Feeding History and Obese-Prone Genotype Increase Survival of Rats Exposed to a Challenge of Food Restriction and Wheel Running




We hypothesized that obese-prone genotype and history of food restriction confer a survival advantage to genetically obese animals under environmental challenge. Male juvenile JCR:LA-cp rats, obese-prone and lean-prone, were exposed to 1.5 h daily meals and 22.5-h voluntary wheel running, a procedure inducing activity anorexia (AA). One week before the AA challenge, obese-prone rats were freely fed (obese-FF), or pair fed (obese-PF) to lean-prone, free-feeding rats (lean-FF). Animals were removed from protocol at 75% of initial body weight (starvation criterion) or after 14 days (survival criterion). AA challenge induced weight loss in all rats, but percent weight loss was more rapid and sustained in lean-FF rats than in obese-FF or obese-PF animals (P < 0.04). Weight loss was significantly higher in obese-FF rats than obese-PF rats, 62% of which achieved survival criterion and stabilized with zero weight loss. Obese-PF rats survived longer, on average (12.0 ± 1.1 day) than obese-FF (8.2 ± 1.1 day) and lean-FF rats (3.5 ± 0.2 day) (P < 0.02). Wheel running increased linearly in all groups; lean-FF increased more rapidly than obese-FF (P < 0.05); obese-PF increased at an intermediate rate (P < 0.02), and those rats that survived stabilized daily rates of wheel running. Prior food restriction of juvenile obese-prone rats induces a survival benefit beyond genotype, that is related to achievement of homeostasis. This metabolic adaptive process may help explain the development of human obesity in the presence of an unstable food environment which subsequently transitions to an abundant food supply.


Excess adiposity is linked to immediate and long-term health risks, including increased risk of cardiovascular disease, type 2 diabetes, and increased middle-age mortality (1,2). Genetic differences between individuals explain a major proportion of the within-population variation in BMI (3). However, genetic susceptibility alone does not result in obesity without specific environmental influences (4). A positive energy balance results in energy storage as fat, body mass increase and/or growth. In this sense, obesity can be viewed as a result of physiological dysfunction, or as a natural response to the environment. This adaptive hypothesis is an extension of the “thrifty gene” theory of obesity. The theory explains how genes favoring the efficient use of energy stores in periods of feast and famine result in obesity in the food-rich environment of prosperous societies (5).

A major problem for living organisms is to maintain energy balance when faced with environmental challenges, such as food depletion or famine (6). Such challenges activate a complex interplay of stress, metabolic and motivational processes that initiate and maintain food-related behavior and homeostasis (7). These processes, however, have limitations—a fact demonstrated when rodents are placed on time-limited daily meals (1–2 h) and provided with unlimited access to running wheels (22–23 h). Under these conditions, animals reduce food intake, lose body weight and escalate wheel running, essentially becoming decompensated. The excessive physical activity further suppresses food consumption, body weight plummets, and animals die of this vicious feedback cycle (activity anorexia (AA)) (8,9). In addition, genetic variations can influence an organism's responses to acute food depletion (10). Our ongoing research is based on the JCR:LA-cp rat incorporating the autosomal recessive cp gene (a Tyr763Stop mutation for the ObR leptin receptor) (11). We have shown that juvenile (40 day old) obese-prone (cp/cp) rats, lacking the ObR leptin receptor exhibit a fitness advantage over lean-prone (+/?) when exposed to the challenge of restricted food and access to wheel running (AA challenge) (12). Understanding this gene/environment interaction on neurophysiological and metabolic determinants of the survival benefit of the obese-prone genotype would provide new insights into the underlying mechanisms of obesity.

During periods of diminished food supply, obese-prone rats mobilize and use their energy stores more efficiently than lean-prone, allowing prolonged survival and food-related travel (as represented by ability to run extensively). In addition to genotype, the feeding history of an organism plays an important role in survival under the combination of food restriction and extensive food-related travel (13). In an early study of AA, prior exposure to a 1 h/day feeding schedule allowed 75% of rats to survive an AA challenge, whereas 90% of the animals without this feeding history died (14). Subsequent studies showed that prior exposure to the food schedule (1 or 1.5 h/day) decreased body weight loss, and prolonged survival, even while wheel running continued to increase over days (15). Notably, these studies used mature rats with high initial energy reserves. Adaptation to the restricted feeding seems to aid survival when animals have sufficient energy reserves, yet the physiological mechanisms underlying this effect are unclear.

Wang et al. (16) have shown that animals with an ontogenetic history of food restriction will physiologically reduce basal metabolic rate (BMR), increasing their ability to survive subsequent food shortages. This BMR adjustment to food restriction allows animals, with sufficient fat mass, to maintain behavioral functions such as foraging and travel. Another adaptation to food restriction in non-obese humans involves reduction in biomarkers of survival such as fasting insulin and core body temperature (17). Food restriction also induces deprivation stress (14) activating the hypothalamic—pituitary—adrenal axis, increasing plasma corticosterone levels (10), and hypothalamic neuropeptide-Y (NPY) expression (18). The NPY system, that is implicated in the control of food intake under conditions of food abundance, participate in the regulation of food-related behavior such as wheel running when food is scarce (19). This physiological shift from feeding to locomotion makes sense during starvation, when obtaining additional food is critical. In addition, genetic background can influence the differential use of fuels (carbohydrates, fat, and protein) to counteract an excessive body weight loss when food is scarce (20,21).

In the present study, we used juvenile JCR:LA-cp obese-prone and lean-prone rats as an experimental model, since we can manipulate ontogenetic experiences with food deprivation in rats within the same strain, but with different genotypes. We imposed a period of food restriction by pair feeding juvenile obese-prone rats to similar free-feeding lean-prone animals. We hypothesized that this prior food restriction schedule would enhance adaptation of obese-prone rats to the AA challenge, allowing for greater locomotion and longer survival. Lean-prone rats were expected to rapidly succumb to the AA challenge, as in previous studies.

Methods and Procedures


Two replications of the study were conducted. The first replicate involved 24 male JCR:LA-cp rats (obese-prone cp/cp and lean-prone +/?) ∼40 days of age, in 2 groups of 12, were obtained from the established breeding colony at the University of Alberta (Alberta, Edmonton, Canada) (22). Rats, were assigned to three experimental groups of eight, and housed individually in clear polycarbonate cages (47 cm × 27 cm × 20 cm) with sterile wood chip bedding in a temperature (22 ± 2°C) and humidity-controlled environment. Rats were kept on a 12 h light/dark reverse cycle (lights off 7:00–19:00 h) for 14 days (7 days preacclimatization and 7 days of acclimatization) before we started the AA protocol. The second replicate was run separately to obtain energy expenditure (EE) measures and involved 12 40-days-old male JCR:LA-cp rats (obese-prone cp/cp and lean-prone +/?) divided into three groups of four rats (see Energy Expenditure section). Throughout the experiment, animals had free access to water and were fed as outlined in the feeding schedule below. The care and use of animals were in accordance with Guidelines the Canadian Council of Animal Care and subject to prior review and approval by the Animal Care and Use Committee: Livestock, of the University of Alberta.

Apparatus and materials for AA challenge

Twelve Wahmann running wheels (1.1 m circumference) with metal side cages (25 cm × 15.5 cm × 12.5 cm) were used. A computer recorded the number of wheel turns in 1-min interval as previously described (23). Feeding was conducted either in home cages or in feeding cages with the same materials and dimensions as the home cages but without bedding. All food was laboratory chow (LabDiet 5010, Rodent Diet; PMI Nutrition International, Brentwood, MO). An electronic scale was used to measure food, water and body weight to the nearest gram.

Experimental procedures

One day before arrival at the experimental laboratory (last of a 7-day preacclimatization period in the colony holding room), fat and lean body mass of the rats were measured using nuclear magnetic resonance (Minispec LF90 Body Composition Analyzer; Bruker, East Milton, Ontario, Canada). Blood was taken by tail-bleed for baseline biochemical parameters. Rats were then acclimatized to the feeding schedule, over 7 days, with food intake and body weights recorded daily. Obese-prone rats were matched for body weights and randomly assigned to free-feeding (obese-FF) or pair-feeding (obese-PF) groups (8 rats per group). Pair-feeding involved feeding the obese-prone rats daily in one allocation at 7:00 h, the mean amount of food consumed by the lean-prone free-feeding (lean-FF) group. This pair-feeding induced a 33% daily caloric restriction in obese-PF rats.

On day 8 and subsequent days, all groups of rats were weighed and transferred to feeding cages at 7:00 h (onset of dark period) for 1.5-h free access to food followed by 22.5 h of access to running wheels (AA challenge). Rats were removed from the protocol when their body weight reached 75% of initial weight (starvation criterion) or after 14 days of AA challenge (survival criterion). During the AA challenge, body weight, food intake, and wheel turns were measured daily.

Postmortem analysis

When animals met the criterion for either starvation or survival (none were allowed to become moribund), they were removed from their home cages and anaesthetized with isoflurane. Blood was taken by cardiac puncture and the rats immediately perfused intracardially with ice-cold isotonic saline. Brains were removed as described below; liver and white adipose tissue (WAT) (retroperitoneal, epididymal, and subcutaneous) were dissected and weighed. Blood was collected in 10-ml polyethylene tubes-containing EDTA, and stored on ice until centrifugation. Plasma samples were stored at −80°C until analysis.

NPY-mRNA expression in hypothalamic arcuate nucleus

NPY-mRNA expression in the arcuate nucleus was assessed by in-situ hybridization as described by Richard et al. (24) due to its close proximity to the peripheral blood stream and strong response to peripheral cues informing the brain on the status of energy balance. The brains were prepared as previously described (25). Briefly, after perfusion (no >2 min), brains were removed and kept in paraformaldehyde 4%-phosphate buffer 0.2 mol/l solution for 7 days. They were then transferred to a solution-containing paraformaldehyde 4%-phosphate buffer 0.2 mol/l and sucrose (10%) (12 h) and frozen before being sectioned using a sliding microtome (Histoslide 2000; Reichert-Jung, Heidelberger, Germany). Thirty microgram thick coronal sections were taken from the olfactory bulb to the brainstem stored at −30°C in a cold sterile cryoprotecting solution-containing sodium phosphate buffer (50 mmol/l, pH 7.2), ethylene glycol (30%), and glycerol (20%). In-situ hybridization histochemistry was used to localize NPY mRNA.

Plasma biochemical analysis

Total plasma triglyceride concentrations were measured using direct colorimetric chemical enzymatic reactions (Wako Chemicals USA, Richmond, VA). Plasma glucose was determined using glucose oxidase method (Diagnostic Chemical, Charlottown, PEI, Canada Cat#220-32). Insulin and leptin levels were assessed by enzyme-linked immunosorbent assay kit for rodents (insulin kit, Mercodia AB, Uppsala, Sweden, and leptin kit; Linco Research, St Charles, MO). Plasma corticosterone was measured using Milliplex Map (Millipore, Billerica, MA).

Energy expenditure

To obtain data on metabolic rate, a replicate experiment (eight obese-prone cp/cp and four lean-prone +/?) was performed as a separate study. Obese-prone rats were matched for body weights and randomly assigned to obese-FF or obese-PF groups (four rats per group). As described above, obese-PF rats ate the mean amount of food consumed by the lean-FF group. On day 8 of the acclimatization period, all three groups of rats were submitted to 1.5-h free access to food followed by 22.5 h of access to running wheels (AA challenge) per day. After 3 days of the AA challenge, rats were transferred to metabolic cages for energy balance measures. Indirect calorimetry was performed in open-circuit Oxymax chambers, a component of the Comprehensive Lab Animal Monitoring System (Columbus Instruments, Columbus, OH). These digitally controlled metabolic chambers estimate the rat's energy metabolism using the respiratory exchange ratio (RER), the ratio of carbon dioxide production (VCO2) to oxygen consumption (VO2). Oxymax calculates the heat production in kcal/h, indicating the total EE of the animals. We then expressed this heat production in kcal/h/kg body weight0.75 adjusting for metabolic body size. During the measurement, rats were housed individually and maintained at ∼24°C with food available 1.5 h/day and no access to running wheel. Due to the running wheels dimensions, such that they can not fit in metabolic cages, we were unable to measure the EE of animals with wheel running access. All rats were acclimatized to the monitoring cages for 24 h before the automated recording of the physiological parameters for another 24 h.

Statistical analyses

Data are presented as mean ± s.e.m. Food intake, body weight, wheel turns, were analyzed by a mixed-effects ANOVA with groups (lean-FF, obese-FF, and obese-PF) as the between factor and days (acclimatization or AA challenge) as the within subjects term. We used t-tests (paired or independent samples) and one-way ANOVA to further elucidate the within subject effects and for the analyses of EE and plasma biochemical measures. The between subject effects were explored by post hoc comparisons of the means using a Tukey procedure. Regression analysis was used to compare the slopes of the rate of running with comparisons between groups made in accordance with Bailey (26). We conducted a survival analysis (Kaplan—Meier) and used the standard log-rank test (Mantel—Cox) to assess the effect of groups (lean-FF, obese-FF, and obese-PF) on survival. Alpha was set at 0.05 for all analyses. For all pairwise contrasts and multiple comparison of means tests the least significance alpha level for the several contrasts is given in the text. No data was plotted or analyzed if the number remaining was <2.


Body weight and survival

Body weights, before acclimatization, at day 0 (on exposure to the AA challenge) and on removal from the AA challenge, are shown in Table 1. For the 1-week acclimatization and pair feeding period, obese-FF rats showed a greater increase in body weight than lean-FF rats (P = 0.003) (Table 1). The mean body weights of obese-FF and obese-PF groups were not significantly different on initiation of the AA challenge by ANOVA (P = 0.056). However, a nonparametric test showed that the weights of obese-FF rats were in fact greater than those of the obese-PF rats (Mann—Whitney rank sum test, P = 0.028).

Table 1.  Physiological parameters before and after the AA challenge (n = 8 per group)
inline image

Body weight of rats during the AA challenge as a percentage of initial weight is shown in Figure 1, together with the cumulative weight loss. Over the first 3 days, when all rats remained in the protocol, body weight dropped in all groups, with the relative (%) loss being greater in the lean-FF rats (Figure 1, P < 0.04). Cumulative weight loss was significantly lower in obese-PF than obese-FF rats (P = 0.004) (Figure 1; insert).

Figure 1.

Body weight and cumulative weight loss (insert) of lean-FF, obese-FF, and obese-PF rats under AA challenge. Data are mean ± s.e.m. *P < 0.05 lean-FF vs. obese-FF; #P < 0.05 obese-FF vs. obese-PF. Weight loss was not significantly different between lean-FF and obese-FF groups over days 1 and 2, but was different on day 4. Obese-PF lost less weight than obese FF rats from day 1 (P < 0.001) and cumulative weight loss was significantly lower in obese-PF than obese-FF rats (P = 0.004). AA, activity anorexia; FF, free-feeding; PF, pair-feeding.

Figure 2 (upper panel) shows the mean absolute rate of weight loss for the groups over the period of AA challenge. At day 1, the obese-PF rats had a markedly lower rate of weight loss than the obese-FF rats (13.9 ± 0.9 vs. 31.0 ± 1.4 g/day, P < 0.001), whereas the rate of loss of the lean-FF rats was intermediate at 20.7 ± 0.9 g/day (P < 0.001). This substantial drop in body weight on day 1 of the obese-FF rats probably reflects emptying of the full gut compared to a nearly empty gut in the obese-PF group. The obese-PF rats that survived to the end of the AA challenge all achieved a zero rate of weight loss, and thus a homeostatic status. One obese-FF rat survived to day 14, meeting survival criterion, and this animal also achieved a zero rate of loss (data not shown as N = 1). Thus, survival and achievement of body weight homeostasis were perfectly related.

Figure 2.

Rate of body weight loss (upper panel) and survival plot of rats (lower panel) during the AA challenge. All rats that survived the AA challenge effectively achieved zero rate of weight loss, while none of the animals that failed to reach zero loss survived. AA, activity anorexia; FF, free-feeding; PF, pair-feeding.

The lower panel of Figure 2 shows the survival results: obese-PF rats (mean, 12.6 ± 1.21 days) lasted longer in the AA challenge than obese-FF group (8.75 ± 1.17 days) (P = 0.039); lean-FF rats (3.5 ± 0.2 days) survived the least number of days (P < 0.001). The results of the Kaplan—Meier survival analysis indicated distinct survival functions over days of AA for the three groups of rats (Figure 2; lower panel). Log-rank test showed a significant difference among the groups in survival (χ2 = 26.01, df = 2, P < 0.001). A similar analysis also indicated a significant difference between obese-FF and obese-PF rats (χ2 = 4.81, df = 1, P < 0.03). Linear regression analysis showed that weight loss on day 3 of AA, when all rats were still in the protocol, predicted survival in the AA challenge (r = −0.72, P < 0.001)—greater weight loss on day 3 is associated with shorter survival.

Wheel running

Wheel running increased linearly in all groups over the first 4–6 days of AA challenge (P < 0.001) (Figure 3; insert). There were, however, significant differences in rate of increase (as shown by the slopes of the regression lines, m/day/day; m/d/d) between groups; lean-FF (1,359 ± 334 m/day/day) and obese-FF (576 ± 123 m/day/day) rats (P < 0.04), whereas the obese-PF group (943 ± 165 m/day/day) was intermediate and not statistically different from either of the other groups. In spite of differences in rate of increase, the total distance run over the first 3 days of AA was not statistically different between the three groups; lean-FF, 10,700 ± 1,500 m; obese-FF, 8,800 ± 1,200 m; obese-PF, 9,900 ± 1,200 m (P = 0.88) (Figure 4). By day 8, when only obese-prone rats remained under protocol, obese-PF rats had higher total distance run (41,200 ± 5,070 m) than obese-FF rats (25,400 ± 2,570 m) (P = 0.01) and continued to ∼74,000 m. Most of the obese-PF rats (62.5%) continued wheel running at a steady daily rate without meeting the starvation criterion (Figure 4).

Figure 3.

Running rate of lean-FF, obese-FF, and obese-PF rats under AA challenge. Insert graph shows the first 6 days expanded for clarity. Data are mean ± s.e.m.; see Results section for statistical analysis. AA, activity anorexia; FF, free-feeding; PF, pair-feeding.

Figure 4.

Cumulative distance run by lean-FF, obese-FF, and obese-PF rats under AA challenge. Insert graph shows the first 6 days expanded for clarity. Data beyond day 11 is not shown as a minor computer malfunction resulted in some data for day 12 being not recorded. Data are mean ± s.e.m.; see Results section for statistical analysis. AA, activity anorexia; FF, free-feeding; PF, pair-feeding.

Energy intake

Caloric consumption was significantly different over the acclimatization period; obese-FF showed higher daily intake than lean-FF rats and obese-PF rats (controlled by pair feeding) (P < 0.05) (day 0, Figure 5). Throughout the AA challenge, obese-FF and obese-PF rats had similar caloric intakes (P > 0.05, Fig 5). Lean-FF rats ingested only 11.7 ± 0.7 kcal on day 1 of the protocol (P < 0.001), rising to 19.7 ± 1.4 kcal on day 3. For the first 3 days of AA (Table 1), lean-FF ingested in average 14.1± 1.1 kcal/day, compared to 23.9 ± 2.3 and 25.3 ± 2.2 kcal/day for obese-FF and obese-PF, respectively. Regression analysis shows that the lean-prone rats significantly increased caloric intake over days 1–3 (4.12 ± 0.95 kcal/day/day, P < 0.001), but this was insufficient to achieve survival. Over the last 5 days of the AA challenge, obese-FF rats did not significantly change their caloric intake, days 6–10, 0.56 ± 0.48 kcal/day/day, P = 0.26, while obese-PF rats significantly increased caloric intake, days 9–13, 2.03 ± 0.85 kcal/day/day, P = 0.025.

Figure 5.

Caloric intake of rats, before establishment in the protocol (day 0) and during the protocol. Data are mean ± s.e.m. *P < 0.05 vs. obese-prone groups. FF, free-feeding; PF, pair-feeding.

Plasma biochemical parameters

Biochemical parameters are shown in Table 1. Preacclimatization, lean-FF rats had significantly lower plasma TG (P < 0.04), leptin (P < 0.001) and glucose (P < 0.01) concentrations than the obese-FF and obese-PF rats. On the day that rats met the criterion for removal from the AA challenge, plasma TG and glucose concentrations of obese-FF and obese-PF rats did not differ significantly while lean-FF rats differed significantly from obese-FF rats (P = 0.01) for plasma TG concentration (Table 1). On removal from protocol, plasma leptin was essentially depleted in the lean-FF rats and differed significantly from both obese groups (P < 0.001). The obese-FF rats had significantly higher plasma leptin concentrations than obese-PF rats (P = 0.003).

Preacclimatization, there were no significant differences in plasma corticosterone concentrations between the three groups. On removal from protocol, lean-FF rats had higher corticosterone concentrations than both of the obese groups (P < 0.02), which did not differ from each other (Table 1).

Hypothalamic NPY-mRNA expression

On removal from protocol, lean-FF rats displayed higher NPY-mRNA expression compared to the obese-FF and obese-PF rats (P < 0.05), but there was no significant difference between obese-FF and obese-PF groups (Figure 6).

Figure 6.

NPY-mRNA expression levels in the arcuate nucleus after the rats reached starvation or survival criterion under AA challenge. Bars represent mean ± s.e.m. *P < 0.05 vs. lean-FF. AA, activity anorexia; FF, free-feeding; NPY, neuropeptide-Y; PF, pair-feeding.

Body composition

Preacclimatization, lean-FF rats had less WAT than either obese group (P < 0.001), whereas obese-FF and obese-PF rats did not differ significantly. After the challenge, obese-PF rats had significantly less WAT than the obese-FF rats (P < 0.001). WAT in the lean-FF group was almost completely depleted, consistent with our finding of near depletion of plasma leptin (Table 1). Unlike WAT, lean mass was not significantly different among the three groups, either before or following the AA challenge.

Stepwise multiple regressions were conducted separately for obese-prone and lean-prone rats using preacclimatization measures (WAT, lean mass, body weight, corticosterone, TG, and glucose) as predictors of days lasted in the AA challenge. TG concentration, used as an index of immediately available energy resources, predicted survival (obese-prone: r = 0.61, P = 0.01; lean-prone: r = 0.77, P = 0.04), replicating Pierce et al. findings (12).

Energy expenditure

Figure 7 (panel a), depicts the RER of rats after 3 days of AA challenge. There was no significant difference between the groups of rats during the dark (active) phase of the diurnal cycle. However, during the light (inactive) period RER was significantly lower in obese-FF and obese-PF rats compared to lean-FF rats (P < 0.05), with no difference between obese-prone groups. Lean-FF rats expended more energy than obese-PF rats, regardless of the phase of the diurnal cycle (light or dark) (P < 0.05) and more energy than obese-FF rats, during the dark phase only (P < 0.05) (Figure 7; panel b). Obese-FF and obese-PF rats did not significantly differ in heat production.

Figure 7.

Effect of 3 days of AA challenge on energy expenditure in lean-FF, obese-FF, and obese-PF rats. (a) Respiratory exchange ratio (RER) and (b) heat production in lean-FF rats, obese-FF rats, and obese-PF rats for light and dark periods after 3 days of AA challenge. Data are mean ± s.e.m. *P < 0.05, vs. lean-FF for the same period (dark or light). AA, activity anorexia; FF, free-feeding; PF, pair-feeding.


Our results show that juvenile obese-prone rats gain a survival advantage over lean-prone under famine-like conditions, and this advantage is further enhanced by physiological and behavioral changes induced by prior food restriction. In the wild, this survival advantage in young animals, that are the future breeders, would confer increased reproductive success. At a basic level, these results support the “thrifty gene” hypothesis of obesity (6).

On average, freely fed obese-prone animals lasted twice as long under the AA challenge as lean-prone animals, replicating the recent findings of Pierce et al. (12). In fact, half of the lean-prone rats failed to survive beyond day 3 and the all had failed by day 4. In comparison, freely fed obese-prone rats survived 8.2 days and one animal survived to day 14. This genotype difference appears to be related to a number of factors. The lean-prone animals had significantly lower caloric intake on day 1 and markedly lower WAT reserves on entering the AA challenge. While total distance run was not different between the obese-prone rats and their lean-prone counterparts, the rate of increase was significantly different over the 4-day period. Thus, EE was higher in lean-prone rats and reserves lower.

The lower EE measured in obese-prone rats compared to lean-prone could be related to the absence of a functional leptin receptor. Leptin acts through hypothalamic centers to decrease appetite and increase EE, in part through enhanced thermogenesis (27). At a cellular level, leptin is known to induce expression of uncoupling proteins (UCP-1, UCP-2, UCP-3) in the mitochondria via stimulation of β3-adrenaergic receptors, thereby leading to an increased thermogenesis, one of the EE components (28). Uncoupling proteins uncouple the mitochondrial respiratory chain from oxidative phosphorylation so that energy, instead of being used for ATP production, is dissipated as heat (29). Thus, when leptin signaling is defective, this uncoupled process is impaired and heat production is reduced (30). Obese-prone rats would therefore have lower thermogenesis than lean-prone animals, contributing to longer survival in the AA challenge.

The changes in plasma triglycerides, NPY expression, leptin and RER are all consistent with decompensated weight loss. At day 3 of the AA challenge, the RER value was high in the lean-prone rats during the dark phase (0.8) and remained unchanged in the light phase, indicating continued catabolism of protein to produce energy (31), consistent with the depletion of WAT stores in these animals. The RER value was reduced in obese-prone rats during the light (inactive) phase, suggesting a preferential use of fat to produce energy, an adaptation that curtailed muscle proteolysis and allowing prolonged survival (21). Conservation of protein during food scarcity requires the availability of lipid fuels (32), and the reduced RER is consistent with the success of obese-prone rats in retention of half their initial WAT stores after the AA challenge.

Plasma corticosterone concentration, an index of hypothalamic—pituitary—adrenal-axis activity/reactivity, was higher in lean-prone rats than obese-prone animals after AA challenge, indicating greater stress in lean-prone rats that was associated with lower survival (12). Also, the low-caloric intake of lean-FF rats after the AA challenge coincided with high hypothalamic NPY expression. NPY is typically viewed as an orexigenic peptide that stimulates eating (33). The low-caloric intake despite NPY up-regulation suggests that NPY has other physiological functions beyond energy intake, probably food-related travel (34).

Beyond the findings for genotype our results, for the first time, show an added survival benefit of prior food restriction in the obese-prone animal. Pair fed (obese-PF) rats survived, on average, 1.5 times longer than the obese-FF group, with 5 of 8 animals reaching a homeostatic equilibrium and steady state. Thus, prior caloric restriction prepares animals to overcome sudden and severe food shortage, beyond the advantage conferred by the obese-prone genotype alone. These results are in accord with the findings of Routtenberg (14) that exposure to severe caloric restriction in adult outbred rats allows adaptation to the AA challenge. In the present study, juvenile obese-prone rats were pre-exposed to a moderate food restriction (66% of obese-FF) (35) that was still sufficient to cause significant enhancement of survival, beyond that of equivalent animals without prior exposure to food restriction.

Increased survival of obese-PF rats occurred even though these animals had lower initial body weight than the obese-FF. Boakes and Dwyer (36), in contrast, reported increased starvation of rats with lower initial body weights. This discrepancy is probably due to design and genotype differences between the two studies. Boakes and Dwyer used outbred rats categorized into heavy and light groups before the AA challenge while in our study the lower weight of JCR-LA-cp obese-prone rats was induced by the caloric restriction, a quite different condition.

Under the AA challenge, obese-prone groups did not differ in caloric intake. The similar caloric intake of obese-FF and obese-PF rats is apparently contrary to previous research. Dwyer and Boakes (13) reported that outbred rats, preadapted to food restriction, ate more during the AA challenge than rats that were not preadapted. A possibility is that overeating in our obese-prone animals obscured the preadaptation effect observed in the Dwyer and Boakes study. This similar intake of both obese-prone groups, however, indicates that the survival advantage of the obese-PF rats in the AA challenge is not related to energy intake.

We did find, however, differences in the pattern of wheel running between the obese-prone groups. Obese-PF increased their rate of wheel running more rapidly, over the initial 6 days of AA, than obese-FF animals and stabilized the running at a higher rate. The wheel running results indicate that the rate of increase in running is inversely related to survival and depends on the prior feeding regime of obese-prone rats, obese-FF rats displayed a gradual increase in wheel running and survived a mean 8.2 days. Obese-PF animals, in contrast, had a higher rate of increase in wheel running but survived ∼12.2 days in the challenge, with 62.5% meeting the survival criterion. Our findings indicate that obese-PF rats are able to survive the AA challenge despite higher rates of running or food-related travel.

Paradoxically, when rodents are severely challenged to obtain food by wheel running, high workload requirements lead to negative energy balance, resulting in a reduction of resting metabolic rate (RMR) and lean mass (37). That is, wheel running is negatively related to measures of BMR when the requirement for food is high (37). Measures of BMR were not obtained directly in this study, but body composition findings suggest that the metabolic efficiency of obese-PF rats is a result of BMR reduction related to high levels of wheel running. Instead of increasing food intake on the time-limited meals, obese-PF rats compensated for the high energy cost (workload) of food-related travel by decreasing BMR.

Taken together, the higher rate wheel running, reduced body weight loss and similar energy intake of obese-PF rats compared to the obese-FF group, suggests greater metabolic efficiency of the obese-PF animals. The gain in metabolic efficiency is probably due to the caloric restriction imposed by pair-feeding obese-prone rats to free-feeding lean-prone animals. Similar to the absence of the leptin receptor, caloric restriction is known to reduce thermogenesis via the UCP-1 uncoupling process (38). Caloric restriction and high rate wheel running also independently reduce BMR (37,39). Thus, prior pair-feeding and leptin receptor deficiency synergistically reduce thermogenesis, allowing high metabolic efficiency of obese-PF rats in the AA challenge. These animals also would gain in metabolic efficiency through their lower BMR compared to obese-FF rats. The additive effects of reduced thermogenesis and low BMR might explain the greater survival of obese-PF rats in the AA challenge.

We found similar EE in the obese-PF and obese-FF groups as measured in the metabolic cages, despite the high rate of wheel running in the obese-PF rats. This lack of difference in EE is probably due to the removal of the rats from the AA challenge after 3 days to take the EE measure. Our metabolic cages do not have attached running wheels; thus, all animals were maintained on the 1.5-h time limited meal, but no longer had running wheel activity. Without the continuation of access to wheel running, our metabolic data may not reflect exactly the EE during the AA challenge. The metabolic data, however, do provide a indication of EE during AA as rats maintain negative energy balance following 2 days without wheels (15).

The lower weight loss (percent and absolute values), despite similar caloric intake during the AA challenge and greater distances run, supports the metabolic efficiency in the obese-PF group. Under negative energy challenge of AA, obese-FF (no history of food restriction) lost more body weight, increased their daily wheel running and starved in about 8 days. In contrast, the majority of obese-PF rats were able to overcome the nonhomeostatic challenge of AA. These surviving animals achieved homeostasis (zero weight loss) after 10 days when their rate of running had stabilized and caloric intake began to increase. This homeostatic process is likely initiated at the onset of the AA challenge as the consequences of both preconditioning by food restriction and lack of leptin signaling effects on thermogenesis and BMR reductions, but only is apparent after 10 days in our study. Achievement of homeostasis may occur in more or less days depending on the metabolic efficiency of the animals. The attainment of homeostasis by obese-PF animals prevented further energy loss and allowed for survival. Notably, zero weight loss or homeostasis occurred in all animals that survived to the end of the AA challenge. These observations suggest that prior experience with food restriction allows juvenile obese-prone animals to metabolically adjust to a negative energy state, preserving behavioral functions such as physical activity or food-related travel (40). While this finding may be counter intuitive, it is analogous to the well-established phenomenon of cardiac and cerebral ischemic preconditioning (41), where prior mild ischemic challenges allow survival of otherwise fatal ischemic events. These effects may be epigenetic in origin (42), which would be consistent with our results.

Overall, we found that juvenile JCR:LA-cp obese-prone rats pair fed to lean-prone (+/?) animals survived longer in the AA challenge by achieving homeostasis related to an adaptive metabolic efficiency. Thus, caloric restriction at early ages may predispose obese-prone individuals to become more metabolically efficient. An inducible increase in metabolic efficiency may help to explain the increased obesity in low- and middle-income countries where childhood under-nutrition exists in the context of rapid economic development and rural/urban migration (43). Thus, the obese-prone phenotype, that is highly deleterious in a food-rich environment, confers a real survival benefit in an unstable and scarce food environment, that is enhanced by prior caloric restriction. In this study, we were able to isolate both genotype and feeding history effects on behavioral mechanisms related to survival. Subsequent research should not only confirm the interdependent effects of genotype and feeding history, but the specific metabolic and neuroendocrine mechanisms.


This research was funded by a grant from the National Sciences and Engineering Research to Council of Canada (NSERC) to W.D.P. and C.D.H. The Alberta Livestock Industry Development Fund and NSERC supported the research of S.D.P.


The authors declared no conflict of interest.