QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Santos, I. Articles by Pelto, G. Search for Related Content PubMed PubMed Citation Articles by Santos, I. Articles by Pelto, G.

---

Community and International Nutrition

Nutrition Counseling Increases Weight Gain among Brazilian Children¹

Iná Santos^*,2, Cesar G. Victora^*, José Martines, Helen Gonçalves^*, Denise P. Gigante^*, Neiva J. Valle^* and Gretel Pelto^**

Departamento de Medicina Social, Faculdade de Medicina, Universidade Federal de Pelotas, Pelotas, Brazil; Department of Child and Adolescent Health and Development, World Health Organization, Geneva, Switzerland; and the Division of Nutritional Sciences, Cornell University, Ithaca, NY 14853-2801 ^{** *}

²To whom correspondence should be addressed. E-mail: inasantos{at}uol.com.br

ABSTRACT

To assess the impact on child growth of the nutrition-counseling component of the Integrated Management of Childhood Illnesses (IMCI) strategy, a randomized trial was implemented. All 28 government health centers in a Southern Brazil city were paired according to baseline nutritional indicators. One center from each pair was randomly selected and its doctors received 20-h training in nutrition counseling. Thirty-three doctors were included and 12–13 patients < 18 mo of age from each doctor were recruited. The study included testing the knowledge of doctors, observing consultations and visiting the children at home 8, 45 and 180 d after the initial consultation. Maternal knowledge, practices and adherence to nutritional recommendations were assessed, and anthropometric measurements were taken. Day-long dietary intake was evaluated on a subsample of children. Doctors in the intervention group had better knowledge of child nutrition and improved assessment and counseling practices. Maternal recall of recommendations was higher in the intervention than in the control group, as was satisfaction with the consultation. Reported use of recommended foods was also increased. Daily fat intake was higher in the intervention than in the control group; mean daily intakes of energy and zinc also tended to improve. Children 12 mo of age or older had improved weight gain and a positive but nonsignificant improvement in length. Nutrition-counseling training improved doctors’ performances, maternal practices and the diets and weight gain of children. The randomized design with blind outcome evaluation strongly supports a causal link. These results should be replicated in other settings.

KEY WORDS: • child growth • nutrition counseling • randomized trial

The World Health Organization (WHO)³ and United Nations Children’s Fund have developed a modular training course for health workers on the Integrated Management of Childhood Illnesses (IMCI), focusing on the leading causes of mortality in children < 5 y of age (acute respiratory infections, diarrhea, malaria, measles and malnutrition). Altogether, >70% of deaths in children < 5 y old are due to these five causes. Malnutrition is an associated cause of death in most cases. The course is designed to update health workers who care for children in first-level outpatient health facilities on the management of these key conditions. Rather than studying separate guidelines on these diseases, the course provides training on how to use an integrated approach in clinic. The module, "Counsel the Mother," covers nutritional counseling topics, in compliance with WHO feeding recommendations (1 ). It includes advice on frequency of breastfeeding and complementary feeding, quality of complementary foods and on active feeding behaviors. A second module, "Management of the Sick Young Infant," covers the assessment and management of breastfeeding problems (2 ). Assessing the child’s feeding, identifying feeding problems and counseling the mother about feeding problems by focusing on giving relevant advice to each mother and using good communication skills are the main objectives of these modules.

By training health workers to assess nutritional status and feeding problems, to provide appropriate feeding counseling and to give follow-up recommendations, caretaker knowledge and behavior is expected to improve. Consequently, the child’s intake of energy and other nutrients should improve, with an impact on nutritional status and morbidity. However, there are a number of limiting factors that may interfere with the success of nutrition counseling. These factors include the intensity of the intervention, the complexity of the behaviors to be changed, timing considerations, the existence of external constraints such as low food availability and the possibility of breast milk displacement by recommended complementary foods. A proper understanding of the role of local culture in this process is also essential for adapting internationally developed documents to the local environment (3 , 4 ). The general objective of the study was to test, under field conditions, the efficacy (5 ) of the nutrition counseling promoted in the IMCI course and to provide specific advice on how it might be improved.

The setting of the study was chosen to represent a situation in which weight and length faltering are present in the 1st year of life, but in which there is not an obvious lack of food, because testing the intervention where lack of food is an important problem would severely curtail its potential effectiveness.

The study was planned to test the hypotheses that nutrition-counseling training would lead to improvements in the knowledge of health care providers regarding the assessment and management of nutritional and feeding problems; the ability of health care providers to detect such problems and to provide appropriate advice to the child’s caretaker; caretakers’ understanding and recall of health care provider recommendations; caretaker’s attitudes and behaviors regarding the correct nutritional management of the child; child’s energy and nutrient intake; and child growth. Those intermediate outcomes were checked throughout the study to ensure that the expected steps that would link up the medical training with the main outcome (child growth) were warranted, an assumption to be attended by an efficacy study.

MATERIALS AND METHODS

Study design and randomization.

The study was a randomized trial using health centers as the units of randomization. All 28 municipal health centers in the city of Pelotas (Southern Brazil) were paired according to levels of malnutrition and socioeconomic indicators of the neighborhood. One center from each pair was selected for the intervention group (nutrition-counseling training) by flipping a coin. The other center joined the control group (current nutrition advice; Fig. 1 ).

View larger version (29K): [in this window] [in a new window]   Figure 1. Nutrition counseling: trial profile.

The training material "Counsel the Mother" and "Management of the Sick Young Infant" modules) was translated and adapted to local conditions according to the WHO Adaptation Guide. The module and the chart, "Counsel the Mother," contained generic feeding recommendations for children 2 mo to 5 y old. Background information on current practices available from epidemiological and ethnographic researches previously developed in the city (3 , 4 ) identified two groups of problems: those related to breastfeeding and those related to complementary feeding after 4–6 mo of age. Identified problems in breastfeeding included nonexclusive breastfeeding in first 4–6 mo of life due to use of teas, supplementary milk and/or other foods; and early cessation of breastfeeding. Identified problems in complementary feeding concerned use of low energy dense foods, food preparations (recipes) and presentation that reduce density (diluted soup), dilution of formula milk, monotonous diet and a restrict number of different foods; and diet low in iron.

Initially, an exploratory research was conducted aiming to anticipate maternal acceptance of feeding recommendations to be delivered through the trial. In-depth individual interviews were conducted at home with 30 mothers to explore current use of "papa" (high density food instead of diluted soup), chicken liver, meat and chicken, egg yolk, beans (not only the water in which beans have been cooked), green leafy vegetables, fruits and the use of cup and spoon (instead of bottle) to feed the child. A draft of feeding recommendations then was constructed to be tested through household trials (HHT).

Conduction and planning of the HHT followed the WHO protocol (Annex D) for adaptation of feeding recommendations. Four children of each age group that would be addressed by the nutrition-counseling trial (0–3.9, 4–5.9, 6–11.9 and 12–17.9 mo of age) were selected, totaling 16 children. Mothers and children were current users of health facilities and they were located through a search at child health records among those who attended a medical consultation in the previous week. The observed results of the HHT pointed out that the feeding recommendations to be delivered were culturally acceptable, feasible and affordable by mothers.

Intervention.

Based on the adaptation of the IMCI feeding guidelines, the following key feeding recommendations were identified: increase the frequency of breastfeeds and complementary feeds; give animal protein and micronutrient-rich foods (egg, chicken liver, shredded chicken and beef); add oil to food; and increase the energy and nutrient food density by giving mashed beans instead of the broth and by giving thick papa instead of soup.

A Mother’s Card summarizing key feeding recommendation was developed to assist the health care provider in the counseling process. The card was used to guide discussion with mothers and handed to them as a reminder.

The health care providers’ training program lasted 20 h, with 40% of the time spent in practical sessions at a health center not included in the study. The development of counseling skills was emphasized. The performance of all participants was formally assessed at the end of the training course to document its effectiveness.

Study groups.

The study groups included the health care providers, children attending the selected health centers and being treated by these providers and the mothers or caretakers of these children. Recruitment of children in each pair of clinics started on the same date, soon after training. The first 13 children < 18 mo of age treated by each health care provider were recruited after informed consent.

Outcome measures.

The following outcomes were compared between the two groups by observers who were blind to the intervention or control group status of health care providers, mothers and children: health care providers’ knowledge about nutrition counseling and practices (counseling skills); maternal knowledge about basic nutrition-counseling recommendations; maternal ownership of Mother’s Card and understanding of its contents; maternal satisfaction with health care; maternal compliance with recommended feeding practices (including exclusive breastfeeding for 4–6 mo, complementary feeding after 6 mo, types of complementary foods and meal frequency); child’s dietary intake; and weight and length gains.

Data collection.

Health care providers in both groups were tested on their knowledge of nutrition counseling through a questionnaire 30 d after training. Three consultations with each provider (in both groups) were observed using structured instruments, both shortly after the training course and 180 d later, to assess their counseling skills.

Home visits were undertaken 8, 45 and 180 d after the initial clinic attendance. Maternal knowledge about basic nutrition-counseling recommendations and reported compliance with the recommended feeding practices were assessed. Satisfaction with the care received was assessed in both intervention and control groups at the 8- and 45-d visit. Children were weighed naked using portable United Nations Children’s Fund scales and their supine length was measured using specially designed length boards (AHRTAG baby length measures; London, UK). Standardized procedures were used in all measurements (6 ). Dietary intake was assessed during the three home visits with a 24-h semiquantitative recall, including frequencies and approximate food portions. The frequency of consumption of recommended foods was assessed for the preceding 7 d.

For a subsample of 66 children (two per health care provider) a day-long visit was paid immediately after the second follow-up (45 d after enrollment). Child selection was conducted through simple random sampling: according to the order of entrance in the study, the first and the eighth enrolled child of each doctor were visited at home for full-day dietary assessment. All the food consumed by the child was weighed using methods developed in an earlier local study (7 ).

Using local food composition tables, results of the semiquantitative and of the food weighed intake studies were translated to energy and protein intake. Breast milk intake was assessed by the frequency and duration of feeds (8 ).

Sample size.

The units of randomization were the health centers. For sample size calculations, the significance level was set at 5% (one-tailed test) and the statistical power was set at 80%. A one-tailed test was used because nutrition-counseling training was supposed to give either more or the same effect of any training at all on the outcomes of interest, but could not give less. Two major outcomes were considered: the proportion of health care providers with adequate knowledge of nutrition counseling (for example, those who obtained a score of 80% or greater in a questionnaire on nutrition-counseling contents) and the average weight gain in a 6-month period. Concerning the former outcome, the study detected a difference between 30% (in the control group) and 75% (knowledgeable health care providers in the intervention group) (9 ). The study also detected differences between 10 and 53%; 20 and 65%; 40 and 84%; and 50 and 92%. Such large differences were expected because no specific training on nutrition counseling was ever given to health care providers and inappropriate feeding advice seemed to be very common, for example, giving tea and water to very young children, feeding fruit juices from 8 wk of age, giving diluted soups as the first complementary food, etc.

According to the National Center for Health Statistics reference (10 ), the average boy gains 2.4 kg between 6 and 12 mo of age, and the average girl gains 2.3 kg. Weight gain in the first 6 mo is faster. Data on the average weight gain by health center were available from a population-based birth cohort study carried out in Pelotas in 1993 (11 ). The average weight gain for both sexes, per health center, ranged from 1.7 to 2.3 kg, showing that in most areas weight gain was below the National Center for Health Statistics reference. The standard deviation between health centers was 0.17. Fourteen health centers in each group were needed for detecting a difference in weight gain of 200 g (9 ).

Data analysis.

The intervention and control groups were compared in terms of baseline indicators (at the 8-d visit), including child nutrition and socioeconomic variables (9 ). Any differences between the two groups that had occurred despite the randomization were adjusted for in the analyses. As previously defined, the analyses of the effect of the intervention on final outcomes were first conducted with the whole sample and then explored among the specific age groups.

To ensure that the interdependence between children attending the same health center was taken into account, all analyses were carried out using multilevel modeling (12 ). The highest hierarchical level was the health center, the second, level the health care provider, and the third, the mothers or children. Because only four health centers had more than one doctor, the effect of the health center level was negligible. The multilevel modeling, thus, considered only the doctor and the mother/child levels. Analyses were completed on an intention-to-treat basis.

Quality control.

Quality control measures included strict blinding of interviewers and of the field supervisor regarding the intervention or control status of health care providers, mothers and children; pretested, standardized data collection forms and detailed interviewer guides were used; interviewers were carefully selected and evaluated; thorough training on interviewing and anthropometric measurements, followed by standardization sessions with assessment of intra and interobserver variability, were performed; and 10% of all interviews were repeated by a supervisor.

Ethics.

The study complied with the International Ethical Guidelines for Biomedical Research Involving Human Subjects (13 ). Children in the control group continued receiving routine nutrition advice and the research procedures involved minimal risk. Individual informed consent was obtained from the parents of children selected to the study during the first home visit. Children with signs of severe disease were referred to appropriate care. The study protocol was cleared by the ethical committees of the Federal University of Pelotas and WHO.

RESULTS

Sample.

In the 28 health centers included in the study, 33 doctors were involved with child care: 17 in the intervention and 16 in the control group. Child enrollment varied according to patient flow at each health facility. After 45 d of enrollment, one doctor in the intervention group had assisted only 10 children and two doctors from the control group had assisted only 12 children each. A total of 424 children were included in the study (218 in the intervention and 206 in the control group).

Two children were lost at the first home interview (8 d postcounseling), both from the intervention group (Fig. 1) , because their parents refused to participate. At the second follow-up (45 d postcounseling), 12 children were lost (seven from the intervention and five from the control group). There were three refusals at the second follow-up and three refusals at the third. Two of the refusals at each follow-up were of children from the intervention group. Two children were hospitalized by the time of the second follow-up: one from each study groups. At the third follow-up, only one child was hospitalized (from the intervention group). Hospitalized children were excluded from the corresponding follow-up. The main cause of losing study participants was change of address to a different city (7 of 12 at the second follow-up and 15 of 20 at the third follow-up). One child from the control group died during the study.

Table 1 shows the baseline characteristics of the sample. The average age of the children in the intervention and control groups was similar: 7.9 ± 5.4 and 7.5 ± 5.3 mo, respectively. Despite paired randomization, children from the intervention group had lower family income, social class and maternal schooling. These three variables were included as confounding factors in the analyses wherever they were also associated (P < 0.20) with the outcome of interest. There were no differences between the two groups in sex, age and household size, maternal age, employment or marital status. The nutritional status of the children was very similar between groups (Table 1) , except for the mean weight-for-age Z-score that was significantly higher in the control group, when all children were considered together.

Doctors from both groups (17 from the intervention and 16 from the control group) were tested on nutritional counseling for children < 2 y of age. Questions were based on practical situations arising at the primary care level and covered in the IMCI guidelines. In a total of 77 items, doctors from the intervention group, on average, correctly answered 83% (95% CI: 65–100%; minimum of 47 and maximum of 71). After the exclusion of one returned blank test, the mean number of correct answers by doctors from the control group was of 68% (95% CI: 48–88%; range: 41–65 points; P = 0.02).

Medical doctors’ performance immediately after and 180 d after training.

Immediately after training, a total of 104 medical encounters were observed in both groups (54 in the intervention and 50 in the control group). The mean age of the children attending these consultations was not statistically different between groups: 6.2 ± 4.6 mo in the intervention group and 6.9 ± 5.1 mo in the control group. Nearly 40% of children in both groups were girls. More than 40% of the health care contacts were well-baby visits. Return for follow-up visits accounted for 26.1% of the medical contacts. Complaints of acute respiratory infections, skin problems and diarrhea accounted for 13.5, 9.9 and 6.3% of the consultations, respectively.

During this observation, mothers from the intervention group had their feeding practices assessed more frequently than those in the control group (Table 2). Mothers were equally encouraged to talk by doctors in the intervention and in the control groups. However, doctors behaved differently in terms of use of good communication skills (asking and listening, positively reinforcing mothers and checking their proper understanding of the counseling delivered), providing the intervention group a clear advantage (Table 2) .

Table 3 compares the recommendations made by the intervention and control doctors on both occasions (immediately after and 180 d after training). In more than two-thirds of the initial medical encounters observed in the intervention group (68.6%) and in 74.0% of those in the group control, some type of nutritional counseling—adequate or not—was delivered. Doctors in the intervention group were significantly more likely to make appropriate recommendations, whereas the reverse was observed for the recommendations regarded as inadequate (P < 0.001). As for assessment, doctors’ performances tended to worsen over time, although the number of consultations in which only adequate advice was delivered during the first observation (67.6%) was not statistically different (P = 0.43) from the number noticed during the second observation (74.3%).

    First follow-up. Maternal satisfaction with the care provided was higher in the intervention than in the control group (65.0% vs. 53.4%; P = 0.007). As reported by mothers, doctors in the intervention group were more likely to praise the mother when she reported an appropriate feeding behavior (28.9% vs. 17.7%; P = 0.01).

During the home interview conducted 8 d after consultation, 64.4% of the 218 mothers from the intervention group and 45.1% of the 206 mothers from the control group recalled receiving some type of nutritional advice. In the intervention group, 31.1% of the mothers of children < 4 mo of age who still breastfed recalled that the doctor recommended stopping giving teas in favor of exclusive breastfeeding. In the control group, only 10.9% of breastfeeding mothers recalled having received that recommendation (P = 0.02). Table 4 shows that mothers in the intervention group were significantly more likely to recall IMCI recommendations than those in the control group.

As observed during the first follow-up, among children > 4 mo of age, all the nutritional recommendations were more frequently recalled by mothers in the intervention group than from mothers in the control group. After allowing for potential confounders, the effect of the intervention on maternal recall did not change.

    Third follow-up (180 d after consultation). Six months after consultation, only 194 mothers remembered having received some recommendation during the initial visit: maternal recall of nutritional recommendations received during the initial visit, nonetheless, remained higher in the intervention group for most of the recommendations.

Impact on child diet

    Maternal compliance with feeding recommendations. As reported, during the 8-day postconsultation visit, 121 of the 187 children < 6 mo of age were still breastfeeding: 57 from the intervention and 64 from the control group. The mean number of feeds was similar in both groups (9.7 and 10.4 in the intervention and control groups, respectively). In the intervention group, 64.9% of the children were breastfed at least 8 times in 24 h. In the control group, 71.9% of the children were breastfed at least 8 times in 24 h (P = 0.2).

Between the first and second follow-ups, 25% of those in the intervention group and 17% of those in the control group had been weaned (P = 0.3). Among those still breastfeeding, the proportion of children who were breastfed at least 8 times in 24 h was similar in both groups (51.2 and 58.5% in the intervention and control groups, respectively; P = 0.3). The mean number of feeds per day was 7.7 in the intervention group and 8.4 in the control group (P = 0.4).

Six months after enrollment (at the third follow-up) 93 of 424 children were still breastfed: 45 from the intervention group (21.5%) and 48 from the control group (24.6%). After the exclusion of two mothers from the control group whose children breastfed all the time, the mean time at breast was of 97.9 ± 79.8 min/d and 113.7 ± 87.3 min/d in the intervention and control groups, respectively (P = 0.3).

During the 8th-d visit, the reported use of the key recommended foods and feeding practices was higher in the intervention than in the control group: increased food density (14.6% vs. 3.4%; P < 0.001), giving egg yolk (19.2% vs. 8.2%; P < 0.01), shredded chicken and beef (15.5% vs. 6.3%; P < 0.01), chicken liver (20.5% vs. 6.8%; P < 0.001), and oil, margarine or butter (16.9% vs. 0.5%; P < 0.001). At this first household visit, mothers were also asked to report what types of foods the child was given in the preceding week. No information was collected on food quantities at this stage. Both groups reported similar food types and the only significant difference was in adding oil, margarine or butter, which was reported by 24.2% of mothers in the intervention group and 4.4% in the control group (P < 0.001).

    Child’s dietary intake. Taking all the age groups together, 8 d after consultation, the mean energy intake on the day before the interview (as obtained through the 24-h recall) was similar in both groups (3977.06 ± 1737.24 kJ in the intervention group and 3761.62 ± 1509.13 kJ in the control group). For protein, mean intakes were 36.61 ± 23.78 g and 35.13 ± 22.12 g in the intervention and control groups, respectively. These differences were not statistically significant. Daily fat intake, however, was significantly higher in the intervention group (34.08 ± 14.86 g) than in the control group (31.23 ± 11.46 g).

Through the weighed observations over a full-day period on the subsample, mean daily intake of energy and macronutrients for both the intervention and control groups were lower than those obtained through the 24-h recall of the second follow-up (the nearest information gathered in time). In the intervention group, however, daily intake of energy and macronutrients was higher than those observed in the control group. Those in the intervention group consumed 280.7 kJ, 4.3 g of protein and 2.0 g of fat more per day than those in the control group (Table 5 ). There were also favorable differences in carbohydrate and high biologic value proteins in the intervention group. Zinc intake was 10% higher in the intervention group. None of the above-mentioned differences were statistically significant. The amount of food and the energy density of foods consumed were higher in the intervention group.

DISCUSSION

This trial showed that nutrition counseling delivered in outpatient facilities after training was associated with improved child weight gain. This effect was observed even under unfavorable conditions, because children in the intervention group belonged to lower social strata than did control children, despite paired randomization. Another limitation was that nutrition advice in the control group was found to be surprisingly adequate, a fact that is possibly due to the high level of attention given to child health and nutrition in a city with a tradition for research in this field.

The study design allowed the measurement of the sequence of events that led to improved nutrition. The generic nutrition counseling guidelines were carefully adapted to the local environment and household trials allowed the identification of acceptable, affordable and feasible recommended foods. After training, doctors in the intervention group had significantly better knowledge of child nutrition issues. Observation of consultations soon after training showed that the intervention group doctors had improved performance both in terms of assessing the child’s nutrition as well as in terms of communication skills. Six months later, their performance was not as good as immediately after training but was still significantly better than in the control group. Maternal recall of key nutritional recommendations and satisfaction with the quality of care were significantly higher in the intervention group than in the control group. The reported use of recommended foods was significantly higher in the intervention group. This reported adherence to feeding recommendations was consistent with the information provided through 24-h recall and full-day observations in a subsample of children. Daily fat intake as assessed by the 24-h recall was higher in the intervention group than in the control group, particularly by children ages 6–11 mo. Probably because of the small number of children, the full-day study did not show statistically significant differences in mean daily intakes of energy and zinc between the groups. However, these intakes—as well as the amount of food, its energy density and the intake of proteins of high biological value—tended to be higher in the intervention group. According to Butte (14 ), infants 9–12 mo of age require an estimated 146.4 kJ (boys) or 150.6 kJ (girls) a day for growth. The observed difference between the two groups in the present study (280.7 kJ/d), therefore, could possibly explain the improvement in growth. Due to the small number of children included in the full-day study, it was not possible to break down these results by age. Children from the intervention group grew faster, particularly those who entered the study after 1 y of age. Allowing for confounding, 6 mo after enrollment children in the intervention group presented significantly greater gains in weight, in weight-for-age Z-score and in weight-for-height Z-score.

These results show that nutrition-counseling training improved doctors’ performances, maternal practices and some aspects of the diets and growth of children. The fact that this was a randomized study with blind evaluation of the outcomes strongly supports a causal link. Due to the characteristics of the study, it was not possible to perform a double-blind trial.

Dietary outcomes are notoriously difficult to measure with precision; thus, it is not surprising that significant differences were detected for some but not all outcomes. Nutrition counseling may have improved diets in ways that were not directly measured in this study, for example, mothers may become more aware of the importance of nutrition in general, including the role of dietary diversity and of appropriate feeding behaviors.

The fact that improvements in weight gain were observed only for children who entered the study at 6 mo of age or old, and particularly for 1-y-olds, may be explained as follows. For infants < 6 mo of age, promoting breastfeeding is a major component of nutrition counseling. In Pelotas, children who stop breastfeeding between 3 and 6 mo of age tend to gain weight and length more rapidly than those who continue to be breastfed; therefore, nutrition counseling that focuses on the promotion of exclusive breastfeeding is unlikely to promote faster growth in this age range (15 ). However, at older ages complementary foods play a larger role, and nutrition counseling that focuses on complementary feeding can be expected to improve growth. Also, because > 40% of the entire sample was composed of children < 6 y of age, the lack of effect of the intervention on children from that age group may explain the lack of effect of the intervention at the pooled analyses. Another possible reason is that, although most children undergoing the intervention were seen on more than one occasion during the study, the content of nutrition counseling delivered to the oldest child would be adequate for a longer period (even if the child did not attend other medical consultations) than the nutrition counseling delivered to the youngest child, which requires modification as the child grows.

The lack of a significant effect on linear growth may be due to the relatively short duration of the trial. Because weight responds faster to intervention than length (16 ), a longer period might have been required to observe detectable differences.

The greater gains in weight than in height raised the possibility that the intervention might be leading to overweight children. By the time of the third follow-up only 20 children were above the +2 SD of weight/age Z-score (8 from the intervention group and 12 from the control group). For those children who entered the study 1 y of age or older, eight were above the +2 SD weight/age Z-score, and only three of these children belonged to the intervention group, showing that consequent obesity should not be a matter of concern.

Improvement of the nutrition-counseling efficacy promoted by the IMCI could be achieved through the maintenance of doctors qualified performance. The fact that doctors’ performance in the intervention group worsened relative to the first assessment is consistent with the results observed in another intervention study designed to improve counseling on breastfeeding (17 ). The congruence of these data suggests that supervision and continuing education of health care providers are needed as part of the training strategy of the IMCI.

It is important that these results be replicated in other settings, particularly where there are higher prevalences of malnutrition, the duration of breastfeeding is longer, maternal education lower, food availability more likely to be a constraint and where the baseline performance levels of health care workers is less adequate than in Pelotas. Because in our study only 41.2% of the children were breastfed at the time of enrollment, it would also be important to ensure that future trials have enough power to examine the effects of counseling on breastfeeding practices when provided to mothers of very young children.

FOOTNOTES

¹ Supported by the World Health Organization Department of Child and Adolescent Health.

³ Abbreviations used: HHT, household trials; IMCI, Integrated Management of Childhood Illnesses; WHO, World Health Organization.

Manuscript received 5 July 2001. Initial review completed 15 June 2001. Revision accepted 26 July 2001.

LITERATURE CITED

1. World Health Organization (1995) Management of Childhood Illness: Counsel the Mother. WHO/CDR/95.14.E. 1995 World Health Organization, Division of Diarrhoeal and Acute Respiratory Disease Control Geneva, Switzerland .

2. World Health Organization (1995) Management of Childhood Illness: Management of the Sick Young Infant Age 1 Week up to 2 Months. WHO/ Division of Diarrhoeal and Acute Respiratory Disease Control/95.14.F. 1995 World Health Organization, CDR Geneva, Switzerland .

3. Behague, D. (1992) A Contextual Interpretation of the Insufficient Milk Syndrome in Pelotas, Brazil: The Interplay of Economic, Political and Socio-cultural Constraints on Mothers’ Decisions Concerning their Breastfeeding Practices 1992 MA dissertation Bryn Mawr University, Bryn Mawr, PA. .

4. Nash, C. L. (1997) The Role of Culture in Decision-Making: Infant Feeding in Pelotas, Brazil 1997 PhD dissertation Johns Hopkins University, Baltimore, MD. .

5. Habicht, J. P., Mason, J. B. & Tabatabai, H. (1984) Basic concepts for the design of evaluation during program implementation. Sahn, D. E. Lockwood, R. Scrimshaw, N. S. eds. Methods for the Evaluation of the Impact of Food and Nutrition Programs 1984 United Nations University Tokyo, Japan. .

6. Cameron, N. (1984) The Measurement of Human Growth 1984 Croom Helm Ltd Beckenham, UK. .

7. Olinto, M.T.A., Victora, C., Barros, F. C. & Gigante, D. (1995) Validity of 24-hour recall according to the child nutritional status. J. Nutr. 25:880-884.

8. World Health Organization (1998) Complementary Feeding in Developing Countries: A Review of Current Scientific Knowledge. WHO/NOT/98.1. 1998 World Health Organization Geneva, Switzerland .

9. Kirkwood, B. R. (1988) Essentials of Medical Statistics 1988 Blackwell Oxford, UK. .

10. Hamill, P.V.V., Hamill, P.V.V., Drizid, T. A., Johnson, C. L., Reed, R. B., Roche, A. F. & Moore, W. M. (1977) NCHS Growth Curves for Children Birth to 18 Years. Vital and Health Statistics Series 11:165. DHEW Publication No. (PHS) 78-1650 1977 USHEW, Public Health Service Hyattsville, MD .

11. Victora, C. G., Barros, F. C., Halpern, R., Menezes, A.M.B., Horta, B. L., Tomasi, E., Weiderpass, E., César, J. A., Olinto, M. T., Guimarães, P.R.V., Garcia, M. M. & Vaughan, J. P. (1996) Longitudinal study of the mother and child population in an urban region of Southern Brazil, 1993: methodological aspects and preliminary results. J. Public Health 30:34-45.

12. Council for International Organizations of Medical Sciences, World Health Organization (1993) International Ethical Guidelines for Biomedical Research Involving Human Subjects 1993 Council for International Organizations of Medical Sciences World Health Organization, Geneva, Switzerland. .

13. Goldstein, H. (1995) Multilevel Statistical Models 1995 Edward Arnorld London, UK. .

14. Butte, N. F. (1996) Energy requirements of infants. Eur. J. Clin. Nutr. 50(suppl. 1):S24-S36.

15. Victora, C. G., Morris, S. S., Barros, F. C., Horta, B. L., Weiderpass, E. & Tomasi, E. (1998) Breastfeeding and growth in Brazilian infants. Am. J. Clin. Nutr. 67:452-458.[Abstract]

16. World Health Organization, Expert Committee on Nutrition (1995) Physical Status: Uses and Interpretation of Anthropometry 1995 WHO Technical Report Series No. 854. World Health Organization, Geneva, Switzerland. .

17. Rea, M. F., Venancio, S. I., Martines, J. C. & Savage, F. (1999) Counseling on breastfeeding: assessing knowledge and skills. Bull. World Health Organ. 77:492-498.[Medline]

This article has been cited by other articles:

				H. Saloojee, T. De Maayer, M. L. Garenne, and K. Kahn What's new? Investigating risk factors for severe childhood malnutrition in a high HIV prevalence South African setting1 Scand J Public Health, May 1, 2007; 35(3): 96 - 106. [Abstract] [PDF]

				J. L. Leroy, J.-P. Habicht, G. Pelto, and S. M. Bertozzi Current Priorities in Health Research Funding and Lack of Impact on the Number of Child Deaths per Year Am J Public Health, February 1, 2007; 97(2): 219 - 223. [Abstract] [Full Text] [PDF]

				R. C. Robert, J. Gittelsohn, H. M. Creed-Kanashiro, M. E. Penny, L. E. Caulfield, M. R. Narro, and R. E. Black Process Evaluation Determines the Pathway of Success for a Health Center-Delivered, Nutrition Education Intervention for Infants in Trujillo, Peru J. Nutr., March 1, 2006; 136(3): 634 - 641. [Abstract] [Full Text] [PDF]

				N. Bhandari, S. Mazumder, R. Bahl, J. Martines, R. E Black, M. K Bhan, and other members of the infant feeding study group Use of multiple opportunities for improving feeding practices in under-twos within child health programmes Health Policy Plan., September 1, 2005; 20(5): 328 - 336. [Abstract] [Full Text] [PDF]

				N. Bhandari, S. Mazumder, R. Bahl, J. Martines, R. E. Black, and M. K. Bhan An Educational Intervention to Promote Appropriate Complementary Feeding Practices and Physical Growth in Infants and Young Children in Rural Haryana, India J. Nutr., September 1, 2004; 134(9): 2342 - 2348. [Abstract] [Full Text] [PDF]

				P. Mohan, S. D Iyengar, J. Martines, S. Cousens, and K. Sen Impact of counselling on careseeking behaviour in families with sick children: cluster randomised trial in rural India BMJ, July 31, 2004; 329(7460): 266. [Abstract] [Full Text] [PDF]

				A. Wagstaff, F. Bustreo, J. Bryce, M. Claeson, and the Who-World Bank Child Health and Poverty Workin Child Health: Reaching the Poor Am J Public Health, May 1, 2004; 94(5): 726 - 736. [Abstract] [Full Text] [PDF]

				C. G. Victora, J.-P. Habicht, and J. Bryce Evidence-Based Public Health: Moving Beyond Randomized Trials Am J Public Health, March 1, 2004; 94(3): 400 - 405. [Abstract] [Full Text] [PDF]

				G. H. Pelto, I. Santos, H. Goncalves, C. Victora, J. Martines, and J.-P. Habicht Nutrition Counseling Training Changes Physician Behavior and Improves Caregiver Knowledge Acquisition J. Nutr., February 1, 2004; 134(2): 357 - 362. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Santos, I. Articles by Pelto, G. Search for Related Content PubMed PubMed Citation Articles by Santos, I. Articles by Pelto, G.