Close

Search

Abstract

Providing small enteral feedings for parenterally fed preterm infants during the first few weeks of life improves feeding tolerance. However, it is not known whether these feedings achieve this result via stimulation of gut growth and/or maturation of function. The minimal level needed to attain these responses is also critical to identify, because neonatologists often limit feeding volumes to minimize the risk of necrotizing enterocolitis. Thus, we determined the dose–response relationships between enteral feeding volume and gastrointestinal growth and small intestine motor function. Newborn canine pups (n = 51) received 0, 2.5, 5.0, 7.5, 10, 30 or 100% of their daily fluid intake enterally with the remainder given parenterally for 4–5 d. Motor activity was recorded, blood drawn for determination of gastrin and motilin, and intestinal tissue harvested for determination of DNA and protein content and morphology. Feeding volumes that provided 30% or more of daily fluid intake significantly increased small intestinal mucosal growth above that of unfed pups, but feeding volumes that provided as little as 10% of daily fluid intake significantly induced maturation of motor patterns beyond that of unfed pups. Plasma concentrations of gastrin and motilin did not differ among groups. We conclude that small enteral feedings typically used for minimal enteral feeding strategies improve feeding tolerance by triggering maturation of motor function but not gut growth in neonatal dogs. Small feeding volumes trigger this maturation as well as large volumes.

Nutritional support for preterm neonates is limited to parenteral nutrition during the first 1–3 wk of life when these infants are clinically unstable and require ventilator support. An additional concern that limits enteral feeding is the increased risk of necrotizing enterocolitis (NEC)3. However, there are concerns that withholding enteral feedings from neonates may limit intestinal mucosal growth and delay the maturation of hormonal release. To provide a minimal stimulus to the gut while minimizing the risk of NEC, infants are often given small enteral feedings (i.e., volumes comprising 1–10% of daily fluid intake) during the first 2 wk of life. Studies show that infants given these small enteral feedings experience better feeding tolerance when full enteral feeding volumes are later introduced during the second or third postnatal week (1–3). Because these small feeding volumes are believed to induce small intestinal growth, they are called trophic feeds, priming feeds, or minimal enteral nutrition. However, recent studies in neonatal piglets (4) suggest that enteral feeding volumes of at least 20–40% are needed to maintain or stimulate intestinal growth, suggesting that other physiological responses may mediate the improved feeding tolerance. If additional animal studies confirm that larger feeding volumes are needed to trigger mucosal growth, additional clinical trials would be needed to assess the risk of these higher volumes, because all randomized trials to date have compared small feeding volumes with no feedings (1–3).

An alternative explanation for the improvement in feeding tolerance in infants given small feedings is that these small feedings trigger maturation of intestinal function by stimulating the release of hormones, neurotransmitters and other gastrointestinal peptides. For example, infants who are given small enteral feedings have more mature small intestinal motor patterns and higher plasma gastrin and motilin concentrations than do infants who have been given no feedings (1,5,6). These previous studies suggest that small enteral feedings can induce maturation of gastrointestinal motor function, yet the minimal volume of enteral feedings necessary to achieve this response has not been established in either animal or human studies. Therefore, the objective of this study was to determine the minimal feeding volume necessary to induce both maturation of motor function and increase mucosal growth in neonatal dogs. We performed this study in animals to harvest tissue to assess growth, because this would not be possible in human infants. In this study, we used neonatal dogs, rather than pigs, because unlike pigs that have mature motor function at birth (7), small intestinal motor function and hormone release in neonatal dogs are similar to that of preterm human neonates (8).

MATERIALS AND METHODS

Animals and study design.

We studied 51 mixed breed canine pups from 20 litters using a protocol that was approved by the Baylor College of Medicine Animal Care and Use Committee. Pregnant dogs were housed in separate cages, given supplementary dietary intake and provided whelping boxes. Newborn pups were permitted to nurse for 48–72 h before entering this study. A central venous catheter was placed into the left external jugular vein in 44 of the pups. The central line was threaded into a protective flexible spring, which, in turn, was connected to a toggle to permit the pups to move freely when they awakened. Pups were then housed postoperatively in individual incubators at 28–30°C and given a continuous intravenous infusion (200 mL/[(kg · d]) of a nutritional solution. 4

After 24 h, pups were randomly assigned to 1 of 6 treatment groups. Seven pups were given 200 mL/(kg · d) of the nutrient solution alone, while the other 37 pups were given 2.5, 5.0, 7.5, 10 or 30% of their daily fluid intake enterally and the volume of the nutrient solution accordingly reduced (Table 1). In addition to these 44 pups, another 7 pups were removed from their mothers, maintained in the infant incubators, and fed 200 mL/(kg · d) of the study milk described below.

TABLE 1

Parenteral and enteral intake volumes of newborn dogs fed varying volumes of milk for 4 or 5 d 1

Group, fluid Parenteral Enteral Total
% enteral mL/(kg · d)
0 200 0 200
2.5 195 5 200
5.0 190 10 200
7.5 185 15 200
10.0 180 20 200
30 140 60 200
100 0 200 200
Group, fluid Parenteral Enteral Total
% enteral mL/(kg · d)
0 200 0 200
2.5 195 5 200
5.0 190 10 200
7.5 185 15 200
10.0 180 20 200
30 140 60 200
100 0 200 200

1

Groups are shown as the percentage of daily fluid intake given enterally.

TABLE 1

Parenteral and enteral intake volumes of newborn dogs fed varying volumes of milk for 4 or 5 d 1

Group, fluid Parenteral Enteral Total
% enteral mL/(kg · d)
0 200 0 200
2.5 195 5 200
5.0 190 10 200
7.5 185 15 200
10.0 180 20 200
30 140 60 200
100 0 200 200
Group, fluid Parenteral Enteral Total
% enteral mL/(kg · d)
0 200 0 200
2.5 195 5 200
5.0 190 10 200
7.5 185 15 200
10.0 180 20 200
30 140 60 200
100 0 200 200

1

Groups are shown as the percentage of daily fluid intake given enterally.

Enteral feedings for all pups consisted of pooled bitch milk mixed 1:1 with pooled human milk. This mixture was used because insufficient quantities of dog milk were expressed from mother dogs and other donor dogs for use in this study. The decision was made to blend the canine milk with human milk rather than artificial formula in an attempt to the pups more naturally biologic milk. Enteral feedings were given by orogastric tube in five equal volumes at 0200, 0800, 1200, 1600 and 2000 h daily. Fluid intake was changed daily according to each pup's weight to maintain an intake of 200 mL/(kg · d). Because motor patterns in neonatal canines differ during fasting and feeding (8), and the presence of the migrating motor complex (MMC) can only be determined during fasting, motor activity was recorded for 2 h on the study d 4 or 5 after pups had fasted for 6 h. After this assessment was completed, the pup was killed with an intravenous dose of sodium pentobarbital (50 mg/kg), a 2-mL blood sample was obtained for the determination of plasma concentrations of motilin and gastrin, and the abdominal contents were harvested.

Central venous catheter placement.

General anesthesia was induced and maintained with isoflourane via facemask. After the neck and back were shaved, scrubbed and draped, a 2-cm incision was made in the left anterior lateral neck. The external jugular vein was isolated and transected; a 20-gauge silastic catheter was inserted and the proximal tip positioned in the superior vena cava. The distal end of the catheter was tunneled subcutaneously to exit posteriorly between the scapulae and was attached to a tether and infusion system. Postoperatively, pups were given daily antibiotic (enrofloxin, 0.5 mg/kg intramuscularly, as Baytril; Bayer Corp., Shawnee Mission, KS).

Motor activity.

Motor activity was recorded using a low compliance continuous perfusion manometry system designed for use in neonates. The manometric probe was constructed of polyvinyl extrusion tubing and contained six manometric ports spaced 2.5 cm apart. A manometric tube was inserted through the mouth and positioned in the duodenum under fluoroscopic guidance without the use of anesthesia. Two or more ports were placed in the duodenum. The tube was secured with tape and the animal placed prone on a pillow. Motor activity was recorded for 2 h.

A technician who was not aware of the treatment groups analyzed manometry tracings in 30-min segments by hand. While the pups were fasting, the presence or absence of three types of motor activity was noted. As previously defined for neonatal dogs (8), phase 1 was motor quiescence, phase 2 was irregular activity or nonmigrating phasic activity, and phase 3 was characterized by phasic activity that migrated over three ports. The duration of all phases was calculated in min/(h · lead). The presence or absence of migrating motor activity was also noted.

Plasma hormone assays.

All blood samples were placed in cold heparinized tubes containing EDTA, centrifuged at 228 × g for 10 min at 4°C. The resulting plasma was stored stored at −70°C for later determination of motilin and gastrin by radioimmunoassays. All samples were determined in duplicate in a single assay to reduce interassay variation. Motilin concentrations were determined by radioimmunoassay using guinea pig antisera against a bovine serum albumin conjugate of purified motilin isolated from porcine duodenal mucosa (9). The antisera, used at a 1:1200 dilution, has negligible cross-reactivity with gastric inhibitory polypeptide, secretin, glucagon, gastrin, cholecystokinin-pancreozymen, and vasoactive intestinal polypeptide at concentrations up to 10,000 ng/L. Synthetic motilin fragments 9-22 and 13-22 do not appreciably cross-react. The 125I motilin radiolabel was made from synthetic motilin using a chloramine T procedure with HPLC purification. The assay standard was porcine motilin from Peninsula Laboratories (Belmont, CA). The interassay CV was 8.0% at 36.68 pmol/L and 4.6% at 186.74 pmol/L. The intra-assay CV was 14.6% at 237.87 pmol/L.

Gastrin concentrations were measured using a sensitive radioimmunoassay specific for the carboxy-terminal peptide (10). The assay uses a locally developed rabbit antisera specific for the Trp-Met-Asp-Phe-NH2 amino acid sequence, which is common to gastrins and cholecystokinins. The molar cross-reactivities are: 65% for Gastrin-34, 44% for Cholecystokinin-33, 44% for Cholecystokinin-8 II, 32% for Cholecystokinin-8 I, 32% for Cholecystokinin-4, and 21% for Pentagastrin. The I125-Gastrin-17 label was made from synthetic human gastrin using the chloramine-T procedure with HPLC purification. Specimens were analyzed in duplicate. The assay detection limit was 17.63 pmol/L. The intra-assay CV was 10–15% and the interassay CV was 11% at 155.06 pmol/L and 20% at 23.85 pmol/L.

Tissue sampling and analysis.

The abdomen was opened and the small intestine from the pylorus to ileocecal junction was removed, striped of the mesentery, placed in iced saline, flushed, and blotted dry. The intestine was weighed and suspended with a standard weight to measure the length. Then the duodenum was grossly identified by morphologic appearance, removed, weighed and frozen. The next 2-cm segment of intestine was fixed with 10% buffered formaline and later stained with hemotoxylin and eosin for histology. The remaining intestine was then frozen intact. The stomach was removed and split along the greater curvature, emptied, weighed and frozen. Whole liver was removed and weighed. Protein and DNA content of stomach, duodenal and small intestinal tissue homogenates were determined as described previously (11,12), respectively. The crypt depth and villus height were measured in 20 well oriented crypt-villus columns in each sample using a Zeiss Axioplot by an observer blinded to the treatment group.

Statistical analysis.

Somatic and organs weights, length, protein and DNA contents, characteristics of motor patterns, and plasma concentrations of hormones were compared among groups by ANOVA. When the F test was significant, differences among groups were determined by the Fisher's pair-wise comparisons. Trend analysis was used to compare the incidence of MMC among the groups. Differences were considered to be significant when P < 0.05.

RESULTS

Intestinal growth.

All pups gained weight during the study, but there was no difference among the treatment groups. The mean weight gain ranged from 16 to 37 g/d, or ∼16–41% of initial body weight (Table 2). Relative intestinal weight varied among the groups (Fig. 1, upper panel), as did DNA content (Fig. 1, lower panel) and protein content (Fig. 1, middle panel). Relative stomach weight as well as protein and DNA contents did not differ among the groups (Table 2). Relative liver weights (data not shown) and intestinal lengths also did not differ among the groups (Table 2). By histological evaluation, villus height (Table 2) and crypt to villus ratios (data not shown) did not differ among the study groups.

TABLE 2

Body weight, relative stomach weight, intestinal length and villus height of newborn dogs fed varying volumes of milk enterally for 4 or 5 d2 1

Group, fluid n Initial weight Final weight Weight change Stomach Intestine
Relative weight DNA Protein Length Villus height
% enteral g g g/d g/kg mg/kg body cm/10 g μm
0 7 410 ± 114 554 ± 153 37 ± 7 7.4 ± 1.5 53 ± 23 515 ± 77 2.2 ± 0.5 334 ± 167
2.5 7 366 ± 106 467 ± 132 23 ± 11 7.9 ± 1.6 54 ± 13 607 ± 254 2.7 ± 0.5 438 ± 206
5.0 8 386 ± 76 502 ± 96 20 ± 14 6.7 ± 1.1 52 ± 14 615 ± 118 2.6 ± .06 493 ± 118
7.5 10 399 ± 76 550 ± 89 34 ± 39 8.3 ± 1.6 56 ± 16 665 ± 253 2.5 ± 0.3 378 ± 66
10.0 7 361 ± 76 491 ± 108 31 ± 11 7.9 ± 1.6 49 ± 13 663 ± 129 2.5 ± 0.3 522 ± 132
30.0 5 467 ± 51 575 ± 58 29 ± 7 7.8 ± 1.3 43 ± 7 811 ± 105 2.3 ± 0.2 359 ± 134
100 7 378 ± 121 435 ± 137 16 ± 3 9.8 ± 1.6 50 ± 24 762 ± 127 3.1 ± 0.5 475 ± 159
Group, fluid n Initial weight Final weight Weight change Stomach Intestine
Relative weight DNA Protein Length Villus height
% enteral g g g/d g/kg mg/kg body cm/10 g μm
0 7 410 ± 114 554 ± 153 37 ± 7 7.4 ± 1.5 53 ± 23 515 ± 77 2.2 ± 0.5 334 ± 167
2.5 7 366 ± 106 467 ± 132 23 ± 11 7.9 ± 1.6 54 ± 13 607 ± 254 2.7 ± 0.5 438 ± 206
5.0 8 386 ± 76 502 ± 96 20 ± 14 6.7 ± 1.1 52 ± 14 615 ± 118 2.6 ± .06 493 ± 118
7.5 10 399 ± 76 550 ± 89 34 ± 39 8.3 ± 1.6 56 ± 16 665 ± 253 2.5 ± 0.3 378 ± 66
10.0 7 361 ± 76 491 ± 108 31 ± 11 7.9 ± 1.6 49 ± 13 663 ± 129 2.5 ± 0.3 522 ± 132
30.0 5 467 ± 51 575 ± 58 29 ± 7 7.8 ± 1.3 43 ± 7 811 ± 105 2.3 ± 0.2 359 ± 134
100 7 378 ± 121 435 ± 137 16 ± 3 9.8 ± 1.6 50 ± 24 762 ± 127 3.1 ± 0.5 475 ± 159

1

Values are shown as means ± sd. Groups did not differ.

TABLE 2

Body weight, relative stomach weight, intestinal length and villus height of newborn dogs fed varying volumes of milk enterally for 4 or 5 d2 1

Group, fluid n Initial weight Final weight Weight change Stomach Intestine
Relative weight DNA Protein Length Villus height
% enteral g g g/d g/kg mg/kg body cm/10 g μm
0 7 410 ± 114 554 ± 153 37 ± 7 7.4 ± 1.5 53 ± 23 515 ± 77 2.2 ± 0.5 334 ± 167
2.5 7 366 ± 106 467 ± 132 23 ± 11 7.9 ± 1.6 54 ± 13 607 ± 254 2.7 ± 0.5 438 ± 206
5.0 8 386 ± 76 502 ± 96 20 ± 14 6.7 ± 1.1 52 ± 14 615 ± 118 2.6 ± .06 493 ± 118
7.5 10 399 ± 76 550 ± 89 34 ± 39 8.3 ± 1.6 56 ± 16 665 ± 253 2.5 ± 0.3 378 ± 66
10.0 7 361 ± 76 491 ± 108 31 ± 11 7.9 ± 1.6 49 ± 13 663 ± 129 2.5 ± 0.3 522 ± 132
30.0 5 467 ± 51 575 ± 58 29 ± 7 7.8 ± 1.3 43 ± 7 811 ± 105 2.3 ± 0.2 359 ± 134
100 7 378 ± 121 435 ± 137 16 ± 3 9.8 ± 1.6 50 ± 24 762 ± 127 3.1 ± 0.5 475 ± 159
Group, fluid n Initial weight Final weight Weight change Stomach Intestine
Relative weight DNA Protein Length Villus height
% enteral g g g/d g/kg mg/kg body cm/10 g μm
0 7 410 ± 114 554 ± 153 37 ± 7 7.4 ± 1.5 53 ± 23 515 ± 77 2.2 ± 0.5 334 ± 167
2.5 7 366 ± 106 467 ± 132 23 ± 11 7.9 ± 1.6 54 ± 13 607 ± 254 2.7 ± 0.5 438 ± 206
5.0 8 386 ± 76 502 ± 96 20 ± 14 6.7 ± 1.1 52 ± 14 615 ± 118 2.6 ± .06 493 ± 118
7.5 10 399 ± 76 550 ± 89 34 ± 39 8.3 ± 1.6 56 ± 16 665 ± 253 2.5 ± 0.3 378 ± 66
10.0 7 361 ± 76 491 ± 108 31 ± 11 7.9 ± 1.6 49 ± 13 663 ± 129 2.5 ± 0.3 522 ± 132
30.0 5 467 ± 51 575 ± 58 29 ± 7 7.8 ± 1.3 43 ± 7 811 ± 105 2.3 ± 0.2 359 ± 134
100 7 378 ± 121 435 ± 137 16 ± 3 9.8 ± 1.6 50 ± 24 762 ± 127 3.1 ± 0.5 475 ± 159

1

Values are shown as means ± sd. Groups did not differ.

FIGURE 1

Intestinal relative weight (upper panel), protein content (middle panel), and DNA content (lower panel) of pups fed varying volumes of milk enterally. The values are means ± sem, n = 5–10. *Different from the 0% enteral intake group, P < 0.05.

Intestinal relative weight (upper panel), protein content (middle panel), and DNA content (lower panel) of pups fed varying volumes of milk enterally. The values are means ± sem, n = 5–10. *Different from the 0% enteral intake group, P < 0.05.

Motor activity and hormones.

Recordings of duodenal motor contractions were adequate for analysis in 44 of the 51 pups. By visual inspection, the presence of MMC varied among the feeding groups. Representative recordings are shown in Figure 2. The incidence of MMC did not differ among pups fed 0–7.5% of fluid enterally, ranging from 0% to 17%, but was greater among pups fed 10% or greater, ranging from 50% to 80% (Table 3; P < 0.02, trend analysis). As a result, the duration of MMC (phase 3) ranged from 0 to 1.1 min/h among pups fed 0–7.5% of fluid enterally, whereas it ranged from 5.5 to 7.4 min/h among pups fed 10–100% (Table 3). The overall durations of nonmigrating contractile activity (phase 2) and motor quiescence (phase 1) did not differ among the groups, ranging from 28 to 47 min/(h · lead) and 9–23 min/(h · lead), respectively. Plasma motilin and gastrin concentrations did not differ among the feeding groups (data not shown).

FIGURE 2

Representative small intestinal manometry recordings from pups fed 0 (A), 10 (B), and 100% (C) of fluid enterally, as described in the text. The orientation and magnification of each figure is similar. Motor contractions recorded from the proximal duodenum are shown in the upper line in (A) and (C), while contractions from the distal antrum are shown in the upper line in (B). Each lower line in each figure displays motor contractions recorded 2.5 cm distally to the one above it. The arrows denote the start of phasic contractions in a given lead of recording. When they occur sequentially temporally in three or more leads, as is the case in (B) and (C), they indicate the presence of a MMC. Note that very little motor quiescence is present in the pup given 0% enterally, while phasic contractions are present in pups fed 10% and 100% enterally. Furthermore, in both (B) and (C), phasic activity migrates across three and four recording ports, respectively.

Representative small intestinal manometry recordings from pups fed 0 (A), 10 (B), and 100% (C) of fluid enterally, as described in the text. The orientation and magnification of each figure is similar. Motor contractions recorded from the proximal duodenum are shown in the upper line in (A) and (C), while contractions from the distal antrum are shown in the upper line in (B). Each lower line in each figure displays motor contractions recorded 2.5 cm distally to the one above it. The arrows denote the start of phasic contractions in a given lead of recording. When they occur sequentially temporally in three or more leads, as is the case in (B) and (C), they indicate the presence of a MMC. Note that very little motor quiescence is present in the pup given 0% enterally, while phasic contractions are present in pups fed 10% and 100% enterally. Furthermore, in both (B) and (C), phasic activity migrates across three and four recording ports, respectively.

TABLE 3

Characteristics of motor activity in neonatal dogs fed varying volumes of milk enterally for 4–5 d 1, 2

Group, fluid n Phase 3 MMC present Phase 2 Phase 1
% enteral min/(h · lead) % of dogs 3 min/(h · lead)
0 6 0.5 ± 1.2a 17 37 ± 14 23 ± 15
2.5 7 0a None detected 30 ± 8 23 ± 10
5.0 6 1.1 ± 2.7a 17 39 ± 15 17 ± 12
7.5 8 0a None detected 47 ± 8 11 ± 6
10.0 6 5.5 ± 8.4a 50 44 ± 10 12 ± 7
30.0 5 7.4 ± 5.6b 60 28 ± 7 15 ± 7
100 6 5.8 ± 6.3b 83 47 ± 5 9 ± 5
Group, fluid n Phase 3 MMC present Phase 2 Phase 1
% enteral min/(h · lead) % of dogs 3 min/(h · lead)
0 6 0.5 ± 1.2a 17 37 ± 14 23 ± 15
2.5 7 0a None detected 30 ± 8 23 ± 10
5.0 6 1.1 ± 2.7a 17 39 ± 15 17 ± 12
7.5 8 0a None detected 47 ± 8 11 ± 6
10.0 6 5.5 ± 8.4a 50 44 ± 10 12 ± 7
30.0 5 7.4 ± 5.6b 60 28 ± 7 15 ± 7
100 6 5.8 ± 6.3b 83 47 ± 5 9 ± 5

1

Data for phases 3, 2, and 1 expressed as means ± sd.

2

Means in a column with superscripts without a common letter differ, P < 0.05.

3

Percentages of MMC differed among the groups as determined by trend analysis.

TABLE 3

Characteristics of motor activity in neonatal dogs fed varying volumes of milk enterally for 4–5 d 1, 2

Group, fluid n Phase 3 MMC present Phase 2 Phase 1
% enteral min/(h · lead) % of dogs 3 min/(h · lead)
0 6 0.5 ± 1.2a 17 37 ± 14 23 ± 15
2.5 7 0a None detected 30 ± 8 23 ± 10
5.0 6 1.1 ± 2.7a 17 39 ± 15 17 ± 12
7.5 8 0a None detected 47 ± 8 11 ± 6
10.0 6 5.5 ± 8.4a 50 44 ± 10 12 ± 7
30.0 5 7.4 ± 5.6b 60 28 ± 7 15 ± 7
100 6 5.8 ± 6.3b 83 47 ± 5 9 ± 5
Group, fluid n Phase 3 MMC present Phase 2 Phase 1
% enteral min/(h · lead) % of dogs 3 min/(h · lead)
0 6 0.5 ± 1.2a 17 37 ± 14 23 ± 15
2.5 7 0a None detected 30 ± 8 23 ± 10
5.0 6 1.1 ± 2.7a 17 39 ± 15 17 ± 12
7.5 8 0a None detected 47 ± 8 11 ± 6
10.0 6 5.5 ± 8.4a 50 44 ± 10 12 ± 7
30.0 5 7.4 ± 5.6b 60 28 ± 7 15 ± 7
100 6 5.8 ± 6.3b 83 47 ± 5 9 ± 5

1

Data for phases 3, 2, and 1 expressed as means ± sd.

2

Means in a column with superscripts without a common letter differ, P < 0.05.

3

Percentages of MMC differed among the groups as determined by trend analysis.

DISCUSSION

Extremely low birth weight infants (i.e., those infants with birth weights < 1250 g) commonly are given small enteral feedings in an attempt to reduce the risk for NEC while promoting intestinal growth and postnatal adaptation. This study shows when feedings provide < 30% of daily fluid intake in the neonatal dogs, they do not induce growth as assessed by DNA and protein contents. Our findings concerning the dose relationship of feeding volume and mucosal growth in neonatal dogs are consistent with those reported for neonatal pigs (4,13). Burrin et al. (4 and Stoll et al. (13) have also shown in pigs that small enteral feedings increase intestinal protein content more than proliferation, as reflected by DNA content and villus height. Feedings may mediate increased gut growth by direct effects (14–16), by triggering hormone release (16,17), and by triggering cellular messengers or molecular events (18–20).

Neonatal pigs could not be used to determine the ability of enteral feedings to trigger maturation of intestinal motor function, because their small intestinal motor patterns are mature at birth (7). Small intestinal motor function in neonatal dogs, in contrast, is similar to that of preterm infants born at 28 wk gestational age (8). Moreover, maturation of lower esophageal function and antroduodenal motor patterns mature over the first postnatal month to patterns that more closely resemble those of adults (8,21). Using this neonatal canine model, we have confirmed that enteral feeding volumes that are similar to those used for minimal enteral feeding strategies can trigger maturation of motor function as reflected by the presence of the MMC. However, feeding volumes that exceed 10% of daily fluid intakes provide no further benefit with respect to maturation of motor function. Thus, we conclude that there is no advantage in feeding preterm infants volumes that provide > 10% of daily fluid intake to accrue the benefits of enteral feeding on motor patterns.

One limitation of applying our findings to neonates is that healthy term neonatal pups were studied, and human neonates being cared for in the Neonatal Intensive Care Unit may be ill and/or stressed. However, motor patterns of term neonatal pups are similar to those seen in preterm infants born with gestational ages of 28 wk, and they mature postnatally in neonatal pups and they do in preterm infants (8). Moreover, we have recently confirmed out findings concerning characteristics of motor patterns in preterm infants born with gestational ages < 32 wk (22). Another limitation is that pups were permitted to suckle for 48 h, which is not the case for sick preterm infants. These initial small feedings may have triggered the release of gastrointestinal hormones before the feeding intervention was applied. We speculate that differences in feeding care of our neonatal pups vs. that of preterm infants may explain the apparent inability of small enteral feedings to trigger greater release of gastrointestinal hormones, as is the case in preterm infants. Another explanation is that the small numbers of animals in each feeding group may not have provided sufficient power to detect differences among the groups. Another limitation of the study is the use of feedings that contained both canine and human milk. It is possible that the composition of canine and human milk varies with respect to growth factors, and the pups may have received approximately one-half of the content that would have been present if canine milk had been used exclusively. We speculate that this circumstance is analogous to that which occurs when preterm infants are given mixtures of expressed breast milk and artificial formula when their mothers can not provide sufficient volumes of milk for exclusive use. This study was also limited to the assessment of hormone release while the pups were food-deprived. Although it would have been useful to assess responses to feeding, blood volumes of the neonatal pups ethically precluded serial timed sampling of plasma concentrations of hormones.

Small enteral feedings triggered maturation of motor function in this study using neonatal pups, as has been shown to occur in preterm infants (1,23). Motor function is regulated by the enteric nervous system (ENS) and modulated by hormonal release. Compared with unfed preterm infants, plasma concentrations of gastrin and motilin are higher in infants who have been given small feedings (1,6,23). Although plasma concentrations of gastrin and motlin are higher in preterm infants given large feedings [i.e., 140 mL/(kg · d)] than infants given small feedings [i.e., 20 mL/(kg · d)], maturation of motor patterns is similar (23). These findings in neonates are consistent with those in this study, in that the incidence of MMC was similar in pups given any feeding volume that exceeded 10% of daily fluid intake. Although an intimate relationship exists between enteral nutrients and ENS function in adults with gastroesophageal reflux, irritable bowel disease, and ulcerative colitis, the mechanism whereby enteral feedings mediate maturation of motor function in the neonate has not been explored. Zangen and colleagues (24) have shown that gastric compliance and receptive accommodation in neonates increases rapidly the first several days as larger feeding volumes are ingested. We speculate from their and our observations that primary afferent nerves as well as local neural mechanisms may mediate maturation of the ENS that small feedings appear to induce.

Approximately 10% of extremely low birth weight infants will develop NEC, which causes mortality and morbidity (25) Although the cause of NEC is multifactorial, 90% of infants who develop NEC have been fed (25). Furthermore, the incidence is higher among infants who have had their feeding volumes increased rapidly (26,27). Although the use of parenteral nutrition permits normal growth in preterm infants (28), the absence of enteral nutrients in animals and adult humans reduces mucosal mass and function and blunts the release of gastrointestinal hormones (29,30). In an attempt to reduce the risk of feeding, many neonatologists minimize the enteral feeding volumes used to supplement the nutrient intake provided by parenteral nutrition. When infants are nourished in this manner, they have better hepatic function (2), higher plasma concentrations of gastrointestinal hormones (1), and better feeding tolerance (1–3) with no greater incidence of NEC than infants who are given parenteral nutrition alone (1–3). The volume needed to achieve these benefits has not yet been determined because these previous studies tested volumes ranging from 1 to 24 mL/(kg · d). Our data show that it is not necessary to increase feeding volume rapidly to promote gut function in extremely low birth weight infants when using minimal enteral feeding strategies unless there are other clinical reasons to reduce parenteral feeding volumes, such as concerns of infection.

We thank Dr. George G. Klee, Director of the Immunochemical Core Laboratory at Mayo Clinc and Foundation for performing the radioimmunoassays for this study.

LITERATURE CITED

1.

Berseth

,

C. L.

(

1992

)

Effect of early feeding and maturation of the preterm infant's small intestine

.

J. Pediatr.

20

:

947

–

953

.

2.

Dunn

,

L.

,

Hulman

,

S.

,

Weiner

,

J.

 &

Kliegman

,

R.

 (

1988

)

Beneficial effects of early hypocaloric enteral feeding on neonatal gastrointestinal function: preliminary report of a randomized trial

.

J. Pediatr

112

:

622

–

629

.

3.

Slagle

,

T.

&

Gross

,

S.

(

1988

)

Effect of early low-volume substrate on subsequent feeding tolerance in very low birth weight infants

.

J. Pediatr.

113

:

525

–

531

.

4.

Burrin

,

D. G.

,

Stoll

,

B.

,

Jiang

,

R.

,

Chang

,

X.

,

Hartmann

,

B.

,

Holst

,

J. J.

,

Greeley

,

G. H.

, Jr. &

Reeds

,

P. J.

 (

2000

)

Minimal enteral nutrient requirements for intestinal growth in neonatal piglets: how much is enough?

.

Am. J. Clin. Nutr.

71

:

1603

–

1610

.

5.

Lucas

,

A.

,

Bloom

,

S. R.

&

Aynsley-Green

,

A.

 (

1983

)

Metabolic consequences of depriving preterm infants of enteral nutrition

.

Acta Paediatr. Scand.

72

:

245

–

249

.

6.

Shulman

,

D. I.

&

Konterek

,

K.

(

1993

)

Gastrin, motilin, insulin, and insulin-like growth factor-1 concentrations in very low birth weight infants receiving enteral or parenteral nutrition

.

J. Parenter. Enteral Nutr.

17

:

130

–

135

.

7.

Ruckebusch

,

Y.

(

1986

)

Development of digestive motor patterns during perinatal life: mechanism and significance

.

J. Pediatr. Gastroenterol. Nutr.

5

:

523

–

536

.

8.

Baker

,

J.

&

Berseth

,

C. L.

(

1995

)

Postnatal change in inhibitory regulation of intestinal motor activity in human and canine neonates

.

Pediatr. Res.

38

:

133

–

139

.

9.

Dryburgh

,

J. R.

&

Brown

,

J. C.

(

1980

)

Radioimmunoassay for motilin

.

Gastroenterology

9

:

657

–

678

.

10.

Miller

,

L. J.

,

Jardine

,

I.

,

Weissman

,

E.

,

Go

,

V.L.W.

 &

Speicher

,

D.

 (

1984

)

Cholecystokinin from human brain

.

J. Neurochem.

43

:

835

–

840

.

11.

Lowry

,

O. H.

,

Rosebrough

,

N. J.

,

Farr

,

A. C.

 &

Randall

,

R. J.

 (

1951

)

Protein measurement with the folin phenol reagent

.

J. Biol. Chem.

193

:

265

–

275

.

12.

LaBarca

,

C.

&

Paipen

,

K.

(

1980

)

A simple, rapid, sensitive DNA assay procedure

.

Anal. Biochem.

102

:

344

–

352

.

13.

Stoll

,

B.

,

Chang

,

Z.

,

Fan

,

M. Z.

,

Reeds

,

P. J.

 &

Burrin

,

D. G.

 (

2000

)

Enteral nutrient intake level determines intestinal protein synthesis and accretion rates in neonatal pigs

.

Am. J. Physiol.

279

:

G288

–

G294

.

14.

Weser

,

E.

,

Babbitt

,

J.

,

Hoban

,

M.

 &

Vandeventer

,

A.

 (

1986

)

Intestinal adaptation: different growth responses to disaccharides compared with monosaccharides in rat small bowel

.

Gastroenterology

91

:

1521

–

1527

.

15.

Spector

,

M. H.

,

Levine

,

G. M.

&

Deren

,

J. J.

 (

1977

)

Direct and indirect effects of dextrose and amino acid on gut mass

.

Gastroenterology

72

:

706

–

710

.

16.

Weinberg

,

L. M.

,

Pusateri

,

J. P.

, Jr. &

Levine

,

G. M.

 (

1989

)

Comparison of different caloric substrates on intestinal adaptation in the rat

.

Gastroenterology

96

:

151415

–

151420

.

17.

Weser

,

E.

,

Bell

,

D.

&

Tawil

,

T.

 (

1981

)

Effects of octapeptide-cholecystokinin, secretin, and glucagons on intestinal mucosal growth in parenterally nourished rats

.

Dig. Dis. Sci.

26

:

409

–

416

.

18.

Kandil

,

H. M.

,

Argenzio

,

R. A.

,

Chen

,

W.

,

Berschneider

,

H. M.

,

Stiles

,

A. D.

,

Westwick

,

J. K.

,

Rippe

,

R. A.

,

Brenner

,

D. A.

 &

Rhoads

,

J. M.

 (

1995

)

l-Gluatamine and l-asparagine stimulate ODC activity and proliferation in a porcine jejunal enterocyte line

.

Am. J. Physiol.

269

:

591

–

599

.

19.

Higashiguchi

,

T.

,

Hasselgre

,

P. O.

,

Wagner

,

K.

 &

Fischer

,

J. E.

 (

1993

)

Effect of glutamine on protein synthesis in isolated intestinal epithelial cells

.

J. Parenter. Enteral Nutr.

17

:

307

–

314

.

20.

Litvak

,

D. A.

,

Evers

,

B. M.

,

Hwang

,

K. O.

,

Hellmich

,

M. R.

,

Ko

,

T. C.

 &

Townsend

,

C. M.

 (

1998

)

Butyrate-induced differentiation of Caco-2 cells is associated with apoptosis and early induction of p21Waf1/Cip1 and p27Kip1

.

Surgery

124

:

161

–

169

.

21.

Spedale

,

S. B.

,

Weisbrodt

,

N. W.

&

Morriss

,

F. H.

, Jr (

1992

)

Ontogenic studies of gastrointestinal function: lower esophageal sphincter maturation in neonatal beagle puppies

.

Pediatr. Res.

16

:

851

–

855

.

22.

Bisquera

,

J.

&

Berseth

,

C. L.

(

2001

)

Effect of feeding volume on maturation of gastrointestinal function in preterm infants

.

Gastroenterology

120

(Suppl.):

264

(abs.).

23.

Berseth

,

C. L.

&

Nordyke

,

C.

(

1993

)

Enteral nutrients promote postnatal maturation of intestinal motor activity in preterm infants

.

Am. J. Physiol.

264

:

G1046

–

G1051

.

24.

Zangen

,

S.

,

DiLorenzo

,

C.

,

Zangen

,

T.

,

Mertz

,

H.

,

Schwankovsky

,

L.

 &

Hyman

,

P. E.

 (

2001

)

Rapid maturation of gastric relaxation in newborn infants

.

Pediatr. Res.

50

:

629

–

632

.

25.

Stoll

,

B. J.

(

1994

)

Epidemiology of necrotizing enterocolitis

.

Clin. Perinatol.

21

:

205

–

218

.

26.

McKeown

,

R. E.

,

Marsh

,

T. D.

,

Amarnath

,

U.

,

Garrison

,

C. Z.

,

Addy

,

C. L.

,

Thompson

,

S. J.

 &

Austin

,

J. L.

 (

1992

)

Role of delayed feeding and of feeding increments in necrotizing enterolcolitis

.

J. Pediatr

121

:

764

–

770

.

27.

Berseth

,

C. L.

(

2002

)

Incidence of necrotizing enterocolitis in preterm infants is feeding volume dependent while maturation of small intestinal motor patterns is not

.

Gastroenterology

in press.

28.

Gomez

,

M. R.

&

Heird

,

W. C.

(

1994

)

Total parenteral nutrition in necrotizing enterocolitis

.

Clin. Perinatol.

21

:

425

–

436

.

29.

Buchman

,

A. L.

,

Moukarezel

,

A. A.

,

Bhuta

,

S.

,

Belle

,

M.

,

Ament

,

M. E.

,

Eckhert

,

C. D.

,

Hollander

,

D.

,

Gornbein

,

J.

,

Kopple

,

J. D.

 &

Vijayaroghaven

,

S. R.

 (

1995

)

Parenteral nutrition is associated with intestinal morphologic and functional changes in humans

.

J. Parenter. Enteral Nutr.

19

:

453

–

460

.

30.

Groos

,

S.

,

Hunefeld

,

G.

&

Luciano

,

L.

 (

1996

)

Parenteral versus enteral nutrition: morphological changes in human adult intestinal mucosa

.

J. Submicrosc. Cytol. Pathol.

28

:

61

–

74

.

Abbreviations

  • ENS

  • MMC

  • NEC

    necrotizing enterocolitis

FOOTNOTES

4

Containing [unit/(kg·d)]: glucose (28 g, amino acids (8 g) as Trophamine (McGaw, Irvine, CA), lipid emulsion (6 g) as Intralipid (Baxter, Deerfield, IL), NaCl (3.4 mmol), K2HPO4 (6.6 mmol) of PO4, KCl (3.3 mmol), Ca+2gluconate (2.55 nmol) and MgSO4 (0.3 mmol) and multiple vitamins MVI Ped (AstraZenica, Westborough, MA) that provided vitamin A, 230 iu; vitamin D, 40 iu; vitamin E, 0.7 iu; vitamin K, 20 iu; vitamin C, 0.8 μg; thiamine, 0.12 mg; riboflavin, 0.14 mg; vitamin B-3, 1.7 mg; vitamin B-5, 0.5 mg; vitamin B-6, 0.1 mg; vitamin B-12, 0.1 μg; biotin, 2 μg; and folic acid, 14 μg.

Author notes

1

Published in abstract form [Owens, L., Burrin, D. & Berseth, C. L. (1996) Enteral nutrition has a dose-response effect on maturation of neonatal canine motor activity. Gastroenterology 110: 828 (abs.)].

© 2002 The American Society for Nutritional Sciences

© 2002 The American Society for Nutritional Sciences