- Email: [email protected]
Effects of calcium concentration, acetate, and propionate on calcium absorption in the human distal colon
Volume 15
July/August 1999
APPLIED NUTRITIONAL INVESTIGATION
Numbers 7/8 Nutrition Vol. 15, Nos. 7/8, 1999
Effects of Calcium Concentration, Acetate, and Propionate on Calcium Absorption in the Human Distal Colon TRINIDAD P. TRINIDAD, PHD, THOMAS M.S. WOLEVER, MD, PHD, AND LILIAN U. THOMPSON, PHD From the Department of Nutritional Sciences, Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada ABSTRACT
Previous studies have shown that the short-chain fatty acids acetate (Ac) and propionate (Pr) enhance the absorption of calcium (Ca) in the rectum and distal colon of humans, with Pr being more effective than Ac. To investigate the effect of Ac and Pr on the kinetics of Ca absorption from the human rectum and distal colon, six healthy subjects were studied. Solutions containing various concentrations of CaCl2 䡠 H2O with 56.3 mmol/L Ac, Pr, or NaCl were rectally infused to each subject. Rectal fluid was sampled at the end of the infusion (0 min), and 30 min later colonic contents were collected. Ca absorption for all treatments increased linearly with Ca concentration. For Ca ⫹ NaCl, the slope of regression line was 62 mol 䡠 mmol⫺1 䡠 L Ca. With Ac ⫹ Ca, the slope of Ca absorption increased significantly to 113 mol 䡠 mmol⫺1 䡠 L Ca, and with Pr ⫹ Ca, the slope increased to 159 mol 䡠 mmol⫺1 䡠 L (P ⫽ 0.043 versus Ac ⫹ Ca). Ac and Pr absorption were increased by Ca. The data suggest that, over a physiologic range of Ca concentration, in the absence or presence of Ac and Pr, Ca is absorbed in the human rectum and distal colon by a non-saturable diffusion process, and that Ca absorption is enhanced by Ac and Pr. The data also suggest that both Ac and Pr absorption is stimulated by Ca. Nutrition 1999;15:529 –533. ©Elsevier Science Inc. 1999 Key words: calcium absorption, acetate, propionate
INTRODUCTION
The effect of 1,25-dihydroxy cholecalciferol on calcium (Ca) absorption from the small intestine and colon has been studied in rats using different methods.1– 4 Most of these studies have shown that Ca is absorbed by both a saturable, active process and by a
non-saturable diffusion process. Likewise, the effect of the shortchain fatty acids (SCFAs) acetate (Ac) and butyrate (Bu) on Ca absorption from the distal colon of rats has been studied by the perfusion technique.5 It has been speculated in rats that a Ca/H exchange exists in the distal colon similar to the Na/H exchange
Supported by the Dairy Farmers of Canada and the Natural Sciences and Engineering Research Council. Correspondence to: Lilian U. Thompson, PhD, Department of Nutritional Sciences, Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada M5S 3E2. E-mail: [email protected]
Nutrition 15:529 –533, 1999 ©Elsevier Science Inc. 1999 Printed in the USA. All rights reserved.
0899-9007/99/$20.00 PII S0899-9007(99)00110-0
530
CALCIUM ABSORPTION IN THE COLON
present in the proximal colon. Although much information can be gathered from animal studies, extrapolation of their results to humans must be done with caution. There is evidence that SCFA absorption differs in different species.6 Also, most of the methods used in animal studies are difficult to reproduce in humans. We have used a rectal infusion technique to study Ca absorption from the human distal colon and have shown that Ac and propionate (Pr) enhance Ca absorption from this site.7–9 In our most recent study, we used a single concentration of Ca, 50 mmol/L, with various concentrations of Ac and Pr.9 At a concentration of 18.7 mmol/L, Ac and Pr were equally effective in enhancing Ca absorption, but at a concentration of 56.3 mmol/L, Pr was more effective.9 This suggested that Ac and Pr may have different mechanisms in enhancing Ca absorption. Thus, the purpose of the present study was to determine the effect of Ac and Pr on the kinetics of Ca absorption from the human rectum and distal colon. MATERIALS AND METHODS
Six healthy subjects (one male and five females, 29 ⫾ 4 y of age, 100 ⫾ 2% ideal weight) were studied on different mornings after overnight fasts. Solutions for rectal infusion were prepared in double-distilled water; 12 of the solutions contained various concentrations of Ca as CaCl2 䡠 2H2O (4, 11, 24, 46 mmol/L) with either 56.3 mmol/L sodium acetate (Ac), sodium propionate (Pr), or sodium chloride (NaCl). A solution of 56.3 mmol/L NaCl alone was used as a control. Polyethylene glycol (PEG, 0.625 mmol/L) was added to each solution as an unabsorbable marker. To avoid osmotic diarrhea, the osmolality of the solutions was ⬍300 mosm/L. We have shown that differences in osmolality of the test solutions across the range used here do not affect Ca absorption,7 and that the addition of NaCl to correct for differences in osmolality does not affect the absorption of Ca,7 Ac, or Pr.10 Table I shows the composition of the test solutions. The 13 different treatments were given to each subject in random order, with each subject being studied once or twice per week. The 56 mmol/L concentration of Ac and Pr was used in this study because differences in the effect of Ac and Pr on Ca absorption from the rectum and distal colon were observed at this concentration from our previous study.9 Average Ca intake on a usual Western diet is approximately 25 mmol/d11 of which 30 – 40% is absorbed in the small intestine.12,13 Therefore, approximately 15–17 mmol of Ca reaches the colon per day for potential absorption. Thus, the maximum amount of Ca infused into the rectum in this study, 13.8 mmol, is within the physiologic range. After an overnight fast and 10 –20 min before the start of the study, an enema containing 500 mL of double-distilled water was given through 1 m of Tygon flexible plastic tubing (inside:outside diameter, 2.4:4.0 mm; Norton Tubing and Molded Products, Akron, OH, USA) to clean the colon before infusion of the test solution. Then 5 cm of one end of the tubing was reinserted into the rectum. The subjects injected the 300 mL solution themselves, using a 60 mL syringe over 5 min under close supervision by one of the investigators. Before removing the infusion tube, a 5 mL sample was collected at 0 min by mixing (withdrawing and reinfusing) 10 mL of the fluid three times in 2 min. After 30 min, colonic contents were collected in a plastic bag using a fecal collection frame placed under the toilet seat. The 30-min time point was selected because our previous studies showed that serum Ca increased 30 min after rectal infusion of solutions containing Ca plus SCFAs7 and that Ca disappearance from the rectum increased linearly with time between 0 and 30 min for all treatments with and without Ac or Pr or both (r ⫽ ⫺0.96 to ⫺0.99, P ⫽ 0.04 to 0.006).9 Samples were centrifuged and the supernatant analyzed for the following: Ca using an atomic absorption spectrometer (Varian
TABLE I. COMPOSITION (mmol/L) OF THE DIFFERENT TREATMENT SOLUTIONS Treatment*
Ca†
Ac‡
Pr§
NaCl㛳
NaCl Ca ⫹ NaCl
— 4.0 11.0 24.0 46.0 4.0 11.0 24.0 46.0 4.0 11.0 24.0 46.0
— — — — — 56.3 56.3 56.3 56.3 — — — —
— — — — — — — — — 56.3 56.3 56.3 56.3
56.3 56.3 56.3 56.3 56.3 — — — — — — — —
Ca ⫹ Ac
Ca ⫹ Pr
* All treatments contained 0.625 mmol/L polyethylene glycol (PEG 4000), Mallinkrodt OR, Pointe-Claire, Quebec, Canada. † Calcium chloride (CaCl2 䡠 2H2O), Fisher Scientific, Ontario, Canada. ‡ Sodium acetate, NaAc, anhydrous, Sigma S-8750, St. Louis, MO, USA. § Sodium propionate, NaPr, Food Grade, Van Waters and Roger Ltd., London, Ontario, Canada. 㛳 Sodium chloride, NaCl, AR, British Drug House, Darmstadt, Germany.
Model 1275, Varian, Canada Inc., Ontario, Canada) after dilution with 10 mM lanthanum chloride; Ac and Pr by high performance liquid chromatography,14 and PEG by a turbidimetric method.15 Ca, Ac, and Pr disappearance from the distal colon was used as an index of absorption. PEG was used as an unabsorbable marker to correct for incomplete collection and fluid absorption or secretion by the colon. The differences between the ratios of Ca:PEG, Ac:PEG, and Pr:PEG in the baseline sample (0 min) and the respective ratios in the total colonic sample collected at 30 min were calculated and expressed as a percent of the baseline ratio. The resulting values (percent Ca, Ac, or Pr disappearance) were then converted to mmol by multiplying them by the amount of Ca, Ac, or Pr present in the distal colon at baseline (0 time) in mmol. The Ca absorption described in this study represents the rectum and distal colon. The average total length of the colon has been reported to be 110 cm with the distal segment comprising approximately one-third the total length.16 The diameter of the rectum and distal colon is about 4 cm. Therefore, the total volume of the rectum and distal colon are approximately 150 and 460 mL, respectively. Since the total volume of the solution infused for each subject was 300 mL, a substantial amount of the infused solution is expected to have reached the distal colon. The rectal infusion model has also been used to study the absorption of SCFAs from the rectum.17 The major drawback of this model is that absorptive surface area is unknown. Differences in diameter of the colon for each subject could make the surface area disproportional to the length of the colon; thus measurement of absorption rates in this study is relative rather than absolute. Unpublished data in our laboratory has shown that when similar solutions of SCFA-containing radioactive technicium (99Tc) were infused into the rectum of two of the subjects in this study, the infused solution reached the descending colon but did not enter the transverse colon.
CALCIUM ABSORPTION IN THE COLON
mmol/L Ca 4 11 24 46
531
TABLE II.
TABLE III.
CA ABSORPTION (mmol/L 30 min) AT DIFFERENT CONCENTRATIONS OF CA
AC AND PR ABSORPTION (mmol/L 30 min) AT DIFFERENT CONCENTRATIONS OF CA
Ca ⫹ NaCl
Ca ⫹ Ac
Ca ⫹ Pr
0.20 ⫾ 0.04aw 0.52 ⫾ 0.08ax 1.13 ⫾ 0.12ay 2.86 ⫾ 0.76az
0.38 ⫾ 0.07bw 0.75 ⫾ 0.20bx 2.45 ⫾ 0.40by 5.12 ⫾ 0.56bz
0.41 ⫾ 0.08bw 0.72 ⫾ 0.16bx 2.59 ⫾ 0.31by 7.28 ⫾ 0.83cz
abc
mmol/L Ca 4 11 24 46
Ca ⫹ Ac (Ac absorption)
Ca ⫹ Pr (Pr absorption)
7.46 ⫾ 1.40x 7.56 ⫾ 1.43x 7.88 ⫾ 1.23x 10.90 ⫾ 0.59y
6.34 ⫾ 1.01x 10.05 ⫾ 0.57y* 11.92 ⫾ 0.55z* 12.57 ⫾ 0.58z*
Means within rows (same Ca concentration) having different letter superscripts are significantly different; P ⬍ 0.05. wxyz Means within columns (different Ca concentrations) having different letter superscripts are significantly different; P ⬍ 0.05.
* Absorption of Pr significantly greater than Ac at the same Ca concentration; P ⬍ 0.05. xyz Means within columns (different Ca concentrations) having different letter superscripts are significantly different; P ⬍ 0.05.
Results are expressed as means ⫾ SEM. The significant differences between treatments and time were assessed by two-way repeated measures analysis of variance and the Duncan multiple range test. The relationship between Ca absorption and Ca concentration was determined by linear regression analysis. Data analysis was performed using Statistical Analysis System Program (SAS Institute Inc., Gary, NC, USA). The protocol for this study was approved by the Human Subjects Review Committee at the University of Toronto. All solutions were retained by the subjects for 30 min without difficulty. The mean total volume of infused solution recovered for all treatments was 159 mL (53% of the total volume infused) and there were no significant differences between treatments and subjects. The enema collected after infusion of double-distilled water contained 2.0 ⫾ 0.5 mmol/L of Ca and 2–5 mmol/L Ac and Pr. The residue remaining in the colon after the enema was negligible: after infusion of the control solution containing NaCl alone, the concentration of Ca in the infusate recovered from the rectum 30 min later was 2.1 ⫾ 0.8 mol/L, representing only 0.05– 0.005% of the concentrations of Ca in the test solutions. The concentrations of Ca, Ac, and Pr in the infused solutions did not differ
significantly from those in the samples of fluid collected from the rectum immediately after infusion (time 0). Ca absorption was affected significantly by Ca concentration and the presence of Ac and Pr (Table II). In the presence of NaCl, Ca absorption increased linearly as the concentration of Ca increased (r ⫽ 0.991, P ⫽ 0.001) with the slope of regression line being 62 ⫾ 5 mol 䡠 mmol⫺l 䡠 L Ca (Fig. 1). In the presence of Ac, Ca absorption was significantly greater than in the presence of NaCl at all concentrations of Ca (Table II). Ca absorption with Ac increased linearly with Ca concentration (r ⫽ 0.996, P ⬍ 0.001, Fig. 1) with the slope of the regression line being 113 ⫾ 6 mol 䡠 mmol⫺l 䡠 L Ca. Ca absorption in the presence of Pr tended to be greater than that in the presence of Ac, but the difference only reached significance at 46 mmol/L Ca (Table II, P ⬍ 0.05). In the presence of Pr, Ca absorption also increased linearly with Ca concentration (r ⫽ 0.984, P ⫽ 0.003), and the slope of regression line was 159 ⫾ 18 mol 䡠 mmol⫺l 䡠 L Ca. The slopes for the treatments with Ac and Pr were significantly steeper (P ⬍ 0.05) than that of the control (NaCl) and the difference between the slopes of Ca ⫹ Ac and Ca ⫹ Pr was also significant (P ⫽ 0.043). The absorption of Ac and Pr increased significantly as the concentration of Ca increased (Table III, Fig. 2). There was a
FIG. 1. Amount of calcium absorbed from the rectum and distal colon over 30 min after rectally infusing solutions containing different concentrations of calcium chloride plus 56.3 mmol/L sodium chloride (NaCl), sodium acetate (Ac), or sodium propionate (Pr). Points are means ⫾ SEM (n ⫽ 6); lines are linear regressions.
FIG. 2. Amount of acetate and propionate absorbed from the rectum and distal colon over 30 min after rectally infusing solutions containing 56.3 mmol/L sodium acetate or sodium propionate plus different concentrations of calcium chloride. Points are means ⫾ SEM (n ⫽ 6).
RESULTS
532
CALCIUM ABSORPTION IN THE COLON TABLE IV. pH OF THE INFUSED SOLUTIONS AND RECOVERED INFUSATES Ca concentration (mmol/L)
Treatment
Time (min)
4
11
24
46
Ca ⫹ NaCl
soln infused 0 30 soln infused 0 30 soln infused 0 30
5.3 ⫾ .01ax 8.1 ⫾ .10bx 8.2 ⫾ .10bx 7.0 ⫾ .02ax 8.5 ⫾ .10bx 8.3 ⫾ .10bx 7.4 ⫾ .03ax 8.1 ⫾ .08bx 8.1 ⫾ .09bx
5.0 ⫾ .10ay 7.9 ⫾ .10bx 8.0 ⫾ .10bx 7.0 ⫾ .03ax 7.9 ⫾ .02bz 8.0 ⫾ .01bxy 7.3 ⫾ .02ax 8.1 ⫾ .02bx 8.2 ⫾ .07bx
4.9 ⫾ .04ay 7.7 ⫾ .20bx 8.0 ⫾ .30bx 6.9 ⫾ .01ax 7.9 ⫾ .01bz 8.2 ⫾ .07cx 7.2 ⫾ .08axy 8.2 ⫾ .20bx 8.2 ⫾ .08bx
4.9 ⫾ .02ay 6.0 ⫾ .20by 7.2 ⫾ .06cy 7.0 ⫾ .05ax 8.2 ⫾ .08by 7.8 ⫾ .10by 7.1 ⫾ .03ay 7.6 ⫾ .10ay 7.5 ⫾ .30ay
Ca ⫹ Ac Ca ⫹ Pr
abc xyz
For comparisons between times within treatments, means having different letter superscripts are significantly different; P ⬍ 0.05. For comparisons within treatments between Ca concentrations, means having different letter superscripts are significantly different; P ⬍ 0.05.
positive linear relationship between Ca and Ac absorption (r ⫽ 0.53, P ⫽ 0.008) and between Ca and Pr absorption (r ⫽ 0.54, P ⫽ 0.006, Tables II and III). We did not control for the pH of each treatment. For Ca ⫹ NaCl, at 4, 11, and 24 mmol/L Ca, the pH of the infused solution significantly increased by the end of the infusion with no further increase at time 30 (Table IV). Similar results were obtained for Ca ⫹ Pr at 4, 11, and 24 mol/L Ca and for Ca ⫹ Ac at 4, 11, and 46 mmol/L Ca. For Ca ⫹ Ac at 24 mmol/L Ca, and for Ca ⫹ NaCl at 46 mmol/L Ca, pH had increased by the end of the infusion (0 min), and had increased further by 30 min. For Ca ⫹ Pr at 46 mmol/L Ca, no significant change in pH was observed at any time. The pH of the solutions containing Ca ⫹ Ac and Ca ⫹ Pr did not differ significantly. DISCUSSION
This study has shown that Ca absorption in the human distal colon is directly related to the Ca concentration, either alone or in the presence of Ac or Pr, suggesting that a non-saturable diffusion process (paracellular pathway) is involved. Ca absorption via a paracellular pathway in the distal colon is consistent with the fact that the activity of the Ca binding protein that is responsible for the active transport (transcellular pathway) of Ca in the small intestine is reduced in the large intestine.18 In the presence of Ac or Pr, the formation of a stable [CaAc]⫹ or [CaPr]⫹ complex19 may be a mechanism of Ca absorption in the distal colon. The significant relationship between the amounts of Ca and Ac absorbed and the amounts of Ca and Pr absorbed at different concentrations of Ca (Tables II and III) support this hypothesis. Because cell membranes have low permeability to highly charged ions [Ca⫹⫹], the less-charged complex ion may diffuse more readily through the cell membrane.20 Another possible mechanism of Ca absorption in the distal colon is by a solvent drag process. This process involves water movement through the intercellular junctions toward the mucosal side of the colon due to the activity of the sodium pump in the basolateral membrane, inducing a local hydrostatic pressure.21 This recycling of water may lead to a solvent drag effect by carrying solutes such as Ca from the intercellular space to the mucosal compartment. The velocity of this paracellular water flow is related to the amount of water absorbed.21 It has been shown that SCFAs stimulate water and sodium absorption.22 The enhancing effect of Ac or Pr on Ca absorption by the solvent drag process may be due
to increased water flow in the intercellular space, which may drag Ca to the mucosal side of the colon for absorption. Intestinal SCFA transport is mediated by a diffusion process and is concentration dependent.23 Non-ionic diffusion occurs when SCFAs are protonated with hydrogen ions, which are produced from either the luminal hydration of carbon dioxide to form bicarbonate and hydrogen ions, or from the secretion of hydrogen ions produced intracellularly. Ca may have replaced some of the hydrogen ions protonating Ac or Pr with the formation of a complex for its absorption in the colon; thus at higher concentrations of Ac or Pr and at increasing Ca concentration, the absorption of Ac or Pr was stimulated by Ca. It has been shown that SCFA absorption is independent of the overall pH of the colonic lumen due to a constant pH microclimate at the surface of the colonic epithelium.23 In the present study, differences in pH of the infused solutions do not account for the differences in Ca absorption. The solutions containing Ca ⫹ NaCl had the lowest pH and were associated with the lowest Ca absorption. On the other hand, the pH of the solutions containing Ca ⫹ Ac and Ca ⫹ Pr did not differ, but Ca ⫹ Pr resulted in a significantly greater Ca absorption than Ca ⫹ Ac at the highest concentration of Ca. The conclusion that Ca absorption was increased by Ac and Pr depends upon the assumption that Ca disappearance from the infusion solutions is a measure of Ca absorption, rather than the precipitation of Ca out of solution. Although steady-state conditions and chemical interactions that would precipitate Ca in the distal colon were not specifically investigated in this study, we believe that Ca disappearance is a measure of Ca absorption for two reasons. First, precipitation was not observed when the solutions were left overnight at 4°C. Second, we have observed that rectally infusing a solution containing Ca, Ac, and Pr increased serum Ca concentration compared to an infusion of Ca alone,7 a finding that is difficult to explain other than by increased Ca absorption from the rectum and distal colon. This study showed that, at the highest concentration of Ca, Pr increased Ca absorption significantly more than Ac, a finding that is consistent with the results of our previous study.9 This may be because Pr is more lipid soluble than Ac,23 and because of this, Pr is expected to diffuse faster across the colonocyte membrane with Ca. The normal North American diet is relatively high in phosphate and low in dietary fiber. Increased intake of dietary fiber may be beneficial for colonic health24 and in reducing serum
CALCIUM ABSORPTION IN THE COLON
533
cholesterol concentrations,25 but the possibility of reduced mineral absorption with high fiber intakes has been a major concern.26 Both phosphate and fiber bind calcium, making it unavailable for absorption. The difference between them is that unabsorbed phosphate does not release bound calcium in the colon and, thus, is a major determinant of fecal calcium excretion. Dietary fiber, on the other hand, may be fermented in the colon, a process that not only releases the bound Ca but also results in the production of SCFAs, which may promote the absorption of the released Ca. Thus, the
present results may help to explain why high-fiber diets, in the presence of adequate Ca intakes, are not associated with chronic Ca deficiency.26,27 We conclude that Ca absorption from the human rectum and distal colon is linearly related to Ca concentration in the presence and absence of Ac and Pr, indicating that it involves a nonsaturable diffusion process at least over the range of Ca concentration studied. Ca absorption is increased by Ac and Pr, and Ac and Pr absorption are stimulated by Ca.
REFERENCES 1. Bronner F, Pansu D, Stein WD. An analysis of intestinal Ca transport across the rat intestine. Am J Physiol 1986;250:G561 2. Bronner F. Intestinal Ca absorption: mechanisms and applications. J Nutr 1987;117:1347 3. Favus MJ, Langman CB. Effects of 1,25 dihydroxyvitamin D3 on colonic Ca transport in vitamin D deficient and normal rats. Am J Physiol 1984;246:G268 4. Karbach U, Rummel W. Trans and paracellular Ca transport across the colonic mucosa after short and long-term treatment with 1,25 dihydroxyvitamin D3. Eur J Clin Invest 1986;16:347 5. Lutz T, Scharrer F. The effect of SCFA on Ca absorption by the rat colon. Exp Physiol 1991;76:615 6. Stevens CE, Argenzio RA, Roberts MC. Comparative physiology of the mammalian colon and suggestions for animal models of human disorders. Clin Gastroenterol 1986;15:763 7. Trinidad TP, Wolever TMS, Thompson LU. Interactive effects of Ca and SCFA on absorption in the distal colon of man. Nutr Res 1993;13:417 8. Thompson LU, Trinidad TP, Wolever TMS. Ca absorption in the colon of man. Trace Elements Man Anim 1991;97:30 9. Trinidad TP, Wolever TMS, Thompson LU. Effect of acetate and propionate on calcium absorption from the rectum and distal colon of humans. Am J Clin Nutr 1996;63:574 10. Wolever TMS, Trinidad TP, Thompson LU. Short chain fatty acid absorption from the human distal colon: interactions between acetate, propionate and calcium. J Am Coll Nutr 1995;14:393 11. Allen LH. Calcium bioavailability and absorption: a review. Am J Clin Nutr 1982;35:783 12. Thompson LU, Trinidad TP, Jenkins DJA. Methods for determining the minerals available for absorption in the small intestine and colon. Trace Elements Man Anim 1991;7:25 13. Trinidad TP, Wolever TMS, Thompson LU. Availability of Ca for absorption in the small intestine and colon from diets containing available and unavailable carbohydrates: an in vitro assessment. Intl J Food Sci Nutr 1996;47:83 14. McBurney MI, Thompson LU, Cuff J, Jenkins DJA. Comparison of ileal effluents, dietary fibers and whole foods in predicting the physiological importance of colonic fermentation. Am J Gastroenterol 1988;83:536
15. Malawer SJ, Powell DW. An improved turbidemetric analysis of polyethylene glycol utilizing an emulsifier. Gastroenterol 1967;53:250 16. Blankenhom DH, Hirsch J, Ahrens EH Jr. Transintestinal intubation: technique for measurement of gut length and physiologic sampling at known loci. Proc Exp Biol Med 1955;88:356 17. Saunders DR. Absorption of short chain fatty acids in human stomach and rectum. Nutr Res 1991;11:841 18. Armbrecht HJ. Age and the effect of lactose on calcium and phosphorous uptake by rat. Nutr Res 1987;7:1169 19. Nancollas GH. Thermodynamics of ion association. Part II. Alkaline earth acetates and formates. J Chem Soc 1956:744 20. Marshall DA. Calcium and phosphate kinetics. In: Nordin BEC, ed. Calcium, phosphate and magnesium metabolism. New York: Longman Group Ltd, 1976:260 21. Schultz SG. Salt and water absorption by mammalian small intestine. In: Johnson LR, ed. Physiology of the gastrointestinal tract. New York: Raven Press, 1981:983 22. Argenzio RA, Stevens CE. The large bowel—a supplementary rumen? Nutr Soc 1984;43:13 23. Engelhardt W von. Absorption of short chain fatty acids from the large intestine. In: Cummings JH, Rombeau JL, Sakata T, eds. Physiological and clinical aspects of short chain fatty acids. Cambridge: Cambridge University Press, 1995:149 24. Edwards CA. Dietary fiber, fermentation and the colon. In: Cherbut C, Barry JL, Lairon D, Durand M, eds. Dietary fiber: mechanisms of action in human physiology and metabolism. Paris: John Libbey Eurotext, 1995:51 25. Marlett JA. Sites and mechanisms for the hypocholesterolemic actions of soluble dietary fiber sources. In: Kritchevsky D, Bonfield C, eds. Dietary fiber in health and disease. New York: Plenum Press, 1997: 109 26. Gordon DT, Stoops D, Ratliff V. Dietary fiber and mineral nutrition. In: Kritchevsky D, Bonfield C, eds. Dietary fiber in health and disease. St. Paul: Eagan Press, 1995:267 27. Weaver CM, Plawecki KL. Dietary calcium: adequacy of a vegetarian diet. Am J Clin Nutr 1994;59:1238S
July/August 1999
APPLIED NUTRITIONAL INVESTIGATION
Numbers 7/8 Nutrition Vol. 15, Nos. 7/8, 1999
Effects of Calcium Concentration, Acetate, and Propionate on Calcium Absorption in the Human Distal Colon TRINIDAD P. TRINIDAD, PHD, THOMAS M.S. WOLEVER, MD, PHD, AND LILIAN U. THOMPSON, PHD From the Department of Nutritional Sciences, Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada ABSTRACT
Previous studies have shown that the short-chain fatty acids acetate (Ac) and propionate (Pr) enhance the absorption of calcium (Ca) in the rectum and distal colon of humans, with Pr being more effective than Ac. To investigate the effect of Ac and Pr on the kinetics of Ca absorption from the human rectum and distal colon, six healthy subjects were studied. Solutions containing various concentrations of CaCl2 䡠 H2O with 56.3 mmol/L Ac, Pr, or NaCl were rectally infused to each subject. Rectal fluid was sampled at the end of the infusion (0 min), and 30 min later colonic contents were collected. Ca absorption for all treatments increased linearly with Ca concentration. For Ca ⫹ NaCl, the slope of regression line was 62 mol 䡠 mmol⫺1 䡠 L Ca. With Ac ⫹ Ca, the slope of Ca absorption increased significantly to 113 mol 䡠 mmol⫺1 䡠 L Ca, and with Pr ⫹ Ca, the slope increased to 159 mol 䡠 mmol⫺1 䡠 L (P ⫽ 0.043 versus Ac ⫹ Ca). Ac and Pr absorption were increased by Ca. The data suggest that, over a physiologic range of Ca concentration, in the absence or presence of Ac and Pr, Ca is absorbed in the human rectum and distal colon by a non-saturable diffusion process, and that Ca absorption is enhanced by Ac and Pr. The data also suggest that both Ac and Pr absorption is stimulated by Ca. Nutrition 1999;15:529 –533. ©Elsevier Science Inc. 1999 Key words: calcium absorption, acetate, propionate
INTRODUCTION
The effect of 1,25-dihydroxy cholecalciferol on calcium (Ca) absorption from the small intestine and colon has been studied in rats using different methods.1– 4 Most of these studies have shown that Ca is absorbed by both a saturable, active process and by a
non-saturable diffusion process. Likewise, the effect of the shortchain fatty acids (SCFAs) acetate (Ac) and butyrate (Bu) on Ca absorption from the distal colon of rats has been studied by the perfusion technique.5 It has been speculated in rats that a Ca/H exchange exists in the distal colon similar to the Na/H exchange
Supported by the Dairy Farmers of Canada and the Natural Sciences and Engineering Research Council. Correspondence to: Lilian U. Thompson, PhD, Department of Nutritional Sciences, Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada M5S 3E2. E-mail: [email protected]
Nutrition 15:529 –533, 1999 ©Elsevier Science Inc. 1999 Printed in the USA. All rights reserved.
0899-9007/99/$20.00 PII S0899-9007(99)00110-0
530
CALCIUM ABSORPTION IN THE COLON
present in the proximal colon. Although much information can be gathered from animal studies, extrapolation of their results to humans must be done with caution. There is evidence that SCFA absorption differs in different species.6 Also, most of the methods used in animal studies are difficult to reproduce in humans. We have used a rectal infusion technique to study Ca absorption from the human distal colon and have shown that Ac and propionate (Pr) enhance Ca absorption from this site.7–9 In our most recent study, we used a single concentration of Ca, 50 mmol/L, with various concentrations of Ac and Pr.9 At a concentration of 18.7 mmol/L, Ac and Pr were equally effective in enhancing Ca absorption, but at a concentration of 56.3 mmol/L, Pr was more effective.9 This suggested that Ac and Pr may have different mechanisms in enhancing Ca absorption. Thus, the purpose of the present study was to determine the effect of Ac and Pr on the kinetics of Ca absorption from the human rectum and distal colon. MATERIALS AND METHODS
Six healthy subjects (one male and five females, 29 ⫾ 4 y of age, 100 ⫾ 2% ideal weight) were studied on different mornings after overnight fasts. Solutions for rectal infusion were prepared in double-distilled water; 12 of the solutions contained various concentrations of Ca as CaCl2 䡠 2H2O (4, 11, 24, 46 mmol/L) with either 56.3 mmol/L sodium acetate (Ac), sodium propionate (Pr), or sodium chloride (NaCl). A solution of 56.3 mmol/L NaCl alone was used as a control. Polyethylene glycol (PEG, 0.625 mmol/L) was added to each solution as an unabsorbable marker. To avoid osmotic diarrhea, the osmolality of the solutions was ⬍300 mosm/L. We have shown that differences in osmolality of the test solutions across the range used here do not affect Ca absorption,7 and that the addition of NaCl to correct for differences in osmolality does not affect the absorption of Ca,7 Ac, or Pr.10 Table I shows the composition of the test solutions. The 13 different treatments were given to each subject in random order, with each subject being studied once or twice per week. The 56 mmol/L concentration of Ac and Pr was used in this study because differences in the effect of Ac and Pr on Ca absorption from the rectum and distal colon were observed at this concentration from our previous study.9 Average Ca intake on a usual Western diet is approximately 25 mmol/d11 of which 30 – 40% is absorbed in the small intestine.12,13 Therefore, approximately 15–17 mmol of Ca reaches the colon per day for potential absorption. Thus, the maximum amount of Ca infused into the rectum in this study, 13.8 mmol, is within the physiologic range. After an overnight fast and 10 –20 min before the start of the study, an enema containing 500 mL of double-distilled water was given through 1 m of Tygon flexible plastic tubing (inside:outside diameter, 2.4:4.0 mm; Norton Tubing and Molded Products, Akron, OH, USA) to clean the colon before infusion of the test solution. Then 5 cm of one end of the tubing was reinserted into the rectum. The subjects injected the 300 mL solution themselves, using a 60 mL syringe over 5 min under close supervision by one of the investigators. Before removing the infusion tube, a 5 mL sample was collected at 0 min by mixing (withdrawing and reinfusing) 10 mL of the fluid three times in 2 min. After 30 min, colonic contents were collected in a plastic bag using a fecal collection frame placed under the toilet seat. The 30-min time point was selected because our previous studies showed that serum Ca increased 30 min after rectal infusion of solutions containing Ca plus SCFAs7 and that Ca disappearance from the rectum increased linearly with time between 0 and 30 min for all treatments with and without Ac or Pr or both (r ⫽ ⫺0.96 to ⫺0.99, P ⫽ 0.04 to 0.006).9 Samples were centrifuged and the supernatant analyzed for the following: Ca using an atomic absorption spectrometer (Varian
TABLE I. COMPOSITION (mmol/L) OF THE DIFFERENT TREATMENT SOLUTIONS Treatment*
Ca†
Ac‡
Pr§
NaCl㛳
NaCl Ca ⫹ NaCl
— 4.0 11.0 24.0 46.0 4.0 11.0 24.0 46.0 4.0 11.0 24.0 46.0
— — — — — 56.3 56.3 56.3 56.3 — — — —
— — — — — — — — — 56.3 56.3 56.3 56.3
56.3 56.3 56.3 56.3 56.3 — — — — — — — —
Ca ⫹ Ac
Ca ⫹ Pr
* All treatments contained 0.625 mmol/L polyethylene glycol (PEG 4000), Mallinkrodt OR, Pointe-Claire, Quebec, Canada. † Calcium chloride (CaCl2 䡠 2H2O), Fisher Scientific, Ontario, Canada. ‡ Sodium acetate, NaAc, anhydrous, Sigma S-8750, St. Louis, MO, USA. § Sodium propionate, NaPr, Food Grade, Van Waters and Roger Ltd., London, Ontario, Canada. 㛳 Sodium chloride, NaCl, AR, British Drug House, Darmstadt, Germany.
Model 1275, Varian, Canada Inc., Ontario, Canada) after dilution with 10 mM lanthanum chloride; Ac and Pr by high performance liquid chromatography,14 and PEG by a turbidimetric method.15 Ca, Ac, and Pr disappearance from the distal colon was used as an index of absorption. PEG was used as an unabsorbable marker to correct for incomplete collection and fluid absorption or secretion by the colon. The differences between the ratios of Ca:PEG, Ac:PEG, and Pr:PEG in the baseline sample (0 min) and the respective ratios in the total colonic sample collected at 30 min were calculated and expressed as a percent of the baseline ratio. The resulting values (percent Ca, Ac, or Pr disappearance) were then converted to mmol by multiplying them by the amount of Ca, Ac, or Pr present in the distal colon at baseline (0 time) in mmol. The Ca absorption described in this study represents the rectum and distal colon. The average total length of the colon has been reported to be 110 cm with the distal segment comprising approximately one-third the total length.16 The diameter of the rectum and distal colon is about 4 cm. Therefore, the total volume of the rectum and distal colon are approximately 150 and 460 mL, respectively. Since the total volume of the solution infused for each subject was 300 mL, a substantial amount of the infused solution is expected to have reached the distal colon. The rectal infusion model has also been used to study the absorption of SCFAs from the rectum.17 The major drawback of this model is that absorptive surface area is unknown. Differences in diameter of the colon for each subject could make the surface area disproportional to the length of the colon; thus measurement of absorption rates in this study is relative rather than absolute. Unpublished data in our laboratory has shown that when similar solutions of SCFA-containing radioactive technicium (99Tc) were infused into the rectum of two of the subjects in this study, the infused solution reached the descending colon but did not enter the transverse colon.
CALCIUM ABSORPTION IN THE COLON
mmol/L Ca 4 11 24 46
531
TABLE II.
TABLE III.
CA ABSORPTION (mmol/L 30 min) AT DIFFERENT CONCENTRATIONS OF CA
AC AND PR ABSORPTION (mmol/L 30 min) AT DIFFERENT CONCENTRATIONS OF CA
Ca ⫹ NaCl
Ca ⫹ Ac
Ca ⫹ Pr
0.20 ⫾ 0.04aw 0.52 ⫾ 0.08ax 1.13 ⫾ 0.12ay 2.86 ⫾ 0.76az
0.38 ⫾ 0.07bw 0.75 ⫾ 0.20bx 2.45 ⫾ 0.40by 5.12 ⫾ 0.56bz
0.41 ⫾ 0.08bw 0.72 ⫾ 0.16bx 2.59 ⫾ 0.31by 7.28 ⫾ 0.83cz
abc
mmol/L Ca 4 11 24 46
Ca ⫹ Ac (Ac absorption)
Ca ⫹ Pr (Pr absorption)
7.46 ⫾ 1.40x 7.56 ⫾ 1.43x 7.88 ⫾ 1.23x 10.90 ⫾ 0.59y
6.34 ⫾ 1.01x 10.05 ⫾ 0.57y* 11.92 ⫾ 0.55z* 12.57 ⫾ 0.58z*
Means within rows (same Ca concentration) having different letter superscripts are significantly different; P ⬍ 0.05. wxyz Means within columns (different Ca concentrations) having different letter superscripts are significantly different; P ⬍ 0.05.
* Absorption of Pr significantly greater than Ac at the same Ca concentration; P ⬍ 0.05. xyz Means within columns (different Ca concentrations) having different letter superscripts are significantly different; P ⬍ 0.05.
Results are expressed as means ⫾ SEM. The significant differences between treatments and time were assessed by two-way repeated measures analysis of variance and the Duncan multiple range test. The relationship between Ca absorption and Ca concentration was determined by linear regression analysis. Data analysis was performed using Statistical Analysis System Program (SAS Institute Inc., Gary, NC, USA). The protocol for this study was approved by the Human Subjects Review Committee at the University of Toronto. All solutions were retained by the subjects for 30 min without difficulty. The mean total volume of infused solution recovered for all treatments was 159 mL (53% of the total volume infused) and there were no significant differences between treatments and subjects. The enema collected after infusion of double-distilled water contained 2.0 ⫾ 0.5 mmol/L of Ca and 2–5 mmol/L Ac and Pr. The residue remaining in the colon after the enema was negligible: after infusion of the control solution containing NaCl alone, the concentration of Ca in the infusate recovered from the rectum 30 min later was 2.1 ⫾ 0.8 mol/L, representing only 0.05– 0.005% of the concentrations of Ca in the test solutions. The concentrations of Ca, Ac, and Pr in the infused solutions did not differ
significantly from those in the samples of fluid collected from the rectum immediately after infusion (time 0). Ca absorption was affected significantly by Ca concentration and the presence of Ac and Pr (Table II). In the presence of NaCl, Ca absorption increased linearly as the concentration of Ca increased (r ⫽ 0.991, P ⫽ 0.001) with the slope of regression line being 62 ⫾ 5 mol 䡠 mmol⫺l 䡠 L Ca (Fig. 1). In the presence of Ac, Ca absorption was significantly greater than in the presence of NaCl at all concentrations of Ca (Table II). Ca absorption with Ac increased linearly with Ca concentration (r ⫽ 0.996, P ⬍ 0.001, Fig. 1) with the slope of the regression line being 113 ⫾ 6 mol 䡠 mmol⫺l 䡠 L Ca. Ca absorption in the presence of Pr tended to be greater than that in the presence of Ac, but the difference only reached significance at 46 mmol/L Ca (Table II, P ⬍ 0.05). In the presence of Pr, Ca absorption also increased linearly with Ca concentration (r ⫽ 0.984, P ⫽ 0.003), and the slope of regression line was 159 ⫾ 18 mol 䡠 mmol⫺l 䡠 L Ca. The slopes for the treatments with Ac and Pr were significantly steeper (P ⬍ 0.05) than that of the control (NaCl) and the difference between the slopes of Ca ⫹ Ac and Ca ⫹ Pr was also significant (P ⫽ 0.043). The absorption of Ac and Pr increased significantly as the concentration of Ca increased (Table III, Fig. 2). There was a
FIG. 1. Amount of calcium absorbed from the rectum and distal colon over 30 min after rectally infusing solutions containing different concentrations of calcium chloride plus 56.3 mmol/L sodium chloride (NaCl), sodium acetate (Ac), or sodium propionate (Pr). Points are means ⫾ SEM (n ⫽ 6); lines are linear regressions.
FIG. 2. Amount of acetate and propionate absorbed from the rectum and distal colon over 30 min after rectally infusing solutions containing 56.3 mmol/L sodium acetate or sodium propionate plus different concentrations of calcium chloride. Points are means ⫾ SEM (n ⫽ 6).
RESULTS
532
CALCIUM ABSORPTION IN THE COLON TABLE IV. pH OF THE INFUSED SOLUTIONS AND RECOVERED INFUSATES Ca concentration (mmol/L)
Treatment
Time (min)
4
11
24
46
Ca ⫹ NaCl
soln infused 0 30 soln infused 0 30 soln infused 0 30
5.3 ⫾ .01ax 8.1 ⫾ .10bx 8.2 ⫾ .10bx 7.0 ⫾ .02ax 8.5 ⫾ .10bx 8.3 ⫾ .10bx 7.4 ⫾ .03ax 8.1 ⫾ .08bx 8.1 ⫾ .09bx
5.0 ⫾ .10ay 7.9 ⫾ .10bx 8.0 ⫾ .10bx 7.0 ⫾ .03ax 7.9 ⫾ .02bz 8.0 ⫾ .01bxy 7.3 ⫾ .02ax 8.1 ⫾ .02bx 8.2 ⫾ .07bx
4.9 ⫾ .04ay 7.7 ⫾ .20bx 8.0 ⫾ .30bx 6.9 ⫾ .01ax 7.9 ⫾ .01bz 8.2 ⫾ .07cx 7.2 ⫾ .08axy 8.2 ⫾ .20bx 8.2 ⫾ .08bx
4.9 ⫾ .02ay 6.0 ⫾ .20by 7.2 ⫾ .06cy 7.0 ⫾ .05ax 8.2 ⫾ .08by 7.8 ⫾ .10by 7.1 ⫾ .03ay 7.6 ⫾ .10ay 7.5 ⫾ .30ay
Ca ⫹ Ac Ca ⫹ Pr
abc xyz
For comparisons between times within treatments, means having different letter superscripts are significantly different; P ⬍ 0.05. For comparisons within treatments between Ca concentrations, means having different letter superscripts are significantly different; P ⬍ 0.05.
positive linear relationship between Ca and Ac absorption (r ⫽ 0.53, P ⫽ 0.008) and between Ca and Pr absorption (r ⫽ 0.54, P ⫽ 0.006, Tables II and III). We did not control for the pH of each treatment. For Ca ⫹ NaCl, at 4, 11, and 24 mmol/L Ca, the pH of the infused solution significantly increased by the end of the infusion with no further increase at time 30 (Table IV). Similar results were obtained for Ca ⫹ Pr at 4, 11, and 24 mol/L Ca and for Ca ⫹ Ac at 4, 11, and 46 mmol/L Ca. For Ca ⫹ Ac at 24 mmol/L Ca, and for Ca ⫹ NaCl at 46 mmol/L Ca, pH had increased by the end of the infusion (0 min), and had increased further by 30 min. For Ca ⫹ Pr at 46 mmol/L Ca, no significant change in pH was observed at any time. The pH of the solutions containing Ca ⫹ Ac and Ca ⫹ Pr did not differ significantly. DISCUSSION
This study has shown that Ca absorption in the human distal colon is directly related to the Ca concentration, either alone or in the presence of Ac or Pr, suggesting that a non-saturable diffusion process (paracellular pathway) is involved. Ca absorption via a paracellular pathway in the distal colon is consistent with the fact that the activity of the Ca binding protein that is responsible for the active transport (transcellular pathway) of Ca in the small intestine is reduced in the large intestine.18 In the presence of Ac or Pr, the formation of a stable [CaAc]⫹ or [CaPr]⫹ complex19 may be a mechanism of Ca absorption in the distal colon. The significant relationship between the amounts of Ca and Ac absorbed and the amounts of Ca and Pr absorbed at different concentrations of Ca (Tables II and III) support this hypothesis. Because cell membranes have low permeability to highly charged ions [Ca⫹⫹], the less-charged complex ion may diffuse more readily through the cell membrane.20 Another possible mechanism of Ca absorption in the distal colon is by a solvent drag process. This process involves water movement through the intercellular junctions toward the mucosal side of the colon due to the activity of the sodium pump in the basolateral membrane, inducing a local hydrostatic pressure.21 This recycling of water may lead to a solvent drag effect by carrying solutes such as Ca from the intercellular space to the mucosal compartment. The velocity of this paracellular water flow is related to the amount of water absorbed.21 It has been shown that SCFAs stimulate water and sodium absorption.22 The enhancing effect of Ac or Pr on Ca absorption by the solvent drag process may be due
to increased water flow in the intercellular space, which may drag Ca to the mucosal side of the colon for absorption. Intestinal SCFA transport is mediated by a diffusion process and is concentration dependent.23 Non-ionic diffusion occurs when SCFAs are protonated with hydrogen ions, which are produced from either the luminal hydration of carbon dioxide to form bicarbonate and hydrogen ions, or from the secretion of hydrogen ions produced intracellularly. Ca may have replaced some of the hydrogen ions protonating Ac or Pr with the formation of a complex for its absorption in the colon; thus at higher concentrations of Ac or Pr and at increasing Ca concentration, the absorption of Ac or Pr was stimulated by Ca. It has been shown that SCFA absorption is independent of the overall pH of the colonic lumen due to a constant pH microclimate at the surface of the colonic epithelium.23 In the present study, differences in pH of the infused solutions do not account for the differences in Ca absorption. The solutions containing Ca ⫹ NaCl had the lowest pH and were associated with the lowest Ca absorption. On the other hand, the pH of the solutions containing Ca ⫹ Ac and Ca ⫹ Pr did not differ, but Ca ⫹ Pr resulted in a significantly greater Ca absorption than Ca ⫹ Ac at the highest concentration of Ca. The conclusion that Ca absorption was increased by Ac and Pr depends upon the assumption that Ca disappearance from the infusion solutions is a measure of Ca absorption, rather than the precipitation of Ca out of solution. Although steady-state conditions and chemical interactions that would precipitate Ca in the distal colon were not specifically investigated in this study, we believe that Ca disappearance is a measure of Ca absorption for two reasons. First, precipitation was not observed when the solutions were left overnight at 4°C. Second, we have observed that rectally infusing a solution containing Ca, Ac, and Pr increased serum Ca concentration compared to an infusion of Ca alone,7 a finding that is difficult to explain other than by increased Ca absorption from the rectum and distal colon. This study showed that, at the highest concentration of Ca, Pr increased Ca absorption significantly more than Ac, a finding that is consistent with the results of our previous study.9 This may be because Pr is more lipid soluble than Ac,23 and because of this, Pr is expected to diffuse faster across the colonocyte membrane with Ca. The normal North American diet is relatively high in phosphate and low in dietary fiber. Increased intake of dietary fiber may be beneficial for colonic health24 and in reducing serum
CALCIUM ABSORPTION IN THE COLON
533
cholesterol concentrations,25 but the possibility of reduced mineral absorption with high fiber intakes has been a major concern.26 Both phosphate and fiber bind calcium, making it unavailable for absorption. The difference between them is that unabsorbed phosphate does not release bound calcium in the colon and, thus, is a major determinant of fecal calcium excretion. Dietary fiber, on the other hand, may be fermented in the colon, a process that not only releases the bound Ca but also results in the production of SCFAs, which may promote the absorption of the released Ca. Thus, the
present results may help to explain why high-fiber diets, in the presence of adequate Ca intakes, are not associated with chronic Ca deficiency.26,27 We conclude that Ca absorption from the human rectum and distal colon is linearly related to Ca concentration in the presence and absence of Ac and Pr, indicating that it involves a nonsaturable diffusion process at least over the range of Ca concentration studied. Ca absorption is increased by Ac and Pr, and Ac and Pr absorption are stimulated by Ca.
REFERENCES 1. Bronner F, Pansu D, Stein WD. An analysis of intestinal Ca transport across the rat intestine. Am J Physiol 1986;250:G561 2. Bronner F. Intestinal Ca absorption: mechanisms and applications. J Nutr 1987;117:1347 3. Favus MJ, Langman CB. Effects of 1,25 dihydroxyvitamin D3 on colonic Ca transport in vitamin D deficient and normal rats. Am J Physiol 1984;246:G268 4. Karbach U, Rummel W. Trans and paracellular Ca transport across the colonic mucosa after short and long-term treatment with 1,25 dihydroxyvitamin D3. Eur J Clin Invest 1986;16:347 5. Lutz T, Scharrer F. The effect of SCFA on Ca absorption by the rat colon. Exp Physiol 1991;76:615 6. Stevens CE, Argenzio RA, Roberts MC. Comparative physiology of the mammalian colon and suggestions for animal models of human disorders. Clin Gastroenterol 1986;15:763 7. Trinidad TP, Wolever TMS, Thompson LU. Interactive effects of Ca and SCFA on absorption in the distal colon of man. Nutr Res 1993;13:417 8. Thompson LU, Trinidad TP, Wolever TMS. Ca absorption in the colon of man. Trace Elements Man Anim 1991;97:30 9. Trinidad TP, Wolever TMS, Thompson LU. Effect of acetate and propionate on calcium absorption from the rectum and distal colon of humans. Am J Clin Nutr 1996;63:574 10. Wolever TMS, Trinidad TP, Thompson LU. Short chain fatty acid absorption from the human distal colon: interactions between acetate, propionate and calcium. J Am Coll Nutr 1995;14:393 11. Allen LH. Calcium bioavailability and absorption: a review. Am J Clin Nutr 1982;35:783 12. Thompson LU, Trinidad TP, Jenkins DJA. Methods for determining the minerals available for absorption in the small intestine and colon. Trace Elements Man Anim 1991;7:25 13. Trinidad TP, Wolever TMS, Thompson LU. Availability of Ca for absorption in the small intestine and colon from diets containing available and unavailable carbohydrates: an in vitro assessment. Intl J Food Sci Nutr 1996;47:83 14. McBurney MI, Thompson LU, Cuff J, Jenkins DJA. Comparison of ileal effluents, dietary fibers and whole foods in predicting the physiological importance of colonic fermentation. Am J Gastroenterol 1988;83:536
15. Malawer SJ, Powell DW. An improved turbidemetric analysis of polyethylene glycol utilizing an emulsifier. Gastroenterol 1967;53:250 16. Blankenhom DH, Hirsch J, Ahrens EH Jr. Transintestinal intubation: technique for measurement of gut length and physiologic sampling at known loci. Proc Exp Biol Med 1955;88:356 17. Saunders DR. Absorption of short chain fatty acids in human stomach and rectum. Nutr Res 1991;11:841 18. Armbrecht HJ. Age and the effect of lactose on calcium and phosphorous uptake by rat. Nutr Res 1987;7:1169 19. Nancollas GH. Thermodynamics of ion association. Part II. Alkaline earth acetates and formates. J Chem Soc 1956:744 20. Marshall DA. Calcium and phosphate kinetics. In: Nordin BEC, ed. Calcium, phosphate and magnesium metabolism. New York: Longman Group Ltd, 1976:260 21. Schultz SG. Salt and water absorption by mammalian small intestine. In: Johnson LR, ed. Physiology of the gastrointestinal tract. New York: Raven Press, 1981:983 22. Argenzio RA, Stevens CE. The large bowel—a supplementary rumen? Nutr Soc 1984;43:13 23. Engelhardt W von. Absorption of short chain fatty acids from the large intestine. In: Cummings JH, Rombeau JL, Sakata T, eds. Physiological and clinical aspects of short chain fatty acids. Cambridge: Cambridge University Press, 1995:149 24. Edwards CA. Dietary fiber, fermentation and the colon. In: Cherbut C, Barry JL, Lairon D, Durand M, eds. Dietary fiber: mechanisms of action in human physiology and metabolism. Paris: John Libbey Eurotext, 1995:51 25. Marlett JA. Sites and mechanisms for the hypocholesterolemic actions of soluble dietary fiber sources. In: Kritchevsky D, Bonfield C, eds. Dietary fiber in health and disease. New York: Plenum Press, 1997: 109 26. Gordon DT, Stoops D, Ratliff V. Dietary fiber and mineral nutrition. In: Kritchevsky D, Bonfield C, eds. Dietary fiber in health and disease. St. Paul: Eagan Press, 1995:267 27. Weaver CM, Plawecki KL. Dietary calcium: adequacy of a vegetarian diet. Am J Clin Nutr 1994;59:1238S