Secondary Logo

Journal Logo

Original articles

Complementary Feeding Practices in Preterm Infants: An Observational Study in a Cohort of Italian Infants

Fanaro, Silvia; Borsari, Gloria; Vigi, Vittorio

Author Information
Journal of Pediatric Gastroenterology and Nutrition: December 2007 - Volume 45 - Issue - p S210-S214
doi: 10.1097/01.mpg.0000302974.90867.f1
  • Free


It is well known that there are no evidence-based guidelines about the most appropriate time and method of weaning preterm infants, which is a much debated issue (1). This is a matter of serious concern because there is clear evidence that low birth weight (LBW) infants, and especially very low birth weight (VLBW) and extremely low birth weight (ELBW) infants, in the period after discharge from hospital and often during the first year of life, have increased nutritional needs to replenish tissue stores and to maintain and support catch-up growth (1). The optimal duration of exclusive milk feeding has not yet been determined, mainly because there is a lack of reliable scientific data on this topic. The absence of scientific evidence has inevitably led to different and often contrasting approaches by caregivers, who sometimes advocate early and sometimes late weaning, and too often leave the problem entirely to the family. This is a matter of concern because weaning, besides short-term effects such as infection and atopy, may exert its consequences well beyond infancy (2).

This issue is further complicated by the absence of reliable information on current weaning practices of babies born preterm. A few years ago an important study by Norris et al (3) examined current infant feeding practices among caregivers of preterm infants. Two hundred fifty-three infants born preterm were recruited in southeast England over a 2-year period (3). The study showed that the introduction of complementary foods varied widely and that compliance with the few existing general recommendations on weaning was poor (3). To our knowledge, no comparative figures on current complementary feeding in preterm infants are available for Italy. The aim of the present study was to gather information on the weaning practices in the area of Ferrara, a densely populated, prosperous urban area in northeastern Italy, and to compare our data with those of Norris et al. Ferrara is a university town, with a third-level neonatal intensive care unit, identified as a regional hub.


This observational study on feeding habits during the first year was conducted on 230 preterm infants admitted to the Division of Neonatology of the University Hospital of Ferrara, Italy, from January 2004 to December 2005. Mothers were acquainted with the study during the hospital stay and were asked to participate, receiving a detailed structured questionnaire at the moment of discharge. Maternal age, parity, ethnicity, educational degree, and professional status of the mother were collected from the clinical records. Maternal education was classified in 4 different levels, and the professional status was identified according to the Italian Statistical Institute classification (Table 1) (4). Feeding practices (type of milk, time of introduction of the first solid food, meat, and gluten) and weight of the infant at weaning were recorded by the mothers on the questionnaire, which was returned during the follow-up visits at 6 and 12 months corrected age.

Italian Statistical Institute professional classifications

Data were analyzed dividing infants in subgroups, considering sex, GA, BW, and weight for GA. The age at weaning was considered from both birth and term (40 weeks gestation) to allow for the degree of prematurity. GA was calculated on the last menstrual event, corrected on the basis of the first ultrasound examination, when available. Appropriateness of weight for GA was evaluated with the updated fetal-infant chart of Fenton (5).

The statistical analysis was performed by use of SPSS for Windows version 7, using 1-way analysis of variance, for normally distributed values, and the nonparametric Mann-Whitney test. Single linear regression was performed to assess the association between variables. Significance was taken at P < 0.05.


Of the 230 infants originally enrolled, 47 missed the first visit, and an additional 27 did not complete the entire study. Among the 156 infants (74 boys, 82 girls) whose questionnaires were appropriately completed by their mothers, there were 125 singletons, 14 sets of twins, and 1 set of triplets. GA ranged from 24 to 36 weeks + 5 days (mean ± SE: 32.6 ± 0.24 weeks), and BW ranged from 595 to 4060 g (mean ± SE: 1,981 ± 54.1 g). Thirty infants were classified as small-for-gestational-age (SGA) and 126 as not SGA. The complete set of data is given in Table 2.

Clinical data of the infants studied

Sociodemographic Status of the Parents

Of the 140 mothers, 15 originated from foreign countries (5 from north Africa, 4 from northern Europe, and 6 from eastern Europe). The mean maternal age was 31.1 ± 0.39 years (mean ± SE, range 18–44). Mothers older than 30 years of age were 65% of the total. Multiparous women were 36.4% of the total. In 67.8% of cases, mothers had a high grade of education (III: high school, or IV: degree); 27.1% were housewives or unemployed; and 2 mothers were still students. The professional status of the mothers is shown in Table 3.

Professional classification of the mothers (Italian Statistical Institute)

Influences on Weaning Pattern

The introduction of solid foods began at 22.2 ± 0.4 weeks (mean ± SE) from birth, corresponding to 15.1 ± 0.39 weeks from term. In 60.9% of cases, weaning was started before the fourth month of corrected age (<17th week from term). Female infants were significantly lighter both at birth and at weaning, whereas no differences were detected for weaning age between sexes (Table 4).

Influence of sex, GA, and BW on weaning schedule

Infants born with a higher grade of prematurity (GA <33 weeks) were weaned significantly later considering the chronological age (24.1 ± 0.74 vs 21.2 ± 0.44 weeks, P < 0.001) but significantly earlier in relation to corrected age (13.5 ± 0.72 vs 15.9 ± 0.44 weeks, P < 0.005). Infants with a GA <33 weeks were also significantly lighter at weaning (5553 ± 151 g vs 6482 ± 105 g; P < 0.0001) (Table 4).

There was a direct relationship between weight at birth and weight at weaning (r = 0.55; P < 0.001) and an inverse relationship with weaning age from birth (r = −0.40; P < 0.001) (Figs. 1 and 2). As a matter of fact, ELBW and VLBW infants were older but lighter at weaning than were LBW or normal birth weight infants (Table 4). Considering age from term, ELBW infants tended to be weaned earlier, but that fact was without statistical significance (Table 4).

FIG. 1
FIG. 1:
Correlation between BW and weaning weight.
FIG. 2
FIG. 2:
Correlation between BW and weaning age from birth.

The SGA infants were significantly lighter both at birth and at weaning (5381 ± 208 vs 6338 ± 97 g; P < 0.0001), even if weaning age was higher both from birth and from term (Table 4). The type of milk feeding did not play a significant role in the introduction of solid foods, even if formula-fed infants (73.7%) were weaned earlier (14.6 weeks) than were infants receiving mixed milk feeding (16.1 weeks) or breast milk (16.6 weeks), considering age from term. At the time of introduction of solid foods, formula-fed infants weighed significantly less than did infants receiving mixed diet or breast milk (Table 5).

Influence of milk feeding on weaning schedule

Twenty-seven infants (18.1%) were weaned at a weight <5 kg. Infants weighing less at weaning were those with lower GA and BW (Table 6). In addition, they were younger at weaning when age from term was considered (Table 6).

Factors associated with weaning weight

Maternal age significantly influenced the age at weaning in our cohort of infants: older mothers weaned their infants later (Table 7). By contrast, the influence of maternal level of education and professional status on weaning pattern was not significant, although highly educated (IV) and high-income mothers (A) tended to wean later (considering corrected age) and when the infant was at a higher weight (Table 7).

Influence of maternal factors on weaning schedule

The first solid food was represented in 46.8% of infants by mashed fruit and in 29.1% by rice in a vegetable soup. Meat was offered as part of the first meal to only 9.5% of infants; it was generally introduced 5 weeks after the beginning of weaning. Gluten was added on average after 7 weeks (GA ≥33 weeks) and 8 weeks (GA <33 weeks). In the 2 infants with a positive family history of celiac disease, weaning, and in particular the introduction of gluten, was not delayed.


From the comparison of our findings with those of Norris et al (3), some differences emerge. This might indicate that different environmental conditions and differences in food tradition and in eating habits play a role also in the weaning pattern of preterm infants. In particular, the percentage of infants who were weaned before the fourth month of age was significantly lower in our series. In fact, in our cohort, weaning began before the fourth month in only 6.5% of the infants (vs 49%) considering age from birth, and in 60.9% (vs 95%) considering age from term. The introduction of solid foods began on average at 22 weeks from birth, corresponding to 15 weeks from term; thus, our infants received the first solid food 5 weeks later than did British infants (3). In comparison with the study data reported by Marriot et al (6), who analyzed the effects of early weaning, the introduction of solid foods in our population seemed to be particularly delayed. At any rate, from our perspective, in view of the fact that the optimal time of introducing solid foods is still undefined and uncertain, 4 to 6 months seems quite reasonable.

Weight, more than age (from both birth and term), seems to influence the introduction of solid foods in our population. Infants born SGA were, in fact, weaned almost 1 month later than were not-SGA infants, evidently only when their weight was considered adequate by their mothers or caregivers.

In accordance with the data found in the United Kingdom, older mothers tended to wean later, whereas the educational and professional status of Italian mothers marginally influenced the weaning schedule. Another interesting aspect of our survey was that no consistent effect of the type of milk feeding on weaning age was evident. Formula-fed infants were weaned 2 weeks before breast-fed infants, considering only age from term, but this was not statistically significant. Moreover, this little difference can be ascribed to the fact that formula-fed infants were more likely to have been born more prematurely. For the same reason, weight at weaning was paradoxically lower in formula-fed infants.

A serious concern emerging from our study is that in almost half of the cases, the first solid food offered to the baby was represented by mashed fruit, and the second choice was rice in a vegetable soup. Meat was offered as part of the first meal in only 9.5% of cases. Thus, the first solid food offered to most of the infants had a low energy density and negligible protein, iron, and zinc content. This approach is rather worrying because the introduction of solid foods at the right time and in the right way is a crucial event that can strongly contribute to optimize growth and neurodevelopment in preterm infants, especially those who are growth restricted, have accumulated severe nutritional deficits, and have high energy and nutrient needs. From this aspect, our data are different from those reported by Norris et al (3) and may also represent different nutritional approaches in different populations.

There is a state of uncertainty and concern over the absence of precise, specific guidelines for the introduction of complementary foods in preterm infants. The World Health Organization recommends that infants be exclusively breast-fed for at least 6 months and that solid foods be introduced after 6 months of age. These recommendations, as specifically indicated, “apply to normal, term infants,” even though they “can be used as the basis for developing recommendations on complementary feeding” for preterm, ill, or malnourished infants (7). The UK Department of Health still recommends the introduction of complementary feeding at 4 to 6 months of age, when the infant weighs at least 5 kg (8). A general lack of consensus among experts in infant nutrition, insufficient evidence, heterogeneity of studies performed, scarcity of randomized intervention trials, diversity of the populations examined, difficulties in evaluating and adjusting for socioeconomic and cultural influences, and lack of proper endorsement of the current WHO recommendations have led to the absence of evidence-based national and international recommendations for the introduction of solid foods in infants born preterm. Even in the absence of a general consensus, however, it is important that mothers receive individualized instruction and encouragement from family pediatricians and health care workers about infant feeding practices. Too often, controversial cultural patterns, wrongly perceived needs, poor knowledge of the physiological needs of the infant, and confusing information from multiple commercially oriented sources may in fact lead to the use of inadequate complementary foods at the wrong time.

As a general rule, we should try to tune intakes in accordance with the status, the development, and the needs of the single preterm infant, some of whom may benefit from the introduction of particular complementary foods earlier or later than the average population. Of course, this approach does not endorse a general attitude of early or late weaning for all preterm infants.

Given this state of the art, only future research will provide a better understanding of how these infants should be weaned. At present, in the words of Garza and Frongillo (9), it seems reasonable to say that “there is not a single optimal age for the introduction of complementary foods” in these nutritionally vulnerable infants.


1. Fanaro S, Vigi V. Weaning preterm infants: an open issue. J Pediatr Gastroenterol Nutr 2007; 45:S204–S209.
2. Fewtrell MS, Lucas A, Morgan JB. Factors associated with weaning in full term and preterm infants. Arch Dis Child Fetal Neonatal Ed 2003; 88:F296–F301.
3. Norris FJ, Larkin MS, Williams CM, et al. Factors affecting the introduction of complementary foods in the preterm infant. Eur J Clin Nutr 2002; 56:448–454.
5. Fenton TR. A new growth chart for preterm babies: Babson and Benda's chart updated with recent data and a new format. BMC Pediatr 2003; 3:13.
6. Marriott LD, Foote KD, Bishop JA, et al. Weaning preterm infants: a randomised controlled trial. Arch Dis Child Fetal Neonatal Ed 2003; 88:F302–F307.
7. PAHO/WHO, Division of Health Promotion and Protection/Food and Nutrition Program. Guiding Principles for Complementary Feeding of the Breastfed Child. Washington, DC: PAHO/WHO; 2003.
8. Department of Health. Weaning and the Weaning Diet: Report on Health and Social Subjects No. 45. London: Her Majesty's Stationery Office, 1994.
9. Garza C, Frongillo EA Jr. Infant feeding recommendations. Am J Clin Nutr 1998; 67:815–816.

Complementary foods; Preterm infant; Weaning

© 2007 Lippincott Williams & Wilkins, Inc.