Randomized Trial of Nutrient-Enriched Formula Versus Standard Formula for Postdischarge Preterm Infants
Objectives. Preterm infants are frequently discharged from the hospital growth retarded and show reduced growth throughout childhood. In a large efficacy and safety trial, we tested the hypothesis that nutritional intervention in the first 9 months postterm would reverse postdischarge growth deficits and improve neurodevelopment without adverse safety outcomes.
Participants and intervention. Two hundred eighty-four infants (mean gestation: 30.9 weeks) were studied; 229 were randomly assigned a protein, energy, mineral, and micronutrient-enriched postdischarge formula (PDF; N = 113) or standard term formula (TF; N = 116) from discharge (mean 36.5 weeks' postmenstrual age). A reference group (N = 65) was breastfed until at least 6 weeks' postterm.
Outcome measures. Anthropometry was performed at 6 weeks and 3, 6, 9, and 18 months. Development was measured at 9 months (Knobloch, Passamanick, and Sherrard's developmental screening inventory) and 18 months (Bayley Scales of Infant Development II; primary outcome) postterm.
Results. At 9 months, compared with the TF group, those fed PDF were heavier (difference 370 g; 95% confidence interval [CI]: 84–660) and longer (difference 1.1 cm; 95% CI: 0.3–1.9); the difference in length persisted at 18 months (difference 0.82 cm; 95% CI: −0.04–1.7). There was no effect on head circumference. The effect of diet was greatest in males; at 9 months length deficit with TF was 1.5cm (95% CI: 0.3–2.7), and this remained at 18 months (1.5cm [95% CI: 0.3–2.7]). There was no significant difference in developmental scores at 9 or 18 months, although PDF infants had a 2.8 (−1.3–6.8) point advantage in Bayley motor score scales. At 6 weeks' postterm, exclusively breastfed infants were already 513 g (95% CI: 310–715) lighter and 1.6cm (95% CI: 0.8–2.3) shorter than the PDF group, and they remained smaller up to 9 months' postterm.
Conclusions. 1) Improving postdischarge nutrition in the first 9 months may “reset” subsequent growth—at least until 18 months for body length. We intend to follow-up the children at older ages. The observed efficacy of PDF was not associated with adverse safety outcomes. 2) We cannot reject the hypothesis that postdischarge nutrition benefits motor development and this requires additional study. 3) Our data raise the possibility that breastfed postdischarge preterm infants may require nutritional supplementation, currently under investigation.
- PDF =
- postdischarge formula •
- TF =
- term formula •
- MUAC =
- mid-upper arm circumference •
- SD =
- standard deviation •
- CI =
- confidence interval
Considerable attention has been directed toward improving the nutrition of preterm infants while they are in hospital. This has included the development of special nutrient-enriched formulas designed to meet the projected increased nutritional needs of preterm infants, and multinutrient fortifiers, which can be added to human milk. Long-term outcome data now show the neonatal period to be a critical one in terms of the effects of nutrition on later health and developmental outcomes.1–3 However, until recently relatively little attention had been paid to the postnatal nutrition of these infants after discharge from hospital. At this rather arbitrary time point, breast milk fortifiers are usually stopped, or the infant commonly changed onto a formula designed for healthy term infants. Our previous studies suggest that preterm infants are commonly discharged weighing little more than half the appropriate weight for a term infant at the equivalent postmenstrual age. Of more concern, >30% of these infants remained below the tenth percentile for weight at 18 months, and >20% at 7 to 8 year follow-up.4 A key question is whether such infants have special nutritional requirements in the postdischarge period and, in more biological terms, whether this period of nutrition is also a critical one for later health and development.
The observed small size of preterm infants at hospital discharge is associated with reduced body stores of a variety of specific nutrients. The trend toward earlier discharge of preterm infants means that the opportunity to provide mineral supplements is reduced and many are discharged with undermineralized bones because of a poor intake of calcium and phosphorus.5–7 Although some studies suggest that there is catch-up in bone mass during infancy,8 ,9 the effect of early mineral deficiency on final height and peak bone mass is unknown, and we recently demonstrated an association between biochemical evidence of early metabolic bone disease and reduced stature up to 12 years later.10 Calculations of accretion rates of iron, zinc, and copper in hypothetical well, very low birth weight infants fed to current standards show that they have suboptimal status compared with term infants of equivalent postmenstrual age.11 Zinc status is of particular concern given evidence for its role in growth and brain development.12
Over a decade ago, we designed a nutrient-enriched formula for use in preterm infants after hospital discharge (postdischarge formula [PDF]). This formula, which is intermediate in composition between preterm and term formulas, contains predominantly an increase in protein, with sufficient extra energy to permit utilization. It also contains extra calcium and phosphorus and zinc, deficits of which might limit linear growth, together with additional vitamins and trace elements to support the projected increased growth performance. In a pilot study on 32 preterm infants,13 those randomized to receive the PDF up to 9 months' postterm showed significantly greater weight and length gains while on the diet, and had higher bone mineral content in the distal radius than infants who received a standard term infant formula.14 These findings raised the hypothesis that improved nutrition in preterm infants during the first few months after hospital discharge might ameliorate the growth deficits seen in these infants. However, this small pilot study did not address the question of whether the growth benefit would persist beyond the neonatal period, nor whether it would have any long-term consequences for neurodevelopment.
We therefore tested these hypotheses in a large, randomized trial in postdischarge preterm infants comparing the nutrient-enriched formula with standard term formula up to 9 months' postterm.
Preterm infants (<37 weeks' gestation) were recruited from 5 neonatal units in Cambridge, Leicester, and Nottingham between 1993 and 1995, with 18-month follow-up completed in 1997. All had a birth weight below 1750 g and were free from congenital malformations or conditions known to affect growth or development. Infants were eligible if their mother did not intend to provide breast milk or to breastfeed after hospital discharge, if one of their main caretakers spoke English, and if they intended to stay in the United Kingdom for the next 18 months. Infants were excluded from the study if they weighed >3 kg or were >100 days old at discharge (to exclude infants with bronchopulmonary dysplasia who would be expected to have atypical postdischarge growth), or if they had severe brain disease (Grade III or IV intraventricular hemorrhage, meningitis, or hydrocephalus). They were randomized onto the trial formulas (PDF or standard term formula [TF]) in the week before planned hospital discharge and remained on it until 9 months' postterm. Solids were introduced at a time determined by the parents with advice from their health care professionals (current United Kingdom recommendations for term infants are to introduce solids between 4 and 6 months of age; no specific guidelines exist for preterm infants); research staff were not involved in this decision.
Research nurses collected data from the hospital records, including obstetric data, clinical progress, feed intake data, and growth during the immediate postnatal period. Infants were studied at home at 6 weeks', 12 weeks', and 26 weeks' postterm, when anthropometry was performed. Weight was measured using digital electronic scales, length using a horizontal infant stadiometer (with the infant's head held by one observer, and both legs extended and held by a second observer) and occipito-frontal head circumference and mid-upper arm circumference (MUAC) using a nonstretchable encircling tape. Triceps and sub-scapular skinfold thicknesses were measured using Harpenden calipers (Holtain, Dyfed, Wales, UK).
At 9 and 18 months' postterm, infants were invited to attend a follow-up clinic at their local hospital where anthropometry was performed and the neurodevelopmental tests were administered by a trained doctor.
A reference group of preterm infants whose mothers had decided to breastfeed after discharge were recruited over the same period as the formula-fed infants and followed up to 9 months' postterm; all were breastfed for at least 6 weeks after discharge, but they were allowed up to 2 oz of formula milk per day in addition to breast milk. A standard term infant formula was used to supplement or replace breast milk where required. Anthropometry was performed at each follow-up visit.
The randomization schedule was generated by permuted blocks of randomized length, with the assignments in sealed opaque envelopes, and was stratified by birth weight (< or >1200 g), whether or not infants required supplemental oxygen for >28 days, and by the number of fetuses (twins or triplets were randomized as one onto the same diet). A member of the team who was not involved in subsequent aspects of the trial prepared randomization assignments.
The formulas were color-coded; the codes were held by the formula manufacturers and were not revealed to the investigators until after the principal data analyses were performed. Both formulas were identical in color and smell.
The composition of formulas is shown in Table 1. Both were supplied by Farley Health Products (a division of H. J. Heinz Co Ltd). The PDF (Farley's PremCare) contained nearly 20% more protein in relation to energy than the standard TF (Farley's OsterMilk, Uxbridge, UK), together with increased calcium, phosphorus, trace elements, and vitamins. Both formulas comply with the European Community Directive for the composition of formulas for healthy term infants.
Principal efficacy outcome measures were 1) growth (weight, length, and occipito-frontal head circumference) at 9 and 18 months' postterm; 2) neurodevelopment at 18 months' postterm, measured by the mental and motor scales of the Bayley Scales of Infant Development II,15 from which were derived the Mental and Psychomotor Development Indices. Secondary efficacy outcome was neurodevelopment at 9 months' postterm measured using Knobloch, Passamanick, and Sherrards' Developmental Screening Inventory,16comprising 5 subscales (adaptive, gross motor, fine motor, language, and personal–social). An overall developmental score was calculated as the mean of the 5 subscale scores. At 9 and 18 months' postterm, the research doctor was asked to rate the infant's neurologic status as “normal,” “equivocal,” or “abnormal.” Developmental assessments were performed by 3 doctors; 1 in each center. All were trained by a single experienced investigator (R.M.) with regular monitoring of performance throughout the study.
Key safety outcomes were the number of episodes of infection (gastroenteritis [diarrhea with vomiting]), upper respiratory tract infection, and chest infections requiring antibiotic treatment), the presence of atopy (eczema, wheeze, and asthma), the requirement for medical services (general practitioner visits, hospital outpatient visits, or hospital admissions), withdrawals from the study, and adverse events occurring during the study. These were determined by maternal recall with, where appropriate, information from child health records. Tolerance outcomes were maternal reports of vomiting, colic, and stool characteristics (usual frequency and consistency of stools while receiving the trial milk and maternal reports of constipation). Milk volume intake was not measured because our previous trial using these formulas had shown that milk consumption at any age up to 9 months was unaffected by formula composition.
The target sample size per randomized group (113 infants) was calculated to permit detection of a 300-g difference in weight between groups at 9 months and a 400-g difference at 18 months (in each case the difference between the 10th and 25th centiles) at 5% significance and 70% to 80% power, based on our previous growth data from preterm infants. This sample size would also permit detection of a 4-point (approximately 0.3 standard deviation [SD]) difference in Bayley developmental indices.
Analyses were performed on an intention-to-treat basis. Differences between randomized groups were compared by Student's ttest, Mann-Whitney test, or χ2 test. Differences between the breastfed reference group and formula groups were compared using analysis of variance, with posthoc analysis using Dunnett's test.
The trial profile is shown in Fig 1. Two hundred twenty-nine infants were randomized, 116 to the standard TF (55 [47%] boys) and 113 (58 [51%] boys) to the postdischarge formula. One hundred three infants from each group completed 9 months on the trial diet. Eight PDF and 13 TF infants were withdrawn from the study. The numbers withdrawn because of adverse events were similar in both groups; 6 PDF infants (2 withdrawn by clinicians) and 7 TF infants (4 withdrawn by clinicians). Two PDF infants died during the study, both as a result of sudden infant death syndrome. Two infants who did not complete 9 months on the trial diet (both randomized to TF) were seen for follow-up at 9 months' postterm, and 4 (3 randomized to TF and 1 to PDF) at 18 months' postterm.
There were no significant differences between groups in demographic characteristics or in parental size (Table 2). The clinical course during the neonatal period was also similar in both groups and, at randomization, there were no significant differences in age, weight, length, or head circumference. Both groups were randomized at a similar postconceptional age (36.3 [SD 1.9] versus 36.6 [SD 2.1] weeks for PDF and TF groups, respectively).
Principal Efficacy Measures
At 9 months' postterm, infants receiving PDF were 370-g heavier (95% confidence interval [CI]: 84–660 g), and 1.1-cm longer (95% CI: 0.3–1.9) than those on TF (Table 3). The advantage in length was still present, although reduced, at 18 months' postterm (advantage 0.8 cm (95% CI: −0.04–1.7). When weight, length and head circumference were expressed as SD scores to correct for minor imbalances between groups in gender and in the exact age at follow-up, PDF infants had a significant advantage in both weight and length at 9 months' postterm (Table 3), and the advantage in length persisted at 18 months' postterm (advantage for PDF group 0.32 SD [95% CI: 0.025–0.61], P = .03). Mean SD scores for the cohort at recruitment were −1.49 (SD 0.87) for weight and −1.97 (1.16) for length, with no significant difference between groups. There were no significant differences in head circumference at any time point. Inspection of the interim growth data showed that infants fed PDF were already 220 (44–400) g heavier and 1.1 (0.4–1.7) cm longer at 6 weeks' postterm.
We found no evidence that infants who received PDF had become fat; there were no significant differences between groups in skinfold thicknesses (at 9 months; triceps 9.5 [SD 2.5] vs 9.6 [SD 2.5] cm and subscapular 6.7 [1.6] vs 6.4 cm for PDF and TF groups, respectively: at 18 months; triceps 8.0 [2.3] vs 8.1 [2.1] cm and subscapular 5.7 [1.2] vs 5.7 [1.5] cm) or MUAC (at 9 months 14.9 [SD 1.1] vs 14.7 [1.3] cm and 18 months 14.8 [1.2] vs 14.9 [1.2] cm) and at 9 months' postterm, the mean body weight of PDF infants was still below the 50th centile, with an even greater deficit for infants fed TF.
A planned subgroup analysis according to gender (Table 4) showed that the advantage for infants fed the PDF was greater in boys than girls; by 9 months, boys fed the PDF were 575 (150–1000)-g heavier and 1.5 (0.3–2.7)-cm longer than those fed TF. The advantage in length was maintained to 18 months (advantage for PDF boys 1.5 [0.3–2.7] cm). The difference in weight and length between girls fed TF and PDF was not significant at either time point. Interaction terms for the effects of gender and diet on weight or length were not significant (weight at 9 months,P = .1; length at 9 months, P = .2; length at 18 months, P = .1). Thus, subsequent analyses were performed with both sexes combined.
At 18 months' postterm, Bayley developmental scores did not differ significantly between the 2 randomized groups, although infants previously fed PDF had a 2.7-point advantage in Bayley PDI (0.25 SD). There was no evidence that the effect of diet differed in boys and girls, and no interaction between diet and gender on Bayley developmental scores. The number of infants considered to have either a possible or definite neurologic deficit at 18-month follow-up was the same in both groups.
There was no significant interaction between randomized diet and either gestation or birth weight on later growth and neurodevelopmental outcomes.
Secondary Efficacy Measures
At 9 months' postterm, Knobloch, Passamanick, and Sherrards' Developmental Screening Inventory quotient was not significantly different between groups, although there was a nonsignificant trend toward higher scores in the PDF group, particularly for gross motor and fine motor subscales (Table 5). The proportion of infants considered to have a possible or definite neurologic abnormality at 9-month follow-up was the same in the 2 groups.
There were no significant differences between randomized groups; 85% of PDF infants versus 86% of TF infants had had at least 1 upper respiratory tract infection by 18 months' postterm, 56% versus 49% had had a chest infection, and 43% versus 33% had gastroenteritis. There was no difference in antibiotic usage or atopic disease between groups, and the proportions seeking medical advice were also similar.
Similar proportions from both formula groups were reported to have vomiting or colic. Infants fed PDF had harder stools at 12 weeks (58% infants reported to have hard or formed stools compared with 36% in TF group, P = .04). However, a similar proportion (10%) of infants from both groups were perceived by parents as having constipation. PDF infants passed more stools per week at 12 weeks (median 14 (25th, 75th centiles7 ,20) versus 7,7 ,12 P < .001). By 26 weeks, however, stool characteristics were similar in the 2 groups. The proportions of infants receiving solids were not significantly different at 12 weeks (85% PDF vs 78% TF) or 26 weeks (97% vs 99%).
Exclusion of infants who failed to complete 9 months on the trial diet but who were seen at 9 and 18 months postterm did not alter the findings.
Comparison of Formula-Fed Infants With Breastfed Reference Group
Sixty-five infants formed the breastfed reference group. As expected, these infants had older mothers with higher educational qualifications, and their parents were more likely to come from social class 1 or 2 (Table 2). Breastfed infants were significantly less likely than both formula groups to be receiving solids at 12 weeks (67%, P = .04) and 26 weeks (91%, P = .03).
There were no significant differences in weight, length, or head circumference between breastfed and formula-fed infants at the time of enrollment in the study (Table 3), although breastfed infants had significantly lower weight SD scores (−2.03 [SD 1.84] vs −1.49 [0.87], P = .03). By 6 weeks' postterm (a period during which the reference group were almost exclusively breastfed) breastfed infants were significantly lighter and shorter than both formula groups. Compared with PDF infants, the disadvantage in weight for breastfed infants was 513 g (95% CI: 310–715) or 0.82 (95% CI: 0.41–1.22) SD, and in length, 1.6 (95% CI: 0.8–2.3) cm or 0.63 (95% CI: 0.24–1.03) SD. This disadvantage was still present at 26 weeks (weight 512 [95% CI: 225–798] g or 0.68 [95% CI: 0.27–1.10] SD, and length 0.9 [95% CI: 0.01–1.8] cm or 0.42 [95% CI: 0.02–0.82] SD) and was maintained up to 9 months' postterm (the end of the study for breastfed infants), beyond the period when most breastfed infants had started to receive standard TF (disadvantage in weight 439 [95% CI: 105–773] g or 0.45 [95% CI: −0.06–0.95] SD, and in length 1.1 [95% CI: 0.2–2.1] cm) or 0.65 [95% CI: 0.15–1.15] SD. Breastfed infants also had lower MUAC and triceps and subscapular skinfold thicknesses at enrollment than both formula-fed groups. Significantly lower MUAC and triceps skinfold thickness persisted at follow-up to 9 months (complete data available from authors).
We previously reported, in a small pilot study, that preterm infants randomized to receive a nutrient-enriched PDF rather than a TF after hospital discharge showed greater gains in weight and length by 9 months' postterm.13 Our current large clinical trial confirms and extends these findings, showing that the advantage in length for infants fed the PDF persisted for at least 9 months after the milk was discontinued. This suggests, for the first time, that the postdischarge period in preterm infants may be critical for the programming of the subsequent growth trajectory (although clearly, additional follow-up is required to test this). In contrast, it is interesting that we have shown recently that nutrition during the early neonatal period before hospital discharge in preterm infants has no long-term effect on later growth despite producing major differences in growth rates at the time.10
The growth differences between the diet groups seemed greater in boys than in girls; boys fed the PDF had an advantage in length of 1.5cm (around 0.5 SD) at 18 months of age (although the statistical interaction between the effect of diet and gender on this outcome was significant only at the 10% level). This finding is, however, consistent with previous work demonstrating that preterm boys randomized to receive a preterm formula up to 6 months' postterm showed greater weight and length gain than those fed a standard TF, whereas the same intervention had no significant effect in girls.17 In addition, nutritional interventions have been shown to have a greater effect on neurodevelopmental outcome in preterm boys than in preterm girls.2
Postdischarge nutrition in preterm infants has received relatively little attention, yet preterm infants are frequently discharged from the hospital with a body weight well below that of a healthy term infant at birth. Follow-up of preterm infants suggests that growth deficits persist; in our own study, twice as many children than expected were under the 10th centile at 7.5 to 8 years.4Similar findings have emerged from follow-up studies in the United States.18 In-hospital, preterm infants receive specially designed preterm infant formulas designed to meet their increased nutritional requirements, or human milk supplemented with additional nutrients, it is likely that they would continue to benefit from increased nutritional support after the rather arbitrary time point of hospital discharge. One possibility would be to continue the use of preterm infant formula after discharge, an approach adopted in at least 2 randomized trials.17 ,19 However, this approach risks exposing larger ad libitum fed infants in the community to potentially toxic intakes of certain nutrients,20 for example, vitamin D.
For this reason, a special nutrient-enriched PDF was designed, with nutrient levels intermediate between those of a standard term infant formula and a preterm formula. This formula fulfills the European Community Directive requirements for the composition of a term infant formula. It contains predominantly an increase in protein, but with only a modest increase in energy to permit utilization. It also contains an increase in minerals, particularly calcium, phosphorus, and zinc, deficits of which might limit linear growth, and vitamins to support the projected increased growth performance. Our results suggest that, as anticipated, use of this formula resulted predominantly in an increase in lean tissue rather than increasing fat deposition; infants fed the PDF had a mean weight below the 50th centile at 9 months' postterm, and there was no significant increase in skinfold thicknesses compared with infants fed the standard formula. This is consistent with the study by Cooke et al21 in which preterm boys who received a preterm formula until 6 months' postterm showed greater gain in lean body mass, measured by dual radiograph absorptiometry, than those fed a standard TF.
Two other smaller studies have examined the effect of feeding a nutrient-enriched postdischarge formula to preterm infants to 12 months' postterm. Carver et al22 reported greater head circumference up to 12 months postterm in postdischarge formula-fed infants with birth weight below 1250 g, while Atkinson et al23 found higher weights in infants fed postdischarge formula but no differences in length or head circumference.
A potential concern in any attempt to improve nutrition in vulnerable infants such as these is that the infants will simply down-regulate their intake of formula, thus negating any potential effect on outcome. This did not occur in our pilot study using the same formulas,13 presumably because the nutrient-enriched formula contained primarily an increase in protein, with only a modest increase in energy. For this reason, formula intake measurements were not repeated in the current study. Two previous studies22 ,23 comparing a postdischarge formula with term formula during the first year of life in preterm infants found lower formula intakes in infants receiving the postdischarge formula, resulting in similar energy intakes but higher protein intakes in these infants compared with those receiving term formula. The differences in energy contents between the 2 formulas used in these studies were greater than in our study (75 vs 67 kcal/100 mL compared with 72 vs 68 kcal/100 mL), which may explain the discrepancy between their results and ours.
Infants fed the PDF in our study had similar Bayley Mental Development Indices to infants fed TF. However, Bayley Psychomotor Developmental Indices were almost 3 points higher in the PDF group, representing approximately 0.25 SD. Our study was powered to detect a 0.3 SD difference between groups and a larger study or a meta-analysis of several studies would be required to detect a difference of this magnitude as significant. Developmental scores on the Knobloch, Passamanick, and Sherrards' Developmental Screening Inventory at 9 months' postterm similarly showed a trend toward higher scores in the PDF group who were approximately 1 week in advance of infants fed TF, but these differences were too small to be detected as significant in a study of this size. Additional follow-up of this cohort is planned to examine whether advantages emerge at an age when more specific tests of cognitive function can be undertaken. Our previous studies have indicated that the early postnatal period, before hospital discharge, is a critical window for programming of subsequent developmental outcome.2 An important hypothesis, which cannot be rejected on the basis of the current study findings to date, is that this window of sensitivity may extend into the postdischarge period.
Safety and Tolerance
Both formulas were well tolerated, and there were no differences in safety outcomes (the occurrence of infections, atopy, number of courses of antibiotics, and requirement for medical attention) between groups. Infants who received the PDF were reported to have harder (although more frequent) stools at 12 weeks of age than those fed TF. This was also noted in our pilot study,13 and in a study using the same formulas in term growth retarded infants.24The higher overall fat content of the PDF and its higher calcium content could well have accounted for this difference, resulting in greater formation of insoluble calcium soap formation in the stools. This could potentially be addressed by altering the fat blend and reducing the palmitic acid content. However, the difference seemed to not be problematic in practice, because the reported incidence of constipation was the same in both formula groups and the difference between groups resolved with time.
Comparison of Breastfed and Formula-Fed Infants
Breast-milk may confer important short and longer term benefits for preterm infants, including protection against necrotizing enterocolitis25 and infection26 and promotion of mental development at follow-up.27 Yet human milk does not meet the nutritional needs of these infants and requires supplementation. Because of modern counseling on the benefits of breast milk, an increasing number of mothers of preterm infants take their infants home breastfed. However, it is not clear whether unsupplemented breast milk meets the nutrient requirements of preterm infants after hospital discharge. We therefore compared growth in a group of preterm infants who were predominantly breastfed for at least 6 weeks after discharge with that of the formula-fed infants in the randomized trial. Our results show that breastfed infants had significantly lower weight SD at enrollment than formula-fed infants and that, by 6 weeks' postdischarge, they had a significant disadvantage in both weight and length, and in weight and length SD scores, compared with both formula groups, and particularly compared with infants fed the PDF. This disadvantage persisted at 9 months' postterm, a time by which most of the breastfed group were receiving partly or exclusively formula milk, and despite the fact that the parents of breastfed infants were larger than those of formula-fed infants, indicating a potentially greater genetic growth potential. Breastfed infants also had significantly lower MUAC and skinfold thickness, both at discharge from hospital and up to 9 months postterm compared with formula-fed groups. Breastfed infants were not seen at 18 months' postterm in this study, and it is possible that some or all of the deficit in weight and length would have disappeared by this time point. However, even if the growth deficits are eventually abolished, it is possible that the poor early growth will have had long-term effects on other aspects of health or development. Follow-up is planned to examine these issues.
The use of a nutrient-enriched PDF in preterm infants was associated with improved short-term growth, and, potentially, with improved linear growth beyond the period of dietary randomization. The possibility that improved postdischarge nutrition has beneficial effects on neurodevelopment requires additional investigation. In this context, the use of standard term infant formulas for formula-fed postdischarge preterm infants seems difficult to defend. Indeed, the American Academy of Pediatrics has recently recognized the importance of PDF in the first 9 months in these infants.28 Given that there was a “population shift” in growth performance toward the growth pattern of healthy term infants in the PDF group, we would suggest the use of this formula could be justified as a population policy. As a corollary, if an individual preterm infant fed the PDF were to show catch-up to the 50th centile by, say 3 months, it would be illogical to switch to a standard TF at this point because the growth potential of that infant is unknown and the infant might have had the biological potential to grow on the 90th centile given optimal nutrition. It is also illogical for those managing preterm infants who are growing poorly after discharge to continue using ad hoc, untested methods of supplementing these infants' diets (for example, the common use of energy supplements), when tested and specifically designed products are available.
Our study raises the possibility that breastfed preterm infants may require concomitant nutritional supplements after hospital discharge. Additional outcome data are needed to explore the value of such a policy. Nevertheless, because the majority of mothers breastfeeding preterm infants electively switch their infant to partial or exclusive formula feeding during the first year, it would make sense for this formula to be a postdischarge rather than a standard TF. This policy is defendable in the light of our findings of a continued growth deficit in breastfed infants up to 9 months' postterm, by which time most were receiving TF.
As our knowledge of nutrition becomes more sophisticated, we are likely to identify an increasing number of population subgroups with special nutrient needs. It would seem that postdischarge preterm infants constitute such a group.
This study was supported by Farley Health Products (a division of H. J. Heinz Company Ltd, Stockley Park, Uxbridge, United Kingdom) which also supplied the infant formulas.
We thank the research staff who collected data in the study (Helen Clough, Corina Adams, Ann Humphries, Sharon Collier, Penny Lucas, Julie Owen, Geraldine McHugh, Mary Quinn, Dawn Rodd, Emma Sutton, Julie Stammers, and Catherine Leeson-Payne) and the parents who allowed their infants to participate. We also thank Farley Health Products for their collaboration, contributory funding, and supply of the trial formulas.
- Received July 8, 2000.
- Accepted February 12, 2001.
- Address correspondence to Mary S. Fewtrell, MD, FRCPCPH, MRC Childhood Nutrition Research Centre, Institute of Child Health, 30 Guilford St, London WC1N 1EH, United Kingdom. E-mail:
- Lucas A,
- Brooke OG,
- Morley R,
- Cole TJ,
- Bamford MF
- Lucas A,
- Morley R,
- Cole TJ
- ↵Lucas A. Nutrition, Growth and Development of Postdischarge Preterm Infants. In: Post-hospital Nutrition of the Preterm Infant.Report of the 106th Ross Conference on Pediatric Research, Colorado Springs, Colorado, August 18–20, 1995. Columbus, OH: Ross Products Div, Abbott Labs; 1996:81–89
- James JR,
- Congdon PJ,
- Truscott J,
- Horsman A,
- Arthur R
- Congdon PJ,
- Horsman A,
- Ryan SW,
- Truscott JG,
- Durward H
- ↵Ehrenkranz RA. Response: Micronutrients–Zinc, Copper and Iron In:Post-hospital Nutrition of the Preterm Infant. Report of the 106th Ross Conference on Pediatric Research, Colorado Springs, Colorado, August 18–20, 1995. Columbus, OH: Ross Products Div, Abbott Labs; 1996:38–42
- ↵Reifen RM, Zlotkin S. Microminerals. In: Tsang R, Lucas A, Uauy R, Zlotkin S, eds. Nutritional Needs of the Preterm Infant. Pawling, NY: Caduceus Medical Publishers; 1993:195–207
- Lucas A,
- Bishop NJ,
- King FJ,
- Cole TJ
- Bishop NJ,
- King FJ,
- Lucas A
- ↵Bayley N. Manual for the Bayley Scales of Infant Development. 2nd ed. San Antonio, TX: Psychological Corporation; 1993
- ↵Hack MD, Borawski-Clark E. Postdischarge Growth of Very-Low-Birth-Weight Children. In: Post-hospital Nutrition of the Preterm Infant. Report of the 106th Ross Conference on Pediatric Research, Colorado Springs, Colorado, August 18–20, 1995. Columbus, OH: Ross Products Division, Abbott Labs; 1996:58–63
- Lucas A,
- King FJ,
- Bishop NJ
- Carver JD,
- Wu PY,
- Hall RT,
- et al.
- Atkinson SA,
- Randall-Simpson J,
- Chang M,
- Bosco P
- ↵Fewtrell MS, Morley R, Abbott RA, Singhal A, Stephenson T, MacFadyen UM, Clements H, Lucas A. Catch-up growth in small for gestational age term infants: a randomized trial. Am J Clin Nutr; in press
- ↵American Academy of Pediatrics, Committee on Nutrition. Nutritional needs of preterm infants. In: Pediatric Nutrition Handbook. 4th ed. Elk Grove Village, IL: American Academy of Pediatrics; 1998:72–73
- Copyright © 2001 American Academy of Pediatrics