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 Ntab, B. Articles by Simondon, F. Search for Related Content PubMed PubMed Citation Articles by Ntab, B. Articles by Simondon, F.

---

Community and International Nutrition

A Young Child Feeding Index Is Not Associated with Either Height-for-Age or Height Velocity in Rural Senegalese Children¹

Balthazar Ntab^*, Kirsten B. Simondon^*,2, Jacqueline Milet^*, Badara Cissé^*,,**, Cheikh Sokhna, Denis Boulanger^* and François Simondon^*

^* IRD (Institut de Recherche pour le Développement), Epidemiology and Prevention Research Unit (UR 24), Montpellier, France and Dakar, Senegal; IRD, Tropical African Malaria Research Unit (UR 77), Dakar, Senegal; and ^** LSHTM (London School of Hygiene and Tropical Medicine), London, UK

²To whom correspondence should be addressed. E-mail: kirsten{at}mpl.ird.fr.

	ABSTRACT

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
Ruel and Menon recently published a young child feeding index based on characteristics taken from 24-h and 7-d recalls. A strong positive association was found in 7 Latin American Demographic and Health Surveys for 12- to 36-mo-old children (1). The aim of this study was to test for associations of this index with height-for-age and linear growth in African children. Children (n = 500), aged 12–42 mo, living in a rural area of Senegal were visited in their homes in April-May, and 24-h and 7-d food recalls were conducted with their mothers. Height was measured, and measurements taken 7 mo earlier were used to compute linear growth rates. General linear models were used to adjust for potential confounders (child age and sex, maternal height, BMI, and socioeconomic status). The feeding index was not associated with either height-for-age (adjusted means: –1.01, –1.06, and –1.20 Z-scores for the 1st, 2nd, and 3rd tercile, respectively) or with linear growth (6.2, 6.0, and 6.3 cm/7 mo for the 3 terciles, respectively). Continuing breast-feeding was negatively associated with height-for-age (P < 0.05) and positively associated with linear growth (P < 0.01). Frequent consumption of fruit was positively associated with both (P = 0.059 and P = 0.027, respectively, in adjusted models), whereas food consumption from an animal source was not. In conclusion, the composite feeding index was independent of height and linear growth in these rural African children, due in part to reverse causality between breast-feeding duration and stunting.

KEY WORDS: • Africa • infant feeding • growth disorders • complementary feeding • breast-feeding

Low height-for-age Z-scores (HAZ),³ also referred to as linear growth retardation or stunting, is the main form of malnutrition among children living in poor countries. In 2000, 32.5% of preschool children living in developing countries were stunted and 9.4% were wasted (2). Although the prevalence of stunting has decreased over the last 20 y in Latin America (from 25.6% in 1980 to 12.6% in 2000), it has remained constant in West Africa (from 36.2 to 34.9%) and increased slightly in East Africa (from 46.5 to 48.1%) during the same period of time.

Small stature is recognized as a strong indicator of suboptimal development of children; it arises from a variety of factors, including a chronic deficiency in energy and/or various micronutrients because of insufficient intake or absorption and/or increased needs due to repeated infectious diseases and intestinal dysfunction (3). Controlled randomized trials showed that supplementation with some micronutrients, and zinc in particular, positively affected the linear growth of young children, especially stunted children (4). Supplementation with selected foods, particularly milk, also positively affected linear growth (5).

During recent years, considerable attention was paid to initiatives aimed at globally improving feeding of infants and young children in developing countries. Recommendations were developed concerning various aspects of complementary feeding, such as the optimal number of meals per day and age at introduction and termination of foods specifically intended for infants (6).

In this context, Ruel and co-workers conceived a composite index for a global assessment of infant and young child feeding in developing countries, based in part on these recommendations (1). As stressed by the authors, the overall quality of young child feeding is difficult to assess because of the large number of characteristics involved (e.g., type and variety of food, number of meals, continuation of breast-feeding, or hygiene during preparation). Thus, their index was created to summarize the main young child feeding practices, using a limited number of variables collected during Demographic and Health Surveys (DHS) in Latin American countries. Despite some limitations of DHS data, this approach is interesting because DHS data are available from many countries on large samples of 6- to 36-mo-old children. The index could thus be used for comparisons across countries and for monitoring changes over time within a given country.

The composite index was tested for association with child nutrition status using data from 7 Demographic and Health surveys conducted in 5 Latin American countries. Height-for-age was strongly associated with the index in all 5 countries, even after adjustment for multiple variables (child age and sex, socioeconomic status, mother’s height, parity, and education, number of children < 5 y old in the household, rural or urban area). Some of these variables acted as effect modifiers, notably child age (the association was considerably stronger beyond infancy) (1).

Since the creation of this feeding index, new recommendations for complementary feeding of 6- to 23-mo-old breast-fed children were published, in particular with lower energy intake from complementary food and lower recommended numbers of daily meals than previously (i.e., 2–3 at 6–8 mo of age and 3–4 thereafter) (7). Also, a dietary diversity index was recently shown to be associated with height-for-age in 6- to 23-mo-old children in several Latin American, African, and Asian countries (8).

The aim of the present study was to apply the Ruel-Menon index to the feeding data for young children in a homogeneous rural African population and, if necessary, to adapt it to this context and to test whether it was associated with height-for-age and with height gain over the last 7 mo. Various quantitative aspects of young child feeding were also tested for association with height status and linear growth specifically for breast-fed children < 2 y old because of the current focus on this group. The study was nested into a randomized malaria prevention trial, i.e., The Intermittent Preventive malaria Treatment in children (IPTc) Study.

	SUBJECTS AND METHODS

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
Design

A cross-sectional food consumption and anthropometric survey was conducted in April-May 2003 within a sample of 543 children aged 8–42 mo and nested into the randomized IPTc trial, which included measurements of height and weight in September 2002, thus allowing for computation of height increments over the 7 mo before the survey.

Population

The study population consisted of a Sereer ethnic group, living in a rural area of central Senegal. Demographic and socioeconomic information was available from other research and surveillance programs.

Intervention

The IPTc trial compared malaria morbidity and nutritional status of preschool children receiving artesunate plus sulfadoxine pyromethamine with those of children receiving placebo, given at monthly intervals during the malaria transmission season (September–November). Surveillance of malaria cases was active (weekly home visits to all households) and passive (through detection of malaria cases in dispensaries). All cases detected were treated immediately by trained physicians.

Subjects

Criteria for enrolment in the present study were as follows: 1) inclusion in the IPTc trial; 2) age between 6 and 42 mo, 3) residency in 1 of the 10 villages at the time of the study, and 4) oral informed consent by mothers (or another primary caregiver: father or grandmother). Among the 1203 children 1.5–59 mo old who were eligible for the IPTc trial in September 2002, 615 fulfilled the criteria of age, survival, and residence in April 2003. Among these, 72 were on journeys with their mothers for longer periods of time, whereas 52 were present although their mothers were absent. Thus, 543 children and 491 mothers were included in the survey.

Data collection

Food consumption of the child was assessed by 1 well-trained female fieldworker using semiquantitative recalls with the child’s usual caregiver (the mother or grandmother). First, 24-h recalls assessed the number of meals consisting of complementary food during the day preceding the survey and the foods consumed [millet, rice, tubers, animal milk (either liquid or fermented), fresh fish, dried fish, eggs, poultry, other meats, vegetables, fruits, bread or crackers, other foods]. The number of days each food group had been consumed during the preceding week was also assessed. The number of meals during the preceding day was recalled, but no assessment of amount of intake was conducted.

The mother was asked whether the child was still breast-fed and, if yes, how many times the child had been breast-fed during the preceding day and night. The occurrence of bottle-feeding since birth was also investigated.

The occurrence of morbidity (diarrhea, fever, cough, runny nose, infected eyes) during the preceding week was recalled. Height (recumbent length before age 2 y) and weight measurements of children and their mothers were taken as previously described (9). Two height measurements were taken and the mean computed. The left mid-upper arm circumference was measured for identification of malnourished children. Mothers of children with a value < 12.5 cm (for a child 12 mo old) were invited to the local nutritional rehabilitation center with the child.

The following variables were taken from the database of the study area: child age and sex, maternal age and level of school education (none, primary, secondary), number of children < 5 y old in the household, and a number of potential indicators of economic status (quality of housing: type of walls and roofs and the existence of pit latrines), assets (transistor radio, petrol lamp), possession of agricultural devices (sewing machine, hoe, and cart) and of larger animals (cows, sheep, goats, horses, and donkeys).

Ethics

The IPTc study was approved by the Ethical Committees in the Senegalese Ministry of Health, the London School of Hygiene and Tropical Medicine and IRD. It was conducted in accordance with the Helsinki Declaration of 1975, as revised in 1983. For the nutrition survey in April 2003, specific oral information was given to mothers of eligible children, and their oral consent was required for inclusion.

Variables

    Ruel-Menon child feeding index. The feeding index was constructed as described by the authors (1), but only for children aged 12–42 mo (n = 500), because the number of infants was insufficient (n = 43). A score was assigned using 7 variables (current breast-feeding, past bottle-feeding, number of meals of complementary food, and number of food groups consumed on the day preceding the survey, and frequencies of consumption of animal source foods during the preceding week (1). Briefly, current breast-feeding and absence of bottle-feeding scored 1 point each, whereas 2–3 meals on the preceding day added 1 point, compared with 2 points for 4 meals or more. Consumption of 2–3 food groups [among cereals, tubers, animal milk, meat (excluding poultry), fish/eggs/poultry and "other foods"] added 1 point, whereas that of 4 groups or more added 2. Consumption of animal milk, meat, or poultry/eggs/fish on 1–3 d during the preceding week added 1 point for each food group, compared with 2 points for more frequent consumption (4–7 d). Hence, the index varied from a minimum of 0 to a maximum of 12.

A few children had missing values for frequencies of consumption of animal source foods during the preceding week because of maternal recall difficulties (4 for animal milk and 21 for fresh fish). To avoid biases arising from excessive exclusion of children with missing values, missing variables were arbitrarily set at their median value in the sample (6 d/wk for fresh fish and 0 d/wk for animal milk). Analyses conducted on the subsample of children with complete data yielded very similar results (data not shown).

    Other feeding indices. Although the study was not originally designed for this purpose, we also tested the association of height-for-age and linear growth with other indices, among breast-fed children aged 9–23 mo. One was the number of daily meals grouped in accordance with current feeding recommendations (7), i.e., 0–2, 3–4 and 5/d. Another was a 24-h dietary diversity index constructed differently (1 point for each of cereals, fish/meat/poultry/egg, milk, fruits, vegetables/leaves, peanuts, and bread/crackers), separated into 3 groups using terciles.

Thereafter, we tested a slightly modified version of a dietary diversity index based on 7-d recalls, which was recently published by Arimond and Ruel (8). The consumption of the following 6 food groups was considered: cereals and tubers, milk and milk products, other animal source foods (fish, egg, meat, or poultry), fruits, vegetables/leaves, and peanuts. As in the original study, consumption of a given food group on 3 d/wk scored 1 point, whereas less frequent consumption did not score any points. Thus, the index varied from a minimum of 0 to a maximum of 6 and was divided into 3 groups using terciles (separately for 9- to 11-mo-olds and 12- to 23-mo-olds). All breast-fed children aged 9–23 mo with complete data were included in these subanalyses.

Nutritional status and linear growth

Height-for-age was computed according to the NCHS/WHO sex- and age-specific growth reference (10), using Anthro Software (CDC/WHO, version 1). Height increments were computed as the difference between measurements taken during the survey and in September 2002, divided by the precise duration and multiplied by 7 mo (i.e., the average duration of the interval under study). No adjustment was made for the change in measuring technique at 24 mo of age (standing height instead of recumbent length).

Statistical analysis

The general linear model was used for multivariate analyses of the relation between feeding indices and height-for-age or growth rate in height. Among the 500 children 12 mo old and thus eligible for the main analyses, some were excluded due to missing values (1, 47, and 15 children for maternal schooling, maternal height, and child height/length in September 2002, respectively), leaving 437 children with complete data. The Ruel-Menon index was also tested among 12- to 23-mo-olds (n = 167). Alternative feeding indicators were tested within the subgroup of 9- to 23-mo-old breast-fed children, among whom 165 had complete data.

Height-for-age and growth in height over 7 mo were tested for linear associations with the Ruel-Menon index and with each of its components (all tested separately). Similarly, alternative indices were tested for relations with height-for-age and growth in height, as well as each of their components.

Adjustment was made first for child age alone (age was a strong confounder, especially in the relation with linear growth), and thereafter for child age, sex, and malaria study intervention group, and maternal height, BMI, schooling, and number of children < 5 y old (thereafter named "the full model"). These variables were all included into full models, even if they were not significantly associated with HAZ or the child feeding index in bivariate analyses.

Maternal education was grouped into "any" vs. "none" because only 9 women had reached secondary level, and maternal height and BMI were transformed into categorical variables. Maternal age and parity were not used because they were closely associated with the number of preschool children in the household. Several economic status indices were constructed from available indicators (results not shown). However, the material used for construction of the mother’s hut proved to be a reliable and simple indicator, and was used alone in all analyses.

Associations between 2 categorical variables were tested using the ² test, whereas Student’s t test was used for comparisons of means between 2 groups. Distributions of quantitative variables were summarized as means ± SD. All tests were considered significant for P-values < 0.05. SAS (version 8.1) was used for the analysis.

	RESULTS

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
A total of 543 children took part in the survey; 72 children were absent from their homes. The latter group tended to have younger mothers than the included children (28.8 ± 6.4 y vs. 31.1 ± 1.7 y, P = 0.012) and were also slightly younger themselves (23.2 ± 8.7 mo vs. 25.5 ± 9.1 mo, P = 0.046), but they did not differ with regard to sex, religion, place of birth, number of children < 5 y old in the household and level of maternal education.

The height-for-age of included children was –1.02 ± 1.03 Z-scores with no significant differences between sexes (–1.08 and –0.96 Z-scores for boys and girls, respectively), and a significant, negative relationship with child age, i.e., height-for-age decreased with increasing age (P < 0.01). Mean weight-for-height was –0.61 ± 0.92 Z-scores, and the prevalence of stunting and wasting was 17.4 and 6.4%, respectively. Height status was considerably better during this survey than 7 mo earlier (i.e., during the rainy season in September 2002). Indeed, at that time, height-for-age was –1.48 ± 1.09 and the prevalence of stunting was 31.8% (P < 0.0001 for differences between surveys in both variables). Malaria intervention and control groups did not differ in either height-for-age or linear growth from September 2002 to April 2003 (results not shown).

Reported morbidity rates were high; during the preceding week, 14% had experienced diarrhea, 24% fever, 27% coughing, 69% a runny nose, and 9% infected eyes (conjunctivitis).

Economic status was low, i.e., no families had access to electricity and only 3 used gas for cooking (instead of wood). Petrol lamps were found in all families and a transistor radio was owned by 50%; 20% had access to pit latrines.

    Infant and young child feeding practices. Breast-feeding was ubiquitous up to 18 mo of age and became rare above 2 y (Table 1). Among the 202 lactating women, 38 (18.8%) and 86 (42.6%) felt able to recall the number of breast feedings during the previous day and night, respectively. Response rates did not vary by child age (results not shown). Mean number of recalled breast feedings were 4.4 ± 1.9 (range: 0–10) and 2.7 ± 1.3 (range: 1–10) for day- and nighttime, respectively, and the mean number of breast feedings per 24 h was 7.5 ± 3.0 (range: 0–15; n = 32).

    Distribution of the Ruel-Menon feeding index. The child feeding index score varied from a minimum of 1 to a maximum of 9 in this sample (for a theoretical maximum of 12). A major peak was evident at a score of 5 (48% of subjects), and 82% of children had scores between 4 and 6. The index score varied significantly with age. The percentage of children with scores < 5 was 32, 8, 22, 16, and 16 for age groups 12–17.9, 18–23.9, 24–29.9, 30–35.9, and 36–42 mo, respectively (P < 0.01, Fig. 1).

View larger version (42K): [in this window] [in a new window]   FIGURE 1 Distribution of scores of the Ruel-Menon child feeding index by child age in rural Senegalese children aged 12–42 mo. Scores of this index theoretically rank from 0 to 12. The percentage of children with a score < 5 differed significantly by child age, P < 0.01.

Among 12- to 23-mo-old children, height-for-age was negatively associated with the child feeding index in age-adjusted analyses (means: –0.88, –0.77, and –1.20 Z-scores for the 1st, 2nd and 3rd terciles, respectively, P = 0.034), whereas no significant association was found with linear growth (7.6, 7.0, and 7.5 cm/7 mo for the 3 terciles, respectively, P = 0.17). Adjustment for other potential confounders did not change these results (data not shown).

    Modified child feeding indices. Several modified feeding indices were tested among breast-fed children aged 9–23 mo. The number of daily meals was not associated with height-for-age in age-adjusted analyses, even when grouping was adapted to current feeding recommendations (P > 0.10), nor was a modified 24-h dietary diversity index (Table 5).

Two components of the index tended to be positively associated with height-for-age, i.e., frequent consumption of fruit (P = 0.051) and vegetables (P = 0.052), in age-adjusted analyses, Table 5). Full adjustment led to similar results (P = 0.059 and P = 0.037, respectively). Fruit consumption was positively associated with linear growth in fully adjusted models (means: 7.9 and 8.7 cm/7 mo for rare and frequent consumptions, respectively, P = 0.027), whereas frequent consumption of vegetables had a reverse relation to linear growth (8.3 and 7.4 cm/7 mo for rare and frequent consumption, respectively, P = 0.041). Frequencies of consumption of milk, fish/meat, and cereals were not associated with either height or linear growth.

	DISCUSSION

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
This study provides data on selected qualitative and semiquantitative aspects of young child feeding in a rural West African setting, using a recently published child feeding index (1), and attempts to link them to height status and linear growth. We were unable to find any association of height-for-age or linear growth and the Ruel-Menon index in this setting.

The frequency of consumption of animal milk, eggs, meat, and poultry was extremely low among young children, as was described previously for infants (11), and the number of daily meals was low compared with recent recommendations for breast-fed children (7). Although it increased after weaning, no significant increase was noted for animal milk, eggs, or meat. However, fish was consumed almost daily. The degree to which this frequent consumption of fish contributed to the intake of key micronutrients such as zinc, iron, calcium, and phosphorus is difficult to assess without a quantitative estimation of daily intake.

There are a number of possible explanations for the discrepancy between our results and those of Ruel and Menon. First, our study was conducted in a very different setting, that is, Sahelian West Africa rather than Latin America. Clearly, feeding practices differ between continents. Notably, fish is not a common infant food in Latin America, whereas tubers are rare in Senegal. But an alternative 24-h dietary diversity index, which was constructed using locally consumed food groups and more adequate cutoff points, also was not associated with height-for-age.

Second, our sample size was moderate (500 children aged 12–42 mo, compared with 1178–4834 children aged 12–36 mo per data set in Ruel and Menon’s analysis). However, no tendency toward a positive association between the feeding index and height-for-age was noted; thus, insufficient power is unlikely to explain our results.

Third, children aged 36–42 mo were included in our study, whereas only children < 36 mo old were considered by Ruel and Menon. Height increments did not differ between 2- and 3-y-olds in this sample (Table 4), but it cannot be excluded that linear growth of older children is less sensitive to their food consumption. However, subanalyses conducted within the 12- to 23-mo-old children showed a significant reverse relation between the feeding index and height-for-age.

This reverse relation is probably explained by reverse causality between breast-feeding duration and height-for-age. Indeed, prolonged breast-feeding was associated with faster linear growth, as was shown previously in this and other populations (12–14). Despite this positive association, breast-fed children had lower mean height-for-age, because mothers in this community prolong breast-feeding for stunted children (15,16). This custom created a negative relation between height-for-age and the feeding index because breast-feeding added 1 point to the index. When weaned children were excluded from the analysis, the Ruel-Menon index was no longer associated with height-for-age, even among children < 24 mo old (results not shown).

Finally, the present study was nested in a placebo-controlled malaria prevention trial, and the assessment of feeding practices followed the malaria intervention. No difference existed in height-for-age between intervention and control groups, but both groups had a considerably better nutritional status after the intervention. This difference may be due in part to seasonal differences in linear growth rates (14), but close monitoring and early treatment of ill children during the rainy season probably had an effect as well. However, it seems unlikely that such an effect should mask a positive association between the Ruel-Menon index and height-for-age. That would occur only if children with poorer feeding practices benefited more from better health care.

Among potentially confounding variables, only maternal height was significantly associated with child height-for-age (P < 0.0001), a relation that was described consistently in previous studies (1,17,18). The absence of a relation between child height-for-age and maternal schooling seems at odds with findings from the literature (19). However, in most studies, primary schooling is not associated with a better nutritional status of offspring compared with no schooling (19,20), and only 2% of mothers in the present study had attained secondary level (middle school or high school). In addition, schooling is often not associated with child height-for-age in the poorest households, i.e., mothers need some economic power to make use of their knowledge (20,21), and this African population is overall very poor. We found no association between height-for-age and economic status.

Although we cannot formally exclude residual confounding by unmeasured economic factors, it seems very unlikely that such confounding may explain the lack of relation between the Ruel-Menon index and height-for-age. Indeed, confounding by economic factors is much more likely to create a spurious relation than to obscure an existing one because richer children tend to have better feeding practices and better height status.

Although the relation between the Ruel-Menon child feeding index and height-for-age was strong and consistent across 5 Latin American countries and resistant to adjustment for multiple maternal characteristics, the authors warned against interpreting it as causal (1). However, this interpretation deserves consideration. If current child feeding practices were causally associated with height-for-age, the effect of feeding on linear growth should be strong and sustained to affect height status.

We studied the association between the Ruel-Menon index and height increment over the preceding 7 mo. After adjustment for age, there was no association between the feeding index and height velocity. A breakdown of the feeding index showed that neither the number of meals during the preceding day, nor the frequency of consumption of any of the animal source foods was associated with linear growth; as a result, the feeding index was also not related to linear growth.

An important feature of a child feeding index is that it measures usual feeding practices. It is not known whether the previous week’s consumption is representative of long-term feeding patterns in developing countries. In the Senegalese population studied here, the consumption of meat and poultry is very rare because of its high cost; it is often reserved for religious and family feasts (baptisms, marriages). Thus, the previous week’s consumption may have limited predictive value for consumption over longer periods of time. However, meat consumption was significantly associated with indicators of socioeconomic status (housing quality and maternal education, results not shown), suggesting that 1 week’s recall may give some information about the customary intake. Conversely, a 1-week measure of the consumption of seasonally available foods such as fruits and vegetables is likely to have rather low predictive power over longer periods of time. The availability of animal milk and fish are also subject to seasonal variations in this setting. The feeding variable that most clearly separated children into 2 groups was continued breast-feeding, i.e., before weaning, all children were breast-fed frequently and daily.

In Latin America, animal milk intake during the preceding 24 h seemed to be the variable that was most systematically associated with height-for-age. Indeed, the relationship was positive and significant in all 7 data sets from 5 countries, whereas meat and fish/eggs/poultry were associated with height status in only one data set each (22). In the present study, neither fish/meat nor animal milk was associated with height status, perhaps because of low variability in their consumption across children and/or because of larger variability over time for a given child.

A dietary diversity index, constructed from food frequencies over the last week, was positively associated with height-for-age in 6- to 23-mo-old children in a number of countries, including several located in sub-Saharan Africa (Ethiopia, Rwanda, Zimbabwe) (8). However, this association was not systematic; a nonlinear relationship was found in Mali and none existed in Benin. We adapted this index to our context and data and applied it to breast-fed children 9–23 mo old. It tended to be positively associated with height-for-age (P = 0.15). The difference between the groups of lowest and highest scores was 0.34 Z-scores, which is close to the results published by Arimond and Ruel (from 0.24 Z-scores in Columbia to 0.57 Z-scores in Zimbabwe) (8).

Unexpectedly, food frequencies associated with height-for-age were not those of animal source foods (milk or fish) but rather those of fruits and vegetables. Even more surprising, the association with growth in length over the last 7 mo was positive for fruits (P < 0.05) but negative for vegetables (P < 0.05). These results require confirmation in other studies.

In conclusion, we were unable to show any association between the Ruel-Menon young child feeding index and height-for-age or linear growth rate of 12- to 42-mo-old children living in rural West Africa. Some components of the Arimond-Ruel weekly food frequency index (i.e., consumption of fruits and vegetables) tended to be positively associated with height status in younger breast-fed children, but only fruit consumption was also positively associated with recent growth in height. Thus, it is questionable to make inferences about the effect of feeding behaviors from their association with height-for-age. As is the case for the duration of breast-feeding, mothers may provide better feeding to ill or poorly growing children, thereby obscuring any positive effects on growth and nutritional status. Interventions that are successful in improving feeding practices among infants and young children in low-income countries would allow for a better assessment of the relation between dietary quality, growth, and health.

	ACKNOWLEDGMENTS

 
We acknowledge Madeleine Sene for conducting the dietary recalls and taking anthropometric measurements, Adama Marra for managing the Niakhar study area database, and Brian Greenwood and Jean-François Trape for contributing to the conception of the IPTc study.

	FOOTNOTES

 
¹ The Intermittent Preventive malaria Treatment in children (IPTc) Study was supported by the Gates Malaria Partnership and by the Institut de Recherche pour le Développement (IRD).

³ Abbreviations used: DHS, Demographic and Health Survey; HAZ, height-for-age Z-score; IPTc, Intermittent Preventive malaria Treatment in children; NCHS, National Center for Health Statistics.

Manuscript received 12 October 2004. Initial review completed 22 November 2004. Revision accepted 13 December 2004.

	LITERATURE CITED

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 

1. Ruel, M. T. & Menon, P. (2002) Child feeding practices are associated with child nutritional status in Latin America: innovative uses of the Demographic and Health Surveys. J. Nutr. 132:1180-1187.[Abstract/Free Full Text]

2. ACC/SCN (2000) Fourth Report on the World Nutrition Situation: Nutrition throughout the Life Cycle 2000:1-114 ACC Subcommittee on Nutrition Geneva, Switzerland.

3. Campbell, D. I., Elia, M. & Lunn, P. G. (2003) Growth faltering in rural Gambian infants is associated with impaired small intestinal barrier function, leading to endotoxemia and systemic inflammation. J. Nutr. 133:1332-1338.[Abstract/Free Full Text]

4. Brown, K. H., Peerson, J. M., Rivera, J. & Allen, L. H. (2002) Effect of supplemental zinc on the growth and serum zinc concentrations of prepubertal children: a meta-analysis of randomized controlled trials. Am. J. Clin. Nutr. 75:1062-1071.[Abstract/Free Full Text]

5. Bancroft, T. & Bailey, K. V. (1965) Supplementary feeding trial in New Guinea highland infants. J. Trop. Pediatr. 11:28-34.

6. Brown, K. H., Dewey, K. G. & Allen, L. H. (1998) Complementary Feeding of Infants and Young Children 1998 WHO Geneva, Switzerland.

7. Dewey, K. G. (2003) Guiding Principles for Complementary Feeding of the Breastfed Child 2003 Pan American Health Organization Washington, DC.

8. Arimond, M. & Ruel, M. T. (2004) Dietary diversity is associated with child nutritional status: evidence from 11 Demographic and Health Surveys. J. Nutr. 134:2579-2585.[Abstract/Free Full Text]

9. Simondon, K. B., Delaunay, V., Diallo, A., Elguero, E. & Simondon, F. (2003) Lactational amenorrhea is associated with child age at introduction of complementary food. A prospective cohort study in rural Senegal, W. Africa. Am. J. Clin. Nutr. 78:154-161.[Abstract/Free Full Text]

10. Hamill, P.V.V., Drizd, T. A., Johnson, C. L., Reed, R. B., Roche, A. F. & Moore, W. M. (1979) Physical growth: National Center for Health Statistics percentiles. Am. J. Clin. Nutr. 32:607-629.[Abstract/Free Full Text]

11. Simondon, K. B. & Simondon, F. (1995) Infant feeding and nutritional status: the dilemma of mothers in rural Senegal. Eur. J. Clin. Nutr. 49:179-188.[Medline]

12. Marquis, G. S., Habicht, J. P., Lanata, C. F., Black, R. E. & Rasmussen, K. M. (1997) Human milk or animal protein foods improve linear growth of Peruvian toddlers on marginal diets. Am. J. Clin. Nutr. 66:1102-1109.[Abstract/Free Full Text]

13. Onyango, A. W., Esrey, S. A. & Kramer, M. S. (1999) Continued breastfeeding and child growth in the second year of life: a prospective cohort study in western Kenya. Lancet 354:2041-2045.[Medline]

14. Simondon, K. B., Simondon, F., Costes, R., Delaunay, V. & Diallo, A. (2001) Breast-feeding is associated with improved growth in length, but not weight, in rural Senegalese toddlers. Am. J. Clin. Nutr. 73:959-967.[Abstract/Free Full Text]

15. Simondon, K. B. & Simondon, F. (1998) Mothers prolong breastfeeding of undernourished children in rural Senegal. Int. J. Epidemiol. 27:490-494.[Abstract/Free Full Text]

16. Simondon, K. B., Costes, R., Delaunay, V., Diallo, A. & Simondon, F. (2001) Children’s height, health and appetite influence mothers’ weaning decisions in rural Senegal. Int. J. Epidemiol. 30:476-481.[Abstract/Free Full Text]

17. Schmidt, M. K., Muslimatun, S., West, C. E., Schultink, W., Gross, R. & Hautvast, J.G.A.J. (2002) Nutritional status and linear growth of Indonesian infants in West Java are determined more by prenatal environment than by postnatal factors. J. Nutr. 132:2202-2207.[Abstract/Free Full Text]

18. Neumann, C. G. & Harrison, G. G. (1994) Onset and evolution of stunting in infants and children. Examples from the Human Nutrition Collaborative Research Support Program. Kenya and Egypt studies. Eur. J. Clin. Nutr. 48(suppl. 1):S90-S102.

19. Armar-Klemesu, M. A., Ruel, M. T., Maxwell, D. G., Levin, C. E. & Morris, S. S. (2000) Poor maternal schooling is the main constraint to good child care practices in Accra. J. Nutr. 130:1597-1607.[Abstract/Free Full Text]

20. Ruel, M. T., Habicht, J.-P., Pinstrup-Andersen, P. & Gröhn, Y. (1992) The mediating effect of maternal nutrition knowledge on the association between maternal schooling and child nutritional status in Lesotho. Am. J. Epidemiol. 135:904-914.[Abstract/Free Full Text]

21. Reed, B. A., Habicht, J. P. & Niameogo, C. (1996) The effects of maternal education on child nutritional status depend on socio-environmental conditions. Int. J. Epidemiol. 25:585-592.[Abstract/Free Full Text]

22. Ruel, M. T. (2003) Milk intake is associated with better growth in Latin America: evidence from the Demographic and Health Surveys. FASEB J. 17:A1199 (abs.).

This article has been cited by other articles:

				M. M Moursi, Y. Martin-Prevel, S. Eymard-Duvernay, G. Capon, S. Treche, B. Maire, and F. Delpeuch Assessment of child feeding practices using a summary index: stability over time and association with child growth in urban Madagascar Am. J. Clinical Nutrition, May 1, 2008; 87(5): 1472 - 1479. [Abstract] [Full Text] [PDF]

				B. Ntab, B. Cisse, D. Boulanger, C. Sokhna, G. Targett, J. Lines, N. Alexander, J.-F. Trape, F. Simondon, B. M. Greenwood, et al. Impact of Intermittent Preventive Anti-Malarial Treatment on the Growth and Nutritional Status of Preschool Children in Rural Senegal (West Africa) Am J Trop Med Hyg, September 1, 2007; 77(3): 411 - 417. [Abstract] [Full Text] [PDF]

				P. S. Sawadogo, Y. Martin-Prevel, M. Savy, Y. Kameli, P. Traissac, A. S. Traore, and F. Delpeuch An Infant and Child Feeding Index Is Associated with the Nutritional Status of 6- to 23-Month-Old Children in Rural Burkina Faso J. Nutr., March 1, 2006; 136(3): 656 - 663. [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 Ntab, B. Articles by Simondon, F. Search for Related Content PubMed PubMed Citation Articles by Ntab, B. Articles by Simondon, F.