HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Robinson, R. R. Articles by Slavin, J. L. Search for Related Content PubMed PubMed Citation Articles by Robinson, R. R. Articles by Slavin, J. L.

Effects of Dietary Arabinogalactan on Gastrointestinal and Blood Parameters in Healthy Human Subjects

Department of Food Science and Nutrition, University of Minnesota, St. Paul, Minnesota

Address reprint requests to: Joanne Slavin, Ph.D., Department of Food Science and Nutrition, 1334 Eckles Avenue, St. Paul, MN 55108. E-mail: jslavin{at}umn.edu

	ABSTRACT

TOP ABSTRACT INTRODUCTION METHODS BIOLOGICAL SAMPLE COLLECTION RESULTS DISCUSSION REFERENCES

 
Objectives: Arabinogalactan (AG) is a non-digestible soluble dietary fiber that resists hydrolytic enzyme action and enters the large bowel intact where it is fermented by resident microflora. To determine whether AG has similar physiological properties to other soluble dietary fibers, we examined the effect of 15 and 30 g per day of a commercially available AG from Western Larch on several gastrointestinal and blood parameters.

Methods: Gastrointestinal parameters included fecal microflora, fecal enzyme activity, fecal short-chain fatty acids, fecal pH, fecal weight, transit time and bowel frequency. Blood parameters included total cholesterol, HDL cholesterol, LDL cholesterol, triglycerides, Apo-A1, Apo-B, glucose and insulin. The study consisted of two three-week diet treatments with no washout period. Participants (n=20, 11 males, 9 females) consumed their usual diet in addition to 15 or 30 g AG in a beverage sweetened with aspartame as compared to their usual diet with the control beverage.

Results: Significant increases in total fecal anaerobes were observed with 15 g (p=0.01) and 30 g AG (p=0.001). A significant increase (p=0.02) in Lactobacillus spp. was observed when subjects consumed AG for a total of six weeks regardless of dose. There were no significant changes in other microflora, fecal enzyme activity, transit time, frequency, fecal weight, fecal pH and short-chain fatty acids. Fecal ammonia levels decreased with 15 g (p=0.001) and 30 g (p=0.002) AG. No significant changes in blood lipids or blood insulin were observed.

Conclusions: These data suggest that dietary AG is easily incorporated into the diet, well tolerated in subjects and has some positive effects on fecal chemistry.

Key words: arabinogalactan, microflora, ammonia, blood lipids, gastrointestinal transit time

	INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS BIOLOGICAL SAMPLE COLLECTION RESULTS DISCUSSION REFERENCES

 
Arabinogalactan (AG) is a soluble dietary fiber, commonly consumed in such foods as carrots, tomatoes, radishes, pears, maize, wheat and red wine [1]. In addition, several herbs have been found to contain significant amounts of AG, such as Echinacea purpurea, Angelica acutiloba and Curcuma longa [2–4]. The Western Larch (Larix occidentalis) and Mongolian Larch (Larix dahurica) are commercial sources of AG [5]. Arabinogalactan can be extracted from a variety of purified concentrated sources, although the commercial form used in this study was extracted from the butt wood of Western Larch grown in Northern Minnesota. Arabinogalactan derived from trees of the genus Larix (Larch) is a unique hemicellulosic product and is easily extractable by water in a pure form from non-delignified plant tissues. Arabinogalactans have an average molecular weight between 15,000 and 25,000. AG, also known as larch gum, is similar to gum arabic because it is highly branched, extremely water soluble, and high concentrations can be produced with very low viscosities [6].

Arabinogalactan is fermented by human intestinal bacteria and can induce the enzymes necessary for its degradation [7–11]. In addition, arabinogalactan is fermented at a slower rate than other carbohydrates due to its branched structure [12]. Fermentation is evidenced by the ability of human intestinal microflora to degrade arabinogalactan and produce short-chain fatty acids [13,14]. To date, the studies conducted with arabinogalactan are mainly in vitro. While this work contributes to our understanding of how this substrate is degraded, it is important to remember that the human colon is a complex environment and in vitro studies may not accurately represent bacterial activities within the human colon.

In addition to gastrointestinal parameters, blood lipids may be affected by fiber consumption. Increased fiber consumption may decrease blood cholesterol levels. There has not been previous research conducted evaluating the effect of arabinogalactan consumption on blood lipids. Thus, the objective of this study was to examine the physiological effects of a commercially available Larch arabinogalactan on the gut environment, blood lipids and blood glucose in healthy human subjects.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS BIOLOGICAL SAMPLE COLLECTION RESULTS DISCUSSION REFERENCES

 
Subjects
Subjects (11 male, 11 female) were recruited from the Twin Cities community. Subjects were screened for their ability to consume a beverage with or without AG, continue their habitual diet and exercise routines and provide blood samples on four occasions and fecal samples on three occasions. Participants’ baseline cholesterol levels were 196 ± 26 mg/dL (Mean ± SD). Exclusion criteria included pre-existing medical conditions, recent use of antibiotics or lipid altering medications, alcohol or drug abuse, cigarette smoking and extreme diet. The conditions and procedures of the study were reviewed, and written informed consent was obtained from each subject. Twenty subjects completed this study. One subject dropped out due to illness, and the other subject did not comply with protocol. All aspects of this research study were approved by the University of Minnesota Institutional Review Board Human Subjects Committee.

Study Design
The study utilized a crossover design with no washout period. Subjects were given a beverage containing no AG for seven days. Following this control period, subjects were randomly assigned to receive a dose of either 15 g or 30 g arabinogalactan (Larex Inc., St. Paul, MN). Each dose of AG was consumed for three weeks, and then subjects were crossed over to the other dose. AG was incorporated into 16 ounces of an aspartame-sweetened beverage (Crystal Light®). Subjects consumed one 16-ounce beverage per day in addition to their typical diet throughout the entire seven weeks of the study. They were instructed to consume each beverage given to them and to maintain their usual diet and activity level for the duration of the study. Subjects provided three-day diet records and symptom evaluation surveys once during each treatment (0g, 15g, 30g AG).

Assessment of Subjects’ Habitual Diets
During the last three days of baseline and treatment periods, subjects collected detailed three-day diet records. Nutrients were determined with the Nutrition Data System for Research (NDS-R) software version 4.0, developed by the Nutrition Coordinating Center, University of Minnesota, Minneapolis, MN, Food and Nutrient Database 28.

	BIOLOGICAL SAMPLE COLLECTION

TOP ABSTRACT INTRODUCTION METHODS BIOLOGICAL SAMPLE COLLECTION RESULTS DISCUSSION REFERENCES

 
Fecal Collection
Subjects collected fecal samples for the final five days of each treatment period. On days 3, 24 and 45 of the study, each subject swallowed plastic radio opaque pellets to mark intestinal transit time. All feces were subsequently collected into individual containers, defecation times were recorded and samples were weighed and frozen immediately at -20°C until analyzed. Fecal samples were subsequently X-rayed, and pellets per stool were counted. Passage of 80% of the pellets was considered transit time. The first four days of fecal samples for each subject were composited for calculation of stool weight. A fresh fecal sample was obtained from each subject at the conclusion of the transit time collection. Subjects were asked to defecate into sterile bags and include an anaero-pouch sachet (Mitsubishi Gas Chemical Company, Inc., New York, NY), which was sealed to keep the atmosphere reduced until sample analysis. Subjects delivered fresh fecal samples to our laboratory, and within 24 hours of defecation samples were analyzed for microbiological information. Subjects were given symptom evaluation questionnaires to fill out once during each phase of the study. Subjects marked their symptoms on a 145-mm line. Lines were measured and reported as subjective changes in gastrointestinal parameters.

Microbiology
Eleven grams of fresh fecal sample were obtained from the center of each stool and homogenized in 99 mL of pre-reduced 0.1% peptone water to provide a 1% (wt/vol) fecal slurry. One mL of slurry was diluted serially in peptone water and duplicate spread plates were made using 0.1 ml of each dilution. Total anaerobes were counted using Wilkins-Chalgren agar (Difco Laboratories, Detroit, MI) and enterobacteria were counted using MacConkey agar (Difco). Total lactic acid bacteria were counted using Lactobacilli modified MRS medium (Difco) [15]. Bifidobacterium spp. were counted on X--Gal based medium as described by Chevalier and colleagues [16]. Clostridium spp. were isolated on sulfite-polymyxin-milk agar. Plates were incubated at 37°C in the AnaeroPack^TM (Mitsubishi Gas Company) containing 20% CO₂ and read after 72 hours. Stool slurry pH was determined in each sample with a glass pH electrode.

ß-Glucosidase Enzyme Assay
Samples (40 mL) of 1:10 diluted stool from microbial enumeration studies were placed in 50 mL tubes; 4 mL of Oxyrase® For Broth (Oxyrase, Inc., Mansfield, OH) was added to each sample to maintain an anaerobic environment. Samples were stored at -20°C until analyzed. Samples were thawed, sonicated for three minutes and centrifuged for five minutes at 12,000 x g to pellet particulate matter. Samples were transferred to capped microfuge tubes for individual enzyme assays. ß-Glucosidase activity was assayed at 37°C under atmospheric conditions by following the hydrolysis of 3 mM p-nitrophenyl -ß-D-glucopyranoside (Sigma) after one hour and comparing the p-nitrophenol liberated to a standard curve at an absorbance of 405 nm. The pH of the 1 mL samples was adjusted with the addition of 100 µL 1.0 M potassium phosphate, 1.5 M NaCl, pH 5.5. The reaction was stopped with the addition of 100 µL 1M Na₂CO₃.

Short Chain Fatty Acids
After transit time calculations, four-day fecal collections were homogenized in a blender and stored at -20°C for SCFA analysis. Samples were thawed and 5 g aliquots were placed in Centriprep fluid concentrators, MWCO 30,000 kDa (Amicon Inc., Beverly, MA). Samples were centrifuged for 30 minutes at 1000 x g, room temperature and supernatants (total volume 0.75–1.0 mL) were placed in 15 mL polypropylene tubes; 0.3 mL of 25% m-phosphoric acid was added to each tube, and samples were vortexed and incubated at room temperature for 25 minutes. Samples were centrifuged at 5000 x g for 15 minutes at room temperature. Supernatants were decanted and frozen overnight. The following day, samples were thawed, and the pH of each sample was adjusted to 6.5 using 4 N KOH. Oxalic acid was added at a final concentration of 0.03%, and SCFA concentrations were determined by gas chromatography with use of a Hewlett-Packard 5880A gas chromatograph (Hewlet Packard, Palo Alto, CA) containing an 80/120 Carbopack B-DA/4% Carbowax 20M column (Supelco, Inc., Bellefonte, PA) [17].

Ammonia Assay
Fecal ammonia levels were assayed using the CHEMets® Ammonia-Nitrogen Kit (CHEMetrics, Calverton, VA). One-mL fecal supernatant samples were diluted with 24 mL of distilled, deionized water. Glass ampoules containing Nessler’s reagent, an alkaline solution comprising mercuric iodide and sodium hydroxide, were inserted into diluted fecal samples and filled. Ampoules were mixed, allowed to react for one minute and quantified by comparing to a set of colored standards. A yellow color developed in the presence of ammonia.

Blood Parameters
Fasting blood samples were drawn on the last day of baseline diet and on the last day of each three-week feeding treatment. Blood samples were analyzed for total cholesterol, HDL cholesterol, LDL cholesterol, triglycerides, Apo-A1, Apo-B, glucose and insulin.

Statistical Analysis
Statistical evaluation of results was done by analysis of variance with repeated measures using the factors: 0 g fiber vs. the mean of 15 g and 30g AG treatment. Data were evaluated for the effects of treatment, order and time. Values in tables represent means ± standard error of the means (SEM). Data were analyzed using SAS [18].

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS BIOLOGICAL SAMPLE COLLECTION RESULTS DISCUSSION REFERENCES

 
Three-Day Diet Records
Review of the subjects’ habitual diets indicated that the mean carbohydrate intake as a percentage of total kilocalories did not change significantly throughout the study. Mean protein intake as a percentage of total kilocalories increased significantly (p=0.02) between baseline (14.70 ± 0.68%) and 15 g AG (17.06 ± 0.68%), while there were no significant differences between baseline and treatment with 30 g AG. Mean fat gram intake decreased significantly (p = 0.04) between baseline (84.60 ± 4.50) and 30 g AG (70.86 ± 4.50), while there were no significant differences between baseline and treatment with 15 g AG. There were significant increases in fiber intake when baseline was compared to both the 15 g AG and 30 g AG treatment. Total dietary fiber intakes, including the dietary fiber from AG, were 17.8 g ± 9.0 g for control, 30.0 g ± 8.5 g for the 15 g AG treatment and 41.5 g ± 6.2 g for the 30 g AG treatment.

Intestinal Microflora
There were significant differences in levels of total anaerobes and Lactobacillus species following AG consumption (Table 1). Data are expressed in colony forming units (CFUs) on the log 10 scale. Randomization order did not significantly affect bacterial counts. There were significant increases (p=0.01) in total anaerobes between baseline (10.35 ± 0.10) and the two levels of treatment, 15 g AG (10.74 ± 0.10) and 30 g AG (10.74 ± 0.10) respectively. Lactobacillus spp. measured (9.36 ± 0.14) at baseline and for the two levels of treatment, 15 g AG (9.73 ± 0.14) and 30 g AG (9.73 ± 0.14). These increases were not statistically significant. Length of time consuming AG appeared to be more important than dose (Table 2). Mean Lactobacillus spp. increased significantly (p = 0.02) between baseline (9.36 ± 0.14) and following six weeks of AG consumption (9.82 ± 0.14), whereas three weeks of AG consumption did not produce significant increases in Lactobacillus spp. Levels of fecal Bifidobacterium spp., Clostridium spp. and Enterobacteriaceae did not differ significantly between baseline and AG treatments.

SCFA and SCFA Ratios
The SCFA and SCFA ratios did not change after AG administration (Table 3).

Bowel Habit: Composite Fecal Weight, Intestinal Transit and Frequency
Mean fecal weight, transit time and frequency did not differ significantly between baseline and both the 15g and 30g dose of AG (Table 4).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS BIOLOGICAL SAMPLE COLLECTION RESULTS DISCUSSION REFERENCES

There was no increase in Bifidobacteria counts, another colonic microbe found to promote health benefits. This may be due, in part, to the significant increase in the Lactobacilli population. Species of Lactobacillus may compete with Bifidobacterium spp. for available substrate and adhesion sites within the colonic epithelium.

Because the majority of bacterial fermentation is thought to occur in the proximal colon, analysis of fecal instead of colonic flora probably does not best represent activities within the colon. Additionally, short-term feeding studies may not provide the necessary time to produce recognized changes in bacterial populations.

There were no statistically significant changes in fecal SCFAs or SCFA ratios. Vince and colleagues [11] also did not find increases in fecal short-chain fatty acid production following arabinogalactan consumption. However, their work as well as the work of Englyst and colleagues [12] did report increases in SCFA production following arabinogalactan supplementation of fecal incubates.

Short chain fatty acids are believed to be quickly absorbed following their production; therefore, it is difficult to determine the total amount produced in human subjects. At least 95% of SCFAs produced in the colon are absorbed and therefore can not be seen upon evaluation of fecal samples.

Fecal ammonia levels decreased significantly with both 15 g and 30 g AG. This supports the work of Vince and colleagues [11], who found that subjects fed arabinogalactan had decreased fecal ammonia concentrations following AG supplementation of fecal incubates. High colonic ammonia levels may have detrimental health implications. Studies have shown that ammonia levels as low as 5 mmol/L can have cytopathic effects on colonic epithelial cells. Ammonia has been shown to affect the intermediary metabolism and DNA synthesis of mucosal cells [20]. Ammonia is reported to be toxic toward epithelial cells, a circumstance which leads to their increased turnover. Patients with liver disease who are unable to detoxify ammonia have been successfully treated with antibiotics and lactulose. Lactulose is fermented in the colon by bacteria that utilize ammonia as a nitrogen source, thus decreasing colonic ammonia concentration [21]. AG appears to be similar to lactulose in that it decreases fecal ammonia concentrations.

In the current study, ammonia levels may have been reduced due to the significant increases in total anaerobes. Some anaerobic colonic bacteria prefer to utilize ammonia as a nitrogen source rather than amino acids or peptides when fermenting carbohydrates. A strain of Eubacterium species is reported to have a strict requirement for ammonia [22]. Eubacterium was not a bacterial species enumerated in the current study. Undetected increases in this particular bacterial species may have contributed to the increase in total anaerobes

There were no observed changes in fecal wet weight, transit time or frequency following consumption of arabinogalactan. Gum arabic, a fiber similar to AG, also does not affect fecal wet weight, but has been shown to increase transit time [23]. Soluble dietary fibers, such as AG, are largely fermented, so any increase in fecal weight is due to increases in fermentation gasses and bacterial mass resulting from the proliferation of microbes metabolizing the dietary fiber [24].

Subjects reported no significant changes in bloating, flatulence or stool consistency during the consumption of 15 g AG, although they reported increases at the 30 g AG dose. The increase in flatulence was likely due to the increase in bacterial fermentation in the colon and concomitant production of gases such as hydrogen and methane.

Significant decreases in fat consumption were observed when subjects consumed the 30 g dose of AG. A reason for this change may be explained by the increased reports of bloatedness (fullness) when subjects consumed the 30 g dose of AG. A sensation of fullness may have led subjects to avoid high fat foods.

There were no significant changes in blood lipids following AG consumption. Some soluble dietary fibers have been associated with decreases in total plasma cholesterol. There are a variety of potential cholesterol lowering mechanisms associated with the consumption of dietary fiber. These mechanisms are related to viscosity, SCFA production and bacterial proliferation. Arabinogalactan is relatively non-viscous and therefore may not decrease cholesterol levels for this reason. Another mechanism believed to be involved in the cholesterol lowering effects of dietary fibers is elevated levels of short-chain fatty acids. When dietary fibers are fermented, short-chain fatty acids are produced. There is some research to support that propionate may be the hypocholesterolemic short-chain fatty acid. Also, Lactobacilli bacteria may lower serum cholesterol levels, although the mechanisms are unclear. The microflora may be involved in the deconjugation of bile salts and subsequent inefficient cholesterol absorption, or they may possibly assimilate cholesterol and remove it from the colon [19].

Blood glucose significantly increased following the consumption of 30 g AG. Blood samples were taken from fasted subjects, and we therefore did not expect to see increases in blood glucose levels at any time. The reason for these increases during the treatment phases remains unknown, although possible explanation could be associated with its influence on the production of specific fermentation end products. Increased glucose levels might have been due to an undetected increase in the production of the fermentation end product propionate, which is believed to travel to the liver and increase gluconeogenesis.

In conclusion, a 15 g or 30 g per day supplement of AG increased total fecal anaerobes and decreased fecal ammonia concentrations. Consumption of AG for six weeks led to increased Lactobacillus populations. A dose of 30 g AG increased blood glucose levels. A dose of 15 g/day AG appears to be particularly well tolerated by subjects and has some positive effects on fecal chemistry.

	ACKNOWLEDGMENTS

 
The authors wish to thank Ms. Jennifer Causey for her assistance in lab work and technical support, Mr. Richard Flores for help with diet records and drink preparation and Ms. Julie Rapp for microbial enumeration and processing of fecal samples. This work was supported by a grant from Larex, Incorporated.

	FOOTNOTES

 
This work was supported by Larex, Inc., St. Paul, MN.

Received November 1, 2000. Accepted March 2, 2001.

	REFERENCES

TOP ABSTRACT INTRODUCTION METHODS BIOLOGICAL SAMPLE COLLECTION RESULTS DISCUSSION REFERENCES

 

1. D’Adamo PD: Larch arabinogalactan. J Naturopath Med 6: 33–37, 1990.
2. Egert D, Beuscher N: Studies on antigen specificity of immunoreactive arabinogalactan proteins extracted from Baptisia tinctoria and Echinacea purpurea. Planta Med 58: 163–165, 1992.[Medline]
3. Kiyohara H, Cyong JC, Yamada H: Relationship between structure and activity of an anti-complementary arabinogalactan from the roots of Angelica acutiloba Kitagawa. Carbohydr Res 193: 193–200, 1989.[Medline]
4. Gonda R, Tomoda M, Ohara N, Takada K: Arabinogalactan core structure and immunological activities of uknan C, an acidic polysaccharide from the rhizome of Curcuma longa. Biol Pharm Bull 16: 235–238, 1993.[Medline]
5. Odonmazig P, Ebringerova A, Machova E, Alfoldi J: Structural and molecular properties of the arabinogalactan isolated from Mongolian larchwood (Larix dahurica). Carbohydr Res 252: 317–324, 1994.[Medline]
6. Furia TE: Arabinogalactan (larch gum). In Spiller GA (ed): "CRC Handbook of Food Additives," 2nd ed. Boca Raton, FL: CRC Press, pp 316–317, 1972.
7. Salyers AA, Arthur R, Kuritza A: Digestion of larch arabinogalactan by a stain of human colonic Bacteroides growing in continuous culture. J Agric Food Chem 29: 475–480, 1981.[Medline]
8. Salyers AA, Vercelotti JR, West SEH, Wilkins TD: Fermentation of mucin and plant polysaccharides by strains of Bacteroides from the human colon. Appl Environmental Microbiol 33: 319–322, 1976.
9. Macfarlane GT, Macfarlane S, Gibson GR: Coculture of Bifidobacterium adolescentis and Bacteroides thetaiotaomicron in arabinogalactan-limited chemostats: Effects of dilution rate and pH. Anaerobe 1: 275–281, 1995.
10. Crociani F, Alessandrini A, Mucci MM, Biavati B: Degradation of complex carbohydrates by Bifidobacterium spp. Int J Food Microbiol 24: 199–210, 1994.[Medline]
11. Vince A, McNeil NI, Wagner JD, Wrong OM: The effect of lactulose, pectin, arabinogalactan and cellulose on the production of organic acids and metabolism of ammonia by intestinal bacteria in a fecal incubation system. Br J Nutr 63: 17–26, 1990.[Medline]
12. Englyst HN, Hay S, Macfarlane GT: Polysaccharide breakdown by mixed populations of human faecal bacteria. FEMS Microbiol Ecol 95: 163–172, 1987.
13. Mazur AW, Mohlenkamp MJ, Hiller G: Digestibility of selected carbohydrates by anaerobic bacteria. J Agric Food Chem 41: 1925–1930, 1993.
14. Bradburn DM, Mathers JC, Gunn A, Chapman PD, Johnston IDA: Colonic fermentation of complex carbohydrates in patients with familial adenomatous polyposis. Gut 4: 630–636, 1993.
15. Roy D, Ward P: Evaluation of rapid methods of differentiation of Bifidobacterium species. J Appl Bacteriol 69: 739–749, 1990.
16. Chevalier P, Roy D, Savoie L: X--Gal-based medium for simultaneous enumeration of bifidobacteria and lactic acid bacteria in milk. J Microbiol Methods 13: 75–83, 1991.
17. Erwin ES, Marco GT, Emery EM: Volatile fatty acid analysis of blood and rumen fluid by gas chromatography. J Dairy Sci 44: 1768–1774, 1961.[Abstract/Free Full Text]
18. SAS, Release 6.12. Cary NC: SAS Institute Inc, 1999.
19. Pfeifer A, Rosat JP: Probiotics in alimentation; clinical evidence for their enhancement of the natural immunity of the gut. In Hanson, Yolken (eds): "Probiotics, Other Nutritional Factors, and Intestinal Microflora." Philadelphia: Lippencott-Raven, 1999.
20. Visek WJ: Ammonia metabolism, urea cycle capacity and their biochemical assessment. Nutr Rev 37: 273–282, 1979.[Medline]
21. Weber FL: The effect of lactulose on urea metabolism and nitrogen excretion in cirrhotic patients. Gastroenterology 77: 518–523, 1979.[Medline]
22. Bryant MP, Robinson IM: Some nutritional characteristics of predominant culturable ruminal bacteria. J Bacteriol 84: 605–615, 1962.[Abstract/Free Full Text]
23. Ross AH, Eastwood MA, Brydon WG, Anderson JR, Anderson DM: A study of the effects of dietary gum arabic in humans. Am J Clin Nutr 37: 368–375, 1983.[Abstract/Free Full Text]
24. Stephen AM, Cummings JH: Mechanisms of action of dietary fiber in the human colon. Nature 284: 283–284, 1980.[Medline]

This article has been cited by other articles:

				R. L Schnaar and H. H Freeze A "Glyconutrient Sham" Glycobiology, September 1, 2008; 18(9): 652 - 657. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Robinson, R. R. Articles by Slavin, J. L. Search for Related Content PubMed PubMed Citation Articles by Robinson, R. R. Articles by Slavin, J. L.