Calorie restriction can inhibit or delay carcinogenesis, reportedly due to a reduction in calorie intake rather than by concurrent changes in body mass and/or composition. Our objective was to test the hypothesis that body mass and/or composition have an important effect, independent of energy intake, on the benefits or hazards associated with calorie restriction or overeating, respectively. In the first experiment, transgenic mice that spontaneously develop prostate cancer [transgenic adenocarcinoma of mouse prostate (TRAMP)] were housed at 27°C or 22°C and pair fed the same diet for 21 weeks (95% of ad libitum intake at 27°C). In the second experiment, TRAMP mice were housed at 27°C or 22°C and fed the same diet ad libitum for 21 weeks. Despite a similar calorie intake, pair-fed mice at 27°C (PF27) were heavier (28.3 ± 3.3 versus 17.6 ± 1.6 g at 21 weeks; P < 0.001; mean ± SD) and had greater fat (6.4 ± 2.1 versus 1.9 ± 0.3 g; P < 0.001) and lean mass (P < 0.001) than pair-fed mice at 22°C. Furthermore, PF27 mice had greater levels of serum leptin (P < 0.001), lower levels of adiponectin (P < 0.05), and a greater frequency of prostatic adenocarcinoma (P < 0.05). In contrast, ad libitum–fed mice housed at 22°C consumed ∼30% more calories than ad libitum–fed mice at 27°C, but there was no difference between groups in body composition or cancer progression. These results imply that the ability of calorie restriction to inhibit or delay cancer incidence and progression is mediated in part by changes in energy balance, body mass, and/or body composition rather than calorie intake per se, suggesting that excess calorie retention, rather than consumption, confers cancer risk. [Cancer Res 2007;67(1):417–24]

The prevalence of obesity has markedly increased over the past 2 decades in American adults, and excess adiposity has become more and more recognized as a risk factor for cancer incidence and mortality (1, 2). The fundamental cause of obesity is a long-term imbalance in energy intake and expenditure (i.e., positive energy balance). Calorie restriction provides an effective strategy to combat this imbalance by limiting energy intake and, hence, reducing fat stores. In animal models, a 20% to 40% reduction in calorie intake has consistently shown an ability of calorie restriction to reduce body weight and adiposity (3) and delay several age-related diseases, including cancer (4, 5).

Despite the fact that the main phenotypic change observed in calorie-restricted animals is a reduction in body and fat mass, most investigations have concluded that the benefits of calorie restriction are mediated by a reduction in food intake per se rather than changes in body mass and/or composition (69). However, many of the effects of calorie restriction can be alternatively explained by changes in body composition rather than a decrease in food intake (10). For example, a reduction in body weight, and specifically body fat, can lead to improvements in insulin sensitivity (11), reductions in levels of proinflammatory cytokines (12, 13), and oxidative stress (14), all of which are related to cancer incidence and mortality (1517).

Because restricting calories in mammals typically results in a reduction in body and fat mass, attempting to differentiate the benefits attributed to a reduction in food intake from concurrent changes in body mass and composition is extremely difficult. To date, the cancer-obesity link is based on mounting epidemiologic evidence (1, 18, 19), in vitro studies (2022), and the assumption that obesity-related risk factors implicated in the etiology of cardiovascular disease and diabetes also raise cancer risk (10). However, there remains a lack of in vivo evidence that confirms that excess body weight, and specifically body fat, is a direct cause of cancer incidence and/or progression.

In an attempt to test the role of body mass and/or body composition versus food intake on cancer progression, we used small differences in ambient temperature (5°C) as a tool to manipulate energy expenditure in a mouse model of prostate cancer [transgenic adenocarcinoma of mouse prostate (TRAMP)]. Specifically, when ambient temperature is raised closer to the thermoneutral zone for mice (approximately 30–35°C), thermoregulatory demands, and hence, energy expenditure can be significantly reduced (23). Therefore, we hypothesized that under fixed food intake conditions, mice maintained slightly below thermoneutrality would expend less energy in thermoregulation, have greater gains in body and fat mass, and have more advanced prostate cancer than mice maintained at a lower ambient temperature (27°C versus 22°C). Thus, this experiment would test the hypothesis that body mass and/or composition, but not food intake, are the links between calorie restriction or obesity and cancer.

In the second experiment, we used the same housing temperatures but allowed the animals to feed ad libitum. We hypothesized that TRAMP mice maintained at 22°C would increase food intake in a manner that would offset the increased thermoregulatory demands and that both groups would then have similar body mass and composition throughout the experiment. The purpose of the second experiment was to determine whether increased food intake, without a difference in body mass and composition, would affect cancer progression. The results of these two studies show that a reduction in body, fat, and lean mass, without changing food intake, can significantly inhibit or delay carcinogenesis and that, in the absence of changes in body composition, increasing calorie intake by ∼30% does not accelerate cancer progression. However, to what extent these results may be attributed to energy balance differences is not clear.

Transgenic animals. TRAMP mice on a pure C57BL/6 background were obtained from our breeding colony located at the University of Alabama at Birmingham (Birmingham, AL). Transgenic females were bred with nontransgenic males, and all breeders were maintained on a standard 12-h light/12-h dark photoperiod and fed standard rodent chow (Harlan Teklad, Madison, WI) and water ad libitum. After weaning, male pups were separated from females, and a tail biopsy was collected for determination of transgene incorporation by PCR as described previously (24). Subsequently, all male TRAMP mice were singly housed in cages measuring 28 × 13 × 13 cm on a standard 12-h light/12-h dark photoperiod in air- and humidity-controlled rooms. All work was approved by the Institutional Animal Care and Use Committee of the University of Alabama at Birmingham.

Determination of ad libitum food intake. Initially, ad libitum food intake was determined in 7-week-old male TRAMP mice singly housed at 27°C in a temperature- and humidity-controlled incubator (model IS79SD, Powers Scientific, Inc., Pipersville, PA) for 4 weeks. Food consumption was determined by measuring the loss of food from the hopper of each animal over a 3-day period (one 3-day determination per week). The average 4-week food consumption was determined (2.32 g/d), and 95% of this amount (2.20 g/d) was used for the pair-feeding study.

Experiment 1 design: pair-feeding study. For the pair-feeding experiment, 7-week-old male TRAMP mice were randomly assigned to one of two groups, singly housed in a temperature- and humidity-controlled incubator at 27°C (PF27; n = 54) or singly housed in a temperature- and humidity-controlled room at 22°C (PF22; n = 54) for up to 21 weeks (28 weeks of age). Both groups were fed 2.20 g of a phytoestrogen-free diet (AIN-76A, Harlan Teklad) daily; food was placed in the food hopper of the animal each afternoon at ∼1500 h. Any food remaining in the bedding or hopper after 24 h was removed, weighed, and recorded. In addition, body weights were recorded weekly for all animals.

Mice were removed from the experiment if one or more of the following criteria were met: (a) >50% of food (>1.10 g) remained in the hopper for 2 consecutive days coupled with signs of sickness, lethargy, or a palpable tumor; (b) weight loss exceeding 20% of the body weight from the previous week; or (c) a tumor diameter ≥20 mm in PF27 mice or ≥15 mm in PF22 mice. This smaller diameter was used for PF22 mice because they were considerably smaller than PF27 mice.

Experiment 2 design: ad libitum–feeding study. Seven-week-old male TRAMP mice were randomly assigned to one of two groups, singly housed at 27°C (AL27; n = 24) or singly housed at 22°C (AL22; n = 21) and fed the same diet ad libitum for up to 21 weeks (28 weeks of age).

Mice were removed early from this experiment if one or more of the following criteria were met: (a) weight loss exceeded 20% of the body weight from the previous week or (b) a tumor diameter ≥15 mm was detected. This diameter was chosen as the cutoff for this experiment because ad libitum–fed mice showed a propensity for their condition to deteriorate more quickly than pair-fed mice. Furthermore, a reduction in food intake was not used as an exclusion criterion because significant variability in food intake existed from week to week in these animals. Body weight was recorded weekly, and weekly measures of food intake were determined by measuring loss of food from the hopper of each animal over a 3-day period (one 3-day determination per week).

Body composition. All body composition scans for animals were done early in the morning. The first 60 animals included in the pair-feeding experiment [PF27 (n = 29) and PF22 (n = 31)] were anesthetized with isoflurane (2%), and body composition was assessed in vivo by dual-energy X-ray absorptiometry (DXA; GE Lunar PIXImus software version 1.45, Lunar, Madison, WI) at baseline and at 6, 12, 18, and 21 weeks in study. The remaining 48 pair-fed animals [PF27 (n = 25) and PF22 (n = 23)] and all animals in the ad libitum–feeding experiment were anesthetized i.p. with Nembutal (0.693 mg/g) for the collection of blood and assessment of body composition.

For DXA, fat mass, soft lean tissue mass, bone mineral content (BMC), and bone mineral density (BMD) were determined as described previously (25). Briefly, mice were anesthetized and placed in a prostrate position on the imaging plate, and a total body scan was done in ∼5 min. All analyses were done excluding the head.

Indirect calorimetry. Energy expenditure was assessed by indirect calorimetry at 12 weeks in study as reported by our laboratory previously (26). Energy expenditure was determined from O2 consumption and CO2 production using the equation of Weir (27). Total energy expenditure (TEE) was calculated as the sum of the energy expended over the measurement period, extrapolated to 24 h. Resting energy expenditure (REE) was estimated from the average energy expenditure over three nonconsecutive 10-min intervals during which energy expenditure was minimal. Non-REE (NREE; i.e., activity-related energy expenditure) was calculated as the difference between TEE and REE.

Data were collected for 22 h from a subset of PF27 (n = 5) and PF22 (n = 5) mice, respectively, from the pair-feeding study. A 22-h measurement period was used to allow time for the chambers to be cleaned and for the analyzers to be calibrated between experiments on consecutive days. Mice were placed in the chamber at ∼1600 h, provided a weighed food pellet (2.20 g), and given ad libitum access to water during the measurement period. The mice were measured at the temperature at which they were housed (27°C or 22°C), and the photoperiod was set at 12-h light/12-h dark.

Blood and tissue collection. For mice anesthetized with Nembutal, immediately following the DXA scan, ∼200 μL of blood were collected from the tail under anesthesia at 6 and 12 weeks in study and by cardiac puncture at termination for assessment of hormone and cytokine levels. Serum was separated from blood samples by centrifugation and stored at −70°C until further analysis. In addition, tissue weights were obtained at necropsy for the urogenital system (prostate, bladder, and seminal vesicles) and testes.

Serum measures. Total testosterone was measured in duplicate from serum collected at 6 and 21 weeks in study using a RIA kit (Diagnostic Systems Laboratories, Webster, TX). Serum at 12 weeks in study was assessed in duplicate for adiponectin, interleukin-6 (IL-6), tumor necrosis factor-α (TNF-α), and leptin. Adiponectin was measured by RIA (LINCO Research, St. Louis, MO); IL-6, TNF-α, and leptin were measured by the Luminex100 instrument (Luminex Corp., Austin, TX) using a LINCOplex immunoassay panel (LINCO Research). Serum concentrations of IL-6 and TNF-α are not presented because the values obtained were below the lower detection limit of the assay (12.2 pg/mL). The intra-assay coefficient of variation for reported assays was 7.4% for testosterone, 2.3% for adiponectin, and 4.2% for leptin.

Histopathology. To assess cancer progression and metastasis, a portion of the dorsolateral prostate (DLP) and lymph nodes were collected for histologic analysis. The DLP was placed in an acid-alcohol fixative containing 96% ethanol, 1% glacial acetic acid, and 3% distilled water at 4°C overnight as described by Folkvord et al. (28), and the lymph nodes were placed in phosphate-buffered formalin overnight. The fixed tissue was then embedded in paraffin, and 5 μm sections were mounted onto slides. All DLP sections were stained with Gomori trichrome staining, and lymph node sections were stained with H&E.

Pathologic changes in the DLP were scored as described previously (29, 30) on a 1 to 6 scale by a pathologist who was blinded to the experiment (I.A.E.). Scores 1, 2, and 3 indicated normal tissue, low-grade prostatic intraepithelial neoplasia (PIN), and high-grade PIN, respectively; scores 4, 5, and 6 indicated well-differentiated, moderately-differentiated, and poorly-differentiated prostatic adenocarcinoma, respectively. The presence or absence of cancerous cells in periaortic lymph nodes was used to determine if metastasis of the primary tumor had occurred.

Statistical analyses. All statistical analyses were carried out separately for the pair-feeding study and ad libitum–feeding study because they were not run concurrently and protocols differed slightly. In both experiments, food intake, body weight, and body composition were analyzed by repeated measures ANOVA and followed up with planned contrasts when appropriate. Mice that were removed early from the experiment were not included in the repeated measures analyses. Serum concentrations, energy expenditure, and tissue weights between groups were compared with Student's t test, and χ2 analysis was used to determine if a significant difference existed among groups in the distribution of pathologic scores of the DLP and for the presence or absence of metastasis to lymph nodes. For the χ2 analysis of pathologic score distribution, animals were assigned to one of three groups: normal prostate (score, 1), PIN (score, 2–3), and adenocarcinoma (score, 4–6). All analyses were done using Statistical Analysis System version 8.0 (SAS Institute, Inc., Cary, NC) and considered statistically significant when P < 0.05.

Experiment 1: Pair-Feeding Study

PF27 mice consumed less food but had greater body mass. Despite our intention for pair-fed mice to have the same food intake (2.20 g/d), PF27 mice consumed less food (Fig. 1A). Although this difference was very small (≤0.1 g/d), it was nevertheless significant (P < 0.05). Regardless of this small reduction in food intake, PF27 mice were significantly heavier than PF22 animals, differing by ∼11 g at the end of the study (P < 0.001; Fig. 1B).

Figure 1.

Phenotypic characteristics for pair-fed mice housed at 27°C (PF27; n = 44) or 22°C (PF22; n = 49). Measurements were taken either weekly or at predetermined intervals for (A) food intake, (B) body mass, (C) fat mass, (D) lean mass, and (E) percentage fat. F, TEE was measured in a subset of PF27 (n = 5) and PF22 (n = 5) mice at 12 wks of treatment. Columns, mean; bars, SE. *, P < 0.05, significantly different from PF22 mice; ‡, P < 0.001, significantly different from PF22 mice.

Figure 1.

Phenotypic characteristics for pair-fed mice housed at 27°C (PF27; n = 44) or 22°C (PF22; n = 49). Measurements were taken either weekly or at predetermined intervals for (A) food intake, (B) body mass, (C) fat mass, (D) lean mass, and (E) percentage fat. F, TEE was measured in a subset of PF27 (n = 5) and PF22 (n = 5) mice at 12 wks of treatment. Columns, mean; bars, SE. *, P < 0.05, significantly different from PF22 mice; ‡, P < 0.001, significantly different from PF22 mice.

Close modal

PF27 mice had greater fat mass, lean mass, and bone content. In addition to weighing less than PF27 mice, PF22 animals had less fat mass (Fig. 1C), lean mass (Fig. 1D), and percentage fat (Fig. 1E) as measured by DXA (P < 0.001). Specifically, PF22 mice lost a small but significant amount of lean mass after 6 weeks of treatment but fat mass was unchanged. However, subsequent measures of body composition with DXA indicated that PF22 mice began to gain lean mass whereas fat mass decreased (P < 0.001; Fig. 1C and E).

With regard to bone, despite no differences in measures of BMD and BMC between groups at the beginning of the study, PF27 mice had a more substantial increase in these measures throughout the remainder of the experiment. Specifically, compared with their leaner counterparts, BMD and BMC in PF27 were significantly greater at 6, 12, 18, and 21 weeks in study for BMD (0.044 ± 0.001 versus 0.042 ± 0.002 g/cm2 at 21 weeks; P < 0.001; mean ± SD) and BMC (0.341 ± 0.100 versus 0.320 ± 0.030 g at 21 weeks; P < 0.001), respectively.

Pair-fed mice had similar resting, nonresting, and TEE. At 12 weeks of treatment, PF27 and PF22 mice had similar absolute TEE (P = 0.62; Fig. 1F), REE (3.99 ± 1.65 versus 3.29 ± 1.00 Kcal/d; P = 0.44; mean ± SD), and NREE (3.00 ± 0.23 versus 3.25 ± 0.58 Kcal/d; P = 0.40). Furthermore, after adjusting for lean mass, TEE remained similar between PF27 and PF22 mice (6.84 ± 1.44 versus 6.69 ± 1.44 Kcal/d; P = 0.89) as did REE (4.37 ± 1.46 versus 2.91 ± 1.46 Kcal/d; P = 0.75) and NREE (2.47 ± 0.45 versus 3.77 ± 0.45 Kcal/d; P = 0.48).

PF27 mice had greater levels of serum leptin and lower levels of adiponectin. At 12 weeks of treatment, serum leptin was greater in PF27 mice, which had more body fat (P < 0.001; Fig. 2A), and leptin was significantly related to fat mass in the PF27 group (r = 0.52; P < 0.01) but not in the PF22 group (r = −0.11; P = 0.65). Adiponectin was significantly greater in PF22 mice (P < 0.05; Fig. 2B), but adiponectin was not associated with fat mass in the PF27 group (r = 0.05; P = 0.83) or in the PF22 group (r = −0.07; P = 0.77).

Figure 2.

Serum cytokine measurements obtained from pair-fed mice housed at 27°C (PF27) or 22°C (PF22). A, serum leptin at 12 wks of treatment from PF27 (n = 24) and PF22 (n = 21) mice. B, serum adiponectin at 12 wks of treatment from PF27 (n = 24) and PF22 (n = 21) mice. Columns, mean; bars, SE. *, P < 0.05, significantly different from PF22 mice; ‡, P < 0.001, significantly different from PF22 mice.

Figure 2.

Serum cytokine measurements obtained from pair-fed mice housed at 27°C (PF27) or 22°C (PF22). A, serum leptin at 12 wks of treatment from PF27 (n = 24) and PF22 (n = 21) mice. B, serum adiponectin at 12 wks of treatment from PF27 (n = 24) and PF22 (n = 21) mice. Columns, mean; bars, SE. *, P < 0.05, significantly different from PF22 mice; ‡, P < 0.001, significantly different from PF22 mice.

Close modal

Testes from PF27 mice were heavier, but no difference in serum testosterone. Testes weights were significantly heavier in PF27 mice than in PF22 mice at 21 weeks of treatment (P < 0.05; Fig. 3A). In contrast, the concentration of serum total testosterone was not significantly different between groups at 6 weeks (P = 0.89; Fig. 3B) or 21 weeks of treatment (P = 0.57; Fig. 3B).

Figure 3.

Testes weights and serum testosterone measures obtained from pair-fed mice housed at 27°C (PF27) or 22°C (PF22). A, testes weights obtained from PF27 (n = 44) and PF22 (n = 49) mice at 21 wks of treatment. B, total serum testosterone at 6 wks from PF27 (n = 24) and PF22 (n = 23) mice and 21 wks of treatment from PF27 (n = 42) and PF22 (n = 48) mice. Columns, mean; bars, SE. *, P < 0.05, significantly different from PF22 mice.

Figure 3.

Testes weights and serum testosterone measures obtained from pair-fed mice housed at 27°C (PF27) or 22°C (PF22). A, testes weights obtained from PF27 (n = 44) and PF22 (n = 49) mice at 21 wks of treatment. B, total serum testosterone at 6 wks from PF27 (n = 24) and PF22 (n = 23) mice and 21 wks of treatment from PF27 (n = 42) and PF22 (n = 48) mice. Columns, mean; bars, SE. *, P < 0.05, significantly different from PF22 mice.

Close modal

PF27 mice had more advanced prostate cancer, but no difference in metastasis. Based on the exclusion criterion described in Materials and Methods, 10 of 54 (18.5%) PF27 mice were removed early from the experiment: 7 due to a tumor diameter ≥20 mm and 3 due to a weight loss exceeding 20% of the weight from the previous week. Five of 54 (9.3%) PF22 animals were removed early from the study: all due to a tumor diameter ≥15 mm. Despite the fact that 2-fold more PF27 mice had to be removed early from the study than PF22 mice, there was no significant difference in number of animals removed by χ2 analysis (P = 0.16).

The urogenital system (prostate, bladder, and seminal vesicles) from PF27 mice was significantly heavier than PF22 animals (2.4 ± 2.0 versus 1.0 ± 1.7 g; P < 0.001; Fig. 4A). Furthermore, PF27 mice had a significant shift in the distribution of pathologic scores toward prostatic adenocarcinoma, whereas PF22 mice displayed a higher frequency of normal prostate (score = 1) and lower incidence of prostatic adenocarcinoma (score = 4–6; P < 0.05; Fig. 4B). However, there was no difference in metastasis to lymph nodes between PF27 (92%) and PF22 (90%) mice with poorly differentiated adenocarcinoma (score = 6; P = 0.85).

Figure 4.

Urogenital system weights and histologic evaluation of the DLP from pair-fed mice housed at 27°C (PF27) or 22°C (PF22). A, scatter plot illustrating the urogenital system (prostate, bladder, and seminal vesicles) weight obtained from each PF27 (n = 54) and PF22 (n = 54) animal at necropsy. Line, mean weight for the group. ‡, P < 0.001, significantly different from PF22 mice. B, distribution of pathologic scores for the DLP from PF27 (n = 54) and PF22 (n = 54) mice. Pathologic changes in the DLP were blindly scored using the following 1 to 6 scale: scores 1, 2, and 3 indicated normal tissue, low-grade PIN, and high-grade PIN, respectively; scores 4, 5, and 6 indicated well-differentiated, moderately-differentiated, and poorly-differentiated prostatic adenocarcinoma, respectively. PF27 mice had a significant shift in the distribution of pathologic scores toward PIN and prostatic adenocarcinoma, whereas PF22 mice displayed a greater frequency of normal prostate (P < 0.05).

Figure 4.

Urogenital system weights and histologic evaluation of the DLP from pair-fed mice housed at 27°C (PF27) or 22°C (PF22). A, scatter plot illustrating the urogenital system (prostate, bladder, and seminal vesicles) weight obtained from each PF27 (n = 54) and PF22 (n = 54) animal at necropsy. Line, mean weight for the group. ‡, P < 0.001, significantly different from PF22 mice. B, distribution of pathologic scores for the DLP from PF27 (n = 54) and PF22 (n = 54) mice. Pathologic changes in the DLP were blindly scored using the following 1 to 6 scale: scores 1, 2, and 3 indicated normal tissue, low-grade PIN, and high-grade PIN, respectively; scores 4, 5, and 6 indicated well-differentiated, moderately-differentiated, and poorly-differentiated prostatic adenocarcinoma, respectively. PF27 mice had a significant shift in the distribution of pathologic scores toward PIN and prostatic adenocarcinoma, whereas PF22 mice displayed a greater frequency of normal prostate (P < 0.05).

Close modal

Experiment 2: Ad libitum–Feeding Study

AL22 mice consumed more food but weighed less than AL27 mice. AL22 mice consumed ∼30% more food than AL27 mice (P < 0.05; Fig. 5A). Despite this large difference in food intake, AL22 mice weighed 2 to 3 g less than AL27 mice during the latter half of the study, and this difference was significant at most time points (P < 0.05; Fig. 5B).

Figure 5.

Phenotypic characteristics for ad libitum–fed mice housed at 27°C (AL27; n = 17) or 22°C (AL22; n = 14). Measurements were taken either weekly or at predetermined intervals for (A) food intake, (B) body mass, (C) fat mass, (D) lean mass, and (E) percentage fat. Points, mean; bars, SE. *, P < 0.05, significant difference between groups; ‡, P < 0.001, significant difference between groups.

Figure 5.

Phenotypic characteristics for ad libitum–fed mice housed at 27°C (AL27; n = 17) or 22°C (AL22; n = 14). Measurements were taken either weekly or at predetermined intervals for (A) food intake, (B) body mass, (C) fat mass, (D) lean mass, and (E) percentage fat. Points, mean; bars, SE. *, P < 0.05, significant difference between groups; ‡, P < 0.001, significant difference between groups.

Close modal

No difference in fat or lean mass, but AL27 mice had greater BMC. There was no significant difference in fat mass (P = 0.06; Fig. 5C), lean mass (P = 0.47; Fig. 5D), or percentage fat (P = 0.10; Fig. 5E) between ad libitum–fed mice. BMD was similar between ad libitum–fed mice throughout the experiment (0.045 ± 0.001 versus 0.045 ± 0.001 g/cm2 at 21 weeks; P = 0.30; mean ± SD), but AL27 mice had significantly greater BMC after 12, 18, and 21 weeks of treatment (0.387 ± 0.034 versus 0.355 ± 0.024 g at 21 weeks; P < 0.01).

No difference in testes weight or serum testosterone between ad libitum–fed mice. There was no significant difference for mean testes weights between AL27 and AL22 mice at 21 weeks of treatment (143.0 ± 23.9 versus 143.3 ± 12.5 mg; P = 0.97). Furthermore, the concentration of total testosterone was not significantly different between AL27 and AL22 mice at 6 weeks (1.16 ± 1.61 versus 0.82 ± 2.14 ng/mL; P = 0.54) or 21 weeks of treatment (0.47 ± 0.43 versus 0.51 ± 1.20 ng/mL; P = 0.91).

No difference in cancer progression or metastasis between ad libitum–fed mice. Based on the exclusion criterion described in Materials and Methods, 7 of 24 (29.2%) AL27 mice were removed early from the experiment: 6 due to a tumor diameter ≥15 mm and 1 due to a weight loss exceeding 20% of the weight from the previous week. A similar percentage of mice were removed early from the AL22 group (7 of 21, 33.3%; P = 0.77): 6 due to a tumor diameter ≥15 mm and 1 due to a weight loss exceeding 20% of the weight of the previous week. In addition, there was no significant difference between AL27 and AL22 mice for urogenital system weight (2.8 ± 1.6 versus 3.3 ± 2.0 g; P = 0.34; Fig. 6A) or cancer progression (P = 0.75; Fig. 6B). Furthermore, there was no difference in metastasis to lymph nodes between AL27 (77%) and AL22 (56%) mice with poorly differentiated adenocarcinoma (P = 0.32).

Figure 6.

Urogenital system weights and histologic evaluation of the DLP from ad libitum–fed mice housed at 27°C (AL27) or 22°C (AL22). A, scatter plot illustrating the urogenital system (prostate, bladder, and seminal vesicles) weight obtained from each AL27 (n = 24) and AL22 (n = 21) animal at necropsy. Line, mean weight for the group. There was no significant difference between groups (P = 0.35). B, distribution of pathologic scores for the DLP from AL27 (n = 24) and AL22 (n = 21) mice. Pathologic changes in the DLP were blindly scored using the following 1 to 6 scale: scores 1, 2, and 3 indicated normal tissue, low-grade PIN, and high-grade PIN, respectively; scores 4, 5, and 6 indicated well-differentiated, moderately-differentiated, and poorly-differentiated prostatic adenocarcinoma. There was no significant difference between groups (P = 0.75).

Figure 6.

Urogenital system weights and histologic evaluation of the DLP from ad libitum–fed mice housed at 27°C (AL27) or 22°C (AL22). A, scatter plot illustrating the urogenital system (prostate, bladder, and seminal vesicles) weight obtained from each AL27 (n = 24) and AL22 (n = 21) animal at necropsy. Line, mean weight for the group. There was no significant difference between groups (P = 0.35). B, distribution of pathologic scores for the DLP from AL27 (n = 24) and AL22 (n = 21) mice. Pathologic changes in the DLP were blindly scored using the following 1 to 6 scale: scores 1, 2, and 3 indicated normal tissue, low-grade PIN, and high-grade PIN, respectively; scores 4, 5, and 6 indicated well-differentiated, moderately-differentiated, and poorly-differentiated prostatic adenocarcinoma. There was no significant difference between groups (P = 0.75).

Close modal

The dogma for years has held that a reduction in body weight and body fat from calorie restriction is merely a by-product of consuming fewer calories, with no real impact on outcomes of interest, such as cancer and longevity (6, 7, 9, 10, 31, 32). Indeed, the importance of body fat has been widely discounted due to the classic viewpoint of adipose tissue as only an inert storage depot for triglycerides, and historical studies which suggested that the beneficial component of calorie restriction was the result of a reduction in food intake rather than body fat (6, 7, 9). However, the present study has shown that the retardation of cancer initiation and progression in the mouse prostate is not attributable to the amount of calories consumed and may instead be mediated by changes in energy balance, body weight, and adiposity.

To our knowledge, these data are the first to show that a reduction in food intake is not required for the beneficial effects of calorie restriction with regard to cancer prevention. In agreement with these findings, a previous study by Bluher et al. (33) reported that fat-specific insulin receptor knockout (FIRKO) mice had similar food intake as control animals but were leaner and longer lived, suggesting that body fat may play an important role in longevity. In contrast, Bertrand et al. (6) reported that fat mass was not related to lifespan in ad libitum–fed rats but was positively correlated with lifespan in calorie-restricted rats. A later study by Harrison et al. (7) found that calorie-restricted ob/ob mice were longer lived than ad libitum–fed wild-types despite having nearly twice as much body fat. Indeed, these studies seem to provide strong evidence against a role for body fat in the determination of lifespan, but the validity of the conclusions reached by these studies is one of continual debate (10, 32).

Since the discovery of leptin in 1994 (34), the perception of adipose tissue as an inert storage depot has quickly evolved to that of a highly active endocrine organ, capable of secreting several proinflammatory cytokines, also referred to as “adipokines” (35). Many of these factors secreted from adipocytes and/or associated macrophages have provided viable mechanisms that may directly link obesity to increased cancer incidence and progression and not merely coincidental with the disease (10). Specifically, leptin, TNF-α, IL-6, heparin-binding epidermal growth factor, vascular endothelial growth factor, etc., may play a role in the pathology of cancer by promoting angiogenesis, inflammation, cell proliferation, and insulin resistance (21).

Leptin seems to be the most well studied adipokine due to its established role in the regulation of food intake and energy expenditure (36). More recently, leptin has garnered attention as a potent growth factor involved in the progression of breast (20) and prostate cancer (21, 22). In addition, leptin has been implicated in inducing oxidative stress (37) and aromatase expression (38), which converts androgens to estrogens. It has been clearly shown that the concentration of circulating leptin is positively related to the amount and location of adipose tissue (39), making leptin an attractive target for studying the obesity-cancer link. With regard to this study, leptin was profoundly elevated and significantly associated with fat mass in PF27 mice, whose fat content (g) was >3-fold that of their lean counterparts at 22°C, and may have contributed to the increased cancer progression in this group. Additionally, low concentrations of leptin in PF22 mice may have limited cancer progression.

Adiponectin, which is an adipokine secreted from adipose tissue, is paradoxically decreased in obesity (40, 41). Adiponectin may be unique in that it seems to protect from cancer by opposing proinflammatory mediators and cancer cell growth (42) while enhancing insulin function (40). Indeed, excess total and visceral fat are known to diminish circulating adiponectin and insulin sensitivity, which can lead to a compensatory increase in insulin secretion (43) and the overstimulation of inflammatory and growth-promoting pathways (44). Recent epidemiologic data support a possible inhibitory role of adiponectin on carcinogenesis by showing an inverse association between circulating adiponectin and cancer risk and progression (4547). For example, Goktas et al. (46) found that serum adiponectin levels were lower in patients with prostate cancer and were negatively associated with the histologic grade of the disease. Because serum adiponectin was significantly greater in PF22 mice, which were leaner than PF27 mice, a reduction in body fat may have conferred additional protection from prostate cancer in these animals by diminishing serum leptin and augmenting serum adiponectin levels.

The initiation and progression of prostate cancer in the TRAMP mouse at the outset is an androgen-dependent process (48). Because calorie restriction and fasting have been shown to reduce testosterone levels (49), one possible confounder to our results may have been due to changes in circulating testosterone. Therefore, we measured testes weights and serum testosterone levels to determine if these measures were lower in PF22 mice. We found that the testes from PF22 animals weighed significantly less at 21 weeks of treatment, but this difference was small in relation to PF27 animals (−7%). However, there was no difference between groups in serum testosterone levels measured at 6 weeks in study (∼13 weeks old), a time when testosterone levels would be expected to play a role in initiation, or at 21 weeks of treatment. Thus, the difference in cancer progression observed in experiment 1 was not likely due to a deficiency in circulating testosterone (i.e., castration effect; ref. 48) to stimulate transgene expression in PF22 mice, although a difference in testosterone before 6 weeks of treatment cannot be ruled out. Instead, these data suggest that the retardation of cancer incidence and progression in PF22 mice was related to changes in energy balance, body mass, and/or composition.

The results of this study strongly suggest that the beneficial effects of calorie restriction are mediated by mechanisms other than a reduction in food intake, but to what extent these effects are the direct result of changes in body mass and/or composition versus changes in energy balance remains unclear. In the pair-feeding paradigm, food intake was clamped at 2.2 g/d, an amount that is only 5% less than what TRAMP mice consume ad libitum at 27°C, but is ∼27% less than what they consume ad libitum at 22°C. The disparate amount of voluntary food intake consumed is due to the decreased thermoregulatory demands on mice when housing temperature is raised closer to thermoneutrality (approximately 30–35°C).

Therefore, PF22 mice, which were housed at a lower ambient temperature (22°C) than PF27 mice and had greater thermoregulatory demands, were forced to partition a greater proportion of energy to “defend” body temperature. Because 2.2 g/d was not a sufficient amount of energy consumed by PF22 mice to support initial body mass and thermoregulation, these animals entered a brief period (∼2 weeks) of negative energy balance (energy in < energy out) during which body mass was sufficiently reduced to bring them back into energy balance (energy in = energy out). They then maintained this lower mass for the duration of the study.

In contrast, PF27 mice, who were housed in a more thermoneutral environment (27°C), required less energy to maintain body temperature due to decreased thermoregulatory demands and had greater body (∼8 g) and lean mass (∼5 g) than PF22 mice. In addition, PF27 mice consumed less food than PF22 mice, but energy intake for PF27 mice was still in excess of energy expenditure, which resulted in calorie retention and weight gain. Interestingly, 24-h energy expenditure was not different between groups because animals remained weight stable during the measurement period. However, if weight change over the course of the study is used as a proxy of long-term energy balance, it is evident that PF27 mice expended slightly less energy than PF22 mice because these animals consumed less food but still gained weight.

In the ad libitum–feeding experiment, AL22 animals consumed ∼30% more calories than AL27 mice to compensate for the additional energy expended in thermoregulation. Despite consuming more food, AL22 mice had less body mass and BMC at some time points than AL27 mice, but fat and lean mass were not significantly different. Furthermore, cancer progression was not different between ad libitum–fed animals. Thus, both experiments imply that the prevention of excess calorie retention and obesity are important mediators of the calorie restriction pathway.

Because our results cannot “tease” apart the contribution of body mass and/or composition from energy balance on cancer prevention, more studies working toward a better understanding of the cancer-preventive mechanisms mediated by body mass and/or composition and energy balance are needed. In addition, our study suggests that increased energy expenditure, via thermoregulation, can result in beneficial effects similar to those observed with calorie restriction (i.e., reduced energy intake). Therefore, from a public health standpoint, it would be interesting to determine if a similar benefit could be achieved with exercise, which can also increase energy expenditure and limit weight gain. In support of this notion, a recent study by Colbert et al. (50) found that an exercise-induced negative energy balance reduced polyp burden in male APCMin mice. Taken together, the interaction of obesity, physical activity, and calorie restriction on cancer risk offers an exciting and important challenge for both the clinician and laboratory scientist.

In summary, these data show that despite a similar calorie intake, PF22 mice weighed significantly less, and had lower fat and lean mass than PF27 mice due to increased energy expended in thermoregulation by PF22 mice. PF27 mice, which had more body fat, also had greater levels of serum leptin and lower levels of serum adiponectin. Furthermore, histologic investigation of the DLP indicated a greater frequency of normal prostate in PF22 mice, whereas PF27 mice had a greater frequency of prostatic adenocarcinoma. In contrast, AL22 mice consumed more calories than AL27 mice, but there was no significant difference in fat mass, lean mass, or cancer progression between groups. These results imply that the ability of calorie restriction to inhibit or delay cancer incidence and progression is mediated in part by changes in energy balance, body mass, and/or composition rather than calorie intake per se, suggesting that excess calorie retention, rather than consumption, confers cancer risk.

Grant support: National Cancer Institute (NCI) grant CA101058, University of Alabama at Birmingham Clinical Nutrition Research Unit grant P30DK56336, and NCI Predoctoral Training grant CA47888 (D.M. Huffman).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1
Calle EE, Thun MJ, Petrelli JM, Rodriguez C, Heath CW, Jr. Body-mass index and mortality in a prospective cohort of U.S. adults.
N Engl J Med
1999
;
341
:
1097
–105.
2
Calle EE, Rodriguez C, Walker-Thurmond K, Thun MJ. Overweight, obesity, and mortality from cancer in a prospectively studied cohort of U.S. adults.
N Engl J Med
2003
;
348
:
1625
–38.
3
Berrigan D, Lavigne JA, Perkins SN, et al. Phenotypic effects of calorie restriction and insulin-like growth factor-1 treatment on body composition and bone mineral density of C57BL/6 mice: implications for cancer prevention.
In Vivo
2005
;
19
:
667
–74.
4
Mukherjee P, Abate LE, Seyfried TN. Antiangiogenic and proapoptotic effects of dietary restriction on experimental mouse and human brain tumors.
Clin Cancer Res
2004
;
10
:
5622
–9.
5
Dirx MJ, Zeegers MP, Dagnelie PC, van den Bogaard T, van den Brandt PA. Energy restriction and the risk of spontaneous mammary tumors in mice: a meta-analysis.
Int J Cancer
2003
;
106
:
766
–70.
6
Bertrand HA, Lynd FT, Masoro EJ, Yu BP. Changes in adipose mass and cellularity through the adult life of rats fed ad libitum or a life-prolonging restricted diet.
J Gerontol
1980
;
35
:
827
–35.
7
Harrison DE, Archer JR, Astle CM. Effects of food restriction on aging: separation of food intake and adiposity.
Proc Natl Acad Sci U S A
1984
;
81
:
1835
–8.
8
Kealy RD, Lawler DF, Ballam JM, et al. Effects of diet restriction on life span and age-related changes in dogs.
J Am Vet Med Assoc
2002
;
220
:
1315
–20.
9
Stuchlikova E, Juricova-Horakova M, Deyl Z. New aspects of the dietary effect of life prolongation in rodents. What is the role of obesity in aging?
Exp Gerontol
1975
;
10
:
141
–4.
10
Barzilai N, Gupta G. Revisiting the role of fat mass in the life extension induced by caloric restriction.
J Gerontol A Biol Sci Med Sci
1999
;
54
:
B89
–96; discussion B7–8.
11
Gower BA, Weinsier RL, Jordan JM, Hunter GR, Desmond R. Effects of weight loss on changes in insulin sensitivity and lipid concentrations in premenopausal African American and white women.
Am J Clin Nutr
2002
;
76
:
923
–7.
12
Monzillo LU, Hamdy O, Horton ES, et al. Effect of lifestyle modification on adipokine levels in obese subjects with insulin resistance.
Obes Res
2003
;
11
:
1048
–54.
13
Kopp HP, Krzyzanowska K, Mohlig M, et al. Effects of marked weight loss on plasma levels of adiponectin, markers of chronic subclinical inflammation and insulin resistance in morbidly obese women.
Int J Obes Relat Metab Disord
2005
;
29
:
766
–71.
14
Mohn A, Catino M, Capanna R, et al. Increased oxidative stress in prepubertal severely obese children: effect of a dietary restriction-weight loss program.
J Clin Endocrinol Metab
2005
;
90
:
2653
–8.
15
Iyengar P, Combs TP, Shah SJ, et al. Adipocyte-secreted factors synergistically promote mammary tumorigenesis through induction of anti-apoptotic transcriptional programs and proto-oncogene stabilization.
Oncogene
2003
;
22
:
6408
–23.
16
Schoen RE, Weissfeld JL, Kuller LH, et al. Insulin-like growth factor-I and insulin are associated with the presence and advancement of adenomatous polyps.
Gastroenterology
2005
;
129
:
464
–75.
17
Schiff R, Reddy P, Ahotupa M, et al. Oxidative stress and AP-1 activity in tamoxifen-resistant breast tumors in vivo.
J Natl Cancer Inst
2000
;
92
:
1926
–34.
18
Larsson SC, Permert J, Hakansson N, et al. Overall obesity, abdominal adiposity, diabetes, and cigarette smoking in relation to the risk of pancreatic cancer in two Swedish population-based cohorts.
Br J Cancer
2005
;
93
:
1310
–5.
19
Rapp K, Schroeder J, Klenk J, et al. Obesity and incidence of cancer: a large cohort study of over 145,000 adults in Austria.
Br J Cancer
2005
;
93
:
1062
–7.
20
Hu X, Juneja SC, Maihle NJ, Cleary MP. Leptin—a growth factor in normal and malignant breast cells and for normal mammary gland development.
J Natl Cancer Inst
2002
;
94
:
1704
–11.
21
Frankenberry KA, Somasundar P, McFadden DW, Vona-Davis LC. Leptin induces cell migration and the expression of growth factors in human prostate cancer cells.
Am J Surg
2004
;
188
:
560
–5.
22
Onuma M, Bub JD, Rummel TL, Iwamoto Y. Prostate cancer cell-adipocyte interaction: leptin mediates androgen-independent prostate cancer cell proliferation through c-Jun NH2-terminal kinase.
J Biol Chem
2003
;
278
:
42660
–7.
23
Cannon B, Nedergaard J. Brown adipose tissue: function and physiological significance.
Physiol Rev
2004
;
84
:
277
–359.
24
Greenberg NM, DeMayo F, Finegold MJ, et al. Prostate cancer in a transgenic mouse.
Proc Natl Acad Sci U S A
1995
;
92
:
3439
–43.
25
Nagy TR, Clair AL. Precision and accuracy of dual-energy X-ray absorptiometry for determining in vivo body composition of mice.
Obes Res
2000
;
8
:
392
–8.
26
Powell CS, Blaylock ML, Wang R, et al. Effects of energy expenditure and Ucp1 on photoperiod-induced weight gain in collard lemmings.
Obes Res
2002
;
10
:
541
–50.
27
Weir JB. New methods for calculating metabolic rate with special reference to protein metabolism.
J Physiol
1949
;
109
:
1
–9.
28
Folkvord JM, Viders D, Coleman-Smith A, Clark RA. Optimization of immunohistochemical techniques to detect extracellular matrix proteins in fixed skin specimens.
J Histochem Cytochem
1989
;
37
:
105
–13.
29
Wechter WJ, Leipold DD, Murray ED, Jr., et al. E-7869 (R-flurbiprofen) inhibits progression of prostate cancer in the TRAMP mouse.
Cancer Res
2000
;
60
:
2203
–8.
30
Mentor-Marcel R, Lamartiniere CA, Eltoum IE, Greenberg NM, Elgavish A. Genistein in the diet reduces the incidence of poorly differentiated prostatic adenocarcinoma in transgenic mice (TRAMP).
Cancer Res
2001
;
61
:
6777
–82.
31
Hursting SD, Lavigne JA, Berrigan D, Perkins SN, Barrett JC. Calorie restriction, aging, and cancer prevention: mechanisms of action and applicability to humans.
Annu Rev Med
2003
;
54
:
131
–52.
32
Masoro EJ. Commentary on “Revisiting the role of fat mass in the life extension induced by caloric restriction.”
J Gerontol A Biol Sci Med Sci
1999
;
54A
:
B97
.
33
Bluher M, Kahn BB, Kahn CR. Extended longevity in mice lacking the insulin receptor in adipose tissue.
Science
2003
;
299
:
572
–4.
34
Zhang Y, Proenca R, Maffei M, et al. Positional cloning of the mouse obese gene and its human homologue.
Nature
1994
;
372
:
425
–32.
35
Trayhurn P. Endocrine and signalling role of adipose tissue: new perspectives on fat.
Acta Physiol Scand
2005
;
184
:
285
–93.
36
van Dijk G. The role of leptin in the regulation of energy balance and adiposity.
J Neuroendocrinol
2001
;
13
:
913
–21.
37
Kutlu S, Canpolat S, Aydin M, et al. Exogenous leptin increases lipid peroxidation in the mouse brain.
Tohoku J Exp Med
2005
;
206
:
233
–6.
38
Catalano S, Marsico S, Giordano C, et al. Leptin enhances, via AP-1, expression of aromatase in the MCF-7 cell line.
J Biol Chem
2003
;
278
:
28668
–76.
39
Minocci A, Savia G, Lucantoni R, et al. Leptin plasma concentrations are dependent on body fat distribution in obese patients.
Int J Obes Relat Metab Disord
2000
;
24
:
1139
–44.
40
Arita Y, Kihara S, Ouchi N, et al. Paradoxical decrease of an adipose-specific protein, adiponectin, in obesity.
Biochem Biophys Res Commun
1999
;
257
:
79
–83.
41
Asayama K, Hayashibe H, Dobashi K, et al. Decrease in serum adiponectin level due to obesity and visceral fat accumulation in children.
Obes Res
2003
;
11
:
1072
–9.
42
Bub JD, Miyazaki T, Iwamoto Y. Adiponectin as a growth inhibitor in prostate cancer cells.
Biochem Biophys Res Commun
2006
;
340
:
1158
–66.
43
Kelley DE, Thaete FL, Troost F, Huwe T, Goodpaster BH. Subdivisions of subcutaneous abdominal adipose tissue and insulin resistance.
Am J Physiol Endocrinol Metab
2000
;
278
:
E941
–8.
44
Hursting SD, Switzer BR, French JE, Kari FW. The growth hormone: insulin-like growth factor 1 axis is a mediator of diet restriction-induced inhibition of mononuclear cell leukemia in Fischer rats.
Cancer Res
1993
;
53
:
2750
–7.
45
Mantzoros C, Petridou E, Dessypris N, et al. Adiponectin and breast cancer risk.
J Clin Endocrinol Metab
2004
;
89
:
1102
–7.
46
Goktas S, Yilmaz MI, Caglar K, et al. Prostate cancer and adiponectin.
Urology
2005
;
65
:
1168
–72.
47
Petridou E, Mantzoros C, Dessypris N, et al. Plasma adiponectin concentrations in relation to endometrial cancer: a case-control study in Greece.
J Clin Endocrinol Metab
2003
;
88
:
993
–7.
48
Gingrich JR, Barrios RJ, Kattan MW, et al. Androgen-independent prostate cancer progression in the TRAMP model.
Cancer Res
1997
;
57
:
4687
–91.
49
Chen H, Luo L, Liu J, Brown T, Zirkin BR. Aging and caloric restriction: effects on Leydig cell steroidogenesis.
Exp Gerontol
2005
;
40
:
498
–505.
50
Colbert LH, Mai V, Tooze JA, et al. Negative energy balance induced by voluntary wheel running inhibits polyp development in APCMin mice.
Carcinogenesis
2006
;
27
:
2103
–7.