The Journal of Nutrition Vol. 128 No. 1 January 1998, pp. 79-84

Breath Hydrogen and Methane Expiration in Men and Women after Oat Extract Consumption¹

Kay M. Behall², Daniel J. Scholfield^*, Anna M. C. van der Sluijs, and Judith Hallfrisch^*

Diet and Human Performance Laboratory and the^* Metabolism and Nutrient Interactions Laboratory, Beltsville Human Nutrition Research Center, Agricultural Research Service, U.S. Department of Agriculture, Beltsville, MD 20705; and the  Human Nutrition and Food Science Department, University of Maryland, College Park, MD 20742

	ABSTRACT

Abstract Introduction Methods Results Discussion References

Oat extract has been shown to modify blood glucose response and fasting lipids after dietary incorporation although some abdominal discomfort and increased flatulence were noted. To determine the extent of gas production, hydrogen and methane were determined after tolerance tests containing cooked and uncooked oat extract and after dietary incorporation. Breath gases were determined before and periodically after tolerance tests. Study 1: While consuming a maintenance diet, 24 subjects (55.3-112.5 kg body weight) underwent a tolerance test (1 g carbohydrate/kg body wt) of glucose (GTT, 1700 kJ/100 g) or uncooked, baked, or boiled pudding [2191 kJ/100 g carbohydrate, (0.67 glucose and 0.33 oat extract containing 10 g/100 g -glucan)]. Hydrogen and methane expiration after all tolerance tests with the oat extract puddings, regardless of cooking method, was significantly higher than expirations after the GTT. Cooking the oat extract did not significantly change hydrogen or methane expiration. Study 2: Twenty-three subjects consumed a maintenance diet followed by the incorporation of oat extracts (50 g/8.33 MJ, 1 or 10 g/100 g -glucan) to the diet in a crossover pattern. A GTT and a tolerance test containing 0.67 g glucose and 0.33 g of the respective oat extract/kg body weight were consumed after the maintenance and oat extract diet periods. Breath hydrogen was significantly higher after both oat extract tolerance tests than after the GTT. Hydrogen excretion after the 10% -glucan oat extract was higher at 4, 5 and 6 h than after the 1% -glucan oat extract; breath methane was not significantly different. These data indicate that cooking did not alter the influence of oat extracts on intestinal function, and increased -glucan marginally increased hydrogen expiration.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

Recommendations have been made to increase the dietary intake of carbohydrate as starch and fiber (USDA 1990). Consumption of diets high in complex carbohydrates has been reported to improve blood lipid profiles (Anderson and Akanji 1993, Behall 1990, Glore et al. 1994, Jenkins et al. 1993) and glycemic response (Anderson and Akanji 1993, Behall 1990, Brand et al. 1991), which are metabolic changes beneficial to individuals at risk for atherosclerosis and maturity-onset diabetes. In several studies, oats and oat bran (natural sources high in soluble fibers) were used successfully to lower blood lipid concentrations and have been promoted as lipid-lowering foods (Anderson and Akanji 1993, Jenkins et al. 1993, Schinnick et al. 1991). The soluble fiber (-glucan) in the oats appears to be one of the components responsible for lowering plasma lipids (Behall 1990, Glore et al. 1994, Jenkins et al. 1993). An oat product (Oatrim®) containing concentrated amounts of -glucan was developed and patented by Dr. George Inglett (Agricultural Research Service, USDA, Peoria, IL) for use as a replacement for fat in foods (Inglett 1991).

Approximately 10% of ingested carbohydrate escapes digestion in the small intestine and is available for microbial fermentation in the colon (Cummings and Englyst 1991, Stephen 1991, Wisker and Feldheim 1992). Short-chain fatty acids produced during fermentation may be important in maintaining colonic health (Cummings and Englyst 1991, Livesey and Elia 1995). Hydrogen can be produced during colonic fermentation and is excreted in the breath or as flatus (Cummings and Englyst 1991). Because colonic fermentation is the only source of breath hydrogen (Muir et al. 1994), hydrogen measured in breath can be related to undigested carbohydrates, fiber (Hanson and Winterfeldt 1985) as well as starch (Cummings and Englyst 1991, Muir et al. 1994, Stephen 1991), reaching the colon.

As part of a larger study, we measured breath hydrogen and methane excretion of subjects after consumption of an oat fiber extract (amylodextrin with 1-10% soluble -glucans by weight) in a test meal as an acute tolerance test as well as part of the total diet.

	SUBJECTS AND METHODS

Abstract Introduction Methods Results Discussion References

The study was approved by the Human Studies Committees of the U.S. Department of Agriculture, Georgetown University and Johns Hopkins University. Written informed consent was obtained from the subjects before participation in the study in accordance with institutional guidelines. All subjects were medically evaluated by a physician, and blood was drawn for a routine clinical profile before subjects started the study. None of the subjects was taking drugs known to affect lipid or glucose metabolism, and all medications taken by the subjects during the study were recorded.

Acute tolerance study.  The acute tolerance study was designed to investigate the breath hydrogen and methane responses related to a food containing an oat fiber extract (90% amylodextrin and 10% soluble -glucans) that was served uncooked and prepared by different cooking methods (e.g., boiling and baking). The oat extracts for the tolerance tests and for the dietary incorporation study were provided by Quaker Oats (St. Louis, MO) and Con Agra (Omaha, NE). Subjects were paired, one male and one female, by age and body mass index (kg/m²) as similarly as possible (Table 1).

After an initial screening glucose tolerance test (GTT, 1 g glucose/kg body weight), 24 nondiabetic subjects, 12 men and 12 women, consumed a maintenance diet (15% of the energy from protein, 55% from carbohydrate and 30% from fat) for 3 d, prepared and weighed in the Beltsville Human Nutrition Study Facility (HNSF). Subjects ate breakfast in the facility and took lunch and dinner with them for consumption elsewhere (i.e., work or home). Subjects were instructed to consume all food before 2000 h on d 2 so that they were in a fasting state for sample collection at the start of d 3. Subjects consumed instant pudding containing the oat extract as a tolerance test (1 g carbohydrate/kg body weight). Two thirds of the total carbohydrate consumed was from the pudding and one third was from oat extracts. Men and women consumed an average of 30 g and 23 g of oat extract (dry weight)(3.0 and 2.3 g -glucan), respectively, for each tolerance test. The pudding/oat extract mixture was either uncooked (prepared according to package directions), boiled or baked. The quantity consumed in the tolerance test was based on the prepared weight. The pudding provided 14.62 kJ/kg body weight, and the oat extract was assumed to provide 5.56 kJ/kg body weight. A Latin-square design was used to randomize the order in which subjects were fed the oat extract puddings.

Sample collection and analysis.  Subjects were shown the procedure required for proper collection of breath samples; they were also given written directions. Alveolar air samples were obtained when subjects exhaled through a mouthpiece connected to a dual-bag system by a three-way valve. Breath samples were collected before each load. Samples were collected again 2 h after the load, hourly between h 4 and 10 and 24 h after the load. Subjects were instructed to consume lunch after the 6-h collection. The menu on the day of collection (Table 2) contained no known hydrogen producers, such as lactose in milk products. Subjects were not permitted to exercise during the collections.

Breath hydrogen, methane and carbon dioxide were determined by using a gas chromatograph with a hydrogen detector (MicroLyzer Model SC, Quintron Instrument, Milwaukee, WI). The instrument was calibrated with a reference containing the three measured gases. Breath samples that contained carbon dioxide levels below that expected for alveolar air were discarded. The averages of duplicate measurements at each collection were calculated. Subjects were classified as high or low producers by their methane production after the carbohydrate load was consumed. Methane producers were defined as those producing >10 mmol/L above the nadir observed during the GTT.

Dietary incorporation study.  The dietary incorporation study was designed to investigate the breath hydrogen and methane responses related to chronic consumption of a variety of foods containing the oat extract. Subjects who had fasting plasma cholesterol levels (as determined on an automated spectrophotometric system, Baker Instruments, Allentown, PA)(Behall 1997) between the 50th and 75th percentiles for their age and gender while consuming their self-selected diets were recruited. Twenty-three volunteers, 7 men and 16 women, completed the study. One man was dropped from the study for failure to comply with the study dietary requirements. Within sex, subjects were paired by plasma cholesterol level, age, height and BMI as similarly as possible (Table 1).

Subjects were fed a controlled diet (maintenance, 15% of the energy from protein, 50% from carbohydrate and 35% from fat) consisting of common foods for 1 wk before consuming the diets containing the oat extracts. The same 7-d menus were then fed with the oat extracts [either 1 g/100 g (low) or 10 g/100 g (high) soluble -glucans] replacing 5% of the fat energy with a corresponding increase of the carbohydrate energy. The basic diet provided an average of 1.3 g soluble fiber and 14.0 g total dietary fiber/8.33 MJ (2000 kcal) before the addition the oat extracts (50 g oat extract/8.33 MJ, dry weight); the diet with the low -glucan oat extract averaged 2.1 g soluble fiber and 14.3 total fiber/8.33 MJ, and the diet with the high -glucan oat extract averaged 8.7 g soluble fiber and 26.1 total fiber/8.33 MJ. Each oat fiber diet was fed for 5 wk in a crossover pattern. The soluble oat extract was added to the diet in muffins, cake, brownies, waffles, gelatin, yogurt, spaghetti sauce and meat loaf. Additional information on subjects, study design and menus has been published (Hallfrisch et al. 1995).

View this table: [in this window] [in a new window]   Table 3. Breath hydrogen and methane in 24 human subjects after acute tolerance tests of glucose and pudding with 10% -glucan oat extract, prepared by different methods1,2

View larger version (30K): [in this window] [in a new window]   Fig 1. Breath methane response to a glucose tolerance test (GTT) (1 g glucose/kg body weight) or uncooked and cooked pudding (one third of the carbohydrate from 10% -glucan oat extract) from high methane producer (n = 6) and low producer (n = 18) subjects after acute tolerance tests of each product. Values are least-square means ± SEM (± 3.2 for high producers, ± 1.8 for low producers). Breath methane responses were significantly higher in the high producers with a group × diet × time interaction (P < 0.002). In the high producers, methane expiration was greater after the boiled than the baked pudding.

View larger version (27K): [in this window] [in a new window]   Fig 2. Breath methane response to a glucose tolerance test (GTT) (1 g glucose/kg body weight) after the maintenance diet, to 0.67 g glucose plus 0.33 g oat extract (1 g -glucans/100 g) (low) after the low -glucan oat extract diet, and to 0.67 g glucose plus 0.33 g oat extract (10 g -glucans/100 g) (high) after the high -glucan oat extract diet in high methane producers (n = 6) and low producers (n = 17). Values are least-square means ± SEM (± 2.2 for high producers, ± 1.8 for low producers). Breath methane responses were significantly higher in the high producers (P < 0.001), but not due to diet (P = 0.19) or response × diet interaction (P = 0.70).

View this table: [in this window] [in a new window]   Table 4. Breath hydrogen and methane in 23 human subjects after a tolerance test of glucose with and without oat extract after consumption of a maintenance, low -glucan and high -glucan diet1,2

The maintenance and oat extract diets were prepared and weighed at HNSF. The diets were designed to meet the recommended dietary allowance for all nutrients (NRC 1989). During the controlled-diet periods, morning and evening meals were consumed at HNSF. Lunches and weekend and holiday meals were packaged for consumption elsewhere. Subjects were required to consume only the foods (including sugar or milk added to coffee) provided by the diet facility and to consume all foods given to them. Nonenergy-containing beverages were permitted in addition to the weighed foods. All other energy sources, including candy and alcohol, were proscribed. Subjects' dietary energy levels were based on their initial weight and the Harris-Benedict (1919) equations with individual estimates for activity, or on weight maintenance energy levels previously established for subjects who had participated in previous controlled dietary studies. Subjects were instructed to maintain their usual activities and exercise regimen during the study. During the dietary intervention, energy intake was adjusted during the controlled diet period by decreasing or increasing all foods proportionately when a subject gained or lost 2 kg. Over the 11-wk study, an average of 2.1 kg was lost by the subjects even though energy intake was increased (Behall 1997).

Sample collection and analysis.  Subjects fasted after dinner at the end of each dietary period. The next morning fasting samples were collected as described in the acute tolerance tests. Subjects were then given a carbohydrate solution totaling 1 g/kg body wt, based on the previous day's weight. After the equilibration diet, subjects consumed a GTT. After the two periods in which oat extract was contained in the diet, subjects consumed solutions containing 0.67 g glucose and 0.33 g oat extract/kg body wt. The oat extract in the tolerance solution was that consumed by the subject in the previous 5 wk. Liquid volumes for each tolerance were equivalent, 0.33 mL/kg body weight, but consistency of the tolerance test meals varied with the high -glucan oat extract the most viscous. Breath samples were collected again 2 h after the load, and hourly from h 4 through h 10 after the carbohydrate load. Breath samples were analyzed as described in the acute tolerance tests.

Data analysis.  Data were analyzed statistically using the mixed procedure of SAS (PCSAS, version 6.11, SAS Institute, Cary, NC). Data from the acute tests were evaluated for the fixed effects of cooking (uncooked, boiled, baked), gender, period, time and interactions between cooking, gender and time. Data after the dietary incorporation tolerance tests were evaluated for the fixed effects of diet (level of -glucan), gender, period, time and interactions between diet, gender and time. Data from the acute tests and the dietary incorporation were also evaluated, with a division of subjects between high and low methane producers replacing gender in the analysis. The analysis of repeated measures utilized a compound symmetry covariance matrix. Selected a priori comparisons were performed. Data on physical characteristics are arithmetic means ± SEM. Data for breath hydrogen expirations are reported as least squares means ± SEM. Means comparisons were determined by Least Significant Differences (LSD). The critical level of significance for all tests was set at P < 0.05 (LSD).

	RESULTS

Abstract Introduction Methods Results Discussion References

Acute tolerance tests.  Gender did not affect breath hydrogen or methane expiration. Overall, breath hydrogen and methane expired after all tolerance tests with oat extract pudding were significantly higher (P < 0.001 and 0.01, respectively) than expiration after GTT regardless of cooking method (Table 3). Hydrogen expired at 4, 5 and 6 h was significantly higher after the oat extract puddings than after the GTT (time, P < 0.001). No significant time relationship was observed in methane excretion. Although peak height varied, no significant differences in peak time between the tolerance tests were observed. Boiling or baking oat extract puddings did not significantly alter hydrogen or methane expiration compared with expiration after ingestion of uncooked oat extracts.

Six of the 24 subjects were considered to be methane producers after the GTT. Methane excretion (Fig. 1) by the producers was greatest after the boiled oat extract pudding, significantly higher than levels after the GTT or baked pudding tolerance tests (producer × diet × time, P < 0.001). Consumption of boiled or baked oat extract puddings did not significantly change overall hydrogen or methane expiration compared with that after the consumption of uncooked oat extracts. Only in the high methane producers was there greater excretion after boiled than after baked oat extract pudding (P < 0.002).

Dietary incorporation.  After consumption of each oat extract diet for 5 wk, breath hydrogen and methane expirations after a tolerance test containing glucose plus the dietary oat extract were not significantly different between the men and women. Overall hydrogen expiration was significantly higher when the oat extracts were included in the diet (diet, P < 0.03, diet × time, P < 0.02) than after the maintenance diet (Table 4). Breath hydrogen was higher in samples from fasting subjects after both oat extract tolerance tests than after the maintenance diet (P < 0.05). Hydrogen expiration remained significantly higher (P < 0.03) at the 2- and 4-h collections after both oat extract tolerance tests than after the GTT. Hydrogen expiration remained significantly higher (P < 0.05) after the high -glucan extract through the 7-h collection and after the low -glucan extract through the 6-h collection than after GTT. Hydrogen excretion after the 10% -glucan oat extract was higher at 4, 5 and 6 h than after the 1% -glucan oat extract. Thereafter, hydrogen expiration after the GTT increased to match that observed after the oat extract tolerance tests. No significant differences in peak time between the tolerance tests were observed.

Methane expiration was not significantly different when subjects ingested the two levels of -glucan. There was some fluctuation in excretion with time (P < 0.02). However, most of the fluctuation occurred in a few subjects who were classified as high methane producers (four men, four premenopausal women, and two postmenopausal women) (Fig. 2). Neither the high or low methane producers had greather methane expiration when they consumed the higher level of -glucan (Fig. 2).

	DISCUSSION

Abstract Introduction Methods Results Discussion References

The oat extract used in this study was developed to serve as a fat replacer in food products by developing a viscous gel with liquids. Incorporation of the moderate amount of oat extract (50-75 g/d), containing a mixture of amylodextrin and -glucans, into the diet was determined to be feasible while maintaining acceptability of the diet as a whole. As with the change in any fiber source, some intestinal discomfort was noted by some subjects during the wk 1 when the higher fiber diet was consumed. Complaints ranged from a very full feeling to bloating and flatulence.

After dietary incorporation of the two oat extracts, which differed in the amount of -glucan, only hydrogen expiration was significantly increased. The rise was greater after the oat extract containing 10% -glucan, the more viscous oat extract. Marthinsen and Fleming (1982) fed five men, for 9 d, a controlled diet including xylan, pectin, cellulose or corn bran and compared it with a fiber-free formula diet. Methane and hydrogen excretions were greater with the xylan and pectin diets than with the other diets. Hydrogen and methane excretions after cellulose and corn bran consumption were similar to that observed after the fiber-free diet. As noted in our study, the viscous fibers, rather than the particulate fibers, resulted in the higher gas excretion. With all of the diets, 2-5 d of consumption was necessary to establish a relatively constant level of gas excretion. Adaptation to our diets containing the oat extracts also appeared to occur within wk 1. Marthinsen and Fleming (1982) reported hydrogen production increasing throughout the day, whereas methane production remained relatively constant.

Melcher et al. (1991) screened subjects for methane production. Thirty-six percent of subjects screened were judged to be methane producers, within the range (30-38%) suggested for the average population. Methane excreted was not significantly changed by the addition, for 10 d, of 24 g/d of dietary fiber from cereal (Fiber One) to a self-selected diet in the 15 methane producers or 9 nonproducers consuming the supplement. Subjects for the two studies reported here were not screened for methane production when they consumed their habitual diet. The number of methane producers in the present study averaged 25% in the acute testing study and 43% in the dietary intervention study, near the percentage observed by Melcher and co-workers (1991).

The addition of isolated viscous fibers to an acute tolerance test meal has generally not resulted in increased breath hydrogen excretion. No significant change in hydrogen excretion from that observed after the fiber-free meal was observed after guar gum mixed with bread (15 and 100 g, respectively) (Robb et al. 1991) nor after 20 g of guar or psyllium was added to a polysaccharide-free breakfast (Wolever et al. 1992). Psyllium or pectin mixed with milk reduced the excretion of hydrogen in lactose malabsorbers compared with the response observed after milk alone (Nguyen et al. 1982). However, 40 g of oat or citrus fiber (Hanson and Winterfeldt 1985) and oat gum (2.4 g, equivalent to 1.9 g -glucan) added to rice pudding (Lund and Johnson 1991) resulted in increased hydrogen excretion compared with control tolerances. In this study, hydrogen excretion increased after all of the oat extract tolerance tests (oat extract puddings or glucose/oat extract mixtures) compared with base line.

Colonic fermentation has been noted as the only source of breath hydrogen (Muir et al. 1994). It has been used to estimate the amount of undigested fiber (Hanson and Winterfeldt 1985) and starch (Cummings and Englyst 1991, Muir et al. 1994, Stephen 1991) reaching the colon. A number of mechanisms have been proposed by which fiber may affect the amount of digestible nutrients reaching the colon. These include delaying gastric emptying, decreasing transit time, decreasing pH and increasing viscosity, thereby forming a physical barrier and decreasing enzyme/substrate contact.

The viscous fibers, oat and citrus, used by Hanson and Winterfeldt (1985) resulted in greater hydrogen concentration and slower peak excretion (6.2 and 6.4 h, respectively) than did the particulate fibers, wheat and corn (5.6 h peak). The time delay may have resulted from entrapment of nutrients by the viscous fiber. We observed no significant peak time difference in either study; only peak height varied significantly between the tolerance tests.

Extraction of the fiber from the original food matrix appears to reduce the amount of hydrogen produced even when the amount of fiber was similar. Breath hydrogen after 50 g carbohydrate loads of whole oat flour averaged 2.5 times higher than after refined oat flour; whole wheat yielded ~5 times as much hydrogen as did the refined white wheat flour (Levitt et al. 1987). It seems likely that fiber also interfered with the digestion and/or absorption of the starch moiety of the flour because hydrogen excretion has been observed after ingestion of 100 g of carbohydrate as white wheat flour, which contains minimal fiber (Levitt et al. 1987). Because most of either oat extract used in the dietary incorporation study was amylodextrin, and the breath hydrogen expiration was elevated even with the low -glucan oat extract, an interaction between the fiber and starch moiety is likely.

In this study, breath hydrogen and methane expiration were not significantly different after cooked vs. uncooked oat extract. Hydrogen excretion after the diet and tolerance test with 10% -glucan oat extract was modestly higher than the maintenance or 1% -glucan oat extract diets. Most subjects excreted little methane regardless of the diet. Lund and Johnson (1991) reported that acute meals of 50 g uncooked rolled oats (equivalent to 2.1 g of -glucan) gave a higher average hydrogen production than did equivalent cooked rolled oats [84.1 ± 13.0 and 65.2 ± 11.3 mmol/(L·h), respectively], but the differences were not significant (Lund and Johnson 1991).

The maintenance diet, with 35% of energy from fat, was similar to the average fat intake in the U.S. (Stephen and Wald 1990), whereas the oat extract diet with 30% fat was equivalent to intakes recommended in Dietary Guidelines for Americans (USDA 1990). The modest reduction in dietary fat was accomplished without changing the foods used in the diet. With minimal changes in food selection or preparation, this oat extract may improve the diets of those at risk for cardiovascular disease or diabetes. Further research is necessary to determine whether colonic digestion of the oat extract contributes primarily to short-chain fatty acid production, because breath hydrogen excretion was minimal.

	ACKNOWLEDGMENTS

We thank Willa Mae Clark and Elisa Armero for excellent technical support. We thank Evelyn Lashley and the staff of the Human Study Facility, BHNRC, for assistance with the experimental diet phase of the study, and the subjects who completed the study.

	FOOTNOTES

Manuscript received 7 October 1996. Initial reviews completed 25 November 1996. Revision accepted 15 September 1997.

	LITERATURE CITED

Abstract Introduction Methods Results Discussion References

• Anderson, J. W. & Akanji, A. O. (1993) Treatment of diabetes with high fiber diets. In: CRC Handbook of Dietary Fiber in Human Nutrition (Spiller, G. A., ed.), pp. 443-470. CRC Press, Ann Arbor, MI.
• Behall, K. M. (1990) Effect of soluble fibers on plasma lipids, glucose tolerance and mineral balance. In: New Developments in Dietary Fiber (Furda, I. & Brine, C. J., eds.), pp.7-16. Plenum Press, New York, NY.
• Behall K. M., Scholfield D. J., Hallfrisch J. Effect of beta-glucan level in oat fiber extracts on blood lipids in men and women. J. Am. Coll. Nutr. 1997; 16:46-51 [Abstract]
• Brand J. C., Colagiuri S., Crossman S., Allen A., Truswell A. S. Low glycemic index carbohydrate foods improve glucose control in non-insulin dependent diabetes mellitus (NIDDM). Diabetes Care 1991; 14:95-101 [Abstract]
• Cummings J. H., Englyst H. N. Measurement of starch fermentation in the human large intestine. Can. J. Physiol. Pharmacol. 1991; 69:121-129 [Medline]
• Glore S. R., Van Treeck D., Knehans A. W., Guild M. Soluble fiber and serum lipids. A literature review. J. Am. Diet. Assoc. 1994; 94:425-436 [Medline]
• Hallfrisch J., Scholfield D. J., Behall K. M. Diets containing soluble oat extracts improve glucose and insulin responses of moderately hypercholesterolemic men and women. Am. J. Clin. Nutr. 1995; 61:379-384 [Abstract/Free Full Text]
• Hanson C. F., Winterfeldt E. A. Dietary fiber effects on passage rate and breath hydrogen. Am. J. Clin. Nutr. 1985; 42:44-48 [Abstract/Free Full Text]
• Harris, J. A. & Benedict, F. G. (1919) A Biometric Study of Basal Metabolism in Man, Publication No. 279. Carnegie Institution, Washington, DC.
• Inglett, G. E. (February 26, 1991) Method for making a soluble dietary fiber composition from oats. U.S. Patent No. 4,996,063.
• Jenkins, D.J.A., Spadafora, P. J., Jenkins, A. L. & Rainey-Macdonald, C. G. (1993) Fiber in the treatment of hyperlipidemia. In: CRC Handbook of Dietary Fiber in Human Nutrition (Spiller, G. A., ed.), pp. 419-438. CRC Press, Ann Arbor, MI.
• Levitt M. D., Hirsh P., Fetzer C. A., Sheahan M., Levine A. S. H₂ excretion after ingestion of complex carbohydrates. Gastroenterology 1987; 92:383-389 ^[Medline]
• Livesey, G. & Elia, M. (1995) Short-chain fatty acids as an energy source in the colon: metabolism and clinical implications. In: Physiological and Clinical Aspects of Short-Chain Fatty Acids (Cummings, J. H., Rombeau, J. L. & Sakata, T., eds.), pp. 427-481. Cambridge University Press, New York, NY.
• Lund E. K., Johnson I. T. Fermentable carbohydrate reaching the colon after ingestion of oats in humans. J. Nutr. 1991; 121:311-317
• Marthinsen D., Fleming S. E. Excretion of breath and flatus gases by humans consuming high-fiber diets. J. Nutr. 1982; 112:1133-1143
• Melcher E. A., Levitt M. D., Slavin J. L. Methane production and bowel function parameters in healthy subjects on low- and high-fiber diets. Nutr. Cancer 1991; 16:85-92 [Medline]
• Muir J. G., Lu Z. X., Young G., Gameron-Smith D., Collier G. R., O'Dea K. Resistant starch in the diet increases breath hydrogen and serum acetate in human subjects. Am. J. Clin. Nutr. 1994; 61:792-799 ^{[Abstract/Free Full Text]}
• National Research Council (1989) Recommended Dietary Allowances, 10th ed., p. 799. National Academy Press, Washington, DC.
• Nguyen K. N., Welsh J. D., Manion C. V., Ficken V. J. Effect of fiber on breath hydrogen response and symptoms after oral lactose malabsorbers. Am. J. Clin. Nutr. 1982; 35:1347-1351 [Abstract/Free Full Text]
• Robb P. A., Wolever T.M.S., Hassanein J. Use of breath hydrogen to quantify carbohydrate malabsorption: lack of effect of soluble fibre (guar). Nutr. Res. 1991; 11:1349-1356
• Schinnick F. L., Mathews R., Ink S. Serum cholesterol reduction by oats and other fiber sources. Cereal Foods World 1991; 36:815-821
• Stephen A. M. Starch and dietary fibre: their physiological and epidemiological interrelationships. Can. J. Physiol. Pharmacol. 1991; 69:116-120 [Medline]
• Stephen A. M., Wald N. J. Trends in individual consumption of dietary fat in the United States, 1920-1984. Am. J. Clin. Nutr. 1990; 52:157-169
• U.S. Department of Agriculture & U.S. Department of Health and Human Services (1990) Nutrition for Your Health. Dietary Guidelines for Americans, 3rd ed., pp. 273-293. U.S. Government Printing Office, Washington DC.
• Wisker, E. & Feldheim, W. (1992) Faecal bulking and energy value of dietary fibre. In: Dietary Fibre  a Component of Food. Nutritional Function in Health and Disease (Schweizer, T. F. & Edwards, C. A., eds.), pp. 233-246. Springer-Verlag, London, UK.
• Wolever T.M.S., ter Wal P., Spadafora P., Robb P. Guar, but not psyllium, increases breath methane and serum acetate concentrations in human subjects. Am. J. Clin. Nutr. 1992; 55:719-722 ^{[Abstract/Free Full Text]}

0022-3166/98 $3.00 ©1998 American Society for Nutritional Sciences