The adrenal steroids, dehydroepiandrosterone (DHEA) and its sulfate (DHEAS), have attracted attention for their possible antiaging effects. DHEAS levels in humans decline markedly with age, suggesting the potential importance of this parameter as a biomarker of aging. Here we report that, as seen in humans, male and female rhesus monkeys exhibit a steady, age-related decline in serum DHEAS. This decline meets several criteria for a biomarker of aging, including cross-sectional and longitudinal linear decreases with age and significant stability of individual differences over time. In addition, the proportional age-related loss of DHEAS in rhesus monkeys is over twice the rate of decline observed in humans. Most important is the finding that, in rhesus monkeys, calorie restriction, which extends life span and retards aging in laboratory rodents, slows the postmaturational decline in serum DHEAS levels. This represents the first evidence that this nutritional intervention has the potential to alter aspects of postmaturational aging in a long-lived species.
DIETARY CALORIC restriction (CR) is the most robust and reproducible means of slowing aging, reducing age-related diseases, and extending life span in short-lived species, but biological mechanisms of this nutritional intervention and its relevance to species with maximal life spans greater than 3 yr (especially humans) remain unknown (1, 2). Numerous age-related biological changes are altered by CR. For example, CR reduces oxygen radical damage (3), alters gene expression (4), increases defenses against stresses such as heat shock (5), alters glucose fuel use (6), and retards rates of age-related change in many physiological processes (1, 2).
Since 1987, we have sought to determine whether 30% CR is capable of exerting its antiaging effects in nonhuman primates (7). Other investigators (8, 9) have begun to address this question, as well. A number of physiological effects already have been reported at various biological levels that are consistent with CR studies in rodents. For example, rhesus monkeys, subjected to CR, exhibit delayed sexual (10) and skeletal (11) maturation, altered glucoregulation (9, 12, 13), and reduced body temperature (14). These findings show that physiological responses to CR are strikingly similar in rodents and primate species. However, they do not conclusively show that CR alters basic processes of aging consistent with life span extension, as reported in rodents.
Among the most abundant steroids in the body, dehydroepiandrosterone (DHEA) and dehydroepiandrosterone sulfate (DHEAS) recently have received much attention regarding possible protection against the decrements of aging. Elevated serum levels of DHEAS may be protective against several diseases associated with advancing age, including diabetes (15), heart disease (16), and cancer (17). Hormone therapy using these steroids reportedly also has beneficial effects, such as prevention of obesity (18), reduction of cholesterol (19), and increased resistance to tumors in rodents (20). Having observed elevated levels of DHEA and DHEAS in a subset of young CR monkeys earlier in our study (10, 21), we report here the first evidence that the normal age-associated decline in DHEAS represents a potential biomarker of aging that is slowed by long-term CR, at least during the first half of the life span of a primate species.
A total of 45 male and 29 female rhesus monkeys (Macaca mulatta), in our longitudinal study of aging and CR in nonhuman primates, were used in the present study. To study the effects of normal aging on serum DHEAS, we used control monkeys in three age groups. At the start of the study, groups consisted of 17 young (9 male and 8 female, 1–2 yr), 21 young adults (9 male and 12 female, 3–5 yr), and 18 old (9 male and 9 female, > 17 yr) monkeys. For CR studies, we used 18 male monkeys (8 young and 10 young adult) that had been on CR for 3–6 yr at the time data reported here were collected. All monkeys, except one old male, were born in captivity and had known dates of birth. Housing and feeding conditions were as previously described (7, 11). All monkeys ate the same monkey chow that was supplemented with extra vitamins and minerals to prevent malnutrition in the CR group. Monkeys were fed 2 meals each day at 0700 and 1400 h. Controls were fed approximately ad libitum, based on National Research Council guidelines (22), and CR monkeys received an allotment 30% less than that offered control monkeys of a similar age and body weight.
DHEAS was assayed in a series of blood samples collected during years 1–3 in females and years 3–6 in males. Blood samples were collected under ketamine (7–10 mg/kg, im) or Telazol (3.5 mg/kg, im) anesthesia after an overnight fast. Serum was separated by centrifugation and stored frozen at −80 C for DHEAS analysis. DHEAS was assayed using a competitive-binding RIA in kit form from Diagnostic Products Corporation (Los Angeles, CA).
The relationship between mean annual DHEAS values and chronological age for female (during years 1–3) and male (during years 3–6) ad libitum-fed control rhesus monkeys is summarized in Fig. 1. The maximum life span of rhesus monkeys in captivity is estimated to be 40 yr (23). This presentation allows the graphical depiction of cross-sectional and longitudinal changes, both required to demonstrate the validity of biomarkers of aging (24). It is apparent that during each year, DHEAS levels were generally lower in older, compared with younger, monkeys of both sexes. The cross-sectional age-related decline in DHEAS was confirmed by significant correlations with chronological age calculated for each year for female (P values < 0.01) and male rhesus monkeys (P values < 0.05). r values ranged from 0.39 to 0.71.
Longitudinal changes in DHEAS levels also can be observed in Fig. 1. Focusing on young female monkeys, it is apparent that DHEAS levels generally decline with each successive year. A similar pattern emerged in older female and in male rhesus monkeys. The longitudinal decline was confirmed for all age groups by a 3 (age) by n (measurement occasion) ANOVA and by confirming the longitudinal effect within each age group (ANOVA). All longitudinal comparisons were significant (P values < 0.05).
It is further evident in Fig. 1 that rhesus monkeys, like humans (25, 26), exhibit substantial individual variability in DHEAS at all ages. To assess the interyear reliability of DHEAS values in rhesus monkeys, we computed the correlation between each annual value for an individual monkey and the corresponding annual value of the subsequent year. Stable individual differences were confirmed for both female (r = 0.55–0.61, P < 0.01) and male (r = 0.69–0.81, P < 0.001) monkeys, indicating that the variability observed was generally reliable.
Figure 2 compares the percentage reduction from maximal DHEAS levels in humans and rhesus monkeys over much of their respective life spans. Rhesus monkey raw data were subdivided into age categories to allow direct comparison with the human data. The average DHEAS values in each age category are shown. Maximal DHEAS levels for each gender/species are shown in the inset graph. Figure 2 shows that DHEAS values exhibit a marked age-related decline in both species; and the rate of decline, as indicated by the slopes of the regression lines, is about 2–2.5 times higher in rhesus monkeys, compared with humans, consistent with an approximately 3-fold greater life span of the latter. This species difference in the rate of decline was confirmed by a significant difference in slopes between humans and rhesus monkeys of both genders (P values < 0.05). Examining the inset, it is apparent that maximal DHEAS levels in rhesus monkeys are significantly lower than human levels. In addition, females of both species have lower DHEAS, compared with males. The gender effect was confirmed in rhesus monkeys using a Student’s t test for maximal DHEAS (t = 4.2, P < 0.002).
Figure 3 summarizes the average (\ sem) DHEAS value at a given age for all young adult male monkeys over the course of the study. It is clear that the rate of decline in serum DHEAS is slowed in adult monkeys subjected to long-term CR, compared with age-matched controls. On average, DHEAS declined more than 30% over 4 yr in adult male controls. In male rhesus monkeys subjected to CR, the decline (3%) was markedly less over the same time period. The CR effect was confirmed by a significant difference in the slope of the regression calculated for controls, compared with the slope determined for CR monkeys (t = 3.8, P < 0.005).
Our findings demonstrate the possible utility of serum DHEAS levels as a biomarker of aging in primate species and that CR, a nutritional intervention that extends life span in short-lived species, slows the age-related decline in serum DHEAS in young adult male rhesus monkeys. Serum DHEAS, as a biomarker of aging, is further validated by comparing our findings with humans, in which the rate of decline is much slower (consistent with the differences in life span between rhesus monkeys and humans). The slowing of the normal rate of age-related decline in DHEAS levels in rhesus monkeys provides evidence that aging rate, as indexed by adrenal steroid production, may be retarded by CR.
The age-related decline in serum DHEAS in humans is widely recognized and forms the theoretical basis for potential antiaging effects associated with this hormone. However, the utility of DHEAS as a biomarker of aging in humans has not been evaluated logically, because no previous study has followed the approach outlined in this study. Nonetheless, a review of several human studies shows that serum DHEAS satisfies our criteria for candidate biomarkers of aging (24). In humans, serum DHEAS exhibits both cross-sectional (25, 26, 27) and longitudinal (28, 29) age-related declines. Moreover, despite significant individual variability, interindividual differences remain stable over time (28, 29). These reports, combined with the present results in monkeys, provide strong supportive evidence for serum DHEAS levels as a reliable biomarker of aging.
In the present study, the decline in serum DHEAS levels met our minimum criteria for candidate biomarkers of aging. Specifically, we observed that both cross-sectional and longitudinal declines with age and individual differences remained stable across time. A further validation for DHEAS, or any biomarker, would be to demonstrate differences in rates of change between species consistent with differences in species-specific life span. Our findings (Fig. 2) show that the rate of decline (percent per year) is significantly greater (average, 17.4%) than the rate of change in humans (average 8.3%). These rates are consistent with the nearly 3-fold difference in life span and further validate DHEAS as a biomarker of aging across primate species. Few reports of DHEAS levels during aging in other primate species could be found. In one study, Sapolsky et al. (30) reported cross-sectional age-related declines between 4 and 15 yr of age in a wild population of male and female baboons. Given the limited data available, additional studies are needed to confirm the potential importance of DHEAS as a biomarker of aging in other primate species.
Antiaging and antidisease effects of CR are well documented in laboratory rodents (for review, see Refs. 1 and 2). However, a reliable assay for rodent DHEAS is not available, and as such, the effect of CR on DHEAS in rodents is not known. Studies of weight loss in obese individuals have shown that serum DHEAS levels increased about 125% after a 2-month weight-loss diet in men but not women (31). These authors suggested that insulin is a physiologic regulator of DHEA metabolism and that the increased DHEAS levels might be related to a significant (38%) reduction in insulin levels reported in the same study. Interestingly, several previous studies have shown that fasting insulin is significantly reduced in male rhesus monkeys on CR (9, 12, 13). The role of insulin as a possible physiologic regulator of DHEAS and its relationship to the age changes reported here requires further study.
In summary, the present findings from our longitudinal study of aging and CR in rhesus monkeys demonstrate the possible importance of serum DHEAS as a biomarker of aging in primates, including humans. Moreover, our data show that CR slows the postmaturational decline in serum DHEAS levels. As such, these findings offer the first evidence that the beneficial antiaging effects of CR are not limited to short-lived species and add to an evergrowing body of evidence suggesting that CR induces physiological changes in nonhuman primates (for review, see 32 consistent with results in rodents in which aging rate has been slowed by CR. These findings strengthen the possibility that CR might exert beneficial antiaging effects in long-lived species.
Figure 1. DHEAS declines with age in female and male rhesus monkeys. Each point represents the serum DHEAS level for each monkey at a given age. Legends show r values for cross-sectional comparisons for each year on study (*, P < 0.05; **, P < 0.01). Longitudinal correlations also were significant for both groups of monkeys (P values < 0.05). Figure 2. Rate of serum DHEAS decline in humans (adapted from ref. 26) and rhesus monkeys. Each point represents the percentage reduction from maximal DHEAS levels in a given age group. The inset graph shows the maximal DHEAS levels. *, Maximal DHEAS levels were significantly reduced in female, compared with male rhesus monkeys (P < 0.05). Figure 3. CR slows the rate of decline in serum DHEAS. Each point represents the mean (±sem) DHEAS level at a given age. Ages represent the average (± 0.3 yr) age of young adult male rhesus monkeys for years 3–6 of this longitudinal study. Slopes indicating the rate of decline were calculated from the regressions shown for each monkey group (Control = −3.9 μg/dL per year and Calorie Restricted =− 0.4 μg/dL per year). The rate of change was significantly slower in CR monkeys (P < 0.005).
View Larger
View Larger
View LargerWe recognize the continuing excellence in veterinary care provided by our support staff, especially Edward Tilmont and Lauren Johnson. We also thank Dr. John Sorkin for statistical consultation. Primate Service Poolesville, Veterinary Resources Program, NCRR is fully accredited by the American Association for Accreditation of Laboratory Animal Care, and all procedures described herein were conducted in accordance with NIH guidelines and received full approval of the Animal Care and Use Committee of the Gerontology Research Center, National Institute on Aging, National Institutes of Health.
