SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information

Objective:

High dietary calcium (Ca) in the context of a dairy food matrix has been shown to reduce obesity development and associated inflammation in diet-induced obese (DIO) rodents. The influence of Ca and dairy on these phenotypes in the context of preexisting obesity is not known. Furthermore, interpretations have been confounded historically by differences in body weight gain among DIO animals fed dairy-based protein or high Ca.

Design and Methods:

Adiposity along with associated metabolic and inflammatory outcomes were measured in DIO mice previously fattened for 12 week on a soy protein-based obesogenic high fat diet (45% energy, 0.5% adequate Ca), then fed one of three high fat diets (n = 29-30/group) for an additional 8 week: control (same as lead-in diet), high-Ca (1.5% Ca), or high-Ca + nonfat dry milk (NFDM).

Results and Conclusion:

Mice fed high-Ca + NFDM had modestly, but significantly, attenuated weight gain compared to mice fed high-Ca or versus controls (P < 0.001), whereas mice fed high-Ca alone had increased weight gain compared to controls (P < 0.001). Total measured adipose depot weights between groups were similar, as were white adipose tissue inflammation and macrophage infiltration markers (e.g. TNFα, IL-6, CD68 mRNAs). Mice fed high-Ca + NFDM had significantly improved glucose tolerance following a glucose tolerance test, and markedly lower liver triglycerides compared to high-Ca and control groups. Improved metabolic phenotypes in prefattened DIO mice following provision of a diet enriched with dairy-based protein and carbohydrates appeared to be driven by non-Ca components of dairy and were observed despite minimal differences in body weight or adiposity.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information

Epidemiological or cross-sectional studies have shown an inverse relationship between dietary calcium (Ca) or dairy intake and body weight or fat in humans (1, 2). However, clinical intervention trials investigating this relationship have provided mixed results (3), suggesting that a potential antiobesity effect of increased Ca and/or dairy food consumption may be context-specific (prior Ca intake and status, source of calcium, types of dairy foods consumed, presence or absence of energy restriction, etc.). Metabolic changes promoting adiposity reduction in response to increased Ca or dairy remain to be determined and clarified. Reduced energy absorption resulting from increased fecal fat excretion due to formation of indigestible Ca soaps in the gastrointestinal tract may contribute to the antiobesity properties of high-Ca diets (4). Zemel et al. have hypothesized that low dietary Ca, via associated up-regulation of plasma 1,25-dihydroxyvitamin D (calcitriol), promotes net lipid accumulation (5-7). Thus, according to this model, increasing dietary Ca acts to reduce plasma calcitriol and obesity-associated outcomes. Additional bioactive components of dairy foods (e.g. branched chain amino acids [BCAA], angiotensin converting enzyme [ACE]-inhibitory peptides or other factors) are implicated in antiobesity outcomes, since high-Ca in a dairy matrix yielded greater effects versus elemental Ca alone in aP2-agouti transgenic mice fed an obesogenic diet (8, 9). Studies demonstrating an antiobesity effect of increased dietary Ca in polygenic obesity models have primarily been only in the context of nonfat dry milk (NFDM) or whey providing all dietary protein (10-12). Thus, comparing a high-Ca diet in matrices containing differing protein and carbohydrate sources (dairy and nondairy) is of interest to understand if non-Ca factors in dairy foods play a role in regulation of metabolism and obesity phenotypes.

In addition to potential weight- and adiposity-modulating properties, dietary Ca or other dairy components may directly influence inflammation. Cell culture studies showed increased reactive oxygen species (ROS) generation in adipocytes and macrophages treated with calcitriol, which led to increased pro-inflammatory cytokine gene expression and protein secretion (13, 14). Sun and Zemel showed reduced white adipose tissue (WAT) pro-inflammatory gene expression and circulating cytokines concurrent with reduced plasma calcitriol in aP2-agouti transgenic mice fed high dietary Ca (15, 16), but decreased WAT and systemic inflammation could not be dissociated from reductions in body weight and fat gain in those studies. Recently, reduced inflammation markers in overweight humans was seen in response to feeding a dairy- and Ca-rich weight-maintenance diet, in contrast to no effect by a soy-based low-Ca diet (17). Reduced inflammation occurred without measurable differences in body composition and in conjunction with plasma calcitriol reductions, supporting a possible direct anti-inflammatory role for high dietary Ca. Yet, potential effects of the dairy matrix per se could not be excluded.

Most animal model studies investigating the role of dietary Ca and/or dairy in the reduction of obesity or obesity-associated adipose inflammation have been conducted in the context of obesity prevention, i.e. examining their potential to minimize weight gain over time with provision of a high fat diet (5, 10-12, 18). This preventive model does not reflect typical human intervention scenarios in which the influence of diet on adiposity is assessed in already-overweight persons. Therefore, we have addressed whether high dietary Ca or dairy can attenuate weight gain, adiposity, and associated inflammation and metabolic perturbations in the context of preexisting obesity. These studies leveraged a male C57BL/6J diet-induced obese (DIO) mouse model previously validated to display WAT inflammation and impaired glucose tolerance outcomes when fed a soy protein- and sucrose/cornstarch-based 45% fat diet for 12 week (19). In addition to determining the effect of switching to a high-Ca diet (from 0.5% to 1.5%) in prefattened DIO mice, our study design enabled a comparison of metabolic and inflammatory outcomes in DIO mice fed high-Ca compared to those fed high-Ca in a dairy matrix (NFDM as the sole protein source and containing primarily dairy-based carbohydrates). We hypothesized that high Ca intake would attenuate obesity and WAT inflammation, and further postulated that these effects would be augmented in DIO mice fed a high-Ca diet but with dairy-based protein and carbohydrate sources. To our knowledge, this study represents the first to determine if provision of high Ca or dairy improves metabolism and limits inflammation in the context of ad lib feeding in a polygenic animal model of preexisting obesity.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information

DIO mice

All animal protocols were approved by the University of California at Davis Institutional Animal Care and Use Committee according to Animal Welfare Act guidelines. Four-week-old male C57BL/6J mice were purchased from the Jackson Laboratory and individually housed under standard temperature (20-22°C) and light/dark cycle (12 h:12 h) conditions. Individually-housed mice were fed Picolab® Mouse Diet 20 (Purina LabDiet®) providing ∼22%, ∼22%, and ∼56% energy as protein, fat, and carbohydrate, respectively, for a 1 week acclimation period. At 5 week of age, mice were fed a soy protein-based (0.5% Ca) high fat (45% energy) control diet for 12 weeks. This diet and 12 week timeframe were previously shown to elicit obesity along with increased WAT inflammation and insulin resistance in this model (19). Weight matched DIO mice were randomly assigned to one of three macronutrient matched high fat diets for an additional 8 weeks: the soy protein-based obesogenic control diet (0.5% Ca, n = 29; Control, continued from the prefattening period), soy protein-based high-Ca (1.5% Ca, n = 30; high-Ca), or high-Ca in the context of nonfat dry milk protein and carbohydrates (1.5% Ca, n = 30; high-Ca + NFDM). These diets have been previously described in detail (18). In the case of the NFDM diet, to ensure equal macronutrient energy contribution compared to the other diets while maintaining a dairy nutrient matrix, protein (casein and whey-derived lactalbumin) and carbohydrate (galactose + glucose, the components of lactose) were added. Ca phosphate (CaPO4) was added to the soy-protein based diets to match that naturally-derived from NFDM (to control for type of dietary Ca), with Ca carbonate providing all additional Ca to the experimental diets as typically used in rodent studies comparing different dietary Ca. Cellulose content was adjusted accordingly to account for added Ca in the high-Ca diets to ensure matched macro and micronutrient content of diets, and amounts of fiber are all within the normal range used in purified mouse diets. Animals were given free access to food and water with body weight and food intake (plus spillage) measurements made every 2-3 days. Fecal collections over 48 h were made on a subset of randomly chosen mice (n = 10/group) at week 6 of intervention. Feces were weighed and stored at −80°C until bomb calorimetry was performed by Covance Laboratories Inc. Approximate % fecal energy loss was calculated from each individual animal's 48 h fecal energy loss and energy consumption. Glucose tolerance tests were administered on a subset of randomly chosen mice (n = 16/group) at week 7 of intervention, using a protocol reported elsewhere (19). At week 8 of intervention, mice were briefly food deprived (between 3 and 8 hours starting @ 06:00) prior to tissue and blood collection as described in detail (19).

Total RNA isolation and gene expression analyses

Total RNA was isolated from whole retroperitoneal (RP) fat pads, cDNA was prepared, and 384-well quantitative PCR utilizing gene-specific Taqman® primers and FAM-MGB labeled probes was conducted as previously described in detail (19). In preliminary studies comparing inflammation marker mRNA levels in different WAT depots in a separate study, the RP-WAT was found to display robustly increased markers in response to a high fat diet (data not shown) and was, therefore, used herein for inflammation read-outs.

Plasma and liver analytes, Western blot analysis

Plasma cytokines/chemokines (n = 29-30/group), insulin (n = 29-30/group), glucose (n = 29-30/group), and calcitriol (n = 9-10/group) were measured as previously described (18). Liver lipid was extracted and triglyceride content measured (n = 10/group) as previously described (18). Uncoupling protein (UCP1) protein content of intrascapular brown adipose tissue (BAT) (n = 8/group) was quantified as previously described (18).

Statistical analysis

A one-way ANOVA with Newman-Keuls Multiple Comparison post hoc test was used for three-group comparisons (Prism v. 4.0, Graphpad). Two-way repeated measures ANOVA was used to assess effects of diet, time, and diet × time interactions on body weight gain and cumulative energy intake. To assess the contribution of differences in body weight on markers of RP-WAT macrophage infiltration and inflammation, and to assess the impact of food intake on weight gain, analysis of covariance (ANCOVA) was used to compare group means of the outcome variables controlling for the indicated covariate (SAS for Windows Release 9.2). Data are expressed as mean ± SEM and differences considered to be statistically significant at P ≤ 0.05.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information

Body weight gain, cumulative energy intake, and feed efficiency

Body weight gain over 8 weeks was modestly, but significantly, reduced in previously-fattened DIO mice fed high-Ca + NFDM compared to high-Ca or control mice (Figure 1A). In contrast, body weight gain in mice fed high-Ca was significantly increased versus controls (Figure 1A). There was a significant diet × time interaction on body weight gain (P < 0.0001). Compared to controls, body weight was lower starting on day 14 (P < 0.05) through day 46 (P < 0.05) for mice fed high-Ca + NFDM and heavier starting at day 28 (P < 0.05) of diet onward for mice fed high-Ca. Mice fed high-Ca + NFDM weighed significantly less than mice fed high-Ca at day 7 (P < 0.05) of diet onward.

thumbnail image

Figure 1. Body weights (A) and cumulative energy intake (B) in previously fattened DIO mice fed 0.5% Ca (Control), 1.5% Ca (high-Ca), or 1.5% Ca + nonfat dry milk (high-Ca + NFDM). Values are means ± SE, n = 29-30/treatment. Some error bars are within symbols. Treatments with different letters are significantly different (P < 0.05).

Download figure to PowerPoint

Mice fed high-Ca had significantly increased cumulative energy intake compared to both controls and high-Ca + NFDM fed mice (Figure 1B). Body weight gain was not significantly different between controls and mice fed high-Ca when controlling for energy consumption (ANCOVA), indicating that energy intake largely contributed to the greater weight gain in the high-Ca diet treatment group. This is consistent with our previous observation utilizing these experimental diets in an obesity prevention paradigm (18). There was a significant diet × time interaction on cumulative energy intake (P < 0.0001), with mean cumulative energy intake higher starting at day 21 (P < 0.05) and day 11 (P < 0.05) in mice fed high-Ca compared to controls and high-Ca + NFDM fed mice, respectively. Although cumulative energy intake was significantly increased in mice fed high-Ca + NFDM compared to controls, feed efficiency (mg weight gain/kJ consumed) was significantly decreased in these mice (0.7 ± 0.2; P < 0.001) compared to controls (1.9 ± 0.2). Feed efficiency was significantly increased in mice fed high-Ca (3.1 ± 0.2; P < 0.001) relative to controls.

In a prior obesity prevention experiment comparing these diets, significantly reduced fecal energy density and lower fecal energy loss occurred in DIO mice fed high-Ca without dairy (18). Furthermore, we previously observed increased fecal energy loss in mice fed high-Ca + NFDM due to greater fecal output rather than increased fecal energy density. Consistent with these observations, fecal energy density and energy loss were significantly decreased in mice fed high-Ca compared to controls and high-Ca + NFDM fed mice (Supporting Information Supplemental Figure 1B and C). Fecal energy density did not differ significantly between the latter groups; however, fecal energy loss was increased in mice fed high-Ca + NFDM compared to both controls and mice fed high-Ca due to significantly increased fecal output (Supporting Information Supplemental Figure 1A and C). The net energy losses across the gut corresponded to significantly increased digestible energy availability (kJ/d) in mice fed high-Ca (62.6 ± 1.2) compared to controls (54.0 ± 1.2; P < 0.001) and high-Ca + NFDM fed mice (55.6 ± 1.4: P < 0.01); the latter groups did not differ significantly from each other.

Markers of BAT activation were examined as potential contributors to differences in feed efficiency across diet treatment groups. BAT weight was significantly decreased and increased in mice fed high-Ca + NFDM and high-Ca, respectively, compared to controls (Table 1). UCP1 protein content in BAT was significantly increased in mice fed high-Ca compared to controls, but not compared to high-Ca + NFDM fed mice (Supporting Information Supplemental Figure 2); the latter group was not significantly different than controls. Thus, feed efficiency was increased in mice fed high-Ca despite increased BAT size and UCP1 protein content. These observations suggest that BAT thermogenic potential did not explain feed efficiency differences across groups.

thumbnail image

Figure 2. Absolute (A) and relative (B) WAT depot weights in previously fattened DIO mice fed 0.5% Ca (Control), 1.5% Ca (high-Ca), or 1.5% Ca + nonfat dry milk (high-Ca + NFDM). Values are means ± SE, n = 29-30/treatment. Treatments with different letters are significantly different (P < 0.05).

Download figure to PowerPoint

Table 1. Tissue weights in prefattened DIO mice fed 0.5% Ca (Control), 1.5% Ca (high-Ca), or 1.5% Ca + nonfat dry milk (high-Ca + NFDM)a
TissueControlHigh-CaHigh-Ca + NFDM
  • a

    Values are means + SE, n = 29-30/treatment. Values with different letters are significantly different (P < 0.05).

Liver (g)1.7 ± 0.11a2.4 ± 0.09b1.4 ± 0.05c
Liver (% body weight)4.2 ± 0.19a5.3 ± 0.17b3.7 ± 0.05c
Gastrocnemius muscles (mg)297 ± 8.3295 ± 9.3291 ± 8.7
Gastrocnemius muscles (% body weight)0.7 ± 0.01a0.6 ± 0.02b0.8 ± 0.03a,c
Brown adipose tissue (mg)280 ± 13.4a358 ± 8.8b198 ± 12.7c
Brown adipose tissue (% body weight)0.7 ± 0.02a0.8 ± 0.02b0.5 ± 0.02c
Spleen (mg)90 ± 2.9a99 ± 1.9b77 ± 1.9c
Spleen (% body weight)0.2 ± 0.010.2 ± 0.000.2 ± 0.00

Body composition and metabolic phenotypes

Absolute weights of RP and epididymal (EPI) fat pads were not significantly different between diet treatment groups (Figure 2A), but subcutaneous (SC) fat pads weighed significantly less in mice fed high-Ca + NFDM and weighed more in high-Ca compared to controls (Figure 2A). Total measured WAT depot weight was not significantly different between the diet treatment groups (Figure 2A). Total WAT expressed as a % of body weight (Figure 2B) was significantly decreased in mice fed high-Ca compared to controls, but not high-Ca + NFDM fed mice. The latter two groups were not significantly different from each other. Since total WAT weight was not different between the groups and SC fat pads were heavier in mice fed high-Ca, the reduction in total WAT as a % of body weight observed in mice fed high-Ca was primarily due to heavier body weights in this diet treatment group.

Notably, liver weights were significantly heavier in DIO mice fed high-Ca compared to both controls and mice fed high-Ca + NFDM (Table 1), which corresponded to significantly increased liver triglycerides (mg/g) in mice fed high-Ca (214.2 ± 20.0) compared to controls (120.2 ± 20.6; P < 0.001) and high-Ca + NFDM fed mice (41.4 ± 10.8; P < 0.001). Liver triglycerides were also significantly reduced in mice fed NFDM compared to controls (P < 0.01). Absolute weights of gastrocnemius muscles were not significantly different between groups (Table 1).

DIO mice fed high-Ca + NFDM had improved glucose tolerance (Figure 3A) with significantly reduced glucose area under the curve (AUC) from baseline compared to both controls and mice fed high-Ca (Figure 3B); the latter groups did not differ significantly from each other. Postabsorptive plasma glucose concentrations (mmol/L) measured at the time of tissue collection was also significantly reduced in NFDM fed mice (15.7 ± 0.7) compared to controls (19.0 ± 0.7; P < 0.01) and mice fed high-Ca (19.3 ± 0.9 mmol/L; P < 0.01), which did not differ significantly from each other. Postabsorptive plasma insulin (pmol/L) was significantly lower in mice fed high-Ca + NFDM (113.6 ± 12.5; P < 0.01) and higher in mice fed high-Ca (419.9 ± 37.6; P < 0.001) compared to controls (267.7 ± 42.8).

thumbnail image

Figure 3. Blood glucose concentrations (A) and AUC (B) in previously fattened DIO mice fed 0.5% Ca (Control), 1.5% Ca (high-Ca), or 1.5% Ca + nonfat dry milk (high-Ca + NFDM) following an i.p. glucose tolerance test. Values are means ± SE, n = 16/treatment. Some error bars are within symbols. Treatments with different letters are significantly different (P < 0.05). Fasting blood glucose (mmol/l) in this subset of animals was increased in mice fed high-Ca (15.4 ± 0.5; P < 0.05) and reduced in mice fed high-Ca + NFDM (11.2 ± 0.3; P < 0.01) compared to controls (13.7 ± 0.7).

Download figure to PowerPoint

Inflammatory phenotypes and plasma calcitriol

We have previously shown reduced mRNA expression of genes associated with obesity-induced WAT macrophage infiltration, inflammation, and hypoxic stress in RP fat of DIO mice fed high-Ca + NFDM over 12 week of obesity development (18). Interestingly, in that study mRNA expressions of macrophage markers were highly correlated to body weight and adiposity (over a wide range in 90 mice) regardless of dietary treatment. To control for the potential influence of body weight differences on adipose inflammation outcomes in the current experiment, body weight gain was used as a covariate in comparisons of inflammation markers in RP-WAT. In this context, there were no significant differences in CD68, CD11d, monocyte chemoattractant protein (MCP)-1, tumor necrosis factor (TNF)α, interleukin (IL)-6, or hypoxia inducible factor (HIF)-1α mRNA abundances between diet treatments. In direct comparisons of gene expression across groups by ANOVA, there were no significant differences in RP-WAT mRNA abundances of CD68, CD11d, TNFα, IL-6, or HIF-1α, or of IL-10 (an anti-inflammatory cytokine), found in inflammatory zone (FIZZ1) (a marker of alternatively activated macrophages), or calcitonin gene-related peptide (CGRPα) (an inflammatory modulator thought to be responsive to change in Ca status) (Table 2). Using this analysis, WAT MCP-1 mRNA level was significantly reduced in mice fed high-Ca with or without NFDM compared to controls (Table 2). Plasma markers associated with inflammation were not significantly different between diet treatment groups (Table 3).

Table 2. RP-WAT mRNA levels in diet induced obese mice fed 0.5% Ca (Control), 1.5% Ca (high-Ca), or 1.5% Ca + nonfat dry milk (high-Ca + NFDM)a
GeneDiet
ControlHigh-CaHigh-Ca + NFDM
% of Control
  • a

    Values are means + SE, n = 29 or 30/treatment. Values with different letters are significantly different (P < 0.05).

Cd68100.0 ± 6.884.9 ± 8.777.9 ± 9.4
Itgad (CD11d)100.0 ± 6.598.9 ± 11.080.7 ± 11.4
Ccl2 (MCP-1)100.0 ± 7.1a75.9 ± 6.5b75.1 ± 7.3b
Tnf (TNFα)100.0 ± 4.886.1 ± 7.680.4 ± 7.4
Il6100.0 ± 6.486.0 ± 5.979.2 ± 6.3
Hif1a (HIF-1α)100.0 ± 2.592.9 ± 3.290.0 ± 3.8
Il10100.0 ± 7.990.4 ± 10.774.8 ± 7.8
Retnla (FIZZ-1)100.0 ± 6.7115.8 ± 7.8117.9 ± 7.8
Calca (CGRPα)100.0 ± 5.188.0 ±6.880.5 ± 6.4
Fas (fatty acid synthase)100.0 ± 7.6108.3 ± 7.7114.6 ± 7.2
Table 3. Plasma cytokines/chemokines in prefattened DIO mice fed 0.5% Ca (Control), 1.5% Ca (high-Ca), or 1.5% Ca + nonfat dry milk (high-Ca + NFDM)a
Cytokine/ chemokineDiet
ControlHigh-CaHigh-Ca + NFDM
ng/l
  • a

    Values are means + SE, n = 29-30/treatment.

IL-611.7 ± 4.915.7 ± 7.41.2 ± 0.5
IL-105.7 ± 1.94.2 ± 1.53.5 ± 1.6
IL-12(p70)15.3 ± 4.615.5 ± 5.54.6 ± 1.5
MCP-117.6 ± 4.99.2 ± 2.511.6 ± 3.4
TNF-α8.6 ± 2.34.8 ± 1.55.9 ± 2.1
IL-1564.6 ± 16.667.0 ± 15.638.4 ± 14.1

Elevated circulating calcitriol due to low dietary Ca intake is proposed to contribute to increased adiposity and associated WAT inflammation (20). Plasma calcitriol was significantly decreased in mice fed high-Ca (129.7 ± 6.0 pmol/L: P < 0.001) and NFDM fed mice (194.3 ± 20.5 pmol/L; P < 0.01) compared to controls (279.0 ± 21.0 pmol/L); mice fed high-Ca had significantly lower plasma calcitriol than high-Ca + NFDM fed mice (P < 0.05). These observations indicate that inflammation and obesity phenotypes of mice fed high-Ca with or without dairy are not correlated to changes in circulating calcitriol.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information

Previous animal model studies investigating putative antiobesity properties of dietary Ca or dairy have only been in the context of developing obesity or energy-restricted weight loss (5, 10-12, 18, 21, 22). The influence of these dietary components on body composition and associated phenotypes (metabolic and inflammatory) had not previously been evaluated in the context of preexisting obesity and maintenance on a high fat diet in the ad lib state. Such a paradigm is useful since it allows a more direct comparison of the influence of dietary components on metabolism and inflammation without stark body weight differences or food restriction. The current data do not support the idea that high dietary Ca in isolation provides protection against obesity and associated inflammation, since DIO mice fed a soy protein-based high-Ca diet actually gained more body weight and had higher feed efficiency compared to lower-Ca controls, similar to what was recently observed in an obesity-prevention setting in DIO mice (18). In contrast, DIO mice fed high-Ca in a NFDM matrix gained less body weight and had improved blood sugar control compared to animals fed the high-Ca soy-based diet, and versus lower-Ca control mice. This differs from observations in high-Ca fed aP2-agouti transgenic mice fed an obesity-promoting diet where increased dietary Ca attenuated weight gain, adiposity, and WAT inflammation (5, 9, 15, 16, 23). Experimental variables that may have contributed to these discrepancies include differences in diet composition, energy consumption, genetic variation, and study duration. Since the high-Ca diet in the current study led to hyperphagia, sensory considerations are possible. Notably, studies demonstrating an antiobesity effect of high dietary Ca in polygenic models of rodent obesity have only been in the context of other dairy food components (e.g. NFDM or whey providing all dietary protein) (10-12, 22, 24), suggesting that non-Ca factors explain reduced adiposity under these conditions. Recently, Eller and Reimer provided data indicating that dairy protein type, rather than Ca content, was the primary driver of adiposity reduction in DIO rats (24). In addition to protein, one cannot exclude the possible contribution of NFDM carbohydrates or other factors that might influence metabolism through changing the gut microbiome or via other mechanisms.

The present results are consistent with our previous observation utilizing this DIO mouse model in the context of developing obesity (18), which showed increased body fat accumulation, higher energy consumption, reduced fecal energy loss, and enhanced feed efficiency with high-Ca feeding. In contrast, the high-Ca NFDM-based diet led to less weight gain despite equal net digestible energy compared to controls, indicating higher thermogenesis in the high-Ca NFDM fed mice. The mechanisms for this outcome are not understood. Although suppression of circulating calcitriol following high-Ca feeding has been proposed as a primary mechanism for adiposity reduction (25), body weight differences across diets in the current study or in a previous obesity prevention paradigm (18) are not correlated with differences in plasma calcitriol concentrations. This study further supports the notion that components of dairy other than Ca are primarily responsible for antiobesity properties in polygenic models of rodent obesity, and independent of circulating calcitriol. Dietary whey protein (rich in BCAA) or leucine delivered in drinking water reduces adiposity and increases energy expenditure in DIO mice (26-28). Leucine when provided to muscle cell cultures increased energy consumption and mitochondrial biogenesis markers (e.g. PGC-1 expression, mitochondrial mass), hypothesized to result from activation of mammalian target of rapamycin complex 1 (mTORC1) (29, 30). Whether BCAA or other bioactive components derived from the NFDM diet contributed to metabolic outcomes requires experimental validation.

We have previously observed strong correlations between body weight and RP-WAT mRNA abundance of macrophage markers (e.g. CD68 & CD11d) across a broad range of body weights in growing mice fed 10% or 45% energy as fat for 12 weeks (19), or in DIO mice fed the diets used herein (18). These findings support the concept that macrophage infiltration in WAT is closely yoked to body weight and is not strongly influenced by dietary Ca or NFDM per se. The present study is consistent with this concept, since RP-WAT macrophage mRNA patterns and inflammation markers were not significantly impacted by diets introduced after obesity development had already taken place and when differences in body weight were modest. Altogether, these observations support the hypothesis that—notwithstanding the pathophysiological inflammatory condition of severe obesity—macrophages have a normal, nonpathological role in facilitating WAT expansion, and remodeling during conditions of positive energy balance (e.g. as obesity develops or during sustained periods of overnutrition) to accommodate changing energy storage needs. In this model, WAT macrophage infiltration and activities would necessarily be regulated by as-yet uncharacterized signals linked to body weight, adiposity, and/or energy balance.

Interestingly, despite only modest weight and adiposity differences across dietary treatments, there was a marked improvement in glucose homeostasis and striking reduction in liver fat in DIO mice fed NFDM compared to those fed high-Ca or low-Ca in a soy protein- and sucrose/corn starch-based diet. Improved glucose tolerance and reduced blood insulin coupled to lower liver steatosis in NFDM-fed mice was also seen in our obesity-prevention paradigm (18). These observations imply a direct effect of dairy food factors on these metabolic parameters, but the underlying mechanisms remain to be established in future studies focused on fatty acid metabolite flux in liver, hepatic glucose output, and insulin action. One possibility is that dietary NFDM in this model activates hepatic fatty acid oxidation or promotes export of free fatty acids, thus limiting steatosis despite the high fat diet. Whether or not possible NFDM-associated thermogenesis also plays a role requires further consideration through calorimetry testing. Another consideration is sucrose, since in rodents a high sucrose diet promotes liver steatosis (31), and this sugar was reduced in the NFDM diet (replaced with NFDM-derived lactose plus lactose sugars); furthermore, lactose has been shown to reduce fat accretion in DIO rats (32). Additionally, a BCAA-enriched diet reduced hepatic steatosis in DIO mice (33). Alternatively, metabolic effects emanating from diet-related changes in gut microbiota patterns may be important, and this possibility is currently being investigated.

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information

A role for increased dietary Ca in isolation in attenuating obesity is not supported in the current DIO mouse model since high-Ca fed in a soy protein- and sucrose/cornstarch-based diet increased weight gain, in stark contrast to animals fed high-Ca in a NFDM-based matrix. Identification of the components of milk that modify body weight regulation and metabolism, elucidating their mechanisms-of-action, and determining if learnings from animal models apply to human obesity are interesting avenues for future research.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information

We thank Dr. Trina Knotts for guidance and technical assistance with Western blot analyses and quantification of liver triglycerides, Jan Peerson for statistical analysis, and the UC Davis vivarium staff in helping care for the animals. USDA is an equal opportunity provider and employer.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
OBY_20039_sm_SuppFigs.pdf23KSupporting Information Figures

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.