Effect of Soluble Or Partly Soluble Dietary Fibres Supplementation On

Home , Mineral absorption

European Journal of Clinical Nutrition (1997) 51, 375±380 ß 1997 Stockton Press. All rights reserved 0954±3007/97 $12.00 Effect of soluble or partly soluble dietary ®bres supplementation on absorption and balance of calcium, magnesium, iron and zinc in healthy young men

C Coudray, J Bellanger, C Castiglia-Delavaud*, C ReÂmeÂsy, M Vermorel* and Y Rayssignuier

Center de Recherche, en Nutrition Humaine, d'Auvergne, UniteÂ des Maladies MeÂtaboliques et Micronutriments, INRA de Thiex/ Clermont-Ferrand, 63122 Saint GeneÁs Champanelle, France *UniteÂ des MeÂtabolismes EnergeÂtique et Lipidique, INRA de Theix/ Clermont-Ferrand, 63122 Saint GeneÁs Champanelle, France

Objectives: This study is aimed at investigating the effect of feeding a soluble or partly soluble ®bre rich-diet on the apparent absorption and balance of calcium, magnesium, iron and zinc in healthy young men, by using a chemical balance technique. Study design: Nine healthy young men were given a control diet or the same diet complemented with either inulin (soluble) or sugar beet ®bre (partly soluble) during 28 d periods according to a 3 6 3 latin square design with three repetitions. During the 20 d adaptation period to ®bre ingestion, experimental ®bres were incorporated into bread (60%) and liquid foods (40%) up to a maximum of 40 g/d. Ca, Mg, Fe and Zn were measured in diets and in a 8 d urine and faecal composites to assess mineral absorption and balance. Results: The dietary mineral intake provided (mg/d) 859 Æ 196 of Ca; 311 Æ 43 of Mg; 11.6 Æ 1.7 of Fe; and 11.1 Æ 1.6 of Zn from the control diet. The apparent absorption of minerals from the control diet was (%) Ca: 21.3 Æ 12.5; Mg: 46.3 Æ 10.9; Fe: 21.8 Æ 12.3 and Zn: 14.0 Æ 14.5 (mean Æ s.d.). Ingestion of inulin signi®cantly increased the apparent absorption and the balance of Ca. Sugar beet ®bre ingestion resulted in a signi®cant increase in Ca intake and balance, without modi®cation its apparent absorption. Apparent absorption and balance of Mg, Fe and Zn were not signi®cantly altered by the ingestion of either experimental ®bre. Conclusions: Addition of the two experimental ®bres (inulin or sugar beet ®bre) to normal mixed diets can improve Ca balance without adverse effects on other mineral retention. Sponsorship: This project was supported by the French Ministry of Agriculture, Fisheries and Foods (programme Aliment #2002-Aliment Demain; No. 906335). The authors acknowledge the socieÂteÂ Agro Industries, Recherche et Developpement (Mr R. De Baynast) who supplied them with the experimental ®bres. Discriptors: dietary ®bres; inulin; sugar beet ®bre; intake; absorption; balance; calcium; magnesium; iron; zinc; humans

Introduction sent (Sandstead et al, 1995). However, many studies have indicated that dietary ®bres per se do not appear to affect Many degenerative diseases of Western society, such as trace element absorption (Sunvold et al, 1995). Recently, atherosclerosis, diabetes, hypertension obesity and some attention has been increasingly focused on ®bre isolates, cancers, are directly related to food intake (Jenkins et al, and more especially on fermentable poly- or oligosacchar- 1980; Kritchevsky & Tepper, 1995). There is now over- ides. Inulin, a fructo-oligosaccharide belonging to the whelming evidence that dietary ®bres are a necessary fructan family, is currently used in various agro-food component of human and animal diets and plays a sig- industries, and in dairy and cheese industries (Dysseler & ni®cant role in human health (Briel et al, 1995). According Hoffman, 1995). In our laboratory, we showed that inulin to the statistical data of the French National Institute of enhanced Ca and Mg absorption in animal studies Economic studies, the average daily consumption of dietary (DemigneÂ et al, 1989; Levrat et al, 1991). This observation ®bres by French adults is about 16 g. Bagheri & Debry was recently con®rmed by other workers also in animal (1990) suggest that the recommended daily intake should studies (Delzenne et al, 1995; Ohta et al, 1995). Unfortu- be in the range of 30±40 g for French adults. However, nately, limited information is available for humans, and dietary ®bres and some associated substances, such as only indirect evidence that this effect could occur in phytate, have strong in vitro mineral binding or complexing humans has been reported (Trinidad et al, 1993, 1996). capacities (Persson et al, 1991). For this reason, dietary As part of a larger project concerning the effects of dietary ®bres have been suspected to impair mineral absorption. ®bres in human nutrition, we have studied the conse- Within the last 20 y, several animal and human studies quences for human physiology and nutrition of an increased have shown that foods or diets rich in dietary ®bres may intake of soluble ®bre (inulin) in comparison to partly alter mineral metabolism, especially when phytate is pre- soluble (sugar beet) ®bre. The criteria studied included stool characteristics, nutrient digestibility, energy metabolism (Castiglia-Delavaud et al, 1997) and mineral balance. Correspondence: Dr C Coudray The results of the apparent absorption and the balance of Received 30 December 1996; revised 17 February 1997; accepted 28 calcium, magnesium, iron and zinc are reported in this Dietary ®bres and mineral balance in humans C Coudray et al 376 paper. To our knowledge, this is the ®rst time that the effect Analytical methods of inulin on mineral absorption has been investigated in Food intake was monitored by collection of duplicate humans. meals. Composite samples of food were prepared using metal-free materials. Tubes and receptacles used for the mineral measurements were immersed in a 2% solution of Subjects and methods EDTA for 24 h and then washed thoroughly with deminer- Experimental ®bres alized water. Acids of Suprapur1 quality were used Two types of dietary ®bre were studied. Inulin, a polymer (Merck). Ca, Mg, Fe and Zn were determined in dietary of fructose (15 units) with one unit of glucose, belonging to ®bres, tap water, diets, freeze-dried faeces and urine by the fructan family, was extracted from Chicory roots. It is atomic absorption spectrometry (Perkin Elmer 560 and soluble, fermentable (Nilsson & Bjorck, 1998) and capable 3030). About 0.5 g of ®bres, diets or faeces were dry- of retaining large quantities of water. The second was ashed at 500C for 10 h and the dry residue was taken up in extracted from sugar beet and was composed of hemicellu- HC1 (6 M) and diluted to 50 ml and analysed for Ca and lose (27%), cellulose (23%), pectin (18%) and lignin (3%). Mg. For Zn and Fe analysis, approximately 0.5 g of bread Only the pectin component of the of the last ®bre is soluble or diet, or 0.25 g of ®bres or faeces were wet-ashed with a and the fermenticibility of this ®bre is less and slower than mixture of suprapur1 acids (HNO3/HCl04) prior to that of inulin. Analysis of studied ®bres showed that inulin analysis. Urine was analysed directly with or without contained negligible amounts of phytic acid (<0.3 mg/g), dilution, in 0.1% lanthanum chloride solution. Plasma whereas the level of phytic acid in beet ®bre was as high as was diluted in 0.1% lanthanum chloride (1/50, v/v). Ana- 15.9 Æ 0.3 mg/g (n 6). Both ®bres were prepared by Agro lytical quality was checked using total diet control stan- Industries, Recherche et Developpment, (Compiegne, dards (NIST 1548) for dietary, home built-human faeces for France). The work of FlourieÂ et al (1985) showed that a faecal, and Seronorm (Nycomed) for urinary and serum supplementary intake of 40 g of ®bre per day was the mineral measurements. maximum tolerable quantity to the Western human organ- Phytic acid content in experimental ®bres was deter- ism. In the control diet, ®bre intake was 18 g/d, whereas in mined by colorometric method at 500 nm, after an acid the two experimental diets, the total ®bre intake amounted extraction of HCl 0.67 M as previously described (Latta & to 58 g/d. Soluble or partly soluble ®bres were added by Eskin, 1980). Phytic acid from Sigma was used as external incorporation into bread (24 g/d) and liquid foods (16 g/d). standard. The experimental design was described previously in detail in Castiglia-Delavaud et al (1997). The main aspects dealing Statistical analysis with mineral and trace element balance are presented below. For each experimental treatment, the data are presented as means Æ s.d. (n 9). Data were analysed statis- Subjects and experimental design tically according to a 3 6 3 Latin-square design with Nine male students (21.5 Æ 2.5 y; 174.9 Æ 8.6 cm; 69.3 Æ three repetitions. Comparisons between experimental 5.0 kg) volunteered for this study. All subjects were con- treatments were done by analysis of variance using sidered to be in good health, completed a medical ques- the General Linear Models Procedure (SAS, 1988) includ- tionnaire, and underwent a medical examination by a ing diet, subject, period and repetition. The `LS means' physician. Exclusion criteria were digestive, hepatic, renal statement was used to calculate the adjusted means, and or cardiovascular diseases. The volunteers were fully the `contrast' statement to compare the three treatments. informed of the aims and purposes of the study, and The results were considered signi®cantly different when signed an informed consent. The protocol was approved P < 0.05. by the local ethical committee (CCPPRB No. AV 38) of the Human Nutrition Research Centre at Clermont-Ferrand (France). Results The volunteers were offered the three following diets according to a 3 6 3 Latin square design with three repeti- Daily mineral intake tions: a control diet (18 g dietary ®bre per day); an inulin Table 1 shows the content of Ca, Mg, Fe and Zn in both diet (18 g dietary ®bres 40 g inulin/d); and a sugar beet inulin and sugar beet ®bres. As expected, sugar beet ®bre ®bre diet (18 dietary ®bre 40 g sugar beet ®bre/d). The was especially rich in Ca and Mg compared to inulin. study had a cross-over design, each subject acting as his Consequently, sugar beet ®bre contributed 30%, 18%, 12% own control for the two tested ®bre-rich diets. The control and 2.6% to daily intakes of Ca, Mg, Fe and Zn, respec- diet was composed of four daily diets given alternatively to tively. In contrast, inulin contributed for only about 1% to participating subjects. Meals were selected so that the daily intake of Fe and less than 0.25% to daily intake of any mineral intakes provided by the control diet were compar- other tested minerals. able to intakes from self-selected diets consumed by Subjects received four daily diets in rotation. Actual different groups in industrialised countries. The meal com- daily intake of minerals was determined from the analysis position and the conditions of ®bre supplementation were of these meals and the recording of non-consumed portions previously described by Castiglia-Delavaud et al (1997). which were close to nil (Vernet & Vermoral, 1993). Tap Brie¯y, each of the four daily diets were balanced and water was also analysed for mineral content. The Table 1 composed of some legumes, meat or ®sh, starchy foods or shows the mineral intakes determined over an eight days vegetable, cheese, fruit or compote plus biscuits. Each period of consumption of control diet and tap water. With experimental period lasted for 28 d, and was composed the control and inulin diets, the total daily mineral intake of 2 d of control diet, 14 d of progressive increase in ®bre was 20% lower than the recommended allowances for Ca, intake, and 12 d of constant ®bre consumption. Faeces and Mg and Zn (RDAs, 1989). Addition of sugar beet ®bre urine were collected between d20 and d28 and blood was signi®cantly increased Ca and Mg intakes (P < 0.001) to sampled at d25 for the metabolic mineral balance study. the recommended allowances. Dietary ®bres and mineral balance in humans C Coudray et al 377 Table 1 Mineral daily intake provided by ®bres and alimentation (mg/d)

Ca Mg Fe Zn

Experimental ®bres a,b Inulin 1.53 Æ 0.06 0.46 Æ 0.04 0.16; 0.12 0.03 Æ 0.001 Beet ®bre 324 Æ 1.20 68.4 Æ 0.38 1.33; 1.35 0.274 Æ 0.02 Daily mineral intake from tap waterc Control diet 31.2 Æ 13.2 12.1 Æ 5.11 < 0.005 < 0.03 Inulin diet 30.1 Æ 12.9 11.7 Æ 4.99 < 0.005 < 0.03 Beet ®bre diet 31.8 Æ 11.5 12.3 Æ 4.45 < 0.005 < 0.03 Daily total mineral intakec Control diet 859 Æ 196 311 Æ 4.26 11.6 Æ 1.70 11.1 Æ 1.59 Inulin diet 852 Æ 190 303 Æ 41.7 11.5 Æ 1.77 10.9 Æ 1.61 Beet ®bre diet 1229 Æ 218* 373 Æ 49.6* 12.3 Æ 1.55 11.2 Æ 1.56

aEach value is the mean and s.d. of 2±4 replicates. bCalculated on the basis of 40 g of ®bres per day. cEach value is the mean and s.d. of 9 data. *Signi®cantly different from control or inulin diets (P < 0.001).

Mineral apparent absorption Apparent mineral balance As indicated in the Table 2, the mineral amounts apparently As shown in Table 2, the mean Ca balance was slightly absorbed from the control diet (intake-faeces) were as negative (10 mg/d) in subjects consuming the control diet. follows 179 Æ 115; 144 Æ 38; 2.6 Æ 1.6 and 1.6 Æ 1.6 mg However, there was a large inter-individual variability per day for Ca, Mg, Fe and Zn respectively. Consequently, (from 7274 to 141 mg/d), because urinary Ca excretion the apparent mineral absorption averaged ((intake-faeces)/ was highly variable between subjects (98±289 mg/d). In intake) 6 100 averaged (%) 21.3 Æ 12.5; 46.3 Æ 10.9; contrast, the Mg balance was largely positive (31.5 mgd) 21.8 Æ 12.3; and 14.0 Æ 14.5 for Ca, Mg, Fe and Zn and the inter-individual variability (from 710 to 82 mg/d) respectively with the control diet. With regard to the effects was lower than for Ca. Iron balance was positive in all of inulin compared to the control diet, this soluble ®bre participating subjects (2.52 mg/d) and ranged from 0.55± signi®cantly increased the absorption of Ca (P < 0.01). The 5.91 mg/d. Urinary iron excretion was very low, with low apparent quantity of Ca absorbed from the inulin diet was inter-individual variability (45±73 mg/d). Zinc balance was thus increased signi®cantly (P < 0.05). However, the appar- also positive in most participating subjects and ranged from ent absorption of the other minerals was not signi®cantly 71.97 to 2.78 mg/d. Urinary zinc excretion was from 649 affected by inulin ingestion. In other respects, the absorp- to 1385 mg/d with no large inter-individual variability. tion of all minerals studied was not signi®cantly altered by Ingestion of the soluble ®bre (inulin) did not signi®- sugar beet ®bre ingestion. However, because sugar beet cantly alter urinary Ca excretion (Table 2). However, inulin ®bre is rich in Ca and Mg, the quantity of Ca apparently ingestion resulted in a signi®cant positive Ca balance absorbed from the sugar beet ®bre diet was signi®cantly ( 91.8 mg/d) (compared to control diet, P < 0.05). This higher (P < 0.05) than that from control diet (Table 2). The positive balance re¯ected the enhancement of Ca absorp- same trend was observed for Mg without reaching a tion due to inulin intake. However, inulin ingestion did not statistically signi®cant level. signi®cantly alter urinary excretion of Mg, Fe and Zn, and

Table 2 Effect of inulin and sugar beet ®bres on absorbed quantity, apparent absorption, urinary excretion and balance of Ca, Mg, Fe and Zn in humansa

Ca Mg Fe Zn

Apparently absorbed quantities (mg/d) Control diet 179 Æ 115 144 Æ 37.9 2.58 Æ 1.64 1.58 Æ 1.61 Inulin diet 278 Æ 101* 156 Æ 46.5 1.83 Æ 2.52 2.21 Æ 2.00 Beet ®bre diet 294 Æ 92.4* 169 Æ 42.6 2.25 Æ 0.99 1.59 Æ 0.89 Apparent absorption (%) Control diet 21.3 Æ 12.5 46.3 Æ 10.9 21.8 Æ 12.3 14.0 Æ 14.5 Inulin diet 33.7 Æ 12.1** 51.1 Æ 12.3 14.0 Æ 24.8 20.8 Æ 20.0 Beet ®bre diet 24.7 Æ 9.40 45.0 Æ 7.60 18.1 Æ 7.20 14.1 Æ 7.80 Urinary excretion (mg/d) Control diet 189 Æ 63.6 112 Æ 28.4 0.057 Æ 0.012 0.888 Æ 0.267 Inulin diet 186 Æ 85.0 121 Æ 40.6 0.057 Æ 0.023 0.883 Æ 0.259 Beet ®bre diet 198 Æ 57.1 130 Æ 23.6 0.045 Æ 0.020 0.955 Æ 0.231 Apparent balance (mg/d) Control diet 710.1 Æ 136 31.5 Æ 27.5 2.52 Æ 1.64 0.689 Æ 1.59 Inulin diet 91.8 Æ 115* 35.2 Æ 44.4 1.77 Æ 2.51 1.32 Æ 2.10 Beet ®bre diet 96.1 Æ 64.8* 38.9 Æ 27.5 2.21 Æ 1.00 0.633 Æ 0.95

aEach value is the mean and s.d. of 9 data. *Signi®cantly different from control diet (P < 0.05). **Signi®cantly different from control and sugar beet diets (P < 0.01). Dietary ®bres and mineral balance in humans C Coudray et al 378 only a trend to increase the Mg and Zn balance and to ious electrolyte exchanges (Ca-H). It has been proposed decrease in that of Fe were observed. Similarly, ingestion (Trinidad et al, 1993) that Ca could pass through the cell of the partly soluble ®bre (sugar beet) did not signi®cantly membrane more readily in the form of a complex with alter urinary excretion of Ca. However, sugar beet ®bre lower charge. However, a high rate of Ca absorption in the ingestion resulted in a signi®cantly positive Ca balance large intestine could trigger a feedback mechanism invol- (96.1 mg/d) (P < 0.05 compared to control diet). This ving inhibition of duodenal absorption since there is a positive balance was the consequence of the high mineral control of the digestive balance of Ca by endocrine factors content of sugar beet ®bre which increased Ca intake by (Allen, 1982). about 30% without modifying absorption ef®ciency. The intestinal absorption of Mg has been studied less Finally, sugar beet ®bre ingestion did not signi®cantly extensively than that of Ca. There are several mechanisms alter urinary excretion of Mg, Fe and Zn or their balances. by which Mg can cross the intestinal epithelium: passive Plasma Ca and Mg concentrations were determined in diffusion, solvent drug and active transport. However, the subjects participating in the study towards the end of several investigations in adults have documented that net each test diet period. Plasma Ca was 2.36 Æ 0.134, absorption of Mg is linearly related to dietary Mg over a 2.36 Æ 0.121, 2.34 Æ 0.113 mmol/l; and plasma Mg was normal range of dietary Mg intake. This indicates that 8.07 Æ 0.395, 8.07 Æ 0.712, 7.94 Æ 0.580 mmol/l prior to absorption of Mg in adults primarily occurs by a passive the control, inulin and sugar beet ®bre periods, respec- process. Mg is absorbed primarily in the distal segments of tively. These results indicated that ingestion of either diet- the gastrointestinal tract in humans, and carbohydrates that ary ®bres was without effect on the Ca and Mg status in our can stimulate bacterial fermentation in the intestine are individuals. Plasma Fe and Zn determinations were not known to increase Mg absorption. As animal studies performed because both ingested ®bres showed no effect on showed similar effects of inulin ingestion on Ca and Mg absorption or balance of these minerals. absorption, the same effect of inulin on Mg absorption as that on Ca could be expected in human. The results from the present investigation showed, however, that inulin Discussion ingestion did not signi®cantly affect the absorption and The objective of this investigation was to measure, by the balance of Mg in humans under our experimental condi- chemical balance technique, the effect of feeding soluble tions. Inulin ingestion resulted only in a small increase in and rapid fermentable ®bre (inulin), or partly soluble and Mg absorption, accompanied with a slight increase in Mg slow fermentable ®bre (sugar beet ®bre) -enriched diets on urinary excretion without reaching a signi®cant level. Mg apparent absorption and balance of Ca, Mg, Fe and Zn in balance was also slightly more positive than in the case of healthy young men. Inulin is a polysaccharide degradable the control diet. Several explanations could be advanced for in the ®rst half of large intestine (fermentability 100%) the difference observed in the effect of inulin on Mg whereas the sugar beet ®bre is degraded throughout the absorption between rat and human. It is recognised that large intestine (less and slow fermentability 70%). the digestive tract physiology of rat is different from that of The intestinal absorption of Ca occurs via two processes: human. The caecum segment, absent from human, plays active Ca transport mainly in the duodenum and jejunum indeed an important role in Ca and Mg absorption (Rayssi- and passive transport mainly in the ileum. The active guier & ReÂmeÂsy 1977). Although each experimental period transport of Ca is mediated by calcitriol, the active compo- in the present study lasted for 26 d, it is also possible that nent of vitamin D, but most of the Ca absorbed from food is the effect of inulin on Mg absorption could be slower than passively transported into the intestinal mucosal cells. that on Ca, and a longer adaptation period could be needed. Availability for absorption requires Ca to be solubilised Fe and Zn are essential trace elements and are absorbed either in a free ionic or complexed form. Carbohydrates mainly in the small intestine by active and passive ways. which are able to reach the ileum may stimulate Ca Active transport takes place in the proximal part of intes- intestinal absorption in its part of digestive tract. In fact, tine and is predominant when the element content is low. the results of the present investigation indicate clearly that Passive diffusion can also take place when the element ingestion of an inulin-containing diet increased signi®- level is high in the digestive tract. The results of the present cantly the apparent absorption and balance of Ca in investigation did not show any signi®cant effect of inulin humans. This is the ®rst time, to our knowledge, that ingestion on the apparent absorption and balance of these such an effect has been observed in humans. These results elements. Few studies in the literature have documented the con®rm previously reported data from animal investiga- effect of fermentable soluble ®bre on the absorption and tions in our own laboratory, (DemigneÂ et al, 1989; Levrat et balance of Zn and Fe. Recently, Delazenne et al (1995) al, 1991; Younes et al, 1996), and by other workers reported a signi®cant increase in Fe and Zn absorption in (Delzenne et al, 1995; Ohta et al, 1995). Several hypothesis rats fed with inulin (Raftiline) or its product of hydrolysis about the mechanisms of the effect of inulin on Ca (known as Ra¯itose). Other dietary fermentable com- absorption could be proposed. Carbohydrates that escape pounds, such as polymers of glucose and lactose, are digestion in the small intestine are substrates for the reported to stimulate the intestinal absorption of Zn, but formation of short-chain fatty acids (SCFA) in the large the mechanisms have not been clearly elucidated (Greger et intestine by the intestinal micro¯ora. This fermentation al, 1989). results in a lowering of the luminal pH which raises the As far as the effect of sugar beet ®bre is concerned, the concentration of ionised Ca and accelerates the passive ®rst salient observation is its high content of minerals, in diffusion of Ca (ReÂmeÂsy et al, 1992). The accumulation of particular Ca and Mg. The Ca, provided by this ®bre, calcium phosphate in the large intestine and solubilisation increased the daily Ca intake by more than 30%, and that of this Ca by organic acids (SCFA) probably plays an of Mg by about 20%. This con®rms previous data concern- essential role in the enhancement of Ca absorption (ReÂmeÂsy ing the high participation of some crude ®bres in et al, 1993; Kashimura et al, 1996). It is also possible that mineral intake (Van Dockkum et al, 1982; Behall et al, SCFA directly in¯uence Ca absorption by modifying var- 1987; Fairweather-Tait et al, 1990). Furthermore, sugar Dietary ®bres and mineral balance in humans C Coudray et al 379 beet ®bre addition provided until 450 mg of phytic acid per humans over ®ve weeks. It was concluded that consump- day, 250 mg from beet ®bre added to soup and 200 mg tion of sugar beet ®bre did not constitute any risk with from beet ®bre added to breads, given that phytic acid is respect to trace element nutriture. Interestingly, in an degraded at the level of 50% during baking (Tangkongchitr animal study, Fairweather-Tait et al (1990) reported that et al, 1981). The daily phytic acid intake averages up to Fe and Zn absorption was signi®cantly enhanced by sugar 750 mg/d, but varies largely between countries from beet ®bre intake, when Fe and Zn levels in the control diet 200 mg/d in Sweden to 1000 mg/d in Italy (Reddy et al, were adjusted to those of the test diet. The effect of pectin 1989). In the present study, the sugar beet ®bre diet exerted on Zn retention had been tested in humans and found to be no signi®cant effect on Ca absorption when compared to negligible (Lei et al, 1980). Finally in a similar study to the control diet, but signi®cantly increased the Ca balance. ours, Behall et al (1987) found that re®ned ®bres had no This positive balance was the result of the appropriate effect on mineral balance in subjects consuming recom- intake of Ca provided by the sugar beet ®bre. One could mended dietary allowance levels of Fe and Zn when fed as not exclude that the non effect of sugar beet ®bre on Ca part of their control diet. absorption reported in our study could indeed be the result of two opposite phenomena. The ®rst one is a decrease in Ca absorption due to the high intake of Ca in the sugar beet Conclusions ®bre diet. It is recognised that Ca absorption decreases Our investigations indicate that ingestion of inulin, a when Ca dietary intake is increased (RDAs, 1989). The soluble fermentable ®bre, signi®cantly increases the appar- second is that the sugar beet ®bre could increase Ca ent absorption of Ca and its balance in humans, probably by absorption in our subjects. The measured apparent absorp- solubilizing this element in the colon. Moreover, sugar beet tion of Ca could be the result of these two opposite ®bre, a partly soluble ®bre, did not have any negative effect tendencies and thus not be signi®cantly modi®ed. No on any of the minerals studied in the present work. These similar studies have been reported in relation to the effect results clearly demonstrate that sugar beet ®bre ingestion in of sugar beet ®bre on Ca absorption either in animals or humans can improve the utilization of Ca and Mg and their humans. However, there are numerous studies investigating balances. These two ®bres can thus be added to normal the effects of cereal, vegetable and fruit ®bres on the mixed diets without adverse effects on mineral retention. It absorption of Ca in animals and humans. Most of them is clear that an intake of both soluble and partly soluble have reported no effect on Ca absorption or balance ®bre, having different degrees of fermentability, ensure the (Wisker et al, 1991; Davidsson et al, 1996). However, fermentation in the whole large intestine. Dietary ®bres or some studies described a decrease in Ca absorption with similar compounds could exert positives effects on mineral ingestion of dietary ®bres under certain conditions (Knox et nutrition via the fermentation and the solubilisation of al, 1991; O'brien et al, 1993). It seems that differences in several minerals, or because they are very rich in minerals. the chemical composition, and probably in the physical In the case of inulin, which is not rich in minerals, its properties, of ®bre sources and in other dietary constituents utilisation in the products containing high level of Ca, such such as phytate and oxalic acid, may be responsible for as dairy products, could improve the digestibility of Ca by these contradictory results. solubilising this element in the colon. This will enable Our data show that Mg absorption was not affected by appropriate nutritional recommendations to be made which sugar beet ®bre ingestion when compared with control diet. will enhance mineral absorption. Additional studies are Although sugar beet ®bre ingestion signi®cantly increased needed to evaluate the effect of combining inulin and Mg intake, the absolute quantity of absorbed Mg (mg/d) sugar beet ®bre and to study the effect of other naturally was not signi®cantly affected by sugar beet ®bre intake in fructo-oligosaccharides in various foods on mineral bal- comparison to the control diet. Consequently, Mg balance ance. was not altered by the sugar beet ®bre diet compared to the control diet. We are not aware of any similar studies AcknowledgementsÐWe are grateful for the use of the Human Metabolic concerning the effect of sugar beet ®bre on Mg absorption Facilities in the Centre de Recherche en Nutrition Humaine d'Auvergne either in animals or humans. As for Ca, some studies have and to Dr Y Boirie for medical supervision of all the subjects. We dealt with the effect of vegetable and fruit ®bres on Mg gratefully acknowledge the cooperation of the subjects who suffered so absorption in animal and human. These studies reported conscientiously and cheerfully the restrictions imposed on them by this that there were no differences in Mg balance with these study. Our special appreciation goes to Elisabeth Verdier for the dietetic ®bres (Andersson et al, 1983). Moreover, Bagheri & advices, Jean Claude Tressol for the technical assistance, and Jean Vernet GueÂgen (1985), demonstrated clearly that Mg from wheat for the statistical help. bran was well utilized and bran addition improved con- siderably the balance of Mg. It is worth noting that fruits References and vegetables remain an important source of both dietary ®bres and minerals. Allen LH (1982): Calcium bioavailability and absorption: a review. Am. J. Clin. Nutr. 35, 783±808. With regard to Fe and Zn absorption and balance, the Andersson H, Navert B, Bingham SR, Englust HN & Cummings JH present study did not show any signi®cant effect of partly (1983): The effects of breads containing similar amounts of phytate but soluble dietary ®bres on any of these minerals when different amounts of wheat bran on calcium, zinc and iron balance in compared with the control diet. In agreement with our man. Br. J. Nutr. 50, 503±510. Bagheri S & GueÂgen L (1985): Effect of wheat bran and pectin on the results, Sandstrom et al (1987) reported that sugar beet absorption and retention of phosphorus, calcium, magnesium, and zinc pulp-®bre did not affect the extent of Zn absorption when by the growing pig. Reprod. Nutr. DeÂvelop. 25, 705±716. used as a meat extender in humans. However, it was Bagheri SM & Debry G (1990): Estimation de la consommation moyenne observed that Zn absorption was higher when sugar beet de ®bres alimentaires en France. Ann. Nutr. Metab. 34, 69±75. ®bre was included in bread than when it was used as Behall KM, Schol®eld DJ, Lee K, Powell AS & Moser PB (1987): Minerals Balance in adult men: Effect of four re®ned ®bres. Am. J. muesli. Cossack et al (1992) studied the effects of a Clin. Nutr. 46, 307±314. sugar beet ®bre supplement on Fe and Zn status in Briel F, Achour L & FourieÂ B (1995): Les ®bres alimentaires. Cah. Nutr. Dietary ®bres and mineral balance in humans C Coudray et al 380 Diet. 30, 132±136. A, Holick MF & Yergey AL (1993): High ®ber diets slow bone turnover Castiglia-Delavaud C, Verdier E, Basele JM, Vernet J, Boirie Y, BeaufreÁre in young men but have no effect on ef®ciency of intestinal calcium B & Vermorel M (1997): The impact of sugar beet ®bre or inulin absorption. J. Nutr. 123, 2122±2128. supplementation on nutrient digestive utilisation and net energy value of Ohta A, Ohtsuki M, Baba S, Adachi T, Sakata T & Sakaguchi E (1995): dietary ®bres as determined by whole-body calorimetry in humans. Calcium and Magnesium absorption from the colon and rectum are Submitted simultaneously with the present paper to the Am. J. Clin. increased in rats fed fructo-oligosaccharides. J. Nutr. 125, 2417±2424. Nutr. Persson H, Nyman M, Liljeberg H, Onning G & Frolich W (1991): Cossack ZT, Rojhani A & Musaiger AO (1992): The effects of sugar-beet Binding of mineral elements by dietary ®bres components in cereals. ®ber supplementation for ®ve weeks on zinc, iron and copper status in In vitro (III). Food Chem. 40, 169±178. human subjects. Eur. J. Clin. Nutr. 46, 221±225. Rayssiguier Y & ReÂmeÂsy C (1977): Magnesium absorption in the cecum of Davidsson L, Mackenzie J, Kastnmayer P, Rose A, Golden BE, Aggett PJ rats related to volatile fatty acid production. Ann. Rech. Vet. 8, 105±110. & Hurrell RF (1996): dietary ®ber in weaning cereals: A study of the Reddy NR, Pierson MD, Sathe SK & Salukhe DK (1989): Occurrence, effect on stool characteristics and absorption of energy, nitrogen, and distribution, content, and dietary intake of phytate. In: Phytates in mineral in healthy infants. J. Pediatr. Gastroenterol. Nutr. 22, 167±179. cereals and legumes. Adv. Food Res. 28, 39±56. Delzenne N, Aertssens J, Verplaetse H, Roccaro M & Roberfroid M ReÂmeÂsy C, Behr SR, Levrat MA & DemigneÂ C (1992): Fiber fermentation (1995): Effect of fermentable fructo-oligosaccharides on mineral, nitro- in the cecum and its physiological consequences. Nutr. Res. 12, 1235± gem and energy digestive balance in the rat. Life Sci. 57, 1579±1587. 1244. DemigneÂ C, Levrat AM & ReÂmeÂsy C (1989): Effect of feeding fermen- ReÂmeÂsy C, Levrat AM, Gamet L & DemigneÂ C (1993): Cecal fermentation table carbohydrates on cecal concentration of minerals and their ¯uxes in rats fed fructo-oligosaccharides (Inulin) are modulated by dietary between the cecum and blood plasma in the rat. J. Nutr. 119, 1625± calcium level. Am. J. Physiol. 264, G855±G862. 1630. Sandstead HH (1992): Fiber, phytates and mineral nutrition. Nutr. Rev. 50, Fairweather-Tait SJ & Wright AJA (1990): The effects of sugar-beet ®bre 30±31. and wheat bran on iron and zinc. Br. J. Nutr. 64, 547±552. Sandstrom B, Davidsson L, Kivisto B, Hasselblad C & Cederblad A FlourieÂ B, Etanchaud F & Florent C (1985): Digestion des hydrates de (1987): The effects of vegetables and sugar beet ®bre on the absorption carbone dans le colon humain. Cah. Nutr. DieÂt. 20, 33±36. of zinc from a composite meal. Br. J. Nutr. 58, 49±57. Greger JL, Gutkowski CM & Khazen RR (1989): Interactions of SAS. SAS/STAT user's guide (release 6.03). Cary, N: SAS Institute Inc., lactose with calcium, magnesium and zinc in rats. J. Nutr. 119, 1691± 1988. 1697. Sunvold GD, Titgemeyer EC, Bourquin LD, Hahey GC & Garleb KA Jenkins DJA, Raynolds D, Slavin B, Leeds AR, Jenkins AL & Jepson EH (1995): Alteration of the ®ber and lipid components of a de®ned- (1980): Dietary ®ber and blood lipids; treatment of hypercholesterole- formula diet: Effects on stool characteristics, nutrient digestibility, mia with guar crispbread. Am. J. Clin. Nutr. 33, 575±581. mineral balance and energy metabolism in humans. Am. J. Clin. Nutr. Kashimura J, Kimura M & Itokawa Y (1996): The effects of isomaltulose, 62, 1252±1260. isomalt, and isomaltulose-based oligomers on mineral absorption and Tangkongchitr U, Seib PA & Hoseney RC (1981): Phytic acid. II Its fate retention. Biol. Trace Elem. Res. 54, 239±250. during breadmaking. Cereal Chem. 58, 229±234. Knox TA, Kassarjian Z, Dawson-Hughes B, Golner BB, Dallal GE, Arora Trinidad PT, Wolever TMS & Thompson LU (1993): Interactive effects of S & Russell RM (1991): Calcium absorption in elderly subjects on high- Ca and SCVA on absorption in the distal colon of man. Nutr. Res. 13, and low-®ber diets; effect of gastric acidity. Am. J. Clin. Nutr. 53, 417±425. 1480±1486. Trinidad PT, Wolever TMS & Thompson LU (1996): Effect of acetate and Kritchevsky D & Tepper SA (1995): In¯uence of dietary ®ber on propionate on calcium absorption from the rectum and distal colon of establishment and progression of atherosclerosis in rabbits. J. Nutr. humans. Am. J. Clin. Nutr. 63, 574±578. Biochem. 6, 509±512. Van Dokkum W, Wesstra A & Schippers (1982): Physiological effects of Latta M & Eskin M (1980): A simple and rapid colorimetric method for ®bre-rich types of bread. I. The effect of dietary ®bres from bread on the phytate determination. J. Agric. Food Chem. 28, 1313±1315. mineral balance of young men. Br. J. Nutr. 47, 451±460. Lei KY, Davis MW, Fang MM & Young LC (1980) Effect of pectin on Vernet J & Vermorel M (1993): Base de donnes pour le calcul des zinc, copper and iron balances in humans. Nutr. Rep. Int. 22, 459±466. quantiteÂs d'eÂleÂments nutritifs ingeÂreÂs dans les eÂtudes de nutrition Levrat AM, ReÂmeÂsy C & DemigneÂ C (1991): High propionic acid humaine. Cahier des techinques INRA, 31, 57±76. fermentations and mineral accumulation in the cecum of rats adapted Wisker E, Nagel R, Tanudjaja TK & Feldheim W (1991): Calcium, to different levels of inulin. J. Nutr. 121, 1730±1737. magnesium, zinc and iron balances in young women: effects of a National Research Council (1989) Recommended dietary allowances, 10th low-phytate barley-®ber concentrate. Am. J. Clin. Nutr. 54, 553±559. edition, Washington, DC; National Academy Press. Younes H, DemigneÂ C & ReÂmeÂsy C (1996): Acidic fermentation in the Nilsson U & Bjorck I (1988) Availability of cereal fructans and inulin in caecum increases absorption of calcium in the large intestine of the rat. the rat intestinal tract. J. Nutr. 118, 1482±1486. Br. J. Nutr. 75, 301±314. O'brien KO, Allen LH, Quantromoni P, Siu-Caldera M, Vieira NE, Perez