- Email: [email protected]
Impact of dietary cereal on nutrient absorption and fecal nitrogen loss in formula-fed infants
Impact of dietary cereal on nutrient absorption and fecal nitrogen loss in formula-fed infants Robert J. Shulman, MD, Thomas W. Boutton, PhD,* a n d Peter D. Klein, PhD From the U.S, Department of Agriculture/Agricultural Research Service Children's Nutrition Research Center, Department of Pediatrics, Baylor Cotlege of Medicine, and Texas Children's Hospital, Houston, Texas
Study objective: To determine the c a p a c i t y of infants to digest and absorb rice cereal and to determine the effect of cereal f e e d i n g on total e n e r g y and nitrogen absorption. Setting: Subjects' residences and the Texas Children's Hospital Clinical Research Center, Houston. Patients: Eight healthy 1-month-old bottle-fed infants. Interventions: Infants were fed their usual formula for 3 days. For the subsequent 6 days, they r e c e i v e d 4 gm of rice cereal, l a b e l e d with c a r b o n 13, per 30 ml of the formula. Measurements and main results: Fecal b a l a n c e studies were performed for a 72-hour period while the infants received only formula and a g a i n during the last 3 days of c e r e a l feeding. Breath samples for h y d r o g e n measurement were coll e c t e d before and after the cereal feeding. Nutrient intake was measured and stools were analyzed for 13C a b u n d a n c e , energy, nitrogen, fiber content, and bacterial mass. Cereal absorption was 88 + 9% (mean • SD). Despite a significant increase in energy and nitrogen intake from cereal feeding, the coefficient of absorption fell (energy: 97% to 90%, p -- 0.048; nitrogen: 94% to 74%, p = 0.009). Fecal dry weight increased after the c e r e a l f e e d i n g (p = 0.004), primarily as a result of a sevenfold increase in f e c a l bacterial mass ( p - 0.002). Fecal nitrogen increased primarily because of incorporation of nitrogen into bacteria. No differences were d e t e c t e d in breath h y d r o g e n as a c o n s e q u e n c e of formula f e e d i n g versus formula-cereal feeding. Conclusions: Although cereal was relatively w e l l - a b s o r b e d in this group of infants and increased their intakes of energy and nitrogen, it did not increase the coefficients of e n e r g y and nitrogen absorption. Cereal f e e d i n g increased fecal bacterial mass and bacterial nitrogen. (J PEDIATR1991;118:39-43)
Supported in part with federal funds from the U.S. Department of Agriculture, Agricultural Research Service, under Cooperative Agreement No. 58-7MN1-6-100. The contents of this publication do not necessarily reflect the views or policies of the U.S. Department of Agriculture, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government. Submitted for publication July 20, 1990; accepted Aug. 20, 1990. Reprint requests: Robert J. Shulman, MD, Children's Nutrition Research Center, 1100 Bates Street, Houston, TX 77030. *Now at Texas A&M University, College Station, Tex. 9/20/24946
Cereal is usually the first solid food given to young infants. Although recommendations have been made that solids such as cereal not be added to an infant's diet until 4 to 6 months of age, cereal is often fed within the first 3 months of life. One of the concerns related to the early introduction of cereal into the diet of young infants is the infants' low level of pancreatic amylase before 6 months of age, 1 which has generally been assumed to lead to significant malabsorption of cereal. 2 Our previous studies suggested that digestion of complex carbohydrates, such as cereal, may be more com-
39
40
Shulman. Boutton, and Klein
plete in the young infant than originally believed.2 Evidence suggests that the small intestinal enzyme glucoamylase is important in the digestion of complex carbohydrates in the young infant before the appearance of pancreatic amylase activity) The young infants who are most likely to receive substantial amounts of dietary cereal before 4 to 6 months of age are those with gastroesophageal reflux. Although the addition of cereal to thicken the formula does not appear to reduce the amount of reflux, as determined by esophageal pH probe monitoring, it does reduce the amount of spitting and vomiting. 4 Infants with gastroesophageal reflux may have poor weight gain related to vomiting. The addition of cereal to the formula increases the caloric density of the feeding. 4 However, the extent to which cereal is digested and absorbed and the effects of cereal on the digestion and absorption of other nutrients are unknown. The objectives of our study were to determine the capacity of young infants to digest and absorb rice cereal and the effect of rice cereal feeding on energy and nitrogen absorption. Additionally, we sought to ascertain the effect of cereal feeding on fecal bacterial mass. METHODS
Subjects. Eight infants were selected for the study after informed consent was obtained from their parents. The study was approved by the human investigations committees of Baylor College of Medicine and Texas Children's Hospital. The infants were between 16 and 40 days of age at the start of the study. All infants had normal perinatal histories and were healthy. Design. Breath samples were collected for hydrogen determinations at the start of the study and at 30-minute intervals for 3 hours after the ingestion of 4 ounces (120 ml) of the infant's usual formula, as previously described. 2, 3 The infants were then sent home. During this time, the infants were fed their usual formula on demand, and all feces were collected for 72 hours (days 2 to 5). Preweighed cloths were provided to account for possible losses because of spitting up. The diaper containing the stool was marked with the date and time of collection and was placed in a plastic bag and stored in the freezer. A research nurse visited the homes frequently to ensure that the protocol was followed and to pick up the diapers containing the stools. On day 7, breath samples were again collected for hydrogen determinations before and after the ingestion of 4 ounces (120 ml) of formula containing 4 gm ( ~ 1 tablespoon) of rice cereal (Gerber, Fremont, Mich.) per 30 ml. After the breath test, the infants continued to receive their usual formula containing the rice cereal. On day 10, the infants were readmitted to the clinical re-
The Journal of Pediatrics January 1991
search center for the second phase of the study. Readmission was necessary because the formula and cereal mixture had to be consumed soon after it was prepared. The commercial rice cereal in the formula was replaced by enriched rice cereal labeled with carbon 13 (see discussion of diet, below). Four grams of 13C-enriched rice was added to each 30 ml of formula. All stools were collected for 3 days. On day 13, breath samples were collected after the formula plus cereal mixture was fed, and the infants were discharged from the clinical research center. Diet. On days 7 through 9, the infants were fed rice cereal at home. Preweighed packets of cereal (32 gm) to be added to 8-ounce (240 ml) bottles of formula were provided. Details of the 13C-enriched rice cereal fed to the infants during the 3-day stool collection in the clinical research center have been described previously.5 The 13C-labeled rice was grown in a 13CO2-enriched environment and then processed in a manner identical to that of rice in commercially available Gerber rice cereal. 2 Calculation of nutrient intake and output. Intake was calculated as the difference in weight between the bottles containing formula, or formula and cereal, before and after feeding. Preliminary studies indicated that the proportion of formula to cereal remaining in the bottle after a feeding was similar to that when the formula and cereal were first prepared and mixed in the bottle. Carmine red was used as the colorant to mark the beginning and end of the 72-hour stool collection. A l l stools collected during the 72-hour period were pooled and analyzed. Net absorption was defined as total intake during the 72 hours minus total output over the same period. Coefficient of absorption was defined as follows: ([Total i n t a k e - T o t a l output]/[Total intake]) • 100. Formula and stool analyses. The energy and nitrogen contents of the formula and the formula-cereal mixture were determined by lyophilizing an aliquot of the mixture and performing bomb calorimetry (Parr Instrument Co, Moline, Ill.) and a micro-Kjeldahl procedure (Tecator Inc, Herndon, Va.), respectively. The frozen stool samples were mixed in deionized water, brought to a known volume, and homogenized. Aliquots were weighed, frozen in liquid nitrogen, and lyophilized. The aliquots were weighed again to obtain the dry weight, and then analyzed for energy and nitrogen as described above. Fecal bacterial mass and fiber content were determined from other aliquots by means of the filtration method of Stephen and Cummings. 6 Stools were divided into three fractions; fractions 1 and 2 were primarily fiber, and fraction 3 was an almost pure bacterial isolate.6 The concentration of nitrogen was also measured in the three fractions to determine the amount of fecal nitrogen produced by bacte -
Volume 118 Number 1
Nutrient absorption and nitrogen loss with cereal
Table I. Energy and nitrogen intake and net retention
Table II. Stool dry weight, fiber, bacterial mass, and
Measurement
Energy intake (kcal/kg/day) Energy absorption (kcal/kg/day) Nitrogen intake (mg/kg/day) Nitrogen absorption (mg/kg/day)
Formula
128 125 405 381
+ 25 _+ 24 .+ 84 _+ 75
Formula plus cereal
151 _+ 26* 139 _+ 23 481 _+ 84? 379 ___78
*p = 0.027. "l'p= 0.024.
ria. We determined the digestion of the rice cereal by cambusting other fecal aliquots to obtain the percentage of carbon and the 13C isotope abundance, as we have previously described. 2 Breath hydrogen analysis. Samples for breath hydrogen analysis were stored in gas-tight syringes and analyzed by gas-solid chromatography as previously described, except that results were not corrected to an alveolar reference of 5% CO2 .2' 3 Results were expressed as peak breath hydrogen excretion in parts per million (ppm) of room air, as well as the area under the curve (concentration versus time) calculated by the trapezoidal rule. Statistics. Results are expressed as mean + SD. The Student t test for paired data was used for comparisons. Two-way analysis of variance was used to compare breath hydrogen results. RESULTS The energy density of the formula was 0.67 kcal/gm and the nitrogen concentration was 2.13 mg/gm. For formula plus cereal, these values were 1.0 kcal/gm and 3.2 mg/gm, respectively. The infants' energy intakes were greater while they were receiving formula plus cereal; similarly, nitrogen intake was greater when the infants were fed formula plus cereal (Table I). Despite greater intakes of energy and nitrogen when the infants were fed formula plus cereal, there was no change in net energy or nitrogen absorption (Table I). Thus the coefficients of absorption for energy and nitrogen were less when the infants were fed formula plus cereal (97 _+ 1% to 90 +_ 4%; p -- 0.0048) than when they were fed formula alone (94 + 3% to 79 +_ 7%; p -- 0.0009). The rice feedings precipitated no symptoms of intolerance (e.g., diarrhea, bloating, or constipation) in any of the infants. The net absorption of the cereal was 88 • 9%. Stool excretion (dry weight) increased more than fourfold when the diet of the infants was changed from formula alone to formula.plus cereal (Table II). The increase in stool weight after feeding formula plus cereal was accounted for by an increase in both the fiber content and bacterial mass
41
nitrogen Measurement
Formula
Dry weight (g/kg/day) 0.7 _+ 0.3 Fiber (g/kg/day)* 0.005 ___0.009 Bacterial mass 0.27 + 0.13 (gm/kg/day) t Total nitrogen 24.5 _+ 14.8 (mg/kg/day) Nitrogen from fiber Not detectable (mg/kg/day) * Nitrogen from bacteria 6.7 _+ 4.1 (mg/kg/day)? *Fractions 1 and 2. ~'Fraction3.
Formula plus cereal
p
3.3 + 1.8 0.004 1.161 _+ 1.292 0.04 1.80 ___0.91 0.002 102.0+ 32.5 0.0005 3.9 _+ 3.7 51.8 _+ 16.5 0.0014
of the stool (Table II). Fiber accounted for 0.6 • 0.9% of stool dry weight when the infants were fed formula versus 28.5 • 27.4% (p = 0.026) when they were fed formula plus cereal (Table II). Similarly, bacterial mass accounted for 41.4 + 7.9% versus 55.7 + 8.5% (2 = 0.003) of dry stool weight when the infants were fed formula alone versus formula plus cereal (Table II). Total stool carbon increased from 0.32 to 1.55 gm/kg per day when the diet of the infants was changed from formula alone to formula plus cereal (p = 0.0047). The percentage of total stool carbon derived from cereal when the infants received formula plus cereal was 55% ___ 16%. Thus the increase in stool carbon during the feeding regimen of formula plus cereal was primarily from cereal carbon. When the infants were fed cereal, stool nitrogen excretion was greater than when they were fed formula alone; approximately one third of stool nitrogen excreted during the formula feeding was bacterial (Table II). During the feeding of formula plus cereal, the contribution of bacterial nitrogen to total stool nitrogen increased twofold. The stool fiber fraction contributed minimally to stool nitrogen during both the formula-alone and the formula-plus-cereal feeding periods (Table II). All the infants produced breath hydrogen on at least one of the study days. After the formula feeding (study day 1), three of eight infants produced more than 10 ppm breath hydrogen. After the first feeding of cereal (study day 7), one of eight infants, and after the final day of cereal feeding (study day 13) two of eight infants, produced more than 10 ppm breath hydrogen. No differences were noted in mean peak breath hydrogen produced on study days 1, 7, or 13 (10.9 + 12.6, 4.3 + 3.9, and 6.6 • 5.5 ppm, respectively). Similarly, no differences were noted when the breath hydrogen results were expressed as area under the concen-
42
Shulman, Boutton, and Klein
tration versus time curve (2209 _+ 1567, 1256 + 951, and 2439 _+ 2136 ppm/min, respectively). DISCUSSION As we began our study, we did not know whether the addition of cereal to formula would alter the infants' overall formula intakes. Our data demonstrate that the infants consumed more formula when it was given alone rather than with added cereal. Although the addition of cereal increased the energy density of the mix by 49%, the energy intake of the infants increased by only 18% when they consumed formula plus cereal in comparison with formula alone. T h e results of our study confirm our previous observations that cereal is better absorbed than might be anticipated considering the lack of pancreatic amylase in young infants. 2 The wide range in the ability of infants to digest and absorb the cereal, however, is noteworthy. Our findings indicate that most young infants tolerate the feeding of cereal, although cases Of cereal-induced failure to thrive have been reported. 7 When cereal is fed to healthy young infants, energy and nitrogen absorption are affected. Despite the ability of young infants to digest and absorb cereal, these effects of feeding have been ignored. In our study, the intake of energy and nitrogen increased after the cereal feeding, although the added energy and nitrogen were not absorbed, resulting in a decrease in their coefficient of absorption. The major portion of the increase in stool carbon after the feeding of formula plus cereal was derived from the cereal. These data demonstrate that absorption of the rice cereal in the small intestine was not complete and that a portion reached the colon. Carbohydrate not digested in the small intestine is fermented by colonic bacteria to produce shortchain fatty acids. 8,9 The fatty acids can be absorbed through the colonic mucosa in t h e systemic circulation. 9 Carbohydrat e (e.g., cereal) that would otherwise be excreted in stool is salvaged through this pathway. The colonic bacteria also use carbohydrate as an energy source to support their growth. 10, 11 Carbon from carbohydrate is incorporated into fatty acids produced and into the bacteria themselves. Nitrogen would be available to the bacteria from the nitrogen in the cereal and formula, from sloughed small intestinal mucosal cells and mucus, and in the form of ammonia and urea. 1~ 11 We propose, on the basis of the following findings, that carbon from the cereal was used by colonic bacteria. First, approximately half of the total carbon found in stool was derived from the cereal. Second, there was an almost sevenfold increase in fecal bacterial mass after formula and cereal were fed, in comparison with the amount after formula feeding alone. Previous studies in adults have demon-
The Journal of Pediatrics January 1991
strated an increase in fecal bacterial mass when complex carbohydrates are added to the diet. 1~ 12 Total stool nitrogen excretion increased after cereal was fed; the contribution of bacterial nitrogen to total stool nitrogen also increased. Half of the increase in stool nitrogen after cereal feedings was accounted for by the increase in fecal bacterial nitrogen. These results give furthe r support to the possibility that the malabsorbed carbohydrate in the cereal acted as a substrate for increased bacterial proliferation.!~ It is likely that the fraction of increased stool nitrogen not accounted for by bacterial PrOliferation was accounted for by nitrogen excretion in the soluble fraction of stool. Although the precise amount of cereal that reached the colon cannot be determined with certainty, a rough estimate can be made of the amount fermented to short-chain fatty acids and absorbed versus the amount excreted in stool. Scheppach et ai.1~ reported that ~ 4 gm of carbohydrate is required to produce 1 gm dry weight of bacteria. We found an average 1.53 gm/kg per day increase in the bacterial mass after feeding cereal. This increase in the bacterial mass would require that ~ 6 g m / k g per day of carbohydrate reach the colon. The infants ate ~ 2 0 gm/kg per day (50 kcal/kg per day) of cereal. Therefore 30% of the ingested cereal reached the colon. Because only 12% of the ingested cereal was recovered in stool, we estimate that ~ 18% was converted to fatty acids and absorbed. The fermentation of cardohydrate by colonic bacteria is often accompanied by the production of hydrogen gas, part of which is absorbed and excreted in breath. 8, 9 Although we anticipated that more infants would show a significant rise (>10 ppm) in breath hydrogen after the cereal feeding in comparison with formula alone, this was not the case. In addition, neither peak breath hydrogen nor the area under the breath hydrogen curve of concentration versus time was different after the feeding of formula in comparison with formula plus cereal. We showed previously that peak breath hydrogen production in. young infants is greater after feeding of corn cereal than after feeding of glucose. 2 Breath hydrogen studies in adults, however, indicate that the digestibility of corn is less than that of rice. 13 Although some subjects have a colonic fecal flora that does not produce hydrogen, all the infants we studied produced hydrogen on at least one of the test days. Recent preliminary data may explain the apparent contradiction between the results of the cereal absorption determinations and those of the breath hydrogen tests. 14 Murgatroyd et al. 14 demonstrated that breath hydrogen excretion varies with hydrogen production rates and that different types of carbohydrates result in different hydrogen production rates, even though similar quantities of carbo-
Volume l 18 Humber i
Nutrient absorption and nitrogen loss with cereal
hydrate reach the colon. These data challenge the validity of the quantitative interpretation of breath hydrogen results. ~4 In summary, our data demonstrate that cereal added to the young infant's diet, in an amount comparable to that used in our study, resulted in greater energy and nitrogen intakes but also in a decreased coefficient of absorption; thus no change occurred in apparent energy and nitrogen retention. The increase in stool energy and nitrogen after the feeding of cereal was the result, in part, of a cereal-induced stimulation of colonic bacterial flora growth. The absolute amount of formula intake was reduced. These data suggest that although cereal is generally well absorbed, it should not be considered a supplement to the energy and nitrogen economy of the young infant.
4.
5.
6. 7.
8. 9.
10. We thank C. lmo, S. Sekely, S. Vaidya, and M. Linzel for technical assistance; D. Ferguson, RN, and K. Evans, RN, for assistance in clinical care of the infants; C. H. Lifschitz, MD, and W. C. Heird, MD, for manuscript review; and J. D. Eastman and E. R. Klein for editorial review.
11. 12.
REFERENCES
1. Hadorn B, Zoppi G, Shmerling DH, Prader A, McIntyre I, Anderson CM. Quantitative assessment of exocrine pancreatic function in infants and children. J PEDIATR 1968;73:39-XX. 2. Shulman R J, Wong WW, Irving CS, Nichols BL, Klein PD. Utilization of dietary cereal by young infants. J PEDIATR 1983;103:23-89 3. Shulman R J, Kerzner B, Sloan HR, et al. Absorption and ox-
13.
14.
43
idation of glucose polymers of different lengths in young infants. Pediatr Res 1986;20:740-3. Orenstein SR, Magill HL, Brooks P. Thickening of infant feedings for therapy of gastroesophageal reflux. J PEDIATR 1987;110:181-6. Boutton TW, Bollich CN, Webb BD, Sekely SL, Nichols BL, Klein PD. t3C-labeled rice produced for dietary studies [Abstract]. Am J Clin Nutr 1987;45:844. Stephen AM, Cummings JH. The microbial contribution to human faecal mass. J Med Microbiol 1980;13:45-56. Lilibridge CB, Townes PL. Physiologic deficiency of pancreatic amylase in infancy: a factor in iatrogenic diarrhea. J PEDIATR 1973;82:279-82. Bond JH, Levitt MD. Fate of soluble carbohydrate in the colon of rats and man. J Clin Invest 1976;57:1158-64. Bond JH, Currier BE, Buchwald H, Levitt MD. Colonic conservation of malabsorbed carbohydrate9 Gastroenterology 1980;78:444-7. Schcppach W, Fabian C, Ahrens F, Spengler M, Kasper H. Effect of starch malabsorption on colonic function and metabolism in humans. Gastroenterology 1988;95:1549-55. Stepphen AM. Dietary fibre and colonic nitrogen metabolism. Scand J Gastroenterol 1987;22(supp 129):110-5. Weber FL, Banwell JG, Fresard KM, Cummings JH. Nitrogen in fecal bacterial, fiber, and soluble fractions of patients with cirrhosis: effects of lactulose and lactulose plus neomycin. J Lab Clin Med 1987;110:259-63. Levitt MD, Hirsh P, Fetzer CA, Sheahan FM, Levine AS. H2 excretion after ingestion of complex carbohydrates. Gastroenterology 1987;92:383-9. Murgatroyd PR, Gibson GR, Cummings JH. Quantitative measurement of hydrogen and methane from fermentation using a whole body calorimeter [Abstract]. Gastroenterology 1990;98:A164.
FELLOWSHIPS Available fellowships in pediatric subspecialties and those for general academic pediatric training are listed once a year, in May, in THE JOURNAL OF PEDIATRICS. Each October, forms for listing such fellowships are sent to the Chairman of the Department of Pediatrics at most major hospitals in the United States and Canada. Should you desire to list fellowships, a separate application must be made each year f o r each position. All applications must be returned to M o s b y - Y e a r Book, Inc., by February 15 of the listing year to ensure publication. Additional forms will be supplied on request from the Journal Editing Department, M o s b y - Y e a r Book, Inc., 11830 Westline Industrial Drive, St. Louis, M O 63146-3318/314-453-4317.
Study objective: To determine the c a p a c i t y of infants to digest and absorb rice cereal and to determine the effect of cereal f e e d i n g on total e n e r g y and nitrogen absorption. Setting: Subjects' residences and the Texas Children's Hospital Clinical Research Center, Houston. Patients: Eight healthy 1-month-old bottle-fed infants. Interventions: Infants were fed their usual formula for 3 days. For the subsequent 6 days, they r e c e i v e d 4 gm of rice cereal, l a b e l e d with c a r b o n 13, per 30 ml of the formula. Measurements and main results: Fecal b a l a n c e studies were performed for a 72-hour period while the infants received only formula and a g a i n during the last 3 days of c e r e a l feeding. Breath samples for h y d r o g e n measurement were coll e c t e d before and after the cereal feeding. Nutrient intake was measured and stools were analyzed for 13C a b u n d a n c e , energy, nitrogen, fiber content, and bacterial mass. Cereal absorption was 88 + 9% (mean • SD). Despite a significant increase in energy and nitrogen intake from cereal feeding, the coefficient of absorption fell (energy: 97% to 90%, p -- 0.048; nitrogen: 94% to 74%, p = 0.009). Fecal dry weight increased after the c e r e a l f e e d i n g (p = 0.004), primarily as a result of a sevenfold increase in f e c a l bacterial mass ( p - 0.002). Fecal nitrogen increased primarily because of incorporation of nitrogen into bacteria. No differences were d e t e c t e d in breath h y d r o g e n as a c o n s e q u e n c e of formula f e e d i n g versus formula-cereal feeding. Conclusions: Although cereal was relatively w e l l - a b s o r b e d in this group of infants and increased their intakes of energy and nitrogen, it did not increase the coefficients of e n e r g y and nitrogen absorption. Cereal f e e d i n g increased fecal bacterial mass and bacterial nitrogen. (J PEDIATR1991;118:39-43)
Supported in part with federal funds from the U.S. Department of Agriculture, Agricultural Research Service, under Cooperative Agreement No. 58-7MN1-6-100. The contents of this publication do not necessarily reflect the views or policies of the U.S. Department of Agriculture, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government. Submitted for publication July 20, 1990; accepted Aug. 20, 1990. Reprint requests: Robert J. Shulman, MD, Children's Nutrition Research Center, 1100 Bates Street, Houston, TX 77030. *Now at Texas A&M University, College Station, Tex. 9/20/24946
Cereal is usually the first solid food given to young infants. Although recommendations have been made that solids such as cereal not be added to an infant's diet until 4 to 6 months of age, cereal is often fed within the first 3 months of life. One of the concerns related to the early introduction of cereal into the diet of young infants is the infants' low level of pancreatic amylase before 6 months of age, 1 which has generally been assumed to lead to significant malabsorption of cereal. 2 Our previous studies suggested that digestion of complex carbohydrates, such as cereal, may be more com-
39
40
Shulman. Boutton, and Klein
plete in the young infant than originally believed.2 Evidence suggests that the small intestinal enzyme glucoamylase is important in the digestion of complex carbohydrates in the young infant before the appearance of pancreatic amylase activity) The young infants who are most likely to receive substantial amounts of dietary cereal before 4 to 6 months of age are those with gastroesophageal reflux. Although the addition of cereal to thicken the formula does not appear to reduce the amount of reflux, as determined by esophageal pH probe monitoring, it does reduce the amount of spitting and vomiting. 4 Infants with gastroesophageal reflux may have poor weight gain related to vomiting. The addition of cereal to the formula increases the caloric density of the feeding. 4 However, the extent to which cereal is digested and absorbed and the effects of cereal on the digestion and absorption of other nutrients are unknown. The objectives of our study were to determine the capacity of young infants to digest and absorb rice cereal and the effect of rice cereal feeding on energy and nitrogen absorption. Additionally, we sought to ascertain the effect of cereal feeding on fecal bacterial mass. METHODS
Subjects. Eight infants were selected for the study after informed consent was obtained from their parents. The study was approved by the human investigations committees of Baylor College of Medicine and Texas Children's Hospital. The infants were between 16 and 40 days of age at the start of the study. All infants had normal perinatal histories and were healthy. Design. Breath samples were collected for hydrogen determinations at the start of the study and at 30-minute intervals for 3 hours after the ingestion of 4 ounces (120 ml) of the infant's usual formula, as previously described. 2, 3 The infants were then sent home. During this time, the infants were fed their usual formula on demand, and all feces were collected for 72 hours (days 2 to 5). Preweighed cloths were provided to account for possible losses because of spitting up. The diaper containing the stool was marked with the date and time of collection and was placed in a plastic bag and stored in the freezer. A research nurse visited the homes frequently to ensure that the protocol was followed and to pick up the diapers containing the stools. On day 7, breath samples were again collected for hydrogen determinations before and after the ingestion of 4 ounces (120 ml) of formula containing 4 gm ( ~ 1 tablespoon) of rice cereal (Gerber, Fremont, Mich.) per 30 ml. After the breath test, the infants continued to receive their usual formula containing the rice cereal. On day 10, the infants were readmitted to the clinical re-
The Journal of Pediatrics January 1991
search center for the second phase of the study. Readmission was necessary because the formula and cereal mixture had to be consumed soon after it was prepared. The commercial rice cereal in the formula was replaced by enriched rice cereal labeled with carbon 13 (see discussion of diet, below). Four grams of 13C-enriched rice was added to each 30 ml of formula. All stools were collected for 3 days. On day 13, breath samples were collected after the formula plus cereal mixture was fed, and the infants were discharged from the clinical research center. Diet. On days 7 through 9, the infants were fed rice cereal at home. Preweighed packets of cereal (32 gm) to be added to 8-ounce (240 ml) bottles of formula were provided. Details of the 13C-enriched rice cereal fed to the infants during the 3-day stool collection in the clinical research center have been described previously.5 The 13C-labeled rice was grown in a 13CO2-enriched environment and then processed in a manner identical to that of rice in commercially available Gerber rice cereal. 2 Calculation of nutrient intake and output. Intake was calculated as the difference in weight between the bottles containing formula, or formula and cereal, before and after feeding. Preliminary studies indicated that the proportion of formula to cereal remaining in the bottle after a feeding was similar to that when the formula and cereal were first prepared and mixed in the bottle. Carmine red was used as the colorant to mark the beginning and end of the 72-hour stool collection. A l l stools collected during the 72-hour period were pooled and analyzed. Net absorption was defined as total intake during the 72 hours minus total output over the same period. Coefficient of absorption was defined as follows: ([Total i n t a k e - T o t a l output]/[Total intake]) • 100. Formula and stool analyses. The energy and nitrogen contents of the formula and the formula-cereal mixture were determined by lyophilizing an aliquot of the mixture and performing bomb calorimetry (Parr Instrument Co, Moline, Ill.) and a micro-Kjeldahl procedure (Tecator Inc, Herndon, Va.), respectively. The frozen stool samples were mixed in deionized water, brought to a known volume, and homogenized. Aliquots were weighed, frozen in liquid nitrogen, and lyophilized. The aliquots were weighed again to obtain the dry weight, and then analyzed for energy and nitrogen as described above. Fecal bacterial mass and fiber content were determined from other aliquots by means of the filtration method of Stephen and Cummings. 6 Stools were divided into three fractions; fractions 1 and 2 were primarily fiber, and fraction 3 was an almost pure bacterial isolate.6 The concentration of nitrogen was also measured in the three fractions to determine the amount of fecal nitrogen produced by bacte -
Volume 118 Number 1
Nutrient absorption and nitrogen loss with cereal
Table I. Energy and nitrogen intake and net retention
Table II. Stool dry weight, fiber, bacterial mass, and
Measurement
Energy intake (kcal/kg/day) Energy absorption (kcal/kg/day) Nitrogen intake (mg/kg/day) Nitrogen absorption (mg/kg/day)
Formula
128 125 405 381
+ 25 _+ 24 .+ 84 _+ 75
Formula plus cereal
151 _+ 26* 139 _+ 23 481 _+ 84? 379 ___78
*p = 0.027. "l'p= 0.024.
ria. We determined the digestion of the rice cereal by cambusting other fecal aliquots to obtain the percentage of carbon and the 13C isotope abundance, as we have previously described. 2 Breath hydrogen analysis. Samples for breath hydrogen analysis were stored in gas-tight syringes and analyzed by gas-solid chromatography as previously described, except that results were not corrected to an alveolar reference of 5% CO2 .2' 3 Results were expressed as peak breath hydrogen excretion in parts per million (ppm) of room air, as well as the area under the curve (concentration versus time) calculated by the trapezoidal rule. Statistics. Results are expressed as mean + SD. The Student t test for paired data was used for comparisons. Two-way analysis of variance was used to compare breath hydrogen results. RESULTS The energy density of the formula was 0.67 kcal/gm and the nitrogen concentration was 2.13 mg/gm. For formula plus cereal, these values were 1.0 kcal/gm and 3.2 mg/gm, respectively. The infants' energy intakes were greater while they were receiving formula plus cereal; similarly, nitrogen intake was greater when the infants were fed formula plus cereal (Table I). Despite greater intakes of energy and nitrogen when the infants were fed formula plus cereal, there was no change in net energy or nitrogen absorption (Table I). Thus the coefficients of absorption for energy and nitrogen were less when the infants were fed formula plus cereal (97 _+ 1% to 90 +_ 4%; p -- 0.0048) than when they were fed formula alone (94 + 3% to 79 +_ 7%; p -- 0.0009). The rice feedings precipitated no symptoms of intolerance (e.g., diarrhea, bloating, or constipation) in any of the infants. The net absorption of the cereal was 88 • 9%. Stool excretion (dry weight) increased more than fourfold when the diet of the infants was changed from formula alone to formula.plus cereal (Table II). The increase in stool weight after feeding formula plus cereal was accounted for by an increase in both the fiber content and bacterial mass
41
nitrogen Measurement
Formula
Dry weight (g/kg/day) 0.7 _+ 0.3 Fiber (g/kg/day)* 0.005 ___0.009 Bacterial mass 0.27 + 0.13 (gm/kg/day) t Total nitrogen 24.5 _+ 14.8 (mg/kg/day) Nitrogen from fiber Not detectable (mg/kg/day) * Nitrogen from bacteria 6.7 _+ 4.1 (mg/kg/day)? *Fractions 1 and 2. ~'Fraction3.
Formula plus cereal
p
3.3 + 1.8 0.004 1.161 _+ 1.292 0.04 1.80 ___0.91 0.002 102.0+ 32.5 0.0005 3.9 _+ 3.7 51.8 _+ 16.5 0.0014
of the stool (Table II). Fiber accounted for 0.6 • 0.9% of stool dry weight when the infants were fed formula versus 28.5 • 27.4% (p = 0.026) when they were fed formula plus cereal (Table II). Similarly, bacterial mass accounted for 41.4 + 7.9% versus 55.7 + 8.5% (2 = 0.003) of dry stool weight when the infants were fed formula alone versus formula plus cereal (Table II). Total stool carbon increased from 0.32 to 1.55 gm/kg per day when the diet of the infants was changed from formula alone to formula plus cereal (p = 0.0047). The percentage of total stool carbon derived from cereal when the infants received formula plus cereal was 55% ___ 16%. Thus the increase in stool carbon during the feeding regimen of formula plus cereal was primarily from cereal carbon. When the infants were fed cereal, stool nitrogen excretion was greater than when they were fed formula alone; approximately one third of stool nitrogen excreted during the formula feeding was bacterial (Table II). During the feeding of formula plus cereal, the contribution of bacterial nitrogen to total stool nitrogen increased twofold. The stool fiber fraction contributed minimally to stool nitrogen during both the formula-alone and the formula-plus-cereal feeding periods (Table II). All the infants produced breath hydrogen on at least one of the study days. After the formula feeding (study day 1), three of eight infants produced more than 10 ppm breath hydrogen. After the first feeding of cereal (study day 7), one of eight infants, and after the final day of cereal feeding (study day 13) two of eight infants, produced more than 10 ppm breath hydrogen. No differences were noted in mean peak breath hydrogen produced on study days 1, 7, or 13 (10.9 + 12.6, 4.3 + 3.9, and 6.6 • 5.5 ppm, respectively). Similarly, no differences were noted when the breath hydrogen results were expressed as area under the concen-
42
Shulman, Boutton, and Klein
tration versus time curve (2209 _+ 1567, 1256 + 951, and 2439 _+ 2136 ppm/min, respectively). DISCUSSION As we began our study, we did not know whether the addition of cereal to formula would alter the infants' overall formula intakes. Our data demonstrate that the infants consumed more formula when it was given alone rather than with added cereal. Although the addition of cereal increased the energy density of the mix by 49%, the energy intake of the infants increased by only 18% when they consumed formula plus cereal in comparison with formula alone. T h e results of our study confirm our previous observations that cereal is better absorbed than might be anticipated considering the lack of pancreatic amylase in young infants. 2 The wide range in the ability of infants to digest and absorb the cereal, however, is noteworthy. Our findings indicate that most young infants tolerate the feeding of cereal, although cases Of cereal-induced failure to thrive have been reported. 7 When cereal is fed to healthy young infants, energy and nitrogen absorption are affected. Despite the ability of young infants to digest and absorb cereal, these effects of feeding have been ignored. In our study, the intake of energy and nitrogen increased after the cereal feeding, although the added energy and nitrogen were not absorbed, resulting in a decrease in their coefficient of absorption. The major portion of the increase in stool carbon after the feeding of formula plus cereal was derived from the cereal. These data demonstrate that absorption of the rice cereal in the small intestine was not complete and that a portion reached the colon. Carbohydrate not digested in the small intestine is fermented by colonic bacteria to produce shortchain fatty acids. 8,9 The fatty acids can be absorbed through the colonic mucosa in t h e systemic circulation. 9 Carbohydrat e (e.g., cereal) that would otherwise be excreted in stool is salvaged through this pathway. The colonic bacteria also use carbohydrate as an energy source to support their growth. 10, 11 Carbon from carbohydrate is incorporated into fatty acids produced and into the bacteria themselves. Nitrogen would be available to the bacteria from the nitrogen in the cereal and formula, from sloughed small intestinal mucosal cells and mucus, and in the form of ammonia and urea. 1~ 11 We propose, on the basis of the following findings, that carbon from the cereal was used by colonic bacteria. First, approximately half of the total carbon found in stool was derived from the cereal. Second, there was an almost sevenfold increase in fecal bacterial mass after formula and cereal were fed, in comparison with the amount after formula feeding alone. Previous studies in adults have demon-
The Journal of Pediatrics January 1991
strated an increase in fecal bacterial mass when complex carbohydrates are added to the diet. 1~ 12 Total stool nitrogen excretion increased after cereal was fed; the contribution of bacterial nitrogen to total stool nitrogen also increased. Half of the increase in stool nitrogen after cereal feedings was accounted for by the increase in fecal bacterial nitrogen. These results give furthe r support to the possibility that the malabsorbed carbohydrate in the cereal acted as a substrate for increased bacterial proliferation.!~ It is likely that the fraction of increased stool nitrogen not accounted for by bacterial PrOliferation was accounted for by nitrogen excretion in the soluble fraction of stool. Although the precise amount of cereal that reached the colon cannot be determined with certainty, a rough estimate can be made of the amount fermented to short-chain fatty acids and absorbed versus the amount excreted in stool. Scheppach et ai.1~ reported that ~ 4 gm of carbohydrate is required to produce 1 gm dry weight of bacteria. We found an average 1.53 gm/kg per day increase in the bacterial mass after feeding cereal. This increase in the bacterial mass would require that ~ 6 g m / k g per day of carbohydrate reach the colon. The infants ate ~ 2 0 gm/kg per day (50 kcal/kg per day) of cereal. Therefore 30% of the ingested cereal reached the colon. Because only 12% of the ingested cereal was recovered in stool, we estimate that ~ 18% was converted to fatty acids and absorbed. The fermentation of cardohydrate by colonic bacteria is often accompanied by the production of hydrogen gas, part of which is absorbed and excreted in breath. 8, 9 Although we anticipated that more infants would show a significant rise (>10 ppm) in breath hydrogen after the cereal feeding in comparison with formula alone, this was not the case. In addition, neither peak breath hydrogen nor the area under the breath hydrogen curve of concentration versus time was different after the feeding of formula in comparison with formula plus cereal. We showed previously that peak breath hydrogen production in. young infants is greater after feeding of corn cereal than after feeding of glucose. 2 Breath hydrogen studies in adults, however, indicate that the digestibility of corn is less than that of rice. 13 Although some subjects have a colonic fecal flora that does not produce hydrogen, all the infants we studied produced hydrogen on at least one of the test days. Recent preliminary data may explain the apparent contradiction between the results of the cereal absorption determinations and those of the breath hydrogen tests. 14 Murgatroyd et al. 14 demonstrated that breath hydrogen excretion varies with hydrogen production rates and that different types of carbohydrates result in different hydrogen production rates, even though similar quantities of carbo-
Volume l 18 Humber i
Nutrient absorption and nitrogen loss with cereal
hydrate reach the colon. These data challenge the validity of the quantitative interpretation of breath hydrogen results. ~4 In summary, our data demonstrate that cereal added to the young infant's diet, in an amount comparable to that used in our study, resulted in greater energy and nitrogen intakes but also in a decreased coefficient of absorption; thus no change occurred in apparent energy and nitrogen retention. The increase in stool energy and nitrogen after the feeding of cereal was the result, in part, of a cereal-induced stimulation of colonic bacterial flora growth. The absolute amount of formula intake was reduced. These data suggest that although cereal is generally well absorbed, it should not be considered a supplement to the energy and nitrogen economy of the young infant.
4.
5.
6. 7.
8. 9.
10. We thank C. lmo, S. Sekely, S. Vaidya, and M. Linzel for technical assistance; D. Ferguson, RN, and K. Evans, RN, for assistance in clinical care of the infants; C. H. Lifschitz, MD, and W. C. Heird, MD, for manuscript review; and J. D. Eastman and E. R. Klein for editorial review.
11. 12.
REFERENCES
1. Hadorn B, Zoppi G, Shmerling DH, Prader A, McIntyre I, Anderson CM. Quantitative assessment of exocrine pancreatic function in infants and children. J PEDIATR 1968;73:39-XX. 2. Shulman R J, Wong WW, Irving CS, Nichols BL, Klein PD. Utilization of dietary cereal by young infants. J PEDIATR 1983;103:23-89 3. Shulman R J, Kerzner B, Sloan HR, et al. Absorption and ox-
13.
14.
43
idation of glucose polymers of different lengths in young infants. Pediatr Res 1986;20:740-3. Orenstein SR, Magill HL, Brooks P. Thickening of infant feedings for therapy of gastroesophageal reflux. J PEDIATR 1987;110:181-6. Boutton TW, Bollich CN, Webb BD, Sekely SL, Nichols BL, Klein PD. t3C-labeled rice produced for dietary studies [Abstract]. Am J Clin Nutr 1987;45:844. Stephen AM, Cummings JH. The microbial contribution to human faecal mass. J Med Microbiol 1980;13:45-56. Lilibridge CB, Townes PL. Physiologic deficiency of pancreatic amylase in infancy: a factor in iatrogenic diarrhea. J PEDIATR 1973;82:279-82. Bond JH, Levitt MD. Fate of soluble carbohydrate in the colon of rats and man. J Clin Invest 1976;57:1158-64. Bond JH, Currier BE, Buchwald H, Levitt MD. Colonic conservation of malabsorbed carbohydrate9 Gastroenterology 1980;78:444-7. Schcppach W, Fabian C, Ahrens F, Spengler M, Kasper H. Effect of starch malabsorption on colonic function and metabolism in humans. Gastroenterology 1988;95:1549-55. Stepphen AM. Dietary fibre and colonic nitrogen metabolism. Scand J Gastroenterol 1987;22(supp 129):110-5. Weber FL, Banwell JG, Fresard KM, Cummings JH. Nitrogen in fecal bacterial, fiber, and soluble fractions of patients with cirrhosis: effects of lactulose and lactulose plus neomycin. J Lab Clin Med 1987;110:259-63. Levitt MD, Hirsh P, Fetzer CA, Sheahan FM, Levine AS. H2 excretion after ingestion of complex carbohydrates. Gastroenterology 1987;92:383-9. Murgatroyd PR, Gibson GR, Cummings JH. Quantitative measurement of hydrogen and methane from fermentation using a whole body calorimeter [Abstract]. Gastroenterology 1990;98:A164.
FELLOWSHIPS Available fellowships in pediatric subspecialties and those for general academic pediatric training are listed once a year, in May, in THE JOURNAL OF PEDIATRICS. Each October, forms for listing such fellowships are sent to the Chairman of the Department of Pediatrics at most major hospitals in the United States and Canada. Should you desire to list fellowships, a separate application must be made each year f o r each position. All applications must be returned to M o s b y - Y e a r Book, Inc., by February 15 of the listing year to ensure publication. Additional forms will be supplied on request from the Journal Editing Department, M o s b y - Y e a r Book, Inc., 11830 Westline Industrial Drive, St. Louis, M O 63146-3318/314-453-4317.