Characterizing the phenotypic changes associated with aging in a short-lived primate is necessary in order to develop better translational models for human health, aging, and disease research. A population of conventionally housed marmoset monkeys was assessed to determine if phenotypes of body composition, hematology, and morphometrical measures were associated with age or risk of death. We found that the cause of mortality in older marmosets was more likely to be due to cardiac and chronic kidney disease than in younger marmosets. Older marmosets have decreased fat mass, morphometric measures, and serum albumin. Older marmosets are more likely to show a modified posture while at rest and this modified posture was significantly associated with an increased risk of imminent death. These assessments provide an initial definition of aged health in marmosets and a base for future translational aging research with this species.
The relationships between health, aging, tissue function, and disease in primates and humans is often not well-modeled by rodent studies [
Marmosets are small new world primates that offer a valuable resource as an animal model to examine adult disease risk, aging, and functional decline because they have the shortest average lifespan and fastest reproduction of any anthropoid primate [
Studies of health and physical condition in humans have often focused on a core group of characters including body weight, body composition (fat and lean mass), activity patterns, inflammation, adiposity, gait, overall strength, muscle size, muscle strength, appetite, social function, and cognitive function [
The study population was chosen from the conventional breeding colony of
Seventy-nine male marmosets ranging in age from young adulthood (2 years) to 13.7 years, the maximum lifespan in the SNPRC colony, and 39 nonpregnant females ranging in age from 2 to 10.7 years of age were examined for body composition (body mass, fat mass, and lean mass). In order to assess body composition animals were fasted overnight, weighed, and removed from the home cage via capture in a nest box. Body composition was analyzed using a qualitative magnetic resonance machine (EchoQMR) which has previously been validated for use in the marmoset to assess lean and fat mass [
Additionally, for 46 animals in the colony multiple body composition measurements were made between 1 and 3 years following the first assessment. The average time between assessments was 2.3 years. For this data set there were 19 females ranging in age from 2.2 to 7.3 years of age and 27 males ranging in age from 2.6–8.3 years of age. For this data the change in age and the change in body mass variables were analyzed using correlations in SPSS.
A subset of fifty males was further assessed for morphometrics, basic hematology, CRP concentration, and activity (scored during behavioral observations). For this data collection 2 mL of blood was gathered via the femoral vein from fasted animals. The blood was then sent to the SNPRC clinical pathology laboratory for basic hematology and CRP measurement. Morphological measures were taken using calipers including thigh-knee length and knee-heel length. A measuring tape was used to measure proximal thigh, medial thigh, distal thigh, proximal calf, medial calf, distal calf, and abdominal circumference. All measurements were made in triplicate and then averaged. Home cage behavior data was collected over two thirty-minute periods randomly scheduled over the course of the study using the observer software (Noldus). Males were observed using all occurrences methods and were scored as active (leaping, or quadrupedal motion) or inactive (sleeping, sitting, or stretching). Stretching was defined as a stationary adjusted posture in which the animal supports its weight with two limbs, typically the forelimbs, and the rest of the body is extended. The durations for all behaviors were averaged across observations. These data were then analyzed in SPSS to assess the correlations between the variables and the age, weight, fat, and lean mass of the subjects. In order to evaluate whether the variables measured were associated with a risk of death, the time from measurement to death for each individual was entered into a Cox regression model in SPSS. None of the subjects took part in other experimental or terminal protocols following the measures, and the data were censored at 2.7 years following the initial measure.
The New England Primate Center reported that pathologies and causes of mortality differ by age for marmosets. In animals less than 6 years of age principal causes of death were trauma, inflammatory bowel disease, sepsis, and bacterial infection. In aged animals common causes of death included neoplasia, chronic renal disease, amyloidosis, and diabetes [
Causes of mortality for marmosets at SNPRC 2002–2011 under the age of 6 years (a), and over the age of 6 years (b). IBD: inflammatory bowel disease, iatrogenic: death from a complication associated with anesthesia.
For male marmosets, age was positively associated with the absolute lean mass (
Mass varies in marmosets by both age and gender.
Fat mass
Lean mass
While age and fat mass were not found to be significantly correlated for males, many human population studies find that morphometric and hematological values can be associated with both age and obesity status [
Correlations for aging phenotype variables and age, or fat and lean mass (*partial correlation controlling for age) for 50 male marmosets.
Variable | Age | Fat* | Lean* |
---|---|---|---|
Proximal thigh | |||
Abdominal circ | |||
Albumin | |||
Red blood cell | |||
Hemoglobin | |||
Hematocrit | |||
CRP | |||
Behavior: Rest | |||
Behavior: Stretch |
For male marmosets albumin concentrations are found to be negatively correlated with age (a), CRP is correlated with fat mass (b), and albumin is positively correlated with proximal thigh circumference (c).
Age
CRP concentration
Proximal thigh circumference
In order to determine whether any of the aging phenotype variables were predictive of death within 24 months following the study, data from 47 of the 50 animals that were originally assessed were entered into a Cox regression survival analysis. Three of the original animals were culled and were excluded from the hazards analysis. Twenty-five animals died naturally in the defined time frame, five died within 6 months of the study, 11 more had died within 12 months of the study, and a further 9 animals had died within 24 months following the study; the remaining 22 animals were censored for the analysis. The Cox regression survival analysis was done with the entry method, specifically looking at the variables defined above while controlling for age (Table
Cox regression survival analysis for aging phenotypes in marmosets, *all variables added to the model by entry controlling for age.
Variable | B | Sig ( | Hazard |
---|---|---|---|
Age | 0.293 | 0.000 | 1.341 |
Proximal thigh* | −0.252 | n.s. | 0.777 |
Albumin* | −0.794 | n.s. | 0.452 |
Red blood cell* | 0.248 | n.s. | 1.282 |
Hemoglobin* | −0.461 | n.s. | 0.63 |
Hematocrit* | 0.117 | n.s. | 1.125 |
Fat* | −0.002 | n.s. | 0.998 |
Lean* | −0.015 | n.s. | 0.985 |
Behavior-Stretch* | 0.208 | 0.012 | 1.231 |
The ability to rapidly and easily measure a large number of hematological, morphological and body compositional variables in a small nonhuman primate makes the marmoset an ideal model for future studies of aging and health span. In this study we identified a number of markers that are associated not only with aging but also with risk of death. We found that aged marmosets were more likely to suffer from cardiac and renal failure than were younger marmosets. Rather than lean mass declining with age as would be predicted for sarcopenia-associated changes, in fact the marmosets’ lean mass increases with age while fat mass is lost. Serum albumin was found to be significantly lower in aged marmosets than in young marmosets and was also associated with morphological measures of thigh circumference; this is analogous to results described for elderly humans. While CRP concentrations were not associated with age, they were positively associated with the modified posture noted as stretching, which was the only factor to significantly predict the risk of death when age was removed from the model. Perhaps animals displaying this adjusted posture are suffering from increased discomfort and inflammation. Frailty in humans is often associated with increased inflammatory status and decreased motor abilities; in fact one of the behavioral phenotypes used to evaluate frailty is the ability of the patient to rise from a seated position without modifying their posture [
This paper was the first examination of phenotypes specifically to examine the relationship with age and health status in marmosets. This initial descriptive data is necessary for the further model development and initiation of aging research in marmosets. The results regarding body composition and hematological values and their relation to age closely resemble those previously reported in aging rodents and human populations. Analyses of the relationships among these variables provide operational definitions of “health” in this species that can be related back to human studies and provide a future bridge to examine preventatives, interventions or therapeutics. It will also provide a firm base to which additional phenotypic domains, such as cognitive function or cardiac function, can easily be added in the future.
The authors would like to thank Donna Layne-Colon for her continued dedication to the maintenance of the colony of marmosets at SNPRC and their exceptional care. They thank Joselyn Artavia and Katie Lewis for their help with the data collection for this project. Finally, they thank the anonymous reviewer for insights that enhanced this work. Funding for this project was received from the Keck Foundation Healthspan Grant and the University of Texas Health Science Center San Antonio CTSA IIMS Pilot Project Grant.