerges from the data from 0 0 0 UW NIA NIA NIA J/A females. The CR monkeys consumed signiﬁcantly less Control food than controls for much of the study, but there was no CR apparent difference in bodyweight or adiposity between them. Figure 6 | Morbidity curves for monkeys at NIA and UW shown. Median survival estimates for control and CR J/A females were (a) Graphs represent the ﬁrst occurrence of any age-related disease, not statistically different from each other at B26 and B23 years disorder or condition for combined males and females from UW (top) respectively, and each were lower than that observed for UW CR. and NIA J/A (bottom). Statistics related to this ﬁgure are provided in The lack of response in bodyweight to differences in food intake Supplementary Information, Supplementary Table 4. (b) Incidence of in the female monkeys suggests that there is sexual dimorphism prevalent age-related conditions in nonhuman primates for control and in the relationship between food intake and bodyweight in young CR animals from UW and NIA (J/A and old-onset combined). To compare animals and in how these parameters relate to median survival. studies, cancer and cardiovascular disorders are reported as incidence upon Finally, there is a suggestion that nonhuman primates differ from necropsy and are expressed as a percentage of the animals that are rodents in the suitability of early onset CR. In rodents, early onset deceased. CR is more effective in extending longevity than adult onset CR. For nonhuman primates it appears that CR, while beneﬁcial when implemented in adulthood, does not improve survival when NATURE COMMUNICATIONS | 8:14063 | DOI: 10.1038/ncomms14063 | www.nature.com/naturecommunications 9 Percent incidence Percent incidence Percent without Percent without age-related morbidities age-related morbidities Percent incidence Percent incidence ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14063 implemented in juveniles. Furthermore, it appears that CR is not including cancer, cardiovascular disease and parameters uniformly tolerated in young animals, where the NIA J/A CR associated with diabetes. A lower incidence of cancer was one males and females reached 80% mortality earlier than controls. of the ﬁrst health beneﬁts of CR documented and is considered to Although the numerical values for estimated survival were lower be a hallmark of CR in rodents . The incidence of cancer was in CR groups than in controls, the impact of juvenile onset CR on lower in CR monkeys at both locations indicating that tumour survival remains undetermined, as the mortality curves for both suppression is a conserved feature of mammalian CR. CR also NIA J/A cohorts are incomplete. lowered the incidence of cardiovascular disorders at UW, and Data from both NIA and UW studies highlight sex-dependent NIA monkeys from either diet group appear to have been differences in nutritional modulation of body composition and protected compared to UW Control monkeys. The data shown glucoregulatory function, and prompt further investigation of are based on pathologies identiﬁed at necropsy so that ﬁnal sex-speciﬁcity in the connection between nutrition and disease incidence of both disorders at NIA has yet to be determined. The risk. In rodent studies sexual dimorphism in the response to impact of CR to sustain glucoregulatory function during ageing diet-induced metabolic dysfunction has been reported where was a further commonality between the studies. Monkeys males were more vulnerable than females to increased adiposity presenting with insulin resistance at UW were clinically treated and glucoregulatory dysfunction . It is perhaps not surprising to prevent transition to diabetes, prohibiting a direct comparison then that the separation in fasting glucose curves between UW between the studies. Although the diagnosis employed at the controls and CR was observed in males but not females, and that two study locations represent different time frames in disease females, despite being on the same diet as males, were not progression, the outcome of lower incidence of insulin resistance different in glucose levels between control and CR until after at UW and lower incidence of diabetes at NIA are very much 20 years of age. Although both NIA and UW diets contain in agreement. Given the obvious parallels between human equivalent relative amounts of carbohydrate, the sucrose content and rhesus monkey, it seems highly likely that the beneﬁcial of the puriﬁed UW diet is six times higher than the naturally effects of CR would also be observed in humans. Reports from sourced NIA diet, and the fat content of the UW diet is twice that the multicenter CALERIE study of short-term CR in of the NIA diet. The implication is that elevated sucrose and humans document changes in bodyweight, body composition, increased fat content could lead to increased adiposity, negatively glucoregulatory function and serum risk factors for impacting glucoregulatory function. All the NIA J/A monkeys cardiovascular disease in response to CR (refs 42–47). These maintained adiposity at less than 25% and were in large part outcomes in humans align well with reports on rhesus monkey euglycemic through middle age. As might be predicted, the UW CR (refs 48–51), conﬁrming that the primary response to CR is control males and females had signiﬁcantly greater adiposity conserved between these two species, and suggesting that the compared to NIA controls. In the case of UW males, increased underlying mechanisms may also be conserved. adiposity was signiﬁcantly associated with higher fasting glucose In conclusion, the NIA and UW nonhuman primate ageing levels; however, the association did not hold for the UW female and CR studies address a central concept of relevance to human monkeys where despite differences in adiposity between control ageing and human health: that the age-related increase in disease and CR, fasting glucose levels were equivalent through middle vulnerability in primates is malleable and that ageing itself age, and multilevel analysis failed to identify a signiﬁcant presents a reasonable target for intervention. The last two decades association between adiposity and fasting glucose. Furthermore, have seen considerable advances in ageing research in short-lived there is the suggestion that the effect of diet composition is species and investigations of the mechanisms of CR have been dependent on intake levels. The UW CR male monkeys, on the prominent in this work. It will be particularly informative to same diet as controls but under food limiting conditions, had low determine the degree to which consensus hallmarks of ageing 52,53 adiposity and normal fasting glucose. These data demonstrate described in recent publications also manifest in primate that the relationship between adiposity and glucoregulatory ageing. The tissues and longitudinal data stored over the course of impairment is sex-speciﬁc, and the relationship between dietary these two highly controlled monkey studies present a unique composition and glucoregulatory impairment is dependent on the resource that can be used to identify key pathways responsive level of food intake. The signiﬁcant relationship between to CR in primates, to uncover primate-speciﬁc aspects of adiposity and survival detected in the UW female cohort also the basic biology of ageing, and to determine molecular basis points to sex-dimorphism in the impact of age and the inﬂuence for nutritional modulation of health and ageing. Processes of body composition on mortality risk. This relationship was impacted by CR would be prime targets for the development of identiﬁed for UW females only, presumably because there was a clinical interventions to offset age-related morbidity, and greater range in levels of adiposity among controls and CR for the identiﬁcation of factors involved in the mechanisms of CR will UW females than for NIA females where adiposity of monkeys be pivotal in bringing these ideas to clinical research and human from either diet at either age of onset did not differ. The positive health care. relationship between adiposity and survival in UW females was identiﬁed only after correcting for age and diet, suggesting that Methods the interactions among health, body composition and morbidity Animal care. Speciﬁcs of housing and animal care have been described in detail 20,21 elsewhere . Brieﬂy, all animals are maintained according to the provisions are sensitive to age and, given the lack of association in the male of the Animal Welfare Act of 1966 (Public Law 89–544), plus its subsequent cohort, sex-dependent. amendments, as well as the standards set forth in the document entitled ‘The Guide The catalogue of pathologies identiﬁed in aged monkeys is for the Care and Use of Laboratory Animals’ (NIH Publication No. 85–23). shared with aged humans. The deﬁnitions used to identify Animals rooms are maintained atB21 C and 50–65% humidity. Room lighting is morbidity were determined by veterinary staff and were automatically controlled on a 12-h light, 12-h dark schedule. Animals are monitored daily for general health by dedicated animal care and research staff and essentially equivalent at both sites. A shared feature of both routinely by veterinary staff. Studies at UW and NIA were conducted in accordance studies is the beneﬁcial effect of CR in lowering the risk for age- with protocols approved respectively by the University of Wisconsin–Madison related morbidity by more than two-fold. Factors contributing to Graduate School and National Institute on Aging Institutional Animal Care and these analyses include a range of conditions that are highly Use Committees. prevalent in human geriatric populations, such as sarcopenia, osteoporosis and arthritis. The beneﬁcial effects extended to Statistical analysis. The mortality analysis included all animals with known diseases that are among the most prevalent in human clinical care diagnoses or cause of death prior to July 2015. For time-to-event outcomes, the 10 NATURE COMMUNICATIONS | 8:14063 | DOI: 10.1038/ncomms14063 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14063 ARTICLE three most common methods were used: Kaplan-Meier product-limit method; 21. Mattison, J. A. et al. Impact of caloric restriction on health and Cox proportional hazard regression and parametric survival analysis assuming survival in rhesus monkeys from the NIA study. Nature 489, 318–321 Weibull distributions . Analyses were separated by site (UW; NIA), age-of-onset (2012). and sex. For both Kaplan-Meier and Cox proportional hazards (PH) regression, 22. Weindruch, R. & Walford, R. L. Dietary restriction in mice beginning at 1 year diet group (CR or control) was used as a predictor to test the effect of CR on of age: effect on life-span and spontaneous cancer incidence. Science 215, mortality and morbidity (the onset of age-associated disease). Cox models were 1415–1418 (1982). also used to estimate the HRs and 95% CIs. For the mortality analyses, age at death 23. National Research Council, Committee on Animal Nutrition, Agricultural or current age for censored values was used as the time variable. Estimated median Board. Nutrient requirements of nonhuman primates (National Academy of and mean lifespans were calculated using both Kaplan-Meier analysis and Sciences, 1978). parametric survival analysis assuming a Weibull distribution. Bodyweight and food 24. Lane, M. A. et al. Dietary restriction in nonhuman primates: progress report on intake data were analysed using mixed linear models to account for dependency in the NIA study. Ann. NY Acad. Sci. 673, 36–45 (1992). the observations due to repeatedly measuring each animal over time. In the 25. Kemnitz, J. W. et al. Dietary restriction of adult male rhesus monkeys: design, morbidity analyses, the age at which the animal experienced its ﬁrst age-associated methodology, and preliminary ﬁndings from the ﬁrst year of study. J. Gerontol. diagnosis was used as the time variable. Animals that had had not experienced an 48, B17–B26 (1993). age-associated disease but died of an age-related cause were not censored and their 26. Clarke, M. R. & O’Neil, J. A. Morphometric comparison of Chinese-origin age at death was used as the time variable. For the censored cases where the and Indian-derived rhesus monkeys (Macaca mulatta). Am. J. Primatol. 47, monkeys were alive and had not experienced an age-associated disease, current age 335–346 (1999). was used as the time variable. For censored cases where monkeys had died but had 27. Kanthaswamy, S. et al. Development and validation of a SNP-based assay for not experienced an age-associated disease, age at death was used as the time variable. For all Cox models, the PH assumption was tested by ﬁrst ﬁtting a inferring the genetic ancestry of rhesus macaques (Macaca mulatta). Am. J. Primatol. 76, 1105–1113 (2014). non-PH Cox regression with a CR by time interaction. The CR by time interactions were not signiﬁcant in any model; thus, PH models were considered valid. All 28. Rhesus Macaque Genome, S. et al. Evolutionary and biomedical insights from analyses were performed in SAS 9.3. The statistical signiﬁcance level was set at the rhesus macaque genome. Science 316, 222–234 (2007). a¼ 0.05 for all analyses with no adjustments . Multilevel modelling was 29. Colman, R. J. & Anderson, R. M. Nonhuman primate calorie restriction. conducted to explore relationships among parameters. Antioxid. Redox Signal. 14, 229–239 (2011). 30. Allison, P. D. Survival Analysis Using SAS: A Practical Guide 2nd edn (SAS Institute, 2010). Data availability. The data sets analysed during the current study are available 31. Hadﬁeld, R. M. et al. Risk factors for endometriosis in the rhesus monkey from the corresponding authors on reasonable request. (Macaca mulatta): a case-control study. Hum. Reprod. Update 3, 109–115 (1997). 32. Mattison, J. A., Ottinger, M. A., Powell, D., Longo, D. L. & Ingram, D. K. References Endometriosis: clinical monitoring and treatment procedures in Rhesus 1. Gibbs, R. A. et al. Evolutionary and biomedical insights from the rhesus Monkeys. J. Med. Primatol. 36, 391–398 (2007). macaque genome. Science 316, 222–234 (2007). 33. Kemnitz, J. W. Calorie restriction and aging in nonhuman primates. ILAR j. 52, 2. Zimin, A. V. et al. A new rhesus macaque assembly and annotation for 66–77 (2011). next-generation sequencing analyses. Biol. Direct 9, 20 (2014). 34. Kirkland, J. L. & Peterson, C. Healthspan, translation, and new outcomes for 3. Bowden, D. M. & Williams, D. D. Aging. Adv. Vet. Sci. Comp. Med. 28, 305– animal studies of aging. J. Gerontol. Ser. A, Biol. Sci. Med. Sci. 64, 209–212 341 (1984). (2009). 4. Uno, H. Age-related pathology and biosenescent markers in captive rhesus 35. Hansen, B. C. & Bodkin, N. L. Heterogeneity of insulin responses: phases macaques. Age (Omaha) 20, 1–13 (1997). leading to type 2 (non-insulin-dependent) diabetes mellitus in the rhesus 5. Hudson, J. C., Baum, S. T., Frye, D. M., Roecker, E. B. & Kemnitz, J. W. monkey. Diabetologia 29, 713–719 (1986). Age and sex differences in body size and composition during rhesus monkey 36. Rodriguez, N. A. et al. Clinical and histopathological evaluation of 13 cases of adulthood. Aging 8, 197–204 (1996). adenocarcinoma in aged rhesus macaques (Macaca mulatta). J. Med. Primatol. 6. Pandya, J. D. et al. Decreased mitochondrial bioenergetics and calcium 31, 74–83 (2002). buffering capacity in the basal ganglia correlates with motor deﬁcits in a 37. Uno, H. et al. Colon cancer in aged captive rhesus monkeys (Macaca mulatta). nonhuman primate model of aging. Neurobiol. Aging 36, 1903–1913 (2015). Am. J. Primatol. 44, 19–27 (1998). 7. Ramsey, J. J., Laatsch, J. L. & Kemnitz, J. W. Age and gender differences in body 38. Laakso, M. & Kuusisto, J. Insulin resistance and hyperglycaemia in composition, energy expenditure, and glucoregulation of adult rhesus monkeys. cardiovascular disease development. Nat. Rev. Endocrinol. 10, 293–302 J. Med. Primatol. 29, 11–19 (2000). (2014). 8. Ngwenya, L. B., Heyworth, N. C., Shwe, Y., Moore, T. L. & Rosene, D. L. 39. Dailey, G. New strategies for basal insulin treatment in type 2 diabetes mellitus. Age-related changes in dentate gyrus cell numbers, neurogenesis, and Clin. Therapeut. 26, 889–901 (2004). associations with cognitive impairments in the rhesus monkey. Front. Syst. 40. Polewski, M. A. et al. Plasma diacylglycerol composition is a biomarker of Neurosci. 9, 102 (2015). metabolic syndrome onset in rhesus monkeys. J. Lipid Res. 56, 1461–1470 9. Masoro, E. J., Yu, B. P. & Bertrand, H. A. Action of food restriction in delaying (2015). the aging process. Proc. Natl Acad. Sci. USA 79, 4239–4241 (1982). 41. Grove, K. L., Fried, S. K., Greenberg, A. S., Xiao, X. Q. & Clegg, D. J. 10. Weindruch, R., Walford, R. L., Fligiel, S. & Guthrie, D. The retardation of aging A microarray analysis of sexual dimorphism of adipose tissues in in mice by dietary restriction: longevity, cancer, immunity and lifetime energy high-fat-diet-induced obese mice. Int. J. Obes. (Lond) 34, 989–1000 (2010). intake. J. Nutr. 116, 641–654 (1986). 42. Fontana, L. et al. Calorie restriction or exercise: effects on coronary heart 11. Weindruch, R. H. & Walford, R. L. The retardation of aging and disease by disease risk factors. A randomized, controlled trial. Am. J. Physiol. Endocrinol. dietary restriction (Charles C Thomas, 1988). Metab. 293, E197–E202 (2007). 12. Masoro, E. J. Caloric restriction. Aging 10, 173–174 (1998). 43. Heilbronn, L. K. et al. Effect of 6-month calorie restriction on biomarkers of 13. Gems, D. & Partridge, L. Genetics of longevity in model organisms: debates and longevity, metabolic adaptation, and oxidative stress in overweight individuals: paradigm shifts. Annu. rev. physiol. 75, 621–644 (2013). a randomized controlled trial. JAMA 295, 1539–1548 (2006). 14. Guarente, L. Mitochondria—a nexus for aging, calorie restriction, and sirtuins? 44. Lefevre, M. et al. Caloric restriction alone and with exercise improves CVD risk Cell 132, 171–176 (2008). in healthy non-obese individuals. Atherosclerosis 203, 206–213 (2009). 15. Houthoofd, K. & Vanﬂeteren, J. R. Public and private mechanisms of life 45. Racette, S. B. et al. One year of caloric restriction in humans: feasibility and extension in Caenorhabditis elegans. Mol. Genet. Genomics 277, 601–617 (2007). effects on body composition and abdominal adipose tissue. J. Gerontol. Ser. A, Biol. Sci. Med. Sci. 61, 943–950 (2006). 16. Kennedy, B. K., Steffen, K. K. & Kaeberlein, M. Ruminations on dietary 46. Redman, L. M. et al. Effect of calorie restriction with or without exercise on restriction and aging. Cell. Mol. Life Sci. 64, 1323–1328 (2007). body composition and fat distribution. J. Clin. Endocrinol. Metab. 92, 865–872 17. Mair, W. & Dillin, A. Aging and survival: the genetics of life span extension by dietary restriction. Annu. Rev. Biochem. 77, 727–754 (2008). (2007). 18. Bodkin, N. L., Alexander, T. M., Ortmeyer, H. K., Johnson, E. & Hansen, B. C. 47. Weiss, E. P. et al. Improvements in glucose tolerance and insulin action induced by increasing energy expenditure or decreasing energy intake: Mortality and morbidity in laboratory-maintained Rhesus monkeys and effects a randomized controlled trial. Am. J. Clin. Nutr. 84, 1033–1042 (2006). of long-term dietary restriction. J. Gerontol. Ser. A, Biol. Sci. Med. Sci. 58, 212–219 (2003). 48. Edwards, I. J. et al. Caloric restriction lowers plasma lipoprotein (a) in male but 19. Colman, R. J. et al. Caloric restriction delays disease onset and mortality in not female rhesus monkeys. Exp. Gerontol. 36, 1413–1418 (2001). rhesus monkeys. Science 325, 201–204 (2009). 49. Gresl, T. A. et al. Dietary restriction and glucose regulation in aging rhesus 20. Colman, R. J. et al. Caloric restriction reduces age-related and all-cause monkeys: a follow-up report at 8.5 yr. Am. J. Physiol. Endocrinol. Metab. 281, mortality in rhesus monkeys. Nat. Commun. 5, 3557 (2014). E757–E765 (2001). NATURE COMMUNICATIONS | 8:14063 | DOI: 10.1038/ncomms14063 | www.nature.com/naturecommunications 11 ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14063 Additional information 50. Rezzi, S. et al. Metabolic shifts due to long-term caloric restriction revealed in Supplementary Information accompanies this paper at http://www.nature.com/ nonhuman primates. Exp. Gerontol. 44, 356–362 (2009). naturecommunications 51. Yamada, Y. et al. Long-term calorie restriction decreases metabolic cost of movement and prevents decrease of physical activity during aging in rhesus Competing ﬁnancial interests: R.W. is a member of the board of LifeGen Technologies, monkeys. Exp. Gerontol. 48, 1226–1235 (2013). a company focused on nutritional genomics. G.S.R. is Chief Executive Ofﬁcer of 52. Kennedy, B. K. et al. Geroscience: linking aging to chronic disease. Cell 159, GeroScience, Inc. and Vice President of Prolongevity Technologies. D.K.I. serves as Chief 709–713 (2014). Scientiﬁc Ofﬁcer for GeroScience, Inc., and Prolongevity Technologies, Inc. D.B.A. serves 53. Lopez-Otin, C., Blasco, M. A., Partridge, L., Serrano, M. & Kroemer, G. The on the board of IKEA and has received consulting fees from multiple government, not for hallmarks of aging. Cell 153, 1194–1217 (2013). proﬁt, and for proﬁt organizations with interests in obesity and nutrition. None of these 54. Saville, D. J. Multiple comparison procedures—the practical solution. Am. Stat. activities beneﬁt directly from this research and no competing ﬁnancial interests are 44, 174–180 (1990). declared. The remaining authors declare no competing ﬁnancial interests. Reprints and permission information is available online at http://npg.nature.com/ Acknowledgements Thanks to animal care and research staff at both facilities including Scott Baum, Julie reprintsandpermissions/ Adriansjach, Saverio Capuano, Casey Fitz, Heather Simmons at UW, and Rick Herbert, How to cite this article: Mattison, J. A. et al. Caloric restriction improves health Edward Tilmont, Kelli Vaughan, Mark Szarowicz, Amanda Evans, Danielle Sedlak, Mark and survival of rhesus monkeys. Nat. Commun. 8, 14063 doi: 10.1038/ncomms14063 Bryant and Matt Starost at NIA for their tireless efforts and stalwart commitment over (2017). the years. Funding for the UW study was provided by NIH grants P01AG011915, R01AG040178, R01AG037000, and the Department of Medicine, School of Medicine and Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in Public Health UW Madison. This publication was made possible in part by NCRR/ORIP published maps and institutional afﬁliations. grants P51RR000167/P51OD011106 to the Wisconsin National Primate Research Center, University of Wisconsin-Madison and the use of resources and facilities at the William S. Middleton Memorial Veterans Hospital, Madison, WI. The internet Primate This work is licensed under a Creative Commons Attribution 4.0 Aging Database is funded by contract HHSN-263-2013-00026C from the NIA. The NIA International License. The images or other third party material in this study was supported in part by the Intramural Research Program of the NIH, National article are included in the article’s Creative Commons license, unless indicated otherwise Institute on Aging. in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. Author contributions To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ J.A.M., R.J.C., J.W.K., G.S.R., D.K.I., R.W., R.d.C. and R.M.A. devised the approach, long-term statistical support was provided by D.B.A. and T.M.B., T.M.B. conducted the statistical analysis, J.A.M., R.J.C., R.d.C. and R.M.A. wrote the paper. r The Author(s) 2017 12 NATURE COMMUNICATIONS | 8:14063 | DOI: 10.1038/ncomms14063 | www.nature.com/naturecommunications http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Nature Communications Springer Journals http://www.deepdyve.com/lp/springer-journals/caloric-restriction-improves-health-and-survival-of-rhesus-monkeys-P2yYst85cS

Loading next page...

References (64)

M. Clarke, J. O’Neil (1999)
Morphometric comparison of Chinese‐origin and Indian‐derived rhesus monkeys (Macaca mulatta)
American Journal of Primatology, 47
S. Kanthaswamy, Zachary Johnson, J. Trask, D. Smith, R. Ramakrishnan, Jason Bahk, J. Ng, R. Wiseman, H. Kubisch, E. Vallender, J. Rogers, B. Ferguson (2014)
Development and validation of a SNP‐based assay for inferring the genetic ancestry of rhesus macaques (Macaca mulatta)
American Journal of Primatology, 76
B. Kennedy, Kristan Steffen, M. Kaeberlein (2007)
Ruminations on dietary restriction and aging
Cellular and Molecular Life Sciences, 64
J. Mattison, M. Ottinger, D. Powell, D. Longo, D. Ingram (2007)
Endometriosis: clinical monitoring and treatment procedures in Rhesus Monkeys
Journal of Medical Primatology, 36
A. Zimin, Adam Cornish, Mnirnal Maudhoo, Robert Gibbs, Xiongfei Zhang, Sanjit Pandey, D. Meehan, K. Wipfler, S. Bosinger, Z. Johnson, G. Tharp, G. Marçais, Michael Roberts, B. Ferguson, H. Fox, T. Treangen, S. Salzberg, J. Yorke, R. Norgren (2014)
A new rhesus macaque assembly and annotation for next-generation sequencing analyses
Biology Direct, 9
J. Speakman, S. Mitchell (2011)
Caloric restriction.
Molecular aspects of medicine, 32 3
L. Heilbronn, L. Jonge, M. Frisard, J. DeLany, D. Larson-Meyer, J. Rood, T. Nguyen, Corby Martin, J. Volaufova, Marlene Most, F. Greenway, Steven Smith, Walter Deutsch, D. Williamson, E. Ravussin, for Team (2006)
Effect of 6-month calorie restriction on biomarkers of longevity, metabolic adaptation, and oxidative stress in overweight individuals: a randomized controlled trial.
JAMA, 295 13
Yosuke Yamada, R. Colman, J. Kemnitz, S. Baum, Rozalyn Anderson, R. Weindruch, D. Schoeller (2013)
Long-term calorie restriction decreases metabolic cost of movement and prevents decrease of physical activity during aging in rhesus monkeys
Experimental Gerontology, 48
N. Kurysheva, V. Shlapak, Tamara Ryabova, PhDb (2018)
A case–control study
J. Kirkland, C. Peterson (2009)
Healthspan, translation, and new outcomes for animal studies of aging.
The journals of gerontology. Series A, Biological sciences and medical sciences, 64 2
R. Weindruch, R. Walford (1982)
Dietary restriction in mice beginning at 1 year of age: effect on life-span and spontaneous cancer incidence.
Science, 215 4538
D. Gems, L. Partridge (2013)
Genetics of longevity in model organisms: debates and paradigm shifts.
Annual review of physiology, 75
R. Weindruch, R. Walford, S. Fligiel, D. Guthrie (1986)
The retardation of aging in mice by dietary restriction: longevity, cancer, immunity and lifetime energy intake.
The Journal of nutrition, 116 4
M. Kennelly, K. Ainscough, K. Lindsay, O. Elizabeth, Sullivan, E. Gibney, M. McCarthy, R. Segurado, Giuseppe Devito, O. Maguire, Thomas Smith, M. Hatunic, F. Mcauliffe
A Randomized Controlled Trial
Obstetrics & Gynecology
M. Lane, D. Ingram, R. Cutler, J. Knapka, D. Barnard, G. Roth (1992)
Dietary Restriction in Nonhuman Primates: Progress Report on the NIA Study
Annals of the New York Academy of Sciences, 673
E. Weiss, S. Racette, D. Villareal, L. Fontana, K. Steger-May, K. Schechtman, S. Klein, J. Holloszy (2006)
Improvements in glucose tolerance and insulin action induced by increasing energy expenditure or decreasing energy intake: a randomized controlled trial.
The American journal of clinical nutrition, 84 5
M. Lefevre, L. Redman, L. Heilbronn, J. Smith, Corby Martin, J. Rood, F. Greenway, D. Williamson, Steven Smith, E. Ravussin (2009)
Caloric restriction alone and with exercise improves CVD risk in healthy non-obese individuals.
Atherosclerosis, 203 1
M. Polewski, Maggie Burhans, Minghui Zhao, R. Colman, D. Shanmuganayagam, M. Lindstrom, J. Ntambi, Rozalyn Anderson (2015)
Plasma diacylglycerol composition is a biomarker of metabolic syndrome onset in rhesus monkeys[S]
Journal of Lipid Research, 56
D. Saville (1990)
Multiple Comparison Procedures: The Practical Solution
The American Statistician, 44
(1978)
Committee on Animal Nutrition, Agricultural Board. Nutrient requirements of nonhuman primates
MA Lane (1992)
Dietary restriction in nonhuman primates: progress report on the NIA study
Ann. NY Acad. Sci., 673
DM Bowden (1984)
10.1016/B978-0-12-039228-5.50015-2
Adv. Vet. Sci. Comp. Med., 28
H. Uno, P. Alsum, M. Zimbric, W. Houser, J. Thomson, J. Kemnitz (1998)
Colon cancer in aged captive rhesus monkeys (Macaca mulatta).
American journal of primatology, 44 1
I. Edwards, L. Rudel, J. Terry, J. Kemnitz, R. Weindruch, D. Zaccaro, W. Cefalu (2001)
Caloric restriction lowers plasma lipoprotein (a) in male but not female rhesus monkeys
Experimental Gerontology, 36
L. Guarente (2008)
Mitochondria—A Nexus for Aging, Calorie Restriction, and Sirtuins?
Cell, 132
William Mair, A. Dillin (2008)
Aging and survival: the genetics of life span extension by dietary restriction.
Annual review of biochemistry, 77
T. Gresl, R. Colman, E. Roecker, T. Havighurst, Ze Huang, D. Allison, R. Bergman, J. Kemnitz (2001)
Dietary restriction and glucose regulation in aging rhesus monkeys: a follow-up report at 8.5 yr.
American journal of physiology. Endocrinology and metabolism, 281 4
Author contributions
G. Dailey (2004)
New strategies for basal insulin treatment in type 2 diabetes mellitus.
Clinical therapeutics, 26 6
Barbara Hansen, N. Bodkin (1986)
Heterogeneity of insulin responses: phases leading to Type 2 (non-insulin-dependent) diabetes mellitus in the rhesus monkey
Diabetologia, 29
J. Kemnitz (2011)
Calorie restriction and aging in nonhuman primates.
ILAR journal, 52 1
L. Fontana, D. Villareal, E. Weiss, S. Racette, K. Steger-May, S. Klein, J. Holloszy (2007)
Calorie restriction or exercise: effects on coronary heart disease risk factors. A randomized, controlled trial.
American journal of physiology. Endocrinology and metabolism, 293 1
L. Redman, L. Heilbronn, Corby Martin, A. Alfonso, Steven Smith, E. Ravussin (2007)
Effect of calorie restriction with or without exercise on body composition and fat distribution.
The Journal of clinical endocrinology and metabolism, 92 3
C. López-Otín, M. Blasco, L. Partridge, M. Serrano, G. Kroemer (2013)
The Hallmarks of Aging
Cell, 153
R. Colman, Rozalyn Anderson (2011)
Nonhuman primate calorie restriction.
Antioxidants & redox signaling, 14 2
Koen Houthoofd, J. Vanfleteren (2007)
Public and private mechanisms of life extension in Caenorhabditis elegans
Molecular Genetics and Genomics, 277
Jignesh Pandya, R. Grondin, Heather Yonutas, H. Haghnazar, D. Gash, Zhiming Zhang, P. Sullivan (2015)
Decreased mitochondrial bioenergetics and calcium buffering capacity in the basal ganglia correlates with motor deficits in a nonhuman primate model of aging
Neurobiology of Aging, 36
(2007)
effects on coronary heart disease risk factors
J. Kemnitz, R. Weindruch, E. Roecker, K. Crawford, P. Kaufman, W. Ershler (1993)
Dietary restriction of adult male rhesus monkeys: design, methodology, and preliminary findings from the first year of study.
Journal of gerontology, 48 1
J. Mattison, George Roth, T. Beasley, E. Tilmont, A. Handy, Richard Herbert, Dan Longo, David Allison, Jennifer Young, Mark Bryant, Dennis Barnard, Walter Ward, Wenbo Qi, Donald Ingram, Rafael Cabo (2012)
Impact of caloric restriction on health and survival in rhesus monkeys from the NIA study
Nature, 489
J. Ramsey, Jennifer Laatsch, J. Kemnitz (2000)
Age and gender differences in body composition, energy expenditure, and glucoregulation of adult rhesus monkeys
Journal of Medical Primatology, 29
S. Racette, E. Weiss, D. Villareal, Hassan Arif, K. Steger-May, K. Schechtman, L. Fontana, S. Klein, J. Holloszy (2006)
One year of caloric restriction in humans: feasibility and effects on body composition and abdominal adipose tissue.
The journals of gerontology. Series A, Biological sciences and medical sciences, 61 9
N. Bodkin, T. Alexander, H. Ortmeyer, Elizabeth Johnson, B. Hansen (2003)
Mortality and morbidity in laboratory-maintained Rhesus monkeys and effects of long-term dietary restriction.
The journals of gerontology. Series A, Biological sciences and medical sciences, 58 3
B. Kennedy, S. Berger, A. Brunet, J. Campisi, A. Cuervo, E. Epel, C. Franceschi, G. Lithgow, R. Morimoto, J. Pessin, T. Rando, A. Richardson, E. Schadt, T. Wyss-Coray, F. Sierra (2014)
Geroscience: Linking Aging to Chronic Disease
Cell, 159
KL Grove, SK Fried, A. Greenberg, XQ Xiao, DJ Clegg (2010)
A microarray analysis of sexual dimorphism of adipose tissues in high-fat-diet-induced obese mice
International Journal of Obesity, 34
H Uno (1998)
10.1002/(SICI)1098-2345(1998)44:1<19::AID-AJP2>3.0.CO;2-#
Am. J. Primatol., 44
R. Weindruch, R. Walford (1988)
The Retardation of Aging and Disease by Dietary Restriction
(2010)
Nature Communications
Nature Cell Biology, 12
Laura Ngwenya, Nadine Heyworth, Y. Shwe, T. Moore, D. Rosene (2015)
Age-related changes in dentate gyrus cell numbers, neurogenesis, and associations with cognitive impairments in the rhesus monkey
Frontiers in Systems Neuroscience, 9
S. Rezzi, F. Martin, D. Shanmuganayagam, R. Colman, J. Nicholson, R. Weindruch (2009)
Metabolic shifts due to long-term caloric restriction revealed in nonhuman primates
Experimental Gerontology, 44
John Hudson, S. Baum, D. Frye, E. Roecker, J. Kemnitz (1996)
Age and sex differences in body size and composition during Rhesus monkey adulthood
Aging Clinical and Experimental Research, 8
R. Colman, T. Beasley, Joseph Kemnitz, S. Johnson, R. Weindruch, Rozalyn Anderson (2014)
Caloric restriction reduces age-related and all-cause mortality in rhesus monkeys
Nature Communications, 5
Samia Metwally, J. Rowland, Lizhe Xu, S. Zaki, S. Nichol, P. Rollin, J. Towner, Brigid Batten, Tara Sealy, C. Carrillo, K. Moran, A. Bracht, G. Mayr, Magdalena Sirios-Cruz, Davinio Catbagan, E. Lautner, T. Ksiazek, W. White, M. McIntosh (2009)
Caloric Restriction Delays Disease Onset and Mortality in Rhesus Monkeys
Science, 325
N. Rodríguez, K. Garcia, J. Fortman, T. Hewett, Bunte Rm, Bennett Bt (2002)
Clinical and histopathological evaluation of 13 cases of adenocarcinoma in aged rhesus macaques (Macaca mulatta)
Journal of Medical Primatology, 31
R. Hadfield, P. Yudkin, C. Coe, J. Scheffler, H. Uno, D. Barlow, J. Kemnitz, S. Kennedy (1997)
Risk factors for endometriosis in the rhesus monkey (Macaca mulatta): a case-control study.
Human reproduction update, 3 2
M. Laakso, J. Kuusisto (2014)
Insulin resistance and hyperglycaemia in cardiovascular disease development
Nature Reviews Endocrinology, 10
Richard Gibbs, J. Rogers, M. Katze, Roger Bumgarner, G. Weinstock, E. Mardis, K. Remington, R. Strausberg, J. Venter, Richard Wilson, M. Batzer, C. Bustamante, Evan Eichler, Matthew Hahn, Ross Hardison, K. Makova, W. Miller, A. Milosavljevic, R. Palermo, A. Siepel, J. Sikela, Tony Attaway, S. Bell, Kelly Bernard, C. Buhay, Mimi Chandrabose, M. Dao, Clay Davis, K. Delehaunty, Yan Ding, H. Dinh, S. Dugan-Rocha, L. Fulton, Ramatu Gabisi, Toni Garner, J. Godfrey, A. Hawes, Judith Hernandez, S. Hines, M. Holder, J. Hume, S. Jhangiani, Vandita Joshi, Ziad Khan, E. Kirkness, Andrew Cree, R. Fowler, Sandy Lee, L. Lewis, Zhangwan Li, Yih-shin Liu, S. Moore, D. Muzny, L. Nazareth, Dinh Ngo, Geoffrey Okwuonu, G. Pai, David Parker, H. Paul, C. Pfannkoch, C. Pohl, Y. Rogers, S. Ruiz, A. Sabo, J. Santibanez, Brian Schneider, Scott Smith, E. Sodergren, Amanda Svatek, T. Utterback, Selina Vattathil, Wesley Warren, Courtney White, A. Chinwalla, Yucheng Feng, A. Halpern, L. Hillier, Xiaoqiu Huang, P. Minx, J. Nelson, K. Pepin, X. Qin, G. Sutton, E. Venter, B. Walenz, J. Wallis, K. Worley, Shiaw-Pyng Yang, S. Jones, Marco Marra, Mariano Rocchi, J. Schein, R. Baertsch, Laura Clarke, Miklós Csürös, Jarret Glasscock, R. Harris, P. Havlak, Andrew Jackson, Huaiyang Jiang, Yue Liu, David Messina, Yufeng Shen, H. Song, T. Wylie, Lan Zhang, E. Birney, Kyudong Han, Miriam Konkel, Jungnam Lee, A. Smit, B. Ullmer, Hui Wang, Jinchuan Xing, Richard Burhans, Ze Cheng, J. Karro, Jian Ma, B. Raney, X. She, Michael Cox, Jeffery Demuth, Laura Dumas, Sang-Gook Han, J. Hopkins, A. Karimpour-Fard, Young Kim, J. Pollack, T. Vinař, Charles Addo-Quaye, Jeremiah Degenhardt, A. Denby, M. Hubisz, Amit Indap, C. Kosiol, B. Lahn, Heather Lawson, A. Marklein, R. Nielsen, E. Vallender, A. Clark, B. Ferguson, Ryan Hernandez, Kashif Hirani, H. Kehrer-Sawatzki, J. Kolb, Shobha Patil, Ling-Ling Pu, Yanru Ren, D. Smith, D. Wheeler, Ian Schenck, E. Ball, R. Chen, D. Cooper, B. Giardine, F. Hsu, W. Kent, A. Lesk, D. Nelson, W. O'Brien, Kay Prüfer, P. Stenson, J. Wallace, Hui Ke, Xiao-Ming Liu, Peng Wang, A. Xiang, Fan Yang, G. Barber, D. Haussler, D. Karolchik, Andrew Kern, R. Kuhn, Kayla Smith, Ann Zwieg (2007)
Evolutionary and Biomedical Insights from the Rhesus Macaque Genome
Science, 316
P. Allison (1995)
Survival analysis using sas®: a practical guide
V. Longo (2016)
Faculty Opinions recommendation of Dietary restriction in mice beginning at 1 year of age: effect on life-span and spontaneous cancer incidence.
N. Takahashi (1992)
Aging
Cell, 147
E. Masoro, B. Yu, H. Bertrand (1982)
Action of food restriction in delaying the aging process.
Proceedings of the National Academy of Sciences of the United States of America, 79 13
H. Uno (1997)
Age-related pathology and biosenescent markers in captive rhesus macaques
AGE, 20
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations
E. Masoro (1990)
The Retardation of Aging and Disease by Dietary Restriction
Journal of Nutrition, 120

Publisher: Springer Journals
Copyright: Copyright © 2017 by The Author(s)
Subject: Science, Humanities and Social Sciences, multidisciplinary; Science, Humanities and Social Sciences, multidisciplinary; Science, multidisciplinary
eISSN: 2041-1723
DOI: 10.1038/ncomms14063
Publisher site: See Article on Publisher Site

Abstract

ARTICLE Received 31 May 2016 | Accepted 24 Nov 2016 | Published 17 Jan 2017 DOI: 10.1038/ncomms14063 OPEN Caloric restriction improves health and survival of rhesus monkeys 1, 2, 3,4, 3 2,5 Julie A. Mattison *, Ricki J. Colman *, T. Mark Beasley *, David B. Allison , Joseph W. Kemnitz , 6 7 8,9 1, 8,9, George S. Roth , Donald K. Ingram , Richard Weindruch , Rafael de Cabo ** & Rozalyn M. Anderson ** Caloric restriction (CR) without malnutrition extends lifespan and delays the onset of age-related disorders in most species but its impact in nonhuman primates has been controversial. In the late 1980s two parallel studies were initiated to determine the effect of CR in rhesus monkeys. The University of Wisconsin study reported a signiﬁcant positive impact of CR on survival, but the National Institute on Aging study detected no signiﬁcant survival effect. Here we present a direct comparison of longitudinal data from both studies including survival, bodyweight, food intake, fasting glucose levels and age-related morbidity. We describe differences in study design that could contribute to differences in outcomes, and we report species speciﬁcity in the impact of CR in terms of optimal onset and diet. Taken together these data conﬁrm that health beneﬁts of CR are conserved in monkeys and suggest that CR mechanisms are likely translatable to human health. 1 2 Translational Gerontology Branch, National Institute on Aging, Baltimore, Maryland 21224, USA. Wisconsin National Primate Research Center, University of 3 4 Wisconsin-Madison, Madison, Wisconsin 53715, USA. Department of Biostatistics, University of Alabama, Birmingham, Alabama 35294, USA. Geriatric Research Education and Clinical Center, Birmingham/Atlanta Veterans Administration Hospital, Birmingham, Alabama 35233, USA. Department of Cell and 6 7 Regenerative Biology, University of Wisconsin-Madison, Madison, Wisconsin 53792, USA. GeroScience, Pylesville, Maryland 21323, USA. Pennington Biomedical Research Center, Baton Rouge, Louisiana 70808, USA. Department of Medicine, University of Wisconsin, Madison, Wisconsin 53792, USA. Geriatric Research Education and Clinical Center, William S. Middleton Memorial Veterans Hospital, Madison, Wisconsin 53705, USA. * These authors contributed equally to this work. ** These authors jointly supervised this work. Correspondence and requests for materials should be addressed to R.d.C. (email: [email protected]) or to R.M.A. (email: [email protected]). NATURE COMMUNICATIONS | 8:14063 | DOI: 10.1038/ncomms14063 | www.nature.com/naturecommunications 1 ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14063 he rhesus monkey (Macaca mulatta) is an excellent model CR however, and analysis was based on comparing 109 ad libitum for human ageing. The rhesus monkey genome shares fed males and females from colony records at that facility, 1,2 TB93% sequence identity with the human genome , and including insulin resistant and diabetic animals, to only eight numerous aspects of their anatomy, physiology, neurology, male CR monkeys. Two other studies focused speciﬁcally on the endocrinology and immunology directly parallel those of impact of CR in healthy male and female rhesus monkeys: one at 3,4 humans . Rhesus monkeys have a lifespan measured in the National Institute on Aging (NIA) involving 121 monkeys; decades, and develop, mature and age in similar ways to and the other at the University of Wisconsin Madison (UW) with humans. In terms of ageing this includes greying and thinning 76 monkeys. The same University of Alabama at Birmingham of hair, redistribution of body fat, loss of skin tone, loss of vigour based statistical team was engaged for analysis of data from 4–7 and loss of muscle tone . With age there are increases in clinical both studies. The UW study has reported beneﬁcial effects of CR, manifestations of diseases and disorders that also increase in including signiﬁcant improvements in health and age-related 19 20 prevalence with advancing age in humans, including diabetes, survival , and all-cause survival . In contrast, the NIA study neoplasia, sarcopenia, bone loss, altered immune function and reported no signiﬁcant impact of CR on survival, although 3,4,8 21 cognitive decline . Like humans, nonhuman primates are improvements in health were close to statistical signiﬁcance . genetically heterogeneous so that phenotypes of ageing are The basis for the contrasting outcomes from these two parallel non-uniformly manifested among individual animals. Variance studies has not been established. Analysis of limited published can be somewhat offset through comprehensive life-long physical bodyweight data indicated that the controls were not equivalent and medical health records, where age-related changes are viewed between the two studies , pointing to fundamental differences in from the perspective of each animal individually. Nonhuman study design and implementation. Therefore, to more fully assess primates exhibit feeding patterns and sleeping behaviour similar possible explanations for the discrepant ﬁndings between the two to those of humans, and a key advantage of nonhuman primate studies, we have conducted a comprehensive assessment of studies over human studies is that all variables including the longitudinal data from both sites highlighting differences that environment and dietary intake can be fully deﬁned. In short, may have contributed to the dissimilar outcomes. nonhuman primates are vital models for translating basic research into clinical application. A clear understanding of the biology of ageing, as opposed to Results the biology of individual age-related diseases, could be the critical Intrinsic differences in study design. Most of the early rodent turning point for novel approaches in preventative strategies to CR studies involved very young onset life-long CR initiated post- facilitate healthy human ageing. Caloric restriction (CR) offers a weaning, usually in inbred genetic backgrounds. In the 1980s it powerful paradigm to uncover the cellular and molecular basis for became clear that adult onset CR (12-month-old mice) was also the age-related increase in overall disease vulnerability that is effective in delaying ageing and extending lifespan in rodents, shared by all mammalian species. CR extends median and albeit to a lesser extent than the young onset model . Many maximum lifespan in most strains of laboratory rodents and also rodent CR studies opt to feed control animals ad libitum amounts 9–12 delays the onset of age-related diseases and disorders . of food while others provide less than ad libitum amounts arguing Lifespan is also extended by CR in most short-lived species, that this strategy avoids the confounding effects of obesity including the unicellular yeast, nematodes and invertebrates. and reduces variability in food intake among individuals. There has been rapid progress in identifying potential With the launch of the NIA rhesus monkey study in 1987, the 13–17 mechanisms of CR utilizing these models . These implementation of CR was such that the control monkeys were short-lived species are well suited for the investigation of the not free-fed. Food allotments were determined in accordance with underlying mechanisms of CR due to the relative ease in their data published by the National Research Council to provide genetic manipulation, extensive genetic and developmental approximate ad libitum intake based on their age and bodyweight characterization, low cost, and signiﬁcantly reduced timeframe for the maturing control monkeys without overfeeding . Rations for completion of longevity studies. A key question underpinning were increased to maintain growth and development until full this body of work is whether the biology of CR, and its ability to stature was attained. CR monkeys received 30% less food than delay ageing and the onset of disease, applies to humans and height-, age- and sex-matched control monkeys. The intervention human health. was initiated as young-onset and old-onset groups of males, and To date three independent studies of rhesus monkeys (Macaca young, adult, and old-onset groups of females (Table 1). mulatta) have tackled the question of translatability of CR to Launched in 1989, the UW study initiated the CR diet in adult primate species. The University of Maryland rhesus monkey animals only, after full stature was achieved (B8 years of age for study was the ﬁrst to report a positive association of CR with rhesus monkeys) . Food was provided at levels approximating survival with a 2.6-fold increased risk of death in control animals ad libitum to control animals. To accommodate heterogeneity in compared to restricted . The primary focus of the study was not the feeding behaviours within the cohort, the ad libitum reference Table 1 | Study design. Site Group Control N CR N Age of onset Genetic origin NIA Male Juvenile 10 10 1–2 years Indian/Chinese Adolescent 12 10 3–5 years Indian/Chinese Old 10 10 16–23 years Indian/Chinese Female Juvenile 9 9 1–3 years Indian Adult 15 11 6–14 years Indian/Chinese Old 8 7 16–21 years Indian UW Male 23 23 7–14 years Indian Female 15 15 9–15 years Indian 2 NATURE COMMUNICATIONS | 8:14063 | DOI: 10.1038/ncomms14063 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14063 ARTICLE for each individual was established using baseline food intake deprived overnight. While there were considerable differences in measured over 3–6 months, and CR was implemented on a study design as outlined above, it should be noted that animal per-individual basis. The rationale for these design features at housing and routine animal care were equivalent at NIA and UW UW was to implement a study as it might have been conducted in primate facilities. This included identical housing conditions, humans. temperature and humidity range, light cycles, and the use of tap The source of the monkeys in each cohort and the population water, which was continuously available. Both studies included type represented is also a point of difference for the two studies. animal monitoring several times per day, and a designated The UW monkeys were born and raised at the Wisconsin veterinary staff that inspected the animals routinely and provided National Primate Research Center and were all of Indian origin. outstanding care as needed. The NIA monkeys were sourced from several locations and included monkeys of both Indian and Chinese origin. Chinese male rhesus monkeys are generally heavier and longer than their Impact of CR on survival. The initial goal of both NIA and UW Indian counter parts with the reverse being the case for females, studies was to determine the impact of CR on the health of rhesus and Chinese rhesus monkeys are also thought to exhibit greater monkeys, as it was not a foregone conclusion that CR would be sexual dimorphism . Monkeys of different origin are sufﬁciently an appropriate intervention in long-lived species. The investiga- genetically different that they can be distinguished using a panel tion of the impact of CR on longevity was not considered a of single nucleotide polymorphisms . Apart from population primary outcome at either study location. Even though 121 differences, rhesus monkeys share a similar degree of inter- monkeys were enrolled in the NIA study, the differences in age of individual genetic variation as humans . In this way, the onset (from 1 to 23 years) precluded the animals from being contribution of population type to differences in outcomes of grouped together for data analyses. Although the age range for the two studies as opposed to the contribution of individual time of onset is smaller for the UW study (ages 7–15), with only genetic heterogeneity is difﬁcult to ascertain. 38 outbred genetically distinct monkeys per group (including The diet compositions were another important difference both sexes), it seemed unlikely that the study would have the between the two studies. First, the source of diet components was statistical power required to test CR’s effect on longevity. While different. A naturally sourced diet was employed at the NIA neither study reports longevity data, both studies have yielded facility to ensure that micronutrients such as phytochemicals and survival data. For rhesus monkeys in captivity, the previously trace minerals were provided, acknowledging that there was reported median survival was B26 years of age, 10% survival was potential for seasonal variation. In contrast, a semi-puriﬁed diet B35 years of age and maximal survival was B40 years of age . was employed at UW to ensure that intake could be fully deﬁned Mortality curves were generated separately for UW and NIA and consistent throughout the course of the study. Second, (Fig. 1). Survival estimates for monkeys at both sites were although diets at both locations had a similar caloric density, the calculated based on data captured up to July 2015 using the three relative macronutrient composition of the diets was not most common statistical methods: Kaplan-Meier product-limit equivalent (Table 2). Compared to the UW diet, the NIA diet method; Cox proportional hazard regression and parametric was lower in fat, higher in protein and higher in ﬁbre. Finally, the survival analysis assuming a Weibull distribution (Table 3). nutrient content of the diets was also different. At both locations Because the Weibull distribution is a special case of the diets containedB60% carbohydrates by weight, but sucrose generalized extreme value distribution, it can accommodate comprised less than 7% of total carbohydrates at NIA and 45% estimation of the upper quantiles of a survival distribution and of total carbohydrates at UW. Diets at both locations were replete maximal lifespan, especially when there are censored data due to for vitamins that were provided at or above the recommended animals that remain alive . daily allowance. In the UW adult-onset study, the estimated survival of UW Feeding practices also differed between studies. At NIA, the control animals was close to that of the average recorded for monkeys were fed two meals at B6:30 and 13:00 each day. Any monkeys in captivity (B26 years of age). Considering both males food remaining after the morning meal was removed after about and females together, a statistically signiﬁcant effect of CR in 3 h, and a low calorie treat was provided, typically in the form of a increasing survival was observed (Cox regression P¼ 0.017; small piece of fruit. The afternoon meal was not removed so that Supplementary Table 1). The hazard ratio (HR) of 1.865 (95% monkeys had access to food at night. At UW, all monkeys were conﬁdence interval (CI): 1.119–3.108) indicated that at any time- fed in the morning atB8:00 and any remaining food was point the control monkeys had almost twice the rate of death removed at B16:00 when a treat of fresh fruit or vegetable, which when compared to CR animals. The effect of sex on the response was quickly and completely eaten, was provided. Food allotment to CR was not statistically signiﬁcant. Kaplan-Meier analysis for control animals was adjusted to ensure that there was always showed that median survival estimates were greater for CR some uneaten food to be removed at the end of the day. In this animals for both males and females (Table 3). In the NIA study way UW animals were ad libitum fed during the day but food large differences in ages of monkeys at time of recruitment to the Table 2 | Diet composition at each location. Diet component NIA UW Control/CR Nutrient sources Control CR Nutrient sources (% by weight) (% by weight) (% by weight) Protein 17.3 soybean and ﬁsh meal 13.13 13.13 lactalbumin Carbohydrate 56.9 wheat, corn, sucrose (6.8%) 60.92 58.31 corn, sucrose (45%), dextrin Fat 5.0 soy, corn, ﬁsh oils 10.6 10.6 corn oil Fibre 6.5–9.0 cellulose 5.0 5.0 cellulose Vitamins B140% RDA 100% RDA B130% RDA Calories kcal g 3.9 3.9 3.8 NATURE COMMUNICATIONS | 8:14063 | DOI: 10.1038/ncomms14063 | www.nature.com/naturecommunications 3 ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14063 100 100 Control: 2 CR: 5 Control: 1 CR: 5 Average Average 60 60 onset onset UW females UW males 10 20 30 40 10 20 30 40 Control: 10 CR: 7 Control: 9 CR: 7 Average Average onset onset NIA J/A females NIA J/A males 0 0 10 20 30 40 10 20 30 40 Control: 0 CR: 1 Control: 1 CR: 0 Average Average onset 60 onset 40 40 NIA old females NIA old males 0 0 10 20 30 40 10 20 30 40 Age in years Control Age in years CR Figure 1 | Mortality curves for monkeys at UW and at NIA. These curves depict data for male and female monkeys on the UW study and on the NIA study. Animals are grouped by age where male J/A include juvenile and adolescent onset animals, female J/A include juvenile and adult onset animals, and old include the advanced age onset animals. Inset boxes indicate animals still alive, dashed line marks 50% mortality. Statistics related to this ﬁgure are provided in Supplementary Information, Supplementary Table 1. Table 3 | Survival estimates. Groups KM KM KM KM Weibull Weibull Median IQR Mean SE Median IQR All cause survival estimates Males UW Control 26.11 6.82 25.28 1.03 25.75 5.57 CR 28.32 6.19 26.86 1.23 27.63 8.04 NIA J/A Control 28.78 NE 26.00 1.50 29.22 14.31 CR 26.31 13.06 23.71 1.91 25.86 17.05 NIA old Control 34.78 9.16 34.19 1.89 34.80 7.23 CR 37.10 10.42 34.82 1.95 35.88 8.71 Females UW Control 23.86 7.92 23.56 1.29 24.57 8.62 CR 29.68 15.60 25.78 2.18 27.49 14.37 NIA J/A Control 25.67 12.65 24.58 1.71 25.81 14.21 CR 22.63 NE 19.79 1.20 23.53 15.25 NIA old Control 27.19 8.98 28.57 1.59 29.71 8.91 CR 27.87 11.88 26.13 2.65 26.49 8.80 KM, Kaplan-Meier; IQR, interquartile range. study (Table 1) prompted a separation of data from the early and Although Cox proportional hazard regression indicated that the late onset groups. Here and throughout this report, NIA male differences in survival between J/A control and CR were not juveniles and adolescents (J/A) were grouped and female juveniles statistically signiﬁcant (Supplementary Table 1), CR monkeys and adults were grouped (J/A). The Kaplan-Meier median reached 80% mortality before the controls for both sexes. With estimated survival was not different between NIA control and 38% of the NIA J/A cohort still alive, the survival curves are CR animals for the J/A onset groups of males or females (Fig. 1). incomplete and the impact on survival remains to be determined; 4 NATURE COMMUNICATIONS | 8:14063 | DOI: 10.1038/ncomms14063 | www.nature.com/naturecommunications Percent survival Percent survival Percent survival NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14063 ARTICLE however, the early mortality suggests that for some individuals complications due to endometriosis. A further contributing factor implementation of CR in the very young may confer a survival relates to the policy on treatment of clinical conditions. At UW risk. For old-onset CR, Kaplan-Meier estimated survival was not the policy to treat clinical conditions was implemented from the different between control and CR groups for either males or outset. At NIA, although acute pain and suffering were always females (Table 3), but survival estimates were higher than those of treated, chronic medical conditions, including endometriosis, J/A monkeys and UW controls. For both males and females, were monitored but not medically treated. A policy change was survival estimates for the NIA old-onset cohort were comparable implemented in 2010 due to the high incidence of endometriosis. to or exceeded those for UW CR. The power to assess the impact of CR on survival for NIA J/A Although there were slight discrepancies in the estimated females has been compromised somewhat by this one condition. median survival between the non-parametric Kaplan-Meier and parametric Weibull estimation methods, the survival comparisons between study sites using either analysis were consistent. Biometric and food intake measures from both studies. For A certain degree of sexual dimorphism was observed in survival over a quarter of a century during these studies, bodyweight, body outcomes where incidence of early death appeared to be greater composition and food intake were measured for all 197 monkeys. for females. This observation might be explained in part by Bodyweight was determined in fasted and anesthetized monkeys endometriosis, which is the proliferation of endometrial tissue 2–4 times per year during routine procedures. Longitudinal data outside of the uterus. Endometriosis can occur at relative for all monkeys were averaged by age of the animal (Fig. 2a). As is high incidence in monkeys in captivity (B25%), and risk is the case for humans, monkeys often experience cachexia or 31,32 considerably greater for nulliparous females . Incidence of end-of-life rapid weight loss. To avoid confounding effects of endometriosis was equivalent for control and CR groups. For the weight change that is not related to food intake or diet, data from J/A cohorts in the NIA study, 12 of the 44 females died of the last year of life for each monkey were excluded. To facilitate complications due to endometriosis, and of these the juvenile comparisons among the cohorts, data were grouped into three age onset females were conﬁrmed nulliparous. Females recruited to categories representing young adult (11–13 years of age), late the UW study, in contrast, had at least one but no more than mid-age (18–20 years of age) and advanced age (25–27 years of three healthy infants , and only 2 of 30 females died of age) (Supplementary Tables 2 and 3). 7 7 12 13 12 12 7 9 13 11 9 9 12 11 11 8 7 17 17 11 14 13 12 11 18 18 11 10 19 11 4 4 11 4 11 8 7 11 10 11 8 7 7 7 19 14 12 7 7 5 4 9 5 8 5 4 3 3 3 15 4 16 6 17 21 7 4 3 19 18 18 19 3 19 3 5 18 17 5 5 NIA J/A females UW females NIA old females 15 20 25 20 25 30 10 15 20 25 Age in years Age in years Age in years 16 16 15 21 21 19 17 19 18 14 14 20 14 17 18 15 19 9 19 19 17 11 19 18 19 19 19 18 12 19 12 20 7 17 7 18 20 19 8 12 20 6 7 9 15 12 14 19 6 8 8 15 15 13 12 15 13 10 8 18 19 20 17 9 10 8 10 15 14 10 21 21 9 18 17 9 8 18 10 6 9 15 8 17 8 8 8 7 8 8 8 8 18 18 NIA J/A males UW males NIA old males 10 15 20 25 10 15 20 25 20 25 30 Age in years Age in years Age in years Control CR b Males Females 11–13 years 18–20 years 25–27 years 11–13 years 18–20 years 25–27 years 20 20 0 0 –10 –10 –20 –20 –30 –30 UW control NIA J/A control NIA old control UW CR NIA J/A CR NIA old CR Figure 2 | Bodyweight data for monkeys at NIA and UW. (a) Bodyweight (kg) for male and female monkeys at UW and at NIA grouped by age where male J/A include juvenile and adolescent onset animals, female J/A include juvenile and adult onset animals, and old include the advanced age onset animals. Digits shown in white within the boxes are the numbers of individual animals contributing to each data point, data are shown as mean s.e. of the mean. (b) Comparison of bodyweight averages for monkeys from UW and NIA studies with records of the internet Primate Aging Database (iPAD). Average bodyweight for control and CR monkeys at both study locations were determined by age category including adult (11–13 years of age), late mid-age (18–20 years of age) and advanced age (25–27 years of age). Data are expressed as percent deviation from the iPAD average for females and males from each age category. Statistics related to this ﬁgure are provided in Supplementary Information, Supplementary Tables 2 and 3. NATURE COMMUNICATIONS | 8:14063 | DOI: 10.1038/ncomms14063 | www.nature.com/naturecommunications 5 Weight in kg Weight in kg % difference from iPAD average ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14063 Considering ﬁrst the female monkeys, bodyweight for the NIA average (B18% for males; B19% for females), and CR monkeys J/A was not signiﬁcantly different between control and CR had lower bodyweight than the iPAD average (B12% for males; monkeys for any age categories. UW CR females weighed B11% for females) (Fig. 2b). For NIA J/A, control males were the signiﬁcantly less (17–26%) than controls throughout the study same to slightly heavier (5–10%) than the iPAD average and CR period, and UW female controls weighed signiﬁcantly more than weighed less than the iPAD average (B20%), while control and NIA J/A female controls throughout (Supplementary Table 2). CR female monkeys both weighed less than the iPAD average For NIA old-onset females, bodyweight was not signiﬁcantly throughout the study (B10% and B20% respectively). All NIA different between controls and CR, and was signiﬁcantly lower old-onset monkeys weighed less than the iPAD average for both than bodyweight of UW female controls. In summary, for NIA control (B15% for females;B10% for males) and CR (B22% for J/A and old-onset female cohorts, bodyweight for control and CR females; B21% for males) monkeys. In summary, bodyweights of monkeys was not different from each other and all were UW and NIA control monkeys were not equivalent to each other, signiﬁcantly lower than the UW controls. Considering next the and apart from J/A males, were respectively higher and lower of male monkeys, NIA J/A CR males weighed signiﬁcantly less the iPAD average. (19–22%) than their control counterparts throughout the study. To gain insight into differences in the effect of age and diet on The difference between UW control and CR was slightly greater body composition, dual X-ray absorptiometry measures were (24–35%), with CR males weighing signiﬁcantly less than conducted at intervals throughout the course of the two studies controls. The average peak weight for NIA J/A control males (Fig. 3). Since each animal had multiple measures taken over was B15% lower than that of UW control males, but differences time, estimates of the average percent adiposity (fat/bodyweight in bodyweight were signiﬁcant for the young age category only expressed as percent) were adjusted for age (Supplementary (Supplementary Table 3). Bodyweight of the old-onset NIA Fig. 1). Within groups a main effect of age on adiposity was control and CR males were not signiﬁcantly different at either detected for NIA J/A and UW cohorts. A main effect of diet was mid-age or advanced ages, and old-onset NIA male controls detected for NIA J/A males and for both males and females from weighed signiﬁcantly less than UW controls. In summary, NIA the UW study, where CR was associated with signiﬁcantly lower J/A and UW male cohorts showed a clear bodyweight response to adiposity. The NIA J/A control and CR females did not differ CR, but old-onset NIA control and CR males were not different from each other in adiposity and neither of the NIA old-onset from each other and were signiﬁcantly lower than the UW monkey groups had a main effect of CR on adiposity. Combining controls. the data from NIA J/A and UW, a difference in adiposity was The internet Primate Aging Database (iPAD; http://ipad. detected between controls on the two studies for both males and primate.wisc.edu) is a repository of clinical and biometric data females, where NIA monkeys had signiﬁcantly lower percent from healthy, non-experimental, captive nonhuman primates body fat. Control monkeys from NIA J/A were not statistically housed at research facilities across the USA. Using data from over different from UW CR in percent body fat for both sexes. These 1,200 individual rhesus monkeys of Indian origin, mean body- data show an impact of age on adiposity in all three groups and weights were calculated for the above age categories for males reveal that the impact of CR on adiposity was observed for both (11.6, 12.1, 11.5 kg respectively) and females (7.4, 8.4, 7.8 kg groups of UW monkeys and at NIA for J/A males only. respectively). UW control and CR monkeys fell on either side of Food intake was monitored daily at both sites. At UW daily these averages; control monkeys were heavier than the iPAD measures of food intake were used to calculate means. At NIA 40 40 40 3 7 30 24 10 30 30 4 21 3 24 12 24 20 11 5 1917 3 4 2021 4 6 7 5 17 15 22 3 4 3 8 2 141510 21 2 20 20 11 20 15 17 16 21 15 5 3 16 3 12 23 17 5 21 18 12 4 23 22 2 2 3 7 14 17 6 23 23 2 13 2 13 11 10 11 10 10 2 13 11 12 4 9 5 6 6 5 UW females NIA J/A females NIA old females 10 15 20 25 10 15 20 25 30 35 20 25 30 Age in years Age in years Age in years 40 40 40 34 5 30 30 37 35 2921 10 4 30 NIA old males 21 34 15 39 39 4 5 30 10 18 14 12 6 17 13 20 6 19 20 23 4 20 10 20 13 25 6 7 17 35 31 17 13 23 5 18 32 6 36 33 12 8 24 36 11 16 5 6 12 2 4 14 19 37 15 5 5 5 13 4 32 5 13 36 3 6 4 10 13 10 36 10 7 3 15 3 6 12 13 8 2 3 3 3 NIA J/A males UW males 10 15 20 25 10 15 20 25 30 35 25 30 Age in years Age in years Age in years Control CR Figure 3 | Adiposity data for female and male monkeys at NIA and UW. Percent adiposity (fat (g)/total bodyweight (g)) calculated from DXA (dual energy X-ray absorptiometry) measures conducted during the course of the studies for male and female monkeys at UW and at NIA grouped by age where male J/A include juvenile and adolescent onset animals, female J/A include juvenile and adult onset animals, and old include the advanced age onset animals. Digits shown in white within the boxes are the numbers of individual animals contributing to each data point, data are shown as mean s.e. of the mean. 6 NATURE COMMUNICATIONS | 8:14063 | DOI: 10.1038/ncomms14063 | www.nature.com/naturecommunications Adiposity % Adiposity % NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14063 ARTICLE 750 750 750 NIA J/A females 26 24 25 22 22 NIA old females 650 650 17 650 40 16 22 21 19 17 14 35 13 22 18 550 19 550 550 32 22 17 20 22 21 19 21 22 22 4 13 15 22 9 4 29 17 34 10 14 14 11 23 22 18 9 19 4 450 40 12 15 450 450 6 7 38 6 16 9 7 9 7 25 26 7 20 6 25 6 5 UW females 14 21 8 350 350 10 15 20 15 20 25 20 25 30 Age in years Age in years Age in years 900 15 16 900 40 29 34 30 36 28 13 19 38 15 42 37 42 44 36 20 32 17 800 800 34 NIA old males 15 28 36 28 31 15 19 8 8 8 12 16 12 700 700 12 700 16 14 16 14 27 16 14 12 16 13 22 11 13 31 32 33 13 28 29 7 36 10 14 15 13 31 26 15 11 13 600 20 600 13 16 25 600 500 500 500 13 11 5 15 9 16 NIA J/A males UW males 400 400 400 10 15 20 25 10 15 20 25 25 30 Age in years Age in years Age in years Control CR Figure 4 | Food intake data for monkeys at NIA and UW. Food intake (daily values in Kcalories) for male and female monkeys at UW and at NIA grouped by age where male J/A include juvenile and adolescent onset animals, female J/A include juvenile and adult onset animals, and old include the advanced age onset animals. Digits shown in white within the boxes are the numbers of individual animals contributing to each data point, data are shown as mean s.e. of the mean. food intake means were calculated based on measures conducted an intervention that imparts improved health even in the absence during a single week per year as representative of typical intake. of increased longevity, is viewed as a highly favourable and Longitudinal data for all monkeys were averaged by age of the legitimate example of an ageing intervention. With advancing animal (Fig. 4). Data from the last year of life of each monkey age, rhesus monkeys are vulnerable to many of the same were excluded to avoid confounding effects of end-of-life feeding conditions observed in humans. Among the most prevalent are behaviours that usually include loss of appetite. Considering ﬁrst cancer, cardiac disease, and conditions related to immune the females and using the age categories deﬁned above for both dysfunction and inﬂammation, and examples of each were UW and NIA J/A, the controls consumed signiﬁcantly more identiﬁed in monkeys on the ageing and CR studies at both calories than CR at both young and mid-age, but the difference NIA and UW (Supplementary Table 4). persisted only for UW female monkeys at advanced age. For the Fasting glucose measures were common to both studies and the old-onset NIA, caloric intake was not different between control longitudinal data are shown (Fig. 5). In healthy adult rhesus and CR. Among control monkeys, UW females consumed monkeys fasting glucose levels are 64–68 mg dl (refs 18,35). signiﬁcantly more calories than NIA J/A at mid-age and advanced For NIA J/A, fasting glucose levels were equivalent for controls age and more than old-onset at advanced age. Considering next and CR up to B23 years of age, after which the control and CR the males, the NIA J/A controls consumed signiﬁcantly more males, but not females, began to diverge. Both control and CR calories than CR at young and mid-age and the difference females showed an age-related increase in fasting glucose levels between control and CR was signiﬁcant for UW at mid-age only. after B21 years of age. For UW monkeys, the control males had Old-onset males at NIA differed signiﬁcantly in their caloric higher fasting glucose levels than CR from 15 years of age with a intake between control and CR only at advanced age. Among further divergence of the curves after B23 years of age, while a controls, caloric intake was not different for NIA J/A and UW noticeable difference between control and CR females emerged males at any point in the study, but old-onset males consumed after only B21 years. For the NIA old-onset cohorts, fasting signiﬁcantly less than UW males and NIA J/A males at mid-age. glucose was consistently low for the duration of the study period. In summary, signiﬁcant differences in caloric intake were These data point to an age-related increase in fasting glucose identiﬁed between control and CR monkeys for male and female for rhesus monkeys and single out the UW control males as NIA J/A and UW cohorts, but not for old-onset cohorts until being predisposed to elevated circulating glucose in the fasted advanced age and then for males only. Comparing between sites, state. Using multilevel modelling to investigate the relationship caloric intake for NIA female controls of both J/A and old-onset between adiposity and fasting glucose levels a signiﬁcant was lower than that of UW controls, and for males, caloric intake relationship was identiﬁed for UW males only (P¼ 0.005). of NIA J/A and UW controls were not different from each other A signiﬁcant age by diet interaction was also detected (P¼ 0.014), but old-onset NIA controls were lower than both. suggesting that the impact of age on the relationship between adiposity and glucoregulatory parameters is distinct for control and CR monkeys. Impact of CR on incidence of disease. The concept of health- Veterinarians documented body condition and overall health span is a fairly recent development in ageing research, where a of monkeys biannually at both study locations and indicators of distinction is drawn between chronological age and health diseases or disorders identiﬁed. The age at which a monkey was ﬁrst diagnosed with an age-related condition was used status . Traditionally, an increase in both median and maximum lifespan was considered the hallmark of delayed ageing, and to generate morbidity curves (Fig. 6). Age-related conditions improvements in health were deemed to be a necessary and included sarcopenia, osteoporosis, arthritis, diverticulosis, obvious component of longevity. The perspective has shifted cataracts and persistent heart murmurs, in addition to age- somewhat towards greater emphasis on health and morbidity, so related diseases including cancer, diabetes and cardiovascular NATURE COMMUNICATIONS | 8:14063 | DOI: 10.1038/ncomms14063 | www.nature.com/naturecommunications 7 Daily intake in Kcal Daily intake in Kcal ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14063 130 130 110 110 29 21 14 NIA J/A females NIA J/A males 90 90 19 36 26 13 31 27 34 32 37 30 38 24 36 39 20 38 70 31 39 25 70 37 44 39 29 22 50 37 41 44 39 24 48 41 46 32 59 53 36 45 41 30 60 58 41 35 42 27 24 30 38 32 50 47 50 34 32 42 52 42 44 52 42 50 50 10 15 20 25 30 10 15 20 25 30 Age in years Age in years 130 130 110 110 UW males UW females 40 33 42 42 19 14 36 36 25 36 21 14 17 20 22 33 24 17 16 22 24 38 24 22 17 70 32 34 70 22 19 15 25 22 23 18 12 18 26 21 21 25 40 41 39 14 22 39 39 22 22 17 22 14 10 26 50 50 10 15 20 25 30 10 15 20 25 30 Age in years Age in years 130 130 110 110 90 90 NIA old females NIA old males 13 6 70 14 70 14 18 11 15 10 25 24 26 21 10 9 17 17 17 10 18 13 24 7 21 19 21 7 7 17 16 9 25 22 15 12 11 18 10 6 7 50 50 10 15 20 25 30 10 15 20 25 30 Age in years Age in years Control CR Figure 5 | Fasting glucose values for monkeys at NIA and UW. Circulating levels of glucose (mg dl ) are shown for male and female monkeys at UW and at NIA grouped by age where male J/A include juvenile and adolescent onset animals, female J/A include juvenile and adult onset animals, and old include the advanced age onset animals. Digits shown in white within the boxes are the numbers of observations contributing to each data point, data are shown as mean s.e. of the mean. disease. Cox proportional hazard regression modelling indicated clinical care at both locations, although clinical deﬁnitions that age-related conditions occurred at B2.7 times the rate in representing different disease stages were employed at the two 38,39 control animals compared to CR for UW monkeys (HR: 2.665; sites. Similar to humans , insulin resistance occurs in advance 35,40 CI: 1.527–4.653; P¼ 0.0006). In the NIA J/A cohort, age-related of impaired fasting glucose in rhesus monkeys , which occurs conditions occurred at twice the rate in control monkeys before transition to full diabetes. At UW loss of insulin sensitivity compared to CR (HR: 2.091; CI: 1.169–3.641; P¼ 0.0125) was used to diagnose glucoregulatory impairment and was (Supplementary Table 5; Supplementary Fig. 2). The advanced deﬁned as fasting insulin levels (470mUml ) and an insulin age of the old-onset NIA monkeys precluded detection of the ﬁrst sensitivity index (S )o2(E ) as determined by a frequently occurrence of an age-related condition. sampled intravenous glucose tolerance test. At NIA, fasting The incidence of age-related conditions prevalent in glucose (4100 mg dl ), glucosuria and HbA1c (46.5%) human populations, such as cancer, cardiovascular disease and measures were used to deﬁne diabetes. CR animals had lower glucoregulatory dysfunction/diabetes, was determined for incidence of glucoregulatory dysfunction than controls at both monkeys at UW and NIA, with J/A and old-onset combined. UW and NIA study sites. Multilevel modelling was used to Diagnoses were made clinically by veterinary staff upon investigate possible relationships between adiposity and presentation, and disease-related pathology was subsequently morbidity but no signiﬁcant effects were detected for any group conﬁrmed upon necropsy by a board-certiﬁed pathologist. from either study. Similarly, for all NIA groups and for UW Clinically silent pathologies were identiﬁed at necropsy. To males, no relationship between adiposity and mortality was compare studies, cancer and cardiovascular disorders are reported detected. For UW females, adiposity was associated with a modest as incidence upon necropsy. Adenocarcinoma was the leading reduction in risk for death (HR: 0.927; P¼ 0.01) but only after cause of death at both study sites, consistent with previous reports correcting for age and diet. These data suggest that the inﬂuence 36,37 on cancer incidence in rhesus monkeys . The incidence of of adiposity on survival risk is sexually dimorphic and changes adenocarcinoma or other less common neoplasms was lower in with age. CR monkeys at both locations (Fig. 6). The most common diagnosis of cardiovascular disease in live animals was mitral valve dysfunction, while valvular endocardiosis, cardiomyopathy Discussion and myocardial ﬁbrosis were detected at necropsy. Incidence of Data from both study locations suggest that the CR paradigm is cardiovascular disorders was lower in CR monkeys than controls effective in delaying the effects of ageing in nonhuman primates at UW. There was no apparent impact of diet on incidence of but that the age of onset is an important factor in determining the cardiovascular disease for monkeys at NIA; however, incidence extent to which beneﬁcial effects of CR might be induced. In the for both control and CR monkeys was lower than UW controls. UW study, reduced bodyweight, reduced adiposity and reduced Assessment of glucoregulatory function was part of routine food intake of the CR monkeys were associated with improved 8 NATURE COMMUNICATIONS | 8:14063 | DOI: 10.1038/ncomms14063 | www.nature.com/naturecommunications –1 –1 –1 Glucose (mg dl ) Glucose (mg dl ) Glucose (mg dl ) NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14063 ARTICLE survival, with CR monkeys of both sexes surviving longer than Control controls, B28 and B30 years of age for males and females CR respectively, and longer than the median age for monkeys in captivity (B26 years of age). Although an impact of CR on Average survival was not detected within the NIA old-onset cohort, onset comparison to the UW study shows that bodyweight was signiﬁcantly lower in both control and CR groups of males and females than in their UW control counterparts, and was largely equivalent to that of UW CR. All males and females from the NIA UW old-onset groups consumed fewer calories than their counterpart controls from UW, instead both control and CR were closely aligned with food intake values of UW CR. Importantly, the 10 20 30 median survival estimates for old-onset males were very high, Age in years similar to what has been reported previously as the 90th 100 percentile for this species (B35 years of age). Six of the original 20 monkeys have lived beyond 40 years of age, the previous Control 80 maximal lifespan recorded, and one old-onset CR male monkey is CR currently 43 years old, which is a longevity record for this species. Average 60 onset Median survival estimates for old-onset females, B27 and B28 for controls and CR respectively, were also greater than national median lifespan estimates, with one remaining female currently 38 years of age. The clear beneﬁt in survival estimates for 20 NIA monkeys within the old-onset cohort compared to UW controls suggests that food intake can and does inﬂuence survival. The lack of difference between control and CR old-onset monkeys 10 20 30 suggests that a reduction in food intake beyond that of the Age in years controls brings no further advantage. The minimum degree of restriction that confers maximal beneﬁt in rhesus monkeys has b Cancer Ins. resistance not yet been identiﬁed but is an active topic of investigation. 50 50 60 Taken together, data from both UW and NIA studies support the concept that lower food intake in adult or advanced age is 40 40 associated with improved survival in nonhuman primates. 30 The interpretation of the outcomes from NIA J/A cohorts is more complicated. The J/A cohorts reveal a sex-dependence in 20 20 the relationship between bodyweight and food intake that was not observed in the UW cohorts. The NIA J/A CR males consumed 10 10 signiﬁcantly fewer calories than controls for most of the study, and this was reﬂected in a signiﬁcantly lower bodyweight and a 0 0 0 modest reduction in adiposity. Bodyweight of the NIA J/A UW NIA UW control males was in between that of UW controls and CR Cardiovascular disorder Diabetes monkeys even though caloric intake was equivalent between UW 50 15 and NIA control animals, and age-adjusted average adiposity of NIA J/A males was equivalent to that of UW CR males. Median survival estimates for the NIA J/A male controls and CR were not statistically different at B29 and B26 years, respectively, but estimates for the J/A controls were numerically equivalent to UW CR. These data suggest that in the J/A male cohort, the NIA diet 20 20 5 is associated with lower bodyweight, lower adiposity and improved survival in the absence of CR, and that no further 10 10 advantage is gained by lowering food intake or bodyweight below that of controls. A different picture emerges from the data from 0 0 0 UW NIA NIA NIA J/A females. The CR monkeys consumed signiﬁcantly less Control food than controls for much of the study, but there was no CR apparent difference in bodyweight or adiposity between them. Figure 6 | Morbidity curves for monkeys at NIA and UW shown. Median survival estimates for control and CR J/A females were (a) Graphs represent the ﬁrst occurrence of any age-related disease, not statistically different from each other at B26 and B23 years disorder or condition for combined males and females from UW (top) respectively, and each were lower than that observed for UW CR. and NIA J/A (bottom). Statistics related to this ﬁgure are provided in The lack of response in bodyweight to differences in food intake Supplementary Information, Supplementary Table 4. (b) Incidence of in the female monkeys suggests that there is sexual dimorphism prevalent age-related conditions in nonhuman primates for control and in the relationship between food intake and bodyweight in young CR animals from UW and NIA (J/A and old-onset combined). To compare animals and in how these parameters relate to median survival. studies, cancer and cardiovascular disorders are reported as incidence upon Finally, there is a suggestion that nonhuman primates differ from necropsy and are expressed as a percentage of the animals that are rodents in the suitability of early onset CR. In rodents, early onset deceased. CR is more effective in extending longevity than adult onset CR. For nonhuman primates it appears that CR, while beneﬁcial when implemented in adulthood, does not improve survival when NATURE COMMUNICATIONS | 8:14063 | DOI: 10.1038/ncomms14063 | www.nature.com/naturecommunications 9 Percent incidence Percent incidence Percent without Percent without age-related morbidities age-related morbidities Percent incidence Percent incidence ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14063 implemented in juveniles. Furthermore, it appears that CR is not including cancer, cardiovascular disease and parameters uniformly tolerated in young animals, where the NIA J/A CR associated with diabetes. A lower incidence of cancer was one males and females reached 80% mortality earlier than controls. of the ﬁrst health beneﬁts of CR documented and is considered to Although the numerical values for estimated survival were lower be a hallmark of CR in rodents . The incidence of cancer was in CR groups than in controls, the impact of juvenile onset CR on lower in CR monkeys at both locations indicating that tumour survival remains undetermined, as the mortality curves for both suppression is a conserved feature of mammalian CR. CR also NIA J/A cohorts are incomplete. lowered the incidence of cardiovascular disorders at UW, and Data from both NIA and UW studies highlight sex-dependent NIA monkeys from either diet group appear to have been differences in nutritional modulation of body composition and protected compared to UW Control monkeys. The data shown glucoregulatory function, and prompt further investigation of are based on pathologies identiﬁed at necropsy so that ﬁnal sex-speciﬁcity in the connection between nutrition and disease incidence of both disorders at NIA has yet to be determined. The risk. In rodent studies sexual dimorphism in the response to impact of CR to sustain glucoregulatory function during ageing diet-induced metabolic dysfunction has been reported where was a further commonality between the studies. Monkeys males were more vulnerable than females to increased adiposity presenting with insulin resistance at UW were clinically treated and glucoregulatory dysfunction . It is perhaps not surprising to prevent transition to diabetes, prohibiting a direct comparison then that the separation in fasting glucose curves between UW between the studies. Although the diagnosis employed at the controls and CR was observed in males but not females, and that two study locations represent different time frames in disease females, despite being on the same diet as males, were not progression, the outcome of lower incidence of insulin resistance different in glucose levels between control and CR until after at UW and lower incidence of diabetes at NIA are very much 20 years of age. Although both NIA and UW diets contain in agreement. Given the obvious parallels between human equivalent relative amounts of carbohydrate, the sucrose content and rhesus monkey, it seems highly likely that the beneﬁcial of the puriﬁed UW diet is six times higher than the naturally effects of CR would also be observed in humans. Reports from sourced NIA diet, and the fat content of the UW diet is twice that the multicenter CALERIE study of short-term CR in of the NIA diet. The implication is that elevated sucrose and humans document changes in bodyweight, body composition, increased fat content could lead to increased adiposity, negatively glucoregulatory function and serum risk factors for impacting glucoregulatory function. All the NIA J/A monkeys cardiovascular disease in response to CR (refs 42–47). These maintained adiposity at less than 25% and were in large part outcomes in humans align well with reports on rhesus monkey euglycemic through middle age. As might be predicted, the UW CR (refs 48–51), conﬁrming that the primary response to CR is control males and females had signiﬁcantly greater adiposity conserved between these two species, and suggesting that the compared to NIA controls. In the case of UW males, increased underlying mechanisms may also be conserved. adiposity was signiﬁcantly associated with higher fasting glucose In conclusion, the NIA and UW nonhuman primate ageing levels; however, the association did not hold for the UW female and CR studies address a central concept of relevance to human monkeys where despite differences in adiposity between control ageing and human health: that the age-related increase in disease and CR, fasting glucose levels were equivalent through middle vulnerability in primates is malleable and that ageing itself age, and multilevel analysis failed to identify a signiﬁcant presents a reasonable target for intervention. The last two decades association between adiposity and fasting glucose. Furthermore, have seen considerable advances in ageing research in short-lived there is the suggestion that the effect of diet composition is species and investigations of the mechanisms of CR have been dependent on intake levels. The UW CR male monkeys, on the prominent in this work. It will be particularly informative to same diet as controls but under food limiting conditions, had low determine the degree to which consensus hallmarks of ageing 52,53 adiposity and normal fasting glucose. These data demonstrate described in recent publications also manifest in primate that the relationship between adiposity and glucoregulatory ageing. The tissues and longitudinal data stored over the course of impairment is sex-speciﬁc, and the relationship between dietary these two highly controlled monkey studies present a unique composition and glucoregulatory impairment is dependent on the resource that can be used to identify key pathways responsive level of food intake. The signiﬁcant relationship between to CR in primates, to uncover primate-speciﬁc aspects of adiposity and survival detected in the UW female cohort also the basic biology of ageing, and to determine molecular basis points to sex-dimorphism in the impact of age and the inﬂuence for nutritional modulation of health and ageing. Processes of body composition on mortality risk. This relationship was impacted by CR would be prime targets for the development of identiﬁed for UW females only, presumably because there was a clinical interventions to offset age-related morbidity, and greater range in levels of adiposity among controls and CR for the identiﬁcation of factors involved in the mechanisms of CR will UW females than for NIA females where adiposity of monkeys be pivotal in bringing these ideas to clinical research and human from either diet at either age of onset did not differ. The positive health care. relationship between adiposity and survival in UW females was identiﬁed only after correcting for age and diet, suggesting that Methods the interactions among health, body composition and morbidity Animal care. Speciﬁcs of housing and animal care have been described in detail 20,21 elsewhere . Brieﬂy, all animals are maintained according to the provisions are sensitive to age and, given the lack of association in the male of the Animal Welfare Act of 1966 (Public Law 89–544), plus its subsequent cohort, sex-dependent. amendments, as well as the standards set forth in the document entitled ‘The Guide The catalogue of pathologies identiﬁed in aged monkeys is for the Care and Use of Laboratory Animals’ (NIH Publication No. 85–23). shared with aged humans. The deﬁnitions used to identify Animals rooms are maintained atB21 C and 50–65% humidity. Room lighting is morbidity were determined by veterinary staff and were automatically controlled on a 12-h light, 12-h dark schedule. Animals are monitored daily for general health by dedicated animal care and research staff and essentially equivalent at both sites. A shared feature of both routinely by veterinary staff. Studies at UW and NIA were conducted in accordance studies is the beneﬁcial effect of CR in lowering the risk for age- with protocols approved respectively by the University of Wisconsin–Madison related morbidity by more than two-fold. Factors contributing to Graduate School and National Institute on Aging Institutional Animal Care and these analyses include a range of conditions that are highly Use Committees. prevalent in human geriatric populations, such as sarcopenia, osteoporosis and arthritis. The beneﬁcial effects extended to Statistical analysis. The mortality analysis included all animals with known diseases that are among the most prevalent in human clinical care diagnoses or cause of death prior to July 2015. For time-to-event outcomes, the 10 NATURE COMMUNICATIONS | 8:14063 | DOI: 10.1038/ncomms14063 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14063 ARTICLE three most common methods were used: Kaplan-Meier product-limit method; 21. Mattison, J. A. et al. Impact of caloric restriction on health and Cox proportional hazard regression and parametric survival analysis assuming survival in rhesus monkeys from the NIA study. Nature 489, 318–321 Weibull distributions . Analyses were separated by site (UW; NIA), age-of-onset (2012). and sex. For both Kaplan-Meier and Cox proportional hazards (PH) regression, 22. Weindruch, R. & Walford, R. L. Dietary restriction in mice beginning at 1 year diet group (CR or control) was used as a predictor to test the effect of CR on of age: effect on life-span and spontaneous cancer incidence. Science 215, mortality and morbidity (the onset of age-associated disease). Cox models were 1415–1418 (1982). also used to estimate the HRs and 95% CIs. For the mortality analyses, age at death 23. National Research Council, Committee on Animal Nutrition, Agricultural or current age for censored values was used as the time variable. Estimated median Board. Nutrient requirements of nonhuman primates (National Academy of and mean lifespans were calculated using both Kaplan-Meier analysis and Sciences, 1978). parametric survival analysis assuming a Weibull distribution. Bodyweight and food 24. Lane, M. A. et al. Dietary restriction in nonhuman primates: progress report on intake data were analysed using mixed linear models to account for dependency in the NIA study. Ann. NY Acad. Sci. 673, 36–45 (1992). the observations due to repeatedly measuring each animal over time. In the 25. Kemnitz, J. W. et al. Dietary restriction of adult male rhesus monkeys: design, morbidity analyses, the age at which the animal experienced its ﬁrst age-associated methodology, and preliminary ﬁndings from the ﬁrst year of study. J. Gerontol. diagnosis was used as the time variable. Animals that had had not experienced an 48, B17–B26 (1993). age-associated disease but died of an age-related cause were not censored and their 26. Clarke, M. R. & O’Neil, J. A. Morphometric comparison of Chinese-origin age at death was used as the time variable. For the censored cases where the and Indian-derived rhesus monkeys (Macaca mulatta). Am. J. Primatol. 47, monkeys were alive and had not experienced an age-associated disease, current age 335–346 (1999). was used as the time variable. For censored cases where monkeys had died but had 27. Kanthaswamy, S. et al. Development and validation of a SNP-based assay for not experienced an age-associated disease, age at death was used as the time variable. For all Cox models, the PH assumption was tested by ﬁrst ﬁtting a inferring the genetic ancestry of rhesus macaques (Macaca mulatta). Am. J. Primatol. 76, 1105–1113 (2014). non-PH Cox regression with a CR by time interaction. The CR by time interactions were not signiﬁcant in any model; thus, PH models were considered valid. All 28. Rhesus Macaque Genome, S. et al. Evolutionary and biomedical insights from analyses were performed in SAS 9.3. The statistical signiﬁcance level was set at the rhesus macaque genome. Science 316, 222–234 (2007). a¼ 0.05 for all analyses with no adjustments . Multilevel modelling was 29. Colman, R. J. & Anderson, R. M. Nonhuman primate calorie restriction. conducted to explore relationships among parameters. Antioxid. Redox Signal. 14, 229–239 (2011). 30. Allison, P. D. Survival Analysis Using SAS: A Practical Guide 2nd edn (SAS Institute, 2010). Data availability. The data sets analysed during the current study are available 31. Hadﬁeld, R. M. et al. Risk factors for endometriosis in the rhesus monkey from the corresponding authors on reasonable request. (Macaca mulatta): a case-control study. Hum. Reprod. Update 3, 109–115 (1997). 32. Mattison, J. A., Ottinger, M. A., Powell, D., Longo, D. L. & Ingram, D. K. References Endometriosis: clinical monitoring and treatment procedures in Rhesus 1. Gibbs, R. A. et al. Evolutionary and biomedical insights from the rhesus Monkeys. J. Med. Primatol. 36, 391–398 (2007). macaque genome. Science 316, 222–234 (2007). 33. Kemnitz, J. W. Calorie restriction and aging in nonhuman primates. ILAR j. 52, 2. Zimin, A. V. et al. A new rhesus macaque assembly and annotation for 66–77 (2011). next-generation sequencing analyses. Biol. Direct 9, 20 (2014). 34. Kirkland, J. L. & Peterson, C. Healthspan, translation, and new outcomes for 3. Bowden, D. M. & Williams, D. D. Aging. Adv. Vet. Sci. Comp. Med. 28, 305– animal studies of aging. J. Gerontol. Ser. A, Biol. Sci. Med. Sci. 64, 209–212 341 (1984). (2009). 4. Uno, H. Age-related pathology and biosenescent markers in captive rhesus 35. Hansen, B. C. & Bodkin, N. L. Heterogeneity of insulin responses: phases macaques. Age (Omaha) 20, 1–13 (1997). leading to type 2 (non-insulin-dependent) diabetes mellitus in the rhesus 5. Hudson, J. C., Baum, S. T., Frye, D. M., Roecker, E. B. & Kemnitz, J. W. monkey. Diabetologia 29, 713–719 (1986). Age and sex differences in body size and composition during rhesus monkey 36. Rodriguez, N. A. et al. Clinical and histopathological evaluation of 13 cases of adulthood. Aging 8, 197–204 (1996). adenocarcinoma in aged rhesus macaques (Macaca mulatta). J. Med. Primatol. 6. Pandya, J. D. et al. Decreased mitochondrial bioenergetics and calcium 31, 74–83 (2002). buffering capacity in the basal ganglia correlates with motor deﬁcits in a 37. Uno, H. et al. Colon cancer in aged captive rhesus monkeys (Macaca mulatta). nonhuman primate model of aging. Neurobiol. Aging 36, 1903–1913 (2015). Am. J. Primatol. 44, 19–27 (1998). 7. Ramsey, J. J., Laatsch, J. L. & Kemnitz, J. W. Age and gender differences in body 38. Laakso, M. & Kuusisto, J. Insulin resistance and hyperglycaemia in composition, energy expenditure, and glucoregulation of adult rhesus monkeys. cardiovascular disease development. Nat. Rev. Endocrinol. 10, 293–302 J. Med. Primatol. 29, 11–19 (2000). (2014). 8. Ngwenya, L. B., Heyworth, N. C., Shwe, Y., Moore, T. L. & Rosene, D. L. 39. Dailey, G. New strategies for basal insulin treatment in type 2 diabetes mellitus. Age-related changes in dentate gyrus cell numbers, neurogenesis, and Clin. Therapeut. 26, 889–901 (2004). associations with cognitive impairments in the rhesus monkey. Front. Syst. 40. Polewski, M. A. et al. Plasma diacylglycerol composition is a biomarker of Neurosci. 9, 102 (2015). metabolic syndrome onset in rhesus monkeys. J. Lipid Res. 56, 1461–1470 9. Masoro, E. J., Yu, B. P. & Bertrand, H. A. Action of food restriction in delaying (2015). the aging process. Proc. Natl Acad. Sci. USA 79, 4239–4241 (1982). 41. Grove, K. L., Fried, S. K., Greenberg, A. S., Xiao, X. Q. & Clegg, D. J. 10. Weindruch, R., Walford, R. L., Fligiel, S. & Guthrie, D. The retardation of aging A microarray analysis of sexual dimorphism of adipose tissues in in mice by dietary restriction: longevity, cancer, immunity and lifetime energy high-fat-diet-induced obese mice. Int. J. Obes. (Lond) 34, 989–1000 (2010). intake. J. Nutr. 116, 641–654 (1986). 42. Fontana, L. et al. Calorie restriction or exercise: effects on coronary heart 11. Weindruch, R. H. & Walford, R. L. The retardation of aging and disease by disease risk factors. A randomized, controlled trial. Am. J. Physiol. Endocrinol. dietary restriction (Charles C Thomas, 1988). Metab. 293, E197–E202 (2007). 12. Masoro, E. J. Caloric restriction. Aging 10, 173–174 (1998). 43. Heilbronn, L. K. et al. Effect of 6-month calorie restriction on biomarkers of 13. Gems, D. & Partridge, L. Genetics of longevity in model organisms: debates and longevity, metabolic adaptation, and oxidative stress in overweight individuals: paradigm shifts. Annu. rev. physiol. 75, 621–644 (2013). a randomized controlled trial. JAMA 295, 1539–1548 (2006). 14. Guarente, L. Mitochondria—a nexus for aging, calorie restriction, and sirtuins? 44. Lefevre, M. et al. Caloric restriction alone and with exercise improves CVD risk Cell 132, 171–176 (2008). in healthy non-obese individuals. Atherosclerosis 203, 206–213 (2009). 15. Houthoofd, K. & Vanﬂeteren, J. R. Public and private mechanisms of life 45. Racette, S. B. et al. One year of caloric restriction in humans: feasibility and extension in Caenorhabditis elegans. Mol. Genet. Genomics 277, 601–617 (2007). effects on body composition and abdominal adipose tissue. J. Gerontol. Ser. A, Biol. Sci. Med. Sci. 61, 943–950 (2006). 16. Kennedy, B. K., Steffen, K. K. & Kaeberlein, M. Ruminations on dietary 46. Redman, L. M. et al. Effect of calorie restriction with or without exercise on restriction and aging. Cell. Mol. Life Sci. 64, 1323–1328 (2007). body composition and fat distribution. J. Clin. Endocrinol. Metab. 92, 865–872 17. Mair, W. & Dillin, A. Aging and survival: the genetics of life span extension by dietary restriction. Annu. Rev. Biochem. 77, 727–754 (2008). (2007). 18. Bodkin, N. L., Alexander, T. M., Ortmeyer, H. K., Johnson, E. & Hansen, B. C. 47. Weiss, E. P. et al. Improvements in glucose tolerance and insulin action induced by increasing energy expenditure or decreasing energy intake: Mortality and morbidity in laboratory-maintained Rhesus monkeys and effects a randomized controlled trial. Am. J. Clin. Nutr. 84, 1033–1042 (2006). of long-term dietary restriction. J. Gerontol. Ser. A, Biol. Sci. Med. Sci. 58, 212–219 (2003). 48. Edwards, I. J. et al. Caloric restriction lowers plasma lipoprotein (a) in male but 19. Colman, R. J. et al. Caloric restriction delays disease onset and mortality in not female rhesus monkeys. Exp. Gerontol. 36, 1413–1418 (2001). rhesus monkeys. Science 325, 201–204 (2009). 49. Gresl, T. A. et al. Dietary restriction and glucose regulation in aging rhesus 20. Colman, R. J. et al. Caloric restriction reduces age-related and all-cause monkeys: a follow-up report at 8.5 yr. Am. J. Physiol. Endocrinol. Metab. 281, mortality in rhesus monkeys. Nat. Commun. 5, 3557 (2014). E757–E765 (2001). NATURE COMMUNICATIONS | 8:14063 | DOI: 10.1038/ncomms14063 | www.nature.com/naturecommunications 11 ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14063 Additional information 50. Rezzi, S. et al. Metabolic shifts due to long-term caloric restriction revealed in Supplementary Information accompanies this paper at http://www.nature.com/ nonhuman primates. Exp. Gerontol. 44, 356–362 (2009). naturecommunications 51. Yamada, Y. et al. Long-term calorie restriction decreases metabolic cost of movement and prevents decrease of physical activity during aging in rhesus Competing ﬁnancial interests: R.W. is a member of the board of LifeGen Technologies, monkeys. Exp. Gerontol. 48, 1226–1235 (2013). a company focused on nutritional genomics. G.S.R. is Chief Executive Ofﬁcer of 52. Kennedy, B. K. et al. Geroscience: linking aging to chronic disease. Cell 159, GeroScience, Inc. and Vice President of Prolongevity Technologies. D.K.I. serves as Chief 709–713 (2014). Scientiﬁc Ofﬁcer for GeroScience, Inc., and Prolongevity Technologies, Inc. D.B.A. serves 53. Lopez-Otin, C., Blasco, M. A., Partridge, L., Serrano, M. & Kroemer, G. The on the board of IKEA and has received consulting fees from multiple government, not for hallmarks of aging. Cell 153, 1194–1217 (2013). proﬁt, and for proﬁt organizations with interests in obesity and nutrition. None of these 54. Saville, D. J. Multiple comparison procedures—the practical solution. Am. Stat. activities beneﬁt directly from this research and no competing ﬁnancial interests are 44, 174–180 (1990). declared. The remaining authors declare no competing ﬁnancial interests. Reprints and permission information is available online at http://npg.nature.com/ Acknowledgements Thanks to animal care and research staff at both facilities including Scott Baum, Julie reprintsandpermissions/ Adriansjach, Saverio Capuano, Casey Fitz, Heather Simmons at UW, and Rick Herbert, How to cite this article: Mattison, J. A. et al. Caloric restriction improves health Edward Tilmont, Kelli Vaughan, Mark Szarowicz, Amanda Evans, Danielle Sedlak, Mark and survival of rhesus monkeys. Nat. Commun. 8, 14063 doi: 10.1038/ncomms14063 Bryant and Matt Starost at NIA for their tireless efforts and stalwart commitment over (2017). the years. Funding for the UW study was provided by NIH grants P01AG011915, R01AG040178, R01AG037000, and the Department of Medicine, School of Medicine and Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in Public Health UW Madison. This publication was made possible in part by NCRR/ORIP published maps and institutional afﬁliations. grants P51RR000167/P51OD011106 to the Wisconsin National Primate Research Center, University of Wisconsin-Madison and the use of resources and facilities at the William S. Middleton Memorial Veterans Hospital, Madison, WI. The internet Primate This work is licensed under a Creative Commons Attribution 4.0 Aging Database is funded by contract HHSN-263-2013-00026C from the NIA. The NIA International License. The images or other third party material in this study was supported in part by the Intramural Research Program of the NIH, National article are included in the article’s Creative Commons license, unless indicated otherwise Institute on Aging. in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. Author contributions To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ J.A.M., R.J.C., J.W.K., G.S.R., D.K.I., R.W., R.d.C. and R.M.A. devised the approach, long-term statistical support was provided by D.B.A. and T.M.B., T.M.B. conducted the statistical analysis, J.A.M., R.J.C., R.d.C. and R.M.A. wrote the paper. r The Author(s) 2017 12 NATURE COMMUNICATIONS | 8:14063 | DOI: 10.1038/ncomms14063 | www.nature.com/naturecommunications

Journal

Nature Communications – Springer Journals

Published: Jan 17, 2017

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Caloric restriction improves health and survival of rhesus monkeys

Caloric restriction improves health and survival of rhesus monkeys

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Caloric restriction improves health and survival of rhesus monkeys

Caloric restriction improves health and survival of rhesus monkeys

References (64)

Abstract

Journal

Recommended Articles

There are no references for this article.

Our policy towards the use of cookies