Marmoset Metabolism, Nutrition, and Obesity

Abstract

The use of marmosets as nonhuman primate models of human disease has undergone rapid expansion in the United States in the last decade, with an emphasis in the field of neuroscience. With this expanding need, there has been an increase in the formation of small marmoset colonies. The standardization in care and husbandry techniques for marmosets has historically lagged behind other established nonhuman primate models, resulting in a great deal of variation in practices between colonies. There remains a lack of consensus and evidence-based recommendations regarding best standards for nutrition, enrichment, weight management, and diagnostics for clinical metabolic disease. Marmoset base diets vary broadly in their nutritional value, and therefore the physiological responses to these diets also vary broadly. In this review, we briefly outline what is known about nutrition for captive marmosets and highlight what is known regarding metabolic dysfunction and obesity.

MARMOSET NUTRITION AND METABOLISM

The demand for common marmosets in biomedical research has increased at an extraordinary rate over the past 10 years due to its emergence as a critical model organism in a variety of biomedical disciplines.^1,2 This increased demand has been most acute in neuroscience but is also associated with the development of the marmoset as a model for aging, the continued use of marmosets in infectious disease studies, and increasing interest in the development of transgenic lines of marmosets.^2–4 The increased interest in marmosets as models has grown in part due to recognition of inherent limitations of rodent models in translational studies of human physiology and disease. Among non-human primate (NHP) models, the marmoset provides unique practical advantages, including small size, easy handling, rapid reproductive maturation, high fecundity, compressed life cycle, and social behaviors and communication that more closely resemble those observed in humans.^5,6 Like other NHPs, moreover, the closer genetic similarity of marmosets to humans has great translational value.⁷

There remains variability in the types of nutrition provided to marmosets in captive settings. Diets not only vary in their macronutrient composition (carbohydrates, fat, and proteins) but also in their vitamin and mineral content. The base diets that are fed across colonies also differ broadly in the format in which they are delivered (canned, purified, extruded pellet, mash, or gelled). Often these base diets make up less than the recommended 75% of the daily nutrient intake in marmosets due to the longstanding and continued practice of feeding marmosets a variety of food items in large amounts, either as enrichment or as supplements thought necessary for complete nutrition.⁸ Items range from sources of protein such as canned chicken and tuna, insects, or cheese; to sources of carbohydrates in fruits, vegetables, and legumes; to sources of vitamins such as citrus fruit or fruit drinks. These approaches, often referred to in the literature as a cafeteria-style diet, result in high variation between animals in nutrient intake. This practice persists due to the underlying belief that this variety is needed to ensure adequate food intake and prevent the loss of marmosets due to rapid weight loss. Although the daily variability in the diet may more closely resemble what the primates encounter in natural conditions, it is not based on scientific knowledge regarding daily nutritional intake patterns in the wild. The variability presented to the animals may in fact not only contribute to unnecessary variation in experimental outcomes but may also play a role in some of the most common clinical diseases seen in this species in captivity.

Digestive Disease and Obesity: Paradoxical Common Concerns in Captive Marmosets

Inflammatory bowel disease (IBD) has been a historical limitation in developing the common marmoset as a laboratory animal for biomedical research. Initially, the disease was described as causing colitis; however, data indicate that the disease can become much more severe in the small intestine and has been termed chronic lymphocytic enteritis, or CLE.⁹ The terminology to describe these symptoms varies greatly in the literature to include terms such as marmoset wasting syndrome, colitis, IBD, chronic diarrhea, and CLE.^10–13 The variability in terminology and described pathological changes most likely reflects a variety of diseases with similar presentations and different etiologies. Regardless of presentation, the cause has remained elusive, although treatment to remove specific pathogens such as Giardia from primate colonies had some success in decreasing the prevalence.⁹ The disease is typically characterized by chronic diarrhea and progressive weight loss that leads to decreases in serum albumin.^9,11 Additional secondary findings include osteoporosis, anemia, and ill-thrift. Histologically, CLE is characterized by altered small intestinal morphology with marked lymphocytic infiltrates into the lamina propria accompanied by crypt cell hyperplasia, villous atrophy and fusion, and increased numbers of intraepithelial lymphocytes.^9,14 The frequency of IBD/CLE varies substantially from colony to colony; however, because the disease can be asymptomatic, it has the capacity to compromise a wide variety of marmoset studies.¹⁵ Subclinical gut inflammation may alter effective captive management and research use by resulting in reduced digestive efficiency, calcium homeostasis, and especially lipid malabsorption.^16,17 In addition to health concerns due to intestinal inflammation, the invisible nature of subclinical cases raises concern regarding the potential for significant unaccounted for variation in research studies using common marmosets. This is of particular importance for research investigating pharmacokinetics using oral dosing, but the potential for marmosets to vary widely in nutritional status, for example in the fat-soluble vitamins (A, D, E, and K), also could result in significant variability in metabolism and physiology.¹⁶ Despite the considerable concern regarding CLE and intestinal tract health in marmosets, the underlying causes of this affliction are poorly, if at all, understood; however, multiple hypotheses have been offered, including (a) altered intestinal microbiome; (b) dietary disruptions, including sensitivity to gluten; (c) nutritional inadequacy of captive diets; (d) stress of captivity; (e) immune-mediated disease; and (f) combinations of multiple causes. It is clear that IBD remains a common problem across captive colonies, and the current dietary husbandry regimes for marmosets have not eliminated the risk for digestive dysfunction. It remains unclear whether IBD occurs in wild marmosets and the impact it may have on those animals.

Obesity, paradoxically, is a common affliction in marmoset colonies in stark contrast to the weight loss and cachexia observed in marmosets with progressive IBD. Obesity accompanied by dyslipidemia and altered glucose metabolism^18,19 has emerged as a frequent clinical finding in marmosets. At the Southwest National Primate Research Center (SNPRC), obesity is highly prevalent, particularly in sub-adult marmosets ranging between 1 and 2 years of age, with 46%–52% of animals in this age category demonstrating obesity (body fat >14%) and reduced insulin sensitivity.^19,20 With the increasing prevalence of obesity in marmoset colonies, we now have 2 groups of marmosets at higher risk for mortality: those of low weight and those of high weight. Obesity has generated increases in pathologies rarely seen previously in marmosets, such as hepatomegaly, hepatic steatosis, diabetes, atherosclerosis, cardiomyopathies, and stroke.^18,19 Although these conditions generate considerable interest in marmosets as a model of obesity and metabolic syndrome, this value is limited by our incomplete understanding of what has led to the increases in spontaneous obesity observed in so many marmoset colonies.

Establishing the correct nutritional guidance for the maintenance of captive marmosets is critical to their use as biomedical models. The ability to provide diets with the most appropriate nutrient content to enhance health and maintain homeostasis in the marmoset will increase overall colony health and reduce physiological variability between colonies and between individuals. The prevalence of diseases ranging from irritable bowel disease to obesity in current captive colonies suggests that we have not yet determined the correct nutritional management practices for the animals. In this paper, we will briefly review what is known regarding marmoset nutrition in the wild and in captivity (please see a recent thorough review⁸) and focus on metabolic dysfunction and obesity, which are an emerging problem for captive management of marmosets.

Nutrition in the Wild

Common marmosets are omnivores with a diverse diet in the wild. In Atlantic coastal forests as well as the drier Caatinga habitat, they consume fruit, exudates, and nectar from over 2 dozen plant species.²¹ Their plant diet is balanced with hunting and consuming insects and small vertebrates, with insects presumed to be the primary source of protein in their diet. Their consumption of tree exudates is related to many of their unusual evolved features, including dentition specialized for tree gouging, small size, and claw-like nails that allow for vertical clinging. Gum-based exudates can account for over one-half of their feeding time during certain parts of the year; however, gums appear to be a fallback food, because gum consumption drops considerably during times of high fruit and prey abundance.²² For a detailed discussion of the wild diet of marmosets, consult Arruda et al.²³

Knowledge of the diet of wild marmosets can be informative regarding appropriate nutrient balance and, more importantly, how certain types of foods and feeding opportunities might support a range of natural behaviors in captivity. However, there is no reason to propose that captive marmosets must be fed a diet that specifically mimics the food items they consume in the wild; see Power and Koutsos⁸ for a more thorough description of this topic.

Nutrition in the Captive Setting

Although the common marmoset has been kept in captivity relatively successfully for over 50 years, there is no established standardization of dietary components. Most marmoset colonies feed a commercially available base diet to provide basic nutritional needs, but these base diets vary significantly. A recent comparison of the diets fed at the 2 National Primate Research Centers that house marmosets in the United States revealed significant differences in the nutrition and physiological responses to the diet. The base diet with the highest digestibility fed at the SNPRC was associated with the lowest rate of suspected intestinal inflammation, but it was also associated with the highest prevalence of obesity.¹⁶ The diet fed at the Wisconsin National Primate Research Center was associated with lipid malabsorption in more than one-half of the animals studied. Although the base diets fed at the 2 NPRCs differ in composition, the percent of the total diet constituted by the base diet is similar, as is the enrichment received by the animals. Thus, neither of the base diets being used appears to be an appropriate solution for marmoset dietary husbandry.¹⁶

Captive Marmoset Weight

In an attempt to understand what experienced marmoset investigators believe represents normal body weight in captive common marmosets, we sent a survey to 39 institutions known to house common marmosets for biomedical research. Thirty US institutions and 9 international institutions were included in the contact list. The survey consisted of 29 questions requesting information on current colony census, weighing practices, diets used, and opinions regarding limits for healthy, obese, and underweight determinations. In addition, respondents were asked to provide current weight data for their colony.

Completed surveys were received from 14 institutions, equaling a 36% response rate. Of these, 6 were from non-US institutions. Weight data were received from 16 institutions, equaling a 41% response rate. Similarly, 6 of the respondents were from non-US institutions. Of the 14 institutions returning a completed survey, 2 did not return weight data. Of the 16 institutions that sent weight data, 4 did not return a completed survey. Overall, 18 (46%) institutions returned some information in response to our request.

Colony size of responding institutions ranged from 10 to 542 marmosets (n = 7, <50 animals; n = 1, 50–100 animals; n = 1, 100–200 animals; n = 5200+ animals). All colonies weighed animals at least once every 6 months, with some weighing as frequently as weekly. A range of diets was used from full cafeteria-style diets to purified diets. Respondents expressed a range of opinions regarding what constitutes a healthy weight range in marmosets. All responses fell into the following ranges: adult (2–8 years of age) males, 300–500 g; adult females, 250–500 g; older (>8 years of age) males, 325–450 g; and older females, 300–450 g. Responses regarding weight limits for obesity and underweight are shown in Tables 1 and 2. Figure 1 represents actual data received in response to our request for colony weight information. Average weights (mean ± SEM) for submitted data were: adult males 414.7 ± 2.9 g, adult females 427.8 ± 3.9 g, older males 398.0 ± 3.9 g, and older females 423.3 ± 7.9 g. Analysis of variance with Tukey post-hoc comparisons showed that adult females weighed more (P < .05) than adult males or older males. No other group comparisons were statistically different. Adult females were not known to be pregnant at the time of weight assessment.

Table 1.

What Constitutes an Obese Marmoset

Body Weight (g)	Adult Males	Adult Females	Older Males	Older Females
>350	0%	9%	0%	0%
>400	8%	0%	13%	0%
>450	33%	36%	50%	57%
>500	58%	55%	38%	43%

Values are % of respondents.

Table 2.

What Constitutes an Underweight Marmoset

Body Weight (g)	Adult Males	Adult Females	Older Males	Older Females
<250	8%	18%	13%	13%
<300	42%	27%	38%	38%
<325	33%	36%	38%	38%
<350	17%	18%	13%	13%

Values are % of respondents.

Figure 1 .

Box and whisper plot of actual weight data sent in response to survey request. Box represents interquartile range. Line within box represents median. Whiskers represent range. Adult = 2–8 years of age. Older = 8+ years of age. Adult males, n = 370; adult females, n = 325; older males, n = 57; older females, n = 50. Adult females were not pregnant at time of weight assessment.

Overall, the results of the survey reveal that the lack of a standardized definition of underweight and overweight weight ranges has resulted in a great deal of disparity in how colonies are defining animals potentially at risk. Although we did not directly survey how colony managers and veterinarians are intervening when animals are deemed to be at risk, one could imagine that this implementation would be quite varied when the definitional ranges are more than 100 g different for both underweight and obesity. This 100-g range in definition is approximately 25% of the weight of these animals. One of the goals of this survey was to be able to define normal weight ranges across colonies and to work towards a standardization of clinical weight ranges. Although these data suggest that adult females in captivity tend to be larger than males, the range in weights is substantial. Additionally, although a comparison between body weights and diets fed is of interest, from this small survey-based study, given the wide variety in diets used, we were not able to determine any relationship between the different diets and body weight.

METABOLIC DYSFUNCTION AND OBESITY

Adult Marmoset Obesity

The occurrence of spontaneous obesity in captive marmoset populations and pathological sequelae has been reported.^18,19 Obese adult marmosets have dyslipidemia and altered glucose metabolism with higher fasting glucose concentrations and HbA1c levels.¹⁸ Marmosets fit the human operational definition for metabolic syndrome with atypical values for body weight, body composition, high-density lipoprotein cholesterol (HDL) concentrations, fasting glucose and average HbA1C, or triglyceride concentrations.¹⁸ Marmoset obesity has also been found to be associated with prolonged hyperglycemia, pancreatic islet hyperplasia, and atherosclerotic lesions.¹⁹ Marmosets develop hepatic steatosis with progression towards nonalcoholic steatohepatitis associated with increased lipid deposition and insulin resistance.²⁴ Changes in cholesterol values or blood pressure associated with obesity have not yet been reported for marmosets and need to be further evaluated.

The underlying etiology of obesity in marmoset colonies remains unclear, but it is hypothesized to be associated with an imbalanced caloric intake and output. There are a limited number of studies in which food or energy intake has been measured in marmosets. Marmoset dry matter food intake can be assessed using a 48-hour home cage feeding trial in which all food is weighed in, and dry weight of all waste food is used to determine dry matter intake. Caloric intake is then estimated based on the known caloric values of the basic diets. In a digestion study of individually housed, young adult marmosets (n = 13, 2–5 years old), intake averaged 13.3 g/d dry weight and 50.5 kcal/d and was not related to body mass (ranging from 330 to 434 g, 383 g average).²⁵ An evaluation of older females (n = 4, 7–9 years of age; mean body mass = 420 g) exhibited almost identical results, with average dry matter consumption of 12.9 g/d and 55.14 kcal/d. A more extensive study (n = 81, 2–10 years of age, 316–631 g) found that digestible energy intake, which accounts for difference in digestive ability, was related to body mass but that at any given body mass digestible energy intake could vary widely between individuals.¹⁶ The lack of a definitive relationship between body mass and intake in these studies suggests that, in adult animals, body weight variation is unlikely to be solely explained by variation in food intake.

Wachtman et al¹⁹ fed marmosets diets enriched for fat or glucose to evaluate the impact on weight gain in adult animals. Marmosets fed diets higher in glucose (32% glucose vs 4.5%) had a significant increase in body and fat mass after 16 weeks, whereas marmosets fed a diet higher in fat (22.2% of calories from fat vs 10.4%) displayed no consistent increase in fat mass.¹⁹ The animals exhibited a high inter-individual variability with only 12.5% of animals on high-fat and 33% of animals on high-glucose diets presenting impaired metabolic function after diet manipulation. An understanding of the sources of this variation could provide insights into what features of energy intake influence obesity risk and might also aid in appropriate subject selection and monitoring while on study.

The propensity for obesity in marmosets may be associated with ingestive behaviors, including preferences for high-glucose food items and a tolerance to higher dietary fat, or variability in ingestive patterns. To define ingestive behaviors in marmosets and potential risk for animals to develop obesity, a series of experiments was conducted at SNPRC with all procedures reviewed and approved by the Texas Biomedical Research Institute’s Animal Care and Use Committee. To elucidate whether individuals at risk for obesity were more likely to ingest items that were high in fat, 42 adult common marmosets (21 males and 21 females) were given a dietary choice test and then enrolled on a chronic high-fat diet. The dietary choice test was conducted to determine whether individual responses to a high-fat diet could be predicted, as has previously been demonstrated in rat strains.²⁶ Naïve subjects were presented with dairy products in 2 bottles identical to those used as water bottles. One bottle contained non-fat dairy milk (0.35 kcal/g) and the other contained a high-fat dairy product (Half and Half, 1.3 kcal/g). On the 2 test days, the subject’s regular food was removed and they were provided with 6 hours of exposure to the dairy products. The location of the bottles was randomized among subjects on the first presentation then reversed. The weight and volume consumed were measured at the end of each presentation. Following the choice, test the animals were placed on a fat-enhanced version of their base diet (purified Harlan Teklad marmoset diet) with the addition of lard at 12% dry weight, increasing caloric density to 5.01 kcal/g from 4.47 kcal/g and increasing calories from fat to 33% from 18% for 11 months; both diets were equivalent for sugar content. At the initiation of the study, the subjects were, on average, 4.44 years old (±1.49 SD) and weighed 431.3 g (±48.38 SD). Subjects were weighed monthly, and their body composition including lean and fat masses was determined at the beginning and the end of the high-fat diet challenge by quantitative magnetic resonance. On average, the marmosets on the higher-fat diet challenge did not change in weight or fat mass. However, although the central tendency of the population was no change, individual responses were highly variable, with some animals displaying gains of up to 99.5 g and 29.1% body fat and others showing losses up to 67 g and 16.7% (Table 3). Subjects did not differ in the amount of non-fat milk consumed during the choice test; however, there was a significant difference in the amount of high-fat product consumed (Figure 2). Specifically, those subjects that went on to lose weight during the high-fat diet challenge ate significantly less of the higher-fat dairy product (F = 3.55, P = 0.04). Subjects in this study appear to vary in their attraction to, or tolerance of, high-fat supplements. Both oral and gastrointestinal-tract systems detect and respond to fatty acids and exhibit variable responsiveness in obese-prone vs obese-resistant individuals.²⁷ Studies in both rodents²⁸ and humans²⁹ have demonstrated individual variation in ability to orally detect fatty acids, with those individuals who are more sensitive being less likely to consume high amounts of fats. It is possible that the marmosets with lower high-fat consumption were more sensitive to oral and/or gastrointestinal-tract fatty acid signals.

Table 3.

Weight Changes Following 11-Month High-Fat Diet Challenge

	Weight-Gainers	Weight Constant	Weight Losers
N	14	13	14
Starting age (y)	4.09 ± 1.51	5.07 ± 1.42	4.20 ± 1.47
Starting weight (g)	405.0 ± 37.58	447.95 ±35.61	446.73 ± 53.01
M/F ratio	8/6	4/9	7/7
Weight change (g)	39.53 ± 21.8	−5.49 ± 11.14	−45.73 ± 13.16
Weight change (%)	10.13 ± 6.65	−1.20 ± 2.51	−10.40 ± 3.36
Fat mass change (g)	24.36 ±29.63	1.49 ±19.61	−13.42 ± 17.45
Lean mass change (g)	23.05 ± 22.18	14.59 ± 20.40	−17.58 ± 16.92

Values, with the exception of M/F ratio, are means ± SD.

Figure 2 .

Mean (± SE) of grams consumed in 6 hours of a fat-free and a high-fat dairy product presented in a 2-bottle choice trial. Those subjects that subsequently lost weight while on a high-fat diet challenge consumed significantly less of the high-fat dairy product (F = 3.55, P = .04).

To conduct a finer-grained assessment of daily food intake patterns in marmosets, a Columbus Instruments DM-8 lick counter was adapted to the marmosets,³⁰ which counts the number of times an animal accesses a standard water bottle containing the liquid base purified diet with added xanthum gum as an emulsifier. Twelve subjects—6 weight gainers and 6 non-gainers—were available approximately 2 years after the end of the high-fat feeding study. At the time of the lickometer trials, weight-gainers did not differ from the non-gainers in terms of weight or age. Weight gainers, when given the lickometer, had significantly lower overall lick counts each day (F[1,13] = 12.815, P = .003) and significantly lower average counts per meal than the non-gainers (F[1,13] = 13.327, P = .003); however, the total consumption each day did not differ between weight-gainers and non-gainers (Table 4). The difference in count frequency, with no difference in overall consumption, meant that consumption per lick was higher in the weight gainers.

Table 4.

Comparison of Lickometer Phenotypes in Two Different Studies Involving Obesity-Prone Subjects vs Non–Obesity-Prone Subjects

Phenotypes (Mean)	Obesity-Prone Subjects		Non–Obesity-Prone Subjects
	Obese Juvenile	Weight Gainers	Lean Juveniles	Non-Gainers
Grams/lick	0.015	0.034	0.012	0.020
Licks/meal	98.6	53.6	119.8	104.7
Total licks	2642.48	1824.33	4684.18	3210.92

In the human literature, there is a great deal of debate regarding whether feeding behaviors associated with obesity are consistent over an individual’s lifetime and whether effective interventions may include retraining ingestive behaviors (food choices, portion size, caloric loading, timing of feeding, bite size, and rate of eating). To determine whether ingestive behaviors associated with marmoset obesity are stable over time or reflect specific responses to high-fat diet exposure, a longitudinal assessment of ingestive behaviors in 5 obese adult female marmosets was conducted. The behavioral feeding phenotypes of the females were evaluated every 6 months by using lickometer trials and solid food intake trials for 42 months. Parameters measured for the lickometer trials showed no changes over time for these females. The amount of food consumed, food choice (high fat vs low fat), grams per lick, and length or number of meals did not differ for an individual over time. The solid food intake trials also displayed little individual variance over time. The only time point found to differ significantly during the 42-month assessment was the final data point in which the females displayed a shorter lickometer meal length (18.73 seconds vs 27.32 seconds) and a slight decrease in overall solid food caloric intake. It should be noted that the females were quite a bit older at the end of this study than other animals that have been assessed for obesity and feeding behaviors in the past (7–10 years compared with 4 years old). We believe that these animals displayed stable feeding phenotypes over the course of nearly 2 years, and the decline in feeding at the last time point reflects an age-related change, as seen in humans,^31,32 rather than changes in obesity status or stability of feeding phenotypes.

Overall, these evaluations of responses to chronic higher-fat diet and long-term ingestive pattern suggest that adult marmosets can be assessed using a simple choice test to determine whether they are at higher risk for gaining weight when exposed to items of increased caloric density. These patterns also suggest that animals with a propensity towards obesity intake more food at a single bout due to a difference in bite efficiency and that these patterns are stable across the adult period. This suggests that weight maintenance for adult marmosets classified as weight gainers may be possible by reducing the caloric density of the offered food items, especially exposure to high-fat, highly palatable foods.

Development of Obesity in Infancy

Obesity in marmosets is not limited to an adult onset and presentation of metabolic dysfunction and pathology. In fact, spontaneous obesity may develop very early in marmoset life.³³ Characterization and assessment of pediatric obesity in this species have been greatly enhanced by the ease of handling mother-reared marmoset infants compared with other NHP species.³⁴ An initial examination of marmosets for whom longitudinal weight data from birth through adulthood were available¹⁸ revealed that those individuals classified as obese (using a conservative definition of obesity as those that exceeded the 90th percentile of body weight at 17 months of age) were already significantly heavier at birth, 2 months, and 6 months of age. Obesity in these growing marmosets is not simply defined as differences in weight. Changes in body composition were followed longitudinally in infants from birth through 12 months of age,³³ and obese juveniles were defined as those whose body fat exceeded 14% as measured by quantitative magnetic resonance at 12 months of age. Individuals that were obese at 12 months had already started to differentiate in fat mass from their lean counterparts by 1 month of age, although overall body weight did not differ between the groups until 6 months of age. Body fat composition differed throughout development between these 2 groups, with the lean animals displaying decreasing percent body fat between 1 and 12 months of age whereas the percentage of body fat increased across the same time frame in obese subjects (Figure 3).

Figure 3 .

The percent body fat with SEM for normal and obese animals at 1, 2, 6, and 12 months. Animals defined to be normal had body fat <14% and animals defined to be obese had body fat >14% at 12 months of age. Data from Power et al, 2012.³³

The development of obesity prior to 1 year of age is also associated with impaired metabolic function.²⁰ Obese infants were found to have significantly reduced insulin sensitivity by 6 months of age and by 12 months of age, obese juveniles challenged by an oral glucose tolerance test displayed significantly reduced ability to clear glucose. Additionally, obese individuals displayed higher fasting glucose concentrations and tended to have lower circulating adiponectin at 12 months of age.²⁰ Thus, infants that were trending towards obesity by 6 months of age were beginning to develop metabolic dysfunction and by 1 year of age displayed typical physiological markers of metabolic disease.

The relationship of dry matter and energy intake to body size was different in young, growing marmosets than what was described for adults above.³⁵ Comparing animals that were classified as normal vs obese at 12 months of age, obese individuals had higher dry matter consumption at both 6 (12.6 g vs 10.6 g) and 12 (15.9 g vs 10.7 g) months of age. As opposed to the adults, total dry mass or kcal intake of solid food was strongly related to body mass. These data suggest that variation in intake may be more closely linked to variation in mass growth than it is to maintenance of variable adult mass.

Juveniles age 3, 6, and 12 months were assessed with the lickometer system to define number of meals and duration of meals assessing inter-meal interval criteria.^30,35 No significant differences were found between animals that were determined to be obese vs non-obese at 12 months of age in any of the meal patterning variables typically noted to vary in rodent models. An unexpected and consistent behavioral difference between obese and lean individuals (Table 4) was the total number of licks per meal and per day was significantly lower in the obese subjects, whereas the total intake did not differ. This result indicates that obese infants and juveniles were taking a larger volume per lick of the sipper tube each time they accessed the tube rather than accessing the licksit more frequently or consuming more calories overall. This behavior may be equivalent to the propensity in obese humans to consume larger bites of food rather than increased prevalence of feeding or increased caloric intake.^35,36

Evidence from the examination of obesity development in young marmosets has revealed significant, positive, within-individual correlations between lick counts, grams per meal, grams per lick, and total liquid diet consumption at feeding trials conducted at 3, 6, and 12 months of age.³⁵ Solid food intake was also consistent within an individual over time. Taken together, these results suggest that many of the features measured in ingestive trials, including lickometer and solid food consumption for young marmosets, reflect traits that remain consistent within an individual and act as a marker of their obesity status, propensity towards obesity, and individual feeding style. The ability to assess young marmosets for these markers may allow early detection and intervention to prevent (or induce for study) obesity and the related pathologies.

Physiological Response to Feeding

Work tying behavioral phenotypes to possible endocrine controls has just begun to be explored in marmosets due to the previous limited availability of assays for relevant hormones. Recent assays validated for the marmoset include hormones associated with obesity and food intake such as ghrelin, adiponectin, leptin, and insulin.^20,37 During the validation trials for these hormones, marmosets were found to have a significant negative correlation between body fat and circulating adiponectin concentrations. There was a positive correlation between body weight and leptin values. Finally, low-weight marmosets had significantly higher ghrelin concentrations than high-weight animals.³⁷ With the expansion of these assays, the assessment of these hormones in relation to feeding behaviors and changes in adiposity can now fully begin.

To evaluate postprandial responses in marmosets, 12 adult mixed-sex subjects (6 weight-gainers and 6 non-gainers) from the high-fat study described above were evaluated in a bleed-feed-bleed study. The animals were fasted for 18 hours, a plasma sample was collected, they were allowed access to food for 4 hours, and after this feeding period a final plasma sample was collected. Glucose and triglyceride concentrations were determined using a Unicel DxC 600 (Beckman Coulter). We predicted that weight-gainers would have higher circulating triglycerides in both fasting and refeeding due to decreased clearance via fat oxidation pathways. Weight-gainers and non-gainers did not differ in fasted circulating concentrations of glucose or triglycerides. Those animals that had previously gained weight while on a higher-fat diet displayed an increase in triglyceride concentrations following feeding (Figure 4). Non-gainers generally displayed a modest decline in circulating triglycerides following feeding, with 67% (4/6) subjects showing a decline. In contrast, 83% (5/6) of weight-gainers displayed an increase in triglycerides following feeding, with some subjects showing a relatively large post-feeding increase (Mann Whitney U-test = 7.50, P = .09).

Figure 4 .

The change in circulating triglyceride concentration (post – pre value) following feeding in subjects that gained weight vs those that did not gain weight while on the high-fat diet challenge. The box represents in interquartile range and the line through the box the median value.

Rats susceptible to weight gain on a high-fat diet were less able to up-regulate fat oxidation. They showed increased circulating triglycerides after an 18-hour fast, and when fed a ¹⁴C-labelled fat source produced lower amounts of labelled CO₂ and had greater incorporation of label into adipose tissue.^38,39 Although in our study circulating triglycerides did not differ after an 18-hour fast, the refeeding results are consistent with those subjects classified as responders to a high-fat diet being less able to up-regulate fat oxidation. Both feeding and fasting will increase circulating triglycerides, which will then leave the blood stream via deposition into adipose tissue or oxidation. A poor ability to up-regulate fat oxidation would result in slower clearance of triglycerides and thus higher circulating levels a few hours after feeding, as seen in our study. Our failure to detect a difference after fasting may indicate that 18 hours is not the appropriate time frame for detecting this difference in marmosets. In rats, whereas the difference in fasting triglycerides was significant at 18 hours, it was borderline at 12 hours and was no longer different at 24 hours.³⁹ The availability of these hormone assays will allow further evaluation of physiological responses to feeding, allowing us to make better recommendations regarding diet, feeding, and maintenance to avoid the complications associated with obesity.

WHERE DO WE GO FROM HERE?

There remains a continued critical need for scientific evaluation of marmoset physiological responses to provided nutrition to determine whether there are more appropriate nutritional sources for marmosets in captivity. Further studies are needed to elucidate the diet consumed by wild marmosets to determine critical dietary components that may be integrated into a captive diet. A deeper understanding of feeding choices, feeding behavior, and energetic output in wild marmosets will help to guide further standardization in husbandry and care for captive marmosets. Additionally, studies are needed to determine whether husbandry practices and changes in enrichment that promote activity are able to mitigate the impacts of obesity currently seen in marmoset colonies. The studies outlined above are a starting point to determine standards of care for marmosets, but many of the studies are impacted by variables including small sample size and variability in husbandry between colonies. Standardization is necessary for nutritional management, supplementation and enrichment, weight management, and diagnoses of disease. Although we are beginning to understand variables that are associated with poor digestibility or the propensity of obesity, continued systematic evaluations of responses of the animals to diet are necessary. The ability to design species-appropriate diets and standardized nutritional and husbandry recommendations will reduce variability that contributes to variation in experimental outcomes as well as contributing to the most common clinical diseases seen in this species in captivity.

CONCLUSIONS

Marmosets are a valuable translational biomedical model for a variety of research topics, and a continued expanding demand for these animals in the future is predicted. Current standards of care and understanding of nutritional needs continue to lack the necessary standardization to ensure health and reduce variability across colonies. Irritable bowel disease and obesity remain a concern for captive colonies, and etiologies for these diseases remain unknown. Investigations to evaluate the development of obesity suggest that marmosets spontaneously develop obesity and there are a number of variables that predict the risk of becoming obese, including preference for high-fat milk and lickometer sip size. A great deal of work remains to be done to determine the mechanisms that promote, and therefore could be used to prevent, obesity and metabolic disease in captive marmoset colonies.

Corinna Ross, PhD, is an Associate Professor in the Department of Life Sciences at Texas A&M University, San Antonio, Texas, USA; and Population Health at Texas Biomedical Research Institute, San Antonio, TX, USA.

Ricki J. Colman, PhD, is an Assistant Professor in the Department of Cell & Regenerative Biology in the School of Medicine and Public Health at the University of Wisconsin-Madison, Wisconsin, USA.

Michael Power, PhD, is an Animal Scientist and Curator of the NZP Milk Repository, Nutrition Laboratory, Center for Species Survival, Smithsonian National Zoological Park and Conservation Biology Institute, Washington, DC, USA.

Suzette Tardif, PhD is professor emeritus at the Southwest National Primate Research Center, Texas Biomedical Research Institute, San Antonio, Texas, USA.