Inulin-type fructans are thought to promote bone health through positive actions on mineral retention leading to increased peak bone mass achievement and bone mass conservation during ageing. Experimental studies in animal models have reported beneficial effects on bone mineral content (BMC), bone mineral density (BMD) and gastrointestinal absorption of Ca and other minerals(Reference Weaver1). There is strong evidence of enhanced BMC and BMD in growing male rodents fed inulin-type fructans(Reference Chonan and Watanuki2–Reference Nzeusseu, Dienst and Haufroid6), as well as improved bone health in adult, ovariectomized rats(Reference Taguchi, Ohta and Abe7–Reference Zafar, Weaver and Jones10). Although results of short-term human trials have been mixed, a recent study in adolescent females has shown an increase in Ca absorption as well as BMC and BMD after feeding inulin-type fructans for 8 weeks and 1 year(Reference Abrams, Griffin and Hawthorne11, Reference Abrams, Griffin and Hawthorne12). Effects of inulin on bone mineral retention have not, to the best of our knowledge, been reported in a growing female rodent model. This is an important model to develop, as the attainment of peak bone mass is especially critical in young, growing females in order to protect against osteoporosis.
Precise mechanisms for this effect have not been defined, although it is widely believed that the fermentation of inulin-type fructans in the large intestine results in SCFA production, leading to a decline in intestinal pH(Reference Weaver1). Mineral solubility is thought to increase as pH drops in the large intestine, indirectly increasing mineral absorption through the paracellular pathway(Reference Weaver1). Thus, enhanced mineral bioavailability could explain the increased bone mass accretion and conservation reported with inulin feeding. Alternatively, a fructan-induced trophic effect may also explain increased paracellular mineral absorption in the colon(Reference Topping and Clifton13). Feeding inulin-type fructans to rats has been shown to promote ileal growth as well as mRNA levels of glucagon-like peptide-1 in the ileum and caecum(Reference Topping and Clifton13). Finally, there is preliminary evidence of increased caecal absorptive surface area, increased SCFA concentration and larger intestinal pools of soluble and ionized Ca (to promote the concentration gradient) as potential mechanisms to explain the effect of inulin on mineral absorption(Reference Raschka and Daniel14).
In addition to effects on bone, inulin may positively impact growth quality with regard to body composition. Recent studies have reported decreased feed intake, weight gain and fat pad mass in male rats fed various inulin-type fructans(Reference Cani, Neyrinck and Maton15, Reference Delzenne, Cani and Daubioul16). However, data on the response of whole body (WB) fat and lean mass to inulin is limited. Body weight affects bone mass; however, the relative contributions of lean and fat mass to bone mass accretion have not been fully clarified(Reference Janicka, Wren and Sanchez17). Lean mass is positively associated with increased bone mass in the appendicular and axial skeletons(Reference Janicka, Wren and Sanchez17–Reference Petit, Beck and Shults20). Several studies have reported a negative association between fat mass and bone mass(Reference Janicka, Wren and Sanchez17, Reference Weiler, Janzen and Green21, Reference Young, Hopper and Macinnis22). Thus, a positive shift in body composition from fat to lean mass may promote bone mass accretion.
Inulin-type fructans are considered non-digestible carbohydrates as the β-configuration of the anomeric C2 in the fructose monomer is resistant to enzymatic hydrolysis(Reference Roberfroid23). Both inulin and oligofructose (OLF) are by-products of chicory root, which can be distinguished by their degree of polymerization (DP). Inulin is a linear β(2 → 1) fructan with a DP of 2 to 60 (DPavg 12) and OLF is produced by the partial enzymatic hydrolysis of inulin and has a DP of 2 to 8 (DPavg 4)(Reference Roberfroid23). Long-chain inulin (LC-inulin; DPavg 25) and OLF-enriched inulin can also be produced(Reference Roberfroid23). Beneficial effects on bone mineralization have been demonstrated with inulin, OLF and OLF-enriched inulin, but studies directly comparing these different types of fibre indicate that fructans with the highest DP produce positive effects of greater magnitude with regard to bone density(Reference Roberfroid, Cumps and Devogelaer4–Reference Nzeusseu, Dienst and Haufroid6). Therefore, the objective of the present study was to investigate the effect of dietary LC-inulin on bone mineralization and density, as well as growth quality, in growing, female rats.
Materials and methods
Animals and diets
Forty-eight weanling (3-week old) female Sprague–Dawley rats weighing approximately 55–65 g were obtained from Charles River Laboratories (St Constant, Quebec, Canada). Animals were housed individually in stainless steel hanging cages in a temperature- (21–23°C) and humidity- (55 %) controlled room with a light:dark cycle of 14:10 h. Following a 1 week acclimatization period to a nutritionally complete diet, rats were randomly assigned to treatment groups. Twenty rats were fed a diet containing 10 % cellulose (CL diet) and twenty rats were fed a diet containing 5 % LC-inulin (IN diet) (Beneo HP, Orafti Group, Tienen, Belgium; IN; DPavg>23 and 100 % inulin) and 5 % cellulose. Ten rats were killed from each group at the end of 4 weeks (CL-4 and IN-4) and after 8 weeks (CL-8 and IN-8). Weanling rats were chosen so as to assess a critical period in the development of peak bone mass. In addition, by the end of the dietary treatment, rats had likely not yet reached sexual maturity, thus limiting hormonal effects in the study.
Experimental diets were based on the AIN-93G formulation(Reference Reeves, Nielsen and Fahey24) and provided ad libitum. (With the exception of fibre (cellulose or inulin) the diets contained the following: 54 % carbohydrate (31·3 % maize starch, 13·2 % maltodextrin, 10 % sucrose); 21 % egg white; 3·5 % Zn-free AIN-93G-MX mineral mix; 1 % Zn premix (5·775 g ZnCO3/kg dextrose); 1 % AIN-93-VX vitamin mix; 1 % biotin (200 mg biotin/kg dextrose); 0·54 % potassium phosphate; 0·25 % choline; 0·0014 % tert-butylhydroquinone; 7 % soyabean oil.) Ca concentration was kept constant across diets at 5 mg/kg. Deionized water was also provided ad libitum in plastic bottles with stainless steel sipper tubes. Feed intake was determined daily and body weights were measured weekly. The protocol for animal care procedures received ethics approval from the University of Manitoba Protocol Management and Review Committee.
Tissue collection
All animals were killed by CO2 asphyxiation and exsanguination in keeping with the guidelines of the Canadian Council on Animal Care. Body weights were determined post mortem and trunk blood was collected and stored on ice until centrifuged to obtain serum (2400 rpm at 4°C, 10 min). Parametrial fat pads were excised, rinsed briefly with PBS and blotted dry before weighing and being frozen in liquid nitrogen. Blood samples were stored at − 80°C until analysis.
Dual-energy X-ray absorptiometry scans
The WB, spine, femur and tibia of rat carcasses were analysed for bone area, BMC, BMD, lean body mass, fat body mass and total mass in situ by dual-energy X-ray absorptiometry (DXA, 4500A; Hologic Inc., Bedford, MA, USA; small animal software high resolution option). Animals were placed dorsally, in an anterior–posterior position. DXA has shown high-quality precision and accuracy in measuring BMC and BMD in small animals in situ (Reference Lochmüller, Vung and Weusten25). The precision error (CV %) for triplicate scans of bone area, BMC and BMD was 2·0, 1·4 and < 0·1, respectively, for the WB, 5·9, 7·4 and 6·9, respectively, for the spine, 11·4, 7·8 and 4·8, respectively, for the femur and 8·9, 9·2 and 4·6, respectively, for the tibia.
Lean, fat and total body mass analysis was done on carcasses following removal of trunk blood, spleen, stomach, small and large intestine (including caecum) and parametrial fat pads. Lean fat mass analysis with DXA varies with the extent of evisceration but generally has a CV of < 10 % from anaesthetized rats to completely eviscerated rats, based on studies in our laboratory (unpublished data). The precision error (CV %) for triplicate scans of fat mass, lean mass, total body mass and % fat was 1·7, 0·4, 0·2 and 2·3, respectively.
Bone morphometry
Following in situ DXA analysis, right and left femurs were excised and thoroughly cleaned of soft tissue. Morphometric measures were taken with digital callipers as previously described(Reference Reichling and German26) and recorded to the nearest 0·01 mm. All measures were reproduced in triplicate by the same trained examiner and included length, diaphysis width, femoral neck width and femoral head width.
Mineral analysis
Wet and dry weights of right femurs were determined prior to acid digestion for mineral analysis. Femurs were wet digested with trace-element grade nitric acid (4 ml per sample), as previously described(Reference Clegg, Keen and Lönnerdal27) in disposable DigiPrep tubes (SCP Science, Baie d'Urfé, Quebec, Canada) over 6–7 h at 85°C in a DigiPrep HP heater (24 well; SCP Science). Acid digests were diluted appropriately with double deionized water prior to analysis of Ca and P by inductively coupled plasma optical emission spectrometry analysis (Varian Liberty 200, Varian, Canada). Detection limits were 0·1 mg/ml for Ca and 0·5 mg/ml for P.
Biochemical assays
Osteoblast and osteoclast activity were measured by enzyme-linked immunosorbent assays specific for rat osteocalcin (Rat-Mid Osteocalcin; Osteometer BioTech, Herlev Hovengrade, Denmark) and bone-related plasma degradation products of C-terminal peptides of type I collagen in rats, respectively (CTX-I; Osteometer Biotech). Osteocalcin is considered a marker of osteoblast activity (bone formation), whereas CTX-I is a marker of osteoclast activity (bone resorption). The assay was performed according to the manufacturer's instructions. Samples were analysed in duplicate and agreement was ≧82 %.
Statistical methods
Data were analysed for main effects and interactions of inulin feeding over time by two-way ANOVA with a level of significance of 0·05, using SAS software version 9.1 (SAS Institute, Cary, NC, USA). Data were checked for normality and homogeneity of variance and transformed when necessary. Transformed data are indicated where appropriate; however, only non-transformed means are reported. In the case that data transformation could not correct non-normal data, significant main effects were verified by a non-parametric test (Kruskal–Wallis analysis of ranks). Significant differences were determined among treatment groups with Duncan's multiple range test. All data are reported as means with their standard errors of the mean.
Results
Growth and feed intake
Initial body weights at the start of the study ranged from 96 (sem 4) to 112 (sem 5) g and were not different among treatment groups. End-point body weight was not different between IN and CL treatments at 4 or 8 weeks (Table 1). When calculated as weight gain there was 10 g difference in diet groups but this effect did not reach significance either (CL 146 (sem 7) g; IN 136 (sem 6) g; P = 0·10). There was a trend toward a lower weight gain at weeks 4 and 5 (P = 0·09) for IN-fed rats. Feed intake was 6 % lower in IN-fed rats when calculated as average intake per d (Table 1) and 6 % lower as total feed intake (CL 706 (sem 57) g; IN 663 (sem 52) g; P = 0·02). When analysed on a weekly basis, feed intake was lower in IN-fed rats at weeks 3, 4 and 5 (P = 0·03, P < 0·01 and P < 0·01), with a trend toward a lower intake at week 6 (P = 0·09).
(Values are means with their standard errors)
CL, cellulose diet; IN, inulin diet.
* No interaction between diet and time was found.
† Indicates a significant effect of diet.
‡ Indicates a significant effect of time.
§ For details of animals and procedures, see Materials and methods.
Bone mass
From baseline to end-point at 8 weeks, bone area, BMC and BMD of WB, femur, tibia and spine increased over time (Table 2). However, there were no changes in any parameter due to IN. BMC was also corrected for body weight (data not shown) but results were not different among groups.
(Values are means with their standard errors)
CL, cellulose diet; IN, inulin diet; Wk, week.
* No interaction between diet and time was found.
† Indicates a significant effect of diet.
‡ Indicates a significant effect of time.
§ Data were not normal but had homogeneity of variance; effects were verified by Kruskal–Wallis analysis of ranks for main effect of time and found to be significant (χ2, P < 0·01 whole body BMC; χ2, P = 0·02 Spine BA).
∥ Data were log transformed to achieve normality.
¶ For details of animals and procedures, see Materials and methods.
Femur morphometry and mineralization
Morphometric measures of the excised femur also tended to be higher with time from baseline to 4 or 8 weeks and there were no effects due to IN (Table 3). Femoral dry weight, length and length corrected for body length were higher at week 8 than week 4. In contrast, femoral neck width was 4 % less at week 8 than week 4. Femoral dry weight corrected for body weight increased over time from baseline and was 4 % higher in IN rats.
(Values are means with their standard errors)
CL, cellulose diet; IN, inulin diet; Wk, week.
* No interaction between diet and time were found.
† Indicates a significant effect of diet.
‡ Indicates a significant effect of time.
§ Data were not normal but had homogeneity of variance; effects were verified by Kruskal–Wallis analysis of ranks for main effect of time and found to be significant (χ2, P < 0·01 femur weight; χ2P < 0·01 femur length).
‖ For details of animals and procedures, see Materials and methods.
Femoral Ca and P concentrations were higher with time from baseline (Table 3). There were no effects due to IN feeding.
Serum markers of bone metabolism
Serum osteocalcin concentrations were lower with time from baseline (data not shown). There was a trend toward higher osteocalcin levels in IN rats (CL 323 (sem 27) nmol/l; IN 374 (sem 29) nmol/l; P = 0·02); however, data were not normal and did not have homogeneity of variance. As log transformation was not able to correct these issues, mean values of main effects were analysed by Kruskal–Wallis analysis of ranks and found to be not significantly different (P = 0·15). Serum CTX-I concentrations decreased from baseline over time, but there were no effects associated with IN feeding (CL 31·1 (sem 2·8) nmol/l; IN 35·2 (sem 3·6); P = 0·16 log transformed).
Fat and lean body mass
Parametrial fat pad weight decreased over time and was 28 % lower in IN rats (Table 4). When corrected for body weight, the difference between IN and CL groups was 26 % (Fig. 1 (a)). When analysed by DXA, IN feeding had similar effects on WB fat mass, resulting in 21 % lower fat mass (Fig. 1 (b)) and 17 % lower WB fat mass as a percentage of total body mass (Fig. 1 (c)). Furthermore, fat pad weights and fat mass measured by DXA were highly correlated (R 2 0·73). WB lean mass ( ± BMC) and WB total body mass increased over time but only WB total body mass decreased with IN treatment (5 %, Table 4).
(Values are means with their standard errors)
CL, cellulose diet; IN, inulin diet; WB, whole body; BMC, bone mineral content; Wk, week.
* No interaction between diet and time was found.
† Indicates a significant effect of diet.
‡ Indicates a significant effect of time.
§ Data were not normal but had homogeneity of variance; effects were verified by Kruskal–Wallis analysis of ranks for main effect of time and found to be significant (χ2, P < 0·01 WB lean mass +/ − BMC).
∥ Data were log transformed to achieve normality.
¶ For details of animals and procedures, see Materials and methods.
Effects of inulin on parametrial fat pad mass adjusted for (a) body weight, (b) whole body (WB) fat mass analysed by dual-energy X-ray absorptiometry (DXA) and (c) total fat mass as % total body mass analysed by DXA in growing female rats at 4 and 8 weeks. Values are means with their standard errors of the means for ten rats. Statistical differences among means are indicated by †(P < 0·05), ††(P < 0·01; diet effect) and ‡(P < 0·05), ‡‡(P < 0·01; time effect). Data were log transformed to achieve normality (a), (b). CL, cellulose diet; IN, inulin diet. For details of animals and procedures, see Materials and methods.
Discussion
Previous studies have examined the effects of inulin and OLF on bone density in growing male and adult ovariectomized female rodent models but this is the first study to examine the impact of LC-inulin on bone density in a growing female model. This study shows that LC-inulin did not enhance bone density in either the axial or appendicular skeletons over approximately one and two bone modelling–remodelling stages (4 and 8 weeks) in growing female rats. However, important and novel findings of the present study include the considerable reductions (21–28 %) in parametrial fat pad mass and WB fat mass in IN-fed females (Fig. 1 (a, b)). This result is in partial agreement with studies in male Wistar rats fed OLF and synergy 1 inulin (a 1:1 mixture of OLF and LC-inulin) for 3 weeks, which reported a 30 % reduction in epididymal fat pad mass(Reference Cani, Dewever and Delzenne28), suggesting that inulin feeding may reduce abdominal fat gain. However, fat pad mass of male rats fed only LC-inulin in the Cani et al. (2004) study did not differ from the control after 3 weeks, which does not agree with the present findings in females. This disparity may reflect the differential response in males and females to inulin of varying DP or may be related to the length of the study (3 weeks v. 8 weeks in the present study). Similarly, preliminary studies in our own laboratory have found no change in WB fat mass or lean mass (assessed by DXA) or epididymal fat pad mass in male growing Sprague–Dawley rats fed 10 % LC-inulin for 6 weeks (unpublished data).
Based on the considerable reductions in visceral fat pad weight, WB fat mass and % fat mass (Fig. 1), there appears to be a significant reduction in adiposity associated with inulin feeding. This shift may reflect changes in both subcutaneous and visceral fat mass, although visceral fat is likely to be more amenable to inulin treatment. The reduction in adiposity can be at least partially explained by the decrease in feed intake (Table 1) and WB mass (Table 4), as well as the trend toward less weight gain. It is also interesting that the changes in feed intake (and weight gain) were driven by differences during weeks 3–6, which trailed off during weeks 7 and 8. This is likely due to the decrease in growth rate due to ageing, making any differences due to diet more evident during the rapid growth phase. However, there may also be adaptation to the fibre present in the diet and long-term follow-up studies should address this concern.
There are very few studies reporting the effects of inulin on body composition, as assessed by DXA. A recent study in growing male Wistar rats fed a control diet, 5 % inulin, or 5 % OLF diet for 3 months reported significant improvements in bone density but no change in mean fat mass or lean mass when assessed by DXA or body weight(Reference Nzeusseu, Dienst and Haufroid6). However, specific fat pad weights were not reported by the authors and thus it is difficult to draw comparisons with the present study.
Nonetheless, the reduction in abdominal adiposity reported here has potential implications for the prevention and management of obesity. In fact, there is considerable evidence that inulin-type fructans protect against hepatic steatosis(Reference Delzenne, Daubioul and Neyrinck29), fasting and postprandial triacylglycerolaemia(Reference Roberfroid23), increased energy intake and gain in body weight and fat mass in rats(Reference Cani, Neyrinck and Maton15, Reference Delzenne, Cani and Daubioul16). The mechanism(s) of these effects are thought to work through the modification of gut-associated peptides (involved in the regulation of appetite and body weight) by fermentation products of inulin-type fructans(Reference Cani, Neyrinck and Maton15) and may explain the reduction in feed intake seen in the present study.
In the present study, there was a modest effect of LC-inulin on bone, as femoral weight was higher in IN-fed rats, when corrected for body weight (Table 3). A previous study has reported no change in bone size or length in male rats fed inulin(Reference Nzeusseu, Dienst and Haufroid6). However, these animals were 4 to 5 months of age at the study termination. Thus, it is possible that inulin may positively impact upon bone formation during earlier, more rapid growth stages although this change may only be a transient effect as it was not reflected by increases in bone area when measured by DXA.
The lack of change in the bone resorption marker CTX-I does not agree with previous studies in growing, male rats(Reference Kruger, Brown and Collett5, Reference Nzeusseu, Dienst and Haufroid6), but is not surprising in light of the lack of response in bone density in the present study. These disparities may be related to several factors including sex, age, duration of feeding and DP of inulin. For example, 5·0–5·5 % inulin, OLF and mixed-inulin diets have produced significant increases in BMD and decreases in bone resorption rates in growing male rats from 4 weeks(Reference Kruger, Brown and Collett5) to 22 weeks(Reference Roberfroid, Cumps and Devogelaer4) of feeding. However, similar results have not been reported and/or studied either with LC-inulin or in growing females. Thus, it is possible that LC-inulin, inulin or OLF could improve BMD in growing females over a longer feeding period. Alternatively, inulin or a mixed DP inulin source may be more beneficial than LC-inulin in females. In fact, a recent study has shown enhanced femoral BMD, Ca absorption and femoral Ca concentration and reduced bone turnover in a 9 month old, ovariectomized female rat model fed 5·5 % diets of mixed inulin and OLF or enriched inulin for only 3 weeks(Reference Zafar, Weaver and Jones10). Therefore, the role of inulin-type fructans in achieving peak bone mass in growing female models requires further investigation.
An increase in bone Ca concentration in response to inulin-type fructans has been reported(Reference Takahara, Morohashi and Sano3, Reference Zafar, Weaver and Jones10) and is thought to be mediated through a dose-dependent effect on intestinal Ca absorption(Reference Coxam30). A strong correlation has also been reported between intestinal Ca absorption and bone Ca concentration(Reference Takahara, Morohashi and Sano3). However, inulin has not been shown to affect P absorption, and bone concentrations of this mineral in response to inulin-type fructans have not been widely reported. The impact of LC-inulin on bone mineralization in this study was negligible with regard to femoral Ca and P concentrations (Table 3).
The finding that femoral neck width was 4 % lower at week 8 than week 4 is surprising. The narrowing of the femoral neck width is a concern as this is a site of heightened fracture risk(Reference Bloebaum, Lundeen and Shea31). However, femoral neck is thought to be under strong genetic control(Reference Alam, Sun and Liu32). Thus, this narrowing may be the result of species-specific bone re-modelling for growth and elongation of the femur and not necessarily associated with mineralization and strength.
In summary, the present study demonstrates the effectiveness of LC-inulin in reducing regional and total fat mass in a growing female rat model. This reduction in fat mass has implications for the management of obesity. There were no significant effects of LC-inulin on bone density after 8 weeks in this model. Further studies should investigate the effect of inulin-feeding on bone density in growing females over longer periods of 6 to 12 months. It will also be of interest to compare the effectiveness of LC-inulin with inulin and OLF in both growing female and post-menopausal models. Finally, further work should be carried out on the capacity of inulin to modulate body composition in females in healthy and disease states. Healthy bone development may be compromised in obesity and is therefore a special concern for obese children and teenagers. Inulin may be an effective treatment to promote bone mineralization as well as weight management.
Acknowledgements
We thank the staff of the University of Manitoba Animal Holding Facility, Sarah Cahill, Lisa Rigaux and Nazanin Kazem Moosavi for their assistance. Funding was provided by the Natural Sciences and Engineering Research Council of Canada and the Children's Hospital Foundation of Manitoba to C. G. T. and H. A. W. The Children's Hospital Foundation of Manitoba purchased and maintains the densitometer research facilities of the Manitoba Institute of Child Health used in this study. N. R. R. was supported by a Natural Sciences and Engineering Research Council of Canada Graduate Scholarship. The LC-inulin used in the study was donated by Orafti Group. The authors declare no conflict of interest.
