Print version ISSN 1415-790X
Rev. bras. epidemiol. vol.12 n.2 São Paulo Jun. 2009
Rita Maria Monteiro GoulartI; Ivan França JúniorII; Maria de Fátima Marinho de SouzaIII
- Universidade São Judas Tadeu / USCS - Universidade Municipal de São
Caetano do Sul
IIDepartamento de Saúde Materno-Infantil da Faculdade de Saúde Pública da USP
IIIOrganização Pan-Americana da Saúde, Washington, DC, USA
To identify the factors associated with the nutritional recovery of children
enrolled in the Incentive Program to Combat Malnutrition (ICCN) in the city
of Mogi das Cruzes.
METHODS: An institutional trial without a concomitant control group was conducted with 570 children enrolled in the ICCN, who were observed from July 1999 to July 2001. The nutritional status was evaluated by weight/age scores, considering children with Z score > - 1 as well-nourished children; those with Z score > - 2 and < - 1 as children at risk; those with Z score < - 2 and > - 3 as moderately malnourished, and, finally, those with Z score < - 3 as severely malnourished. The impact of ICCN was analyzed in a multivariate model with the use of estimating equations (GEE - Generalized Estimating Equations) considering p<0.05 as significant.
RESULTS/CONCLUSION: Children's nutritional improvement at the end of the segment shows that the higher the initial nutritional deficiency, the greater the recovery. There were gains in height of 1.12, 0.82, 0.57 and 0.45 standard deviations for severely malnourished, moderately malnourished, at risk for malnutrition, and well-nourished children, respectively. The factors that showed association with the nutritional evolution of malnourished children were 12 to 24 months of age when joining ICCN, birth-weight equal to or greater than 3 kg, and breast-feeding. Factors negatively associated to nutritional recovery were mother's age between 20 to 40 years old and absence of maternal work outside the home. For children at risk, family income was also associated to better nutritional evolution. The ICCN experience in Mogi das Cruzes city suggests that dietary supplementation programs have a relevant role in the nutritional recovery of malnourished individuals.
Keywords: Supplementary feeding. Malnutrition and evaluation.
The macro scenario of child malnutrition in Brazil, evidenced by three population-based cross-sectional studies - National Study on Household Spending (ENDEF) 1974-1975, National Survey on Health and Nutrition (PNSN) 1989, and the National Demographic and Health Survey (PNDS) 1996 - is characterized by a descending trend in the prevalence of the event1. In 1974, prevalence of low weight for the age was 18.4%; it went down to 7.0% in 1989, and reached the percentage of 5.7% in 1996 2.
Recently, the National Survey on Women's and Children's Demographics and Health (PNDS, 2006) showed the same trend of previous studies. The frequency of weight-for-age deficits estimated for Brazilian children less than 5 years of age was 1.7%, therefore, not above expectations based on the reference distribution of healthy, well-nourished children. However, a distribution of the weight-for-age rate close to that expected in well-nourished populations does not necessarily indicate absence of nutritional problems in the population. Indeed, considering the deviation from normality shown for the height-to-age distribution (7.0%) we can conclude that the observation of the weight-for-age rate distribution alone gives us little information about the nutritional status of the Brazilian child population 3.
Decrease of malnutrition in Brazil along recent years was influenced by the improvement of several conditions such as: moderate progresses in family income, expansion of public sanitation, health and education services, reduction in fertility and birth rates, with less numerous families and, finally, growing population movement to the country's urban areas where, as a whole, there are better job and income opportunities 4.
Despite this improvement in children's nutritional condition, child malnutrition continues to be a public health problem, due to its magnitude and negative consequences 5,6, which have kept it as a priority in the government agenda.
Among governmental actions, supplemental feeding is one of the leading strategies to fight nutritional deficiencies, particularly child malnutrition. Brazil has been conducting supplemental feeding programs to malnourished children since the 1940's. However, few assessments have been conducted on the results of these programs among the beneficiary population, since there is not a continuous assessment process of supplemental feeding programs. The few assessments available are conducted by researchers that are interested in the theme but they are neither connected nor systematized in program planning and execution 7.
Assessments are rare elsewhere as well. An analysis of 140 nutritional intervention programs conducted in several countries showed that only 23% had information regarding nutritional impact 8.
Along the past years, some researchers have focused on result assessments and shown the positive effects of supplemental feeding programs developed in different regions of the country 9,10,11,12,13.
However, discussion about the effectiveness of supplemental feeding programs is far from reaching consensus,14 with criticism that goes from aspects of implementation and logistics difficulties in operation to the results reached as the strategy to fight malnutrition.
We understand that assessing programs and policies, when done with technical and technological rigor, enable greater technical and managerial rationality and expand the chances of providing more benefits to population health 15.16.
According to Barreto (2004),15 the construction of a culture to assess public policies may turn the State into a growing consumer of the results of this scientific activity, enabling adjustment and review of the strategies adopted, thus making policies more effective.
In 1999, the Ministry of Health (MS), by means of the ICCN, proposed a set of actions that included supplemental feeding. The initiatives designed to prevent and recover malnourished children combined feeding and nutrition actions and basic health actions. In addition to being assured food distribution, children should be followed up, with monitoring of growth and assessment of nutritional status.17
To follow up program performance, the Ministry of Health established that assessment of nutritional status should be monthly entered in its Food and Nutritional Surveillance System (SISVAN). The city of Mogi das Cruzes joined ICCN and implemented, as of July 1999, the Incentive Program to Combat Malnutrition.
The objective of this study was to identify the factors associated to nutritional recovery among children, in different categories of nutritional status, in the city of Mogi das Cruzes.
The city of Mogi das Cruzes is located 53 km away from the capital. Its population in 2000 was 372,419 inhabitants, with 79,399 permanent private homes and an average of 3.9 members per household. Compared to the national average, Mogi das Cruzes showed high HDI, with 0.81 score 18. However, child mortality was above (21.5) average in the State of São Paulo (17.0)19. With regard to health services, the city had, when the study started, 20 Basic Health Units (UBS) in addition to 2 public hospitals, 3 private hospitals and 1 philanthropic hospital with 503 beds and 4 ICUs.
The intervention assessed was ICCN. The population studied comprised the universe of children seen in UBS (20 units) and enrolled in ICCN between July 1999 to July 2001.
Program enrollment complied with the enrollment criteria defined by the Ministry of Health: from 6 to 24 months of age, in that malnourished children were considered those with weight-for-age below the 3rd percentile and children at nutritional risk were considered those with weight-for-age between the 10th and 3rd percentiles (inclusive). The program also provided for the enrollment of children who in three successive follow-up visits showed weight loss or no weight gain. According to the ICCN, all children should remain in the program for 12 months17.
At a central location, children were weighed and measured again following standardized procedures20.
The supplemental feeding provided included 3.5 kg of whole powdered milk a month (7 cans) and one can of soybean oil per child/month. The energy intake value, in calories, of the powdered milk and oil, reached 580 Kcal/day and protein content was 28.8 g/day, if fully consumed by children at the recommended dilution and quantity.
Some highlights of the education activities conducted in the program included reinforcing medical follow-up and vaccination at UBS, providing guidance on the use of the child card, holding discussions of themes related to child care and breastfeeding with groups of mothers, as described in a previous article 21.
Of the total children enrolled in the period studied (742), 02 children were excluded because they had been referred to the ICCN before six months of age, because their mothers were HIV+; and 16 children were excluded (2.5%) due to consecutive absences, thus totaling 724 children assessed at baseline. Later, 154 children quit the program before completing 12 months. Thus, assessment included the nutritional progress of 570 children, which accounted for a loss of 21.2%.
Assessment of children's nutritional status and progress was conducted based on Z score values.20 The reference standard used was from the National Center for Health Statistics (NCHS). For nutritional diagnosis, z score values defined were: Well-nourished > -1; Risk of malnutrition > -2 and < -1; Moderate malnutrition < -2 and > -3 and Severe malnutrition < -3.
To assess children evolution during intervention, the height-for-age rate was selected due to the fact that linear growth, measured by this rate, is the most affected among Brazilian children 22. The dependent variable was the difference between Z score of height-for-age at start and at the completion of the 12-month program. The explanatory variables related to children were: birth weight, age at program inclusion, maternal breastfeeding, and hospitalization in the period. Whenever possible, birth weight was verified in the birth certificate brought by the mother. Total time of maternal breastfeeding (offer of maternal milk and no other foods concomitantly) and history of hospitalization in the period children stayed with the program were informed by mothers. Gestational age was not collected.
Family-related variables were: mother's nutritional status [Body Mass Index - BMI (FAO/OMS, 1997)], years of education of mother and father, family income, mother's age, mother's job, number of children, family setting, housing conditions, attendance to pre-natal care and if there was any episode of unemployment of a family member during the study period.
Assessment of nutritional status was entered in EPINUT23 program. Data analysis used STATA 7.0 24 program.
Z score averages for height-for-age were described as per the selected explanatory variables comparing the first and last height-for-age measurements.
To assess whether the positive change in nutritional status could be attributed to ICCN, a time-series multivariate analysis model was built with the use of equations to estimate normal distribution (correlation structure - exchangeable) 25. This type of model enables both assessing the effect of each adjusted variable considering time in the program-controlling confounding variables-and, at the same time, controlling data self-correlation. The dependent variable was the initial and final measurement (at 12 months) of the Z score of height-for-age for each child group: malnourished, at nutritional risk and well-nourished. Intervention was assessed based on the number of visits to the program. The number of months in the program was adjusted as a time variable necessary to conduct the model.
At the end of the analysis, model residues were analyzed to assess adjustments.
Three different multivariate models were built: malnourished, at nutritional risk and well-nourished children, to identify the specific factors of each category, since it had already been perceived that recovery presented different intensity and rhythm21. Malnourished children were taken as a single group, since the number of severely malnourished was not enough to support an isolated model.
The variables adjusted in the model were selected by the p value of Wald test 24, calculated in the univariate analysis. A step-by-step strategy was used, starting with a model with all variables adjusted, and removing the variables that scored in Wald test a p value equal or below 0.25. 26
To estimate the importance of a variable in the model, the difference used was deviance from the model with the variable analyzed and without it. With this value, the asymptotic value of Qui square and corresponding probability (p) were estimated.24
The coefficients whose p value in Wald test was below 0.05 were evaluated as a significant trend of Z score, as well as due to the precision of coefficient estimates (95% confidence interval), calculated for each category of variables present in the model.
A variable was considered a confounding variable when, after adjustment to the multivariate model, the coefficient value decreased or increased by 15% of the value attained in the univariate analysis20.
For ethical reasons, a control group was not adopted, because that would imply keeping malnourished children without any intervention. The project was approved by the Ethics Committee of the School of Public Health. Children's legal representatives were invited to take part in the study and sign the consent form.
In the period studied, 724 children were enrolled at the ICCN; along follow-up, 154 children (21.2%) abandoned the study, totaling in the final nutritional status assessment 570 children. However, it seems there has not been a selective loss of follow-up since, when assessing the number of children that joined the program in the different nutritional status categories, the abandonment rate was similar, as can be seen in Table 1.
Additionally, an analysis was performed to detect the leading characteristics presented by the set of children and their families that started follow-up in the program and compared them to those that abandoned the program before completing twelve months. Results indicate that the losses occurred during the follow-up period neither over nor underestimated the results attained in nutritional evolution, once the children that left the program (losses) are similar, in relation to the variables studied, to those that remained in the study, except with regard to the loss of children whose parents had less years of education (1-4 years), which was higher among those that did not remain in the study (Table 2).
A high percentage of severely (6.1%) and moderately malnourished children (20.7%) can be observed, totaling 26.8% (Table 3). Children with Z score above +2 were not identified.
Nutritional recovery, assessed by height-for-age rate according to initial nutritional status and Z score variation, can be seen in Table 4.
The results found in children evolution based on height-for-age rates were positive in all child groups. Groups of moderately malnourished, at nutritional risk, and well-nourished, showed significant differences between the first and last visits (paired t, p<0.001). It is important to highlight that evolution is directly proportional to the severity of nutritional deficit, that is, higher gain for children with higher deficits.
Because of the reduced number of children with severe malnutrition (Z score <-3), these were grouped with children with moderate malnutrition (Z score < -2 and > -3) to investigate the association of nutritional evolution and the selected independent variables.The sample totalled 194 children who, hereinafter, will be called malnourished. The results attained in univariate and multivariate models are shown for the set of children per category of initial nutritional status, in Tables 5 and 6.
After adjusting variables, time of exposure to the program (number of visits) was associated to better nutritional recovery of malnourished children. At each visit, children gained 0.07 Z score of height-for-age.
Also positively associated to nutritional recovery were children's age when they enrolled in the ICCN (12-24 months), birth weight > 3000g and maternal breastfeeding, which showed borderline value (p=0.05); however, the lower limit of the confidence interval of the adjusted average score was above zero.
Children whose mothers did not work evolved worse when compared to the children whose mothers worked. It is interesting to observe that, after adjustment of any possible confounding variables, absence of maternal employment increased negative association with nutritional recovery by 35.6% (- 0.18 to - 0.24).
The degree of malnutrition presented at enrollment in the ICCN (severe and moderate) was also associated. Results show that the recovery of those children who joined the program with higher nutritional deficits (Z score=0.52) was significantly higher.
Such results show that, for malnourished children, attending ICCN was favorable to recovery, regardless of other factors assessed in this study.
When analyzing association between birth weight and initial malnutrition, we can observe that 64.1% of severely malnourished children were born with weight <2500g. Although not adjusted in the model, weight > 3000g kept a significant association with the nutritional improvement of a child exposed to ICCN.
For children at nutritional risk, after adjustment for possible confounding variables, maternal breastfeeding, family income and time of exposure to the program were associated to nutritional improvement.
Absence of maternal employment kept negatively associated to nutritional recovery, with increasing association after multivariate adjustment. The other variables selected for adjustment in the multivariate model, that is, age and number of siblings, did not remain in the model.
For well-nourished children, time of exposure to the program and age (12-24 months) showed association to higher height-for-age gain, regardless of other factors.
Along time of exposure to ICCN, nutritional condition was significantly better for all categories assessed, with higher impact the more intense the initial nutritional deficiency. These results coincide with studies that show that the best responses to supplemental feeding programs are recorded among those children with the highest nutritional deficits9,10,11.
In relation to malnourished children, attending ICCN was associated to nutritional progress in an independent manner, since the coefficient of height-for-age gain was not reduced, even after inclusion of other variables in the statistical model deployed.
Similar results were found by Aitchison et al 28, who studied the effects of a supplemental feeding program in the growth of malnourished children in Indonesia where, after a multivariate analysis, they found a positive effect in the growth of supplemented children when compared to non-supplemented children.
A study conducted with 90 children, 60 of whom malnourished (30 with supplementation and 30 without supplementation) and 30 well-nourished (without supplementation) in Nigeria, found that, at the end of the intervention, supplemented children reached the same weight average as well-nourished children, while malnourished children not supplemented kept the weight disadvantage initially diagnosed, showing that supplemental feeding improves the nutritional status of children with nutritional deficits29.
Despite ethically controversial procedures, such as having malnourished children not receiving nutritional intervention and being merely used for research, supplemental feeding programs conducted in different parts of the world seem to converge to a positive response in this kind of intervention to recover children that show nutritional deficits.
Also identified as variables positively associated to the recovery of malnourished children are child age at the time of enrollment in the program, birth weight, and total maternal breastfeeding offer.
Findings related to better recovery of children between 12 and 24 months of age 30 are consistent with the literature.
According to Schroeder et al.31, few publications have analyzed supplementation response at specific ages. Additionally, more precise understanding of the ages in which children can benefit the most from nutritional supplementation may help to improve the effectiveness of such programs.
In relation to birth weight, although many studies evidence a direct relationship with the child's nutritional future, there are very few publications assessing its influence in the nutritional recovery of children who join supplemental feeding programs. Santos32 refers to the importance of including birth weight in studies that analyze the impact of nutritional interventions in childhood as a possible intervenient variable in the nutritional recovery process.
With regard to maternal breastfeeding, several studies indicate that children that are breastfed have lower risk of malnutrition. However, the influence of maternal breastfeeding in the recovery of children in supplemental feeding programs has been little explored.
The findings in this study, that show positive association of maternal breastfeeding in the nutritional evolution of malnourished children and children at nutritional risk, is more difficult to explain, since what was assessed was the offer of maternal breastfeeding in the past and not the presence of breastfeeding in the period of intervention. Such findings suggest that children that were breastfed are different than those that were not. Perhaps such difference can be explained by the establishment of better attachment between mother and child among those children that were breastfed, or by the fact that maternal breastfeeding can produce long-lasting impact in children's nutritional and health status that enables them to respond better in recovery. This result shows that maternal breastfeeding is a variable that should be explored in studies on the impact of supplemental feeding programs.
Still in the group of malnourished children, children whose mothers did not work had worse evolution when compared to those whose mothers worked. Studies have evidenced that, quite often, mothers of malnourished children do not perform paid jobs. However, when women do have jobs, they improve family income, which reflects in better nutritional status of children 29.
Studies compiled by Carvalhaes and Benício 33 did not find conclusive evidences about the association between maternal work and children's nutritional status. Other researchers say that, depending on the income generated, type of job, suitability of substitute care, and child age-in addition to mother's autonomy to use the resources generated from her job-effects on children's nutritional status may vary from risk to protection.
The findings of this study consistently show association between unemployment and malnutrition, with worse child recovery in the absence of maternal work. A possible explanation may be the less availability of resources for childcare, before and after supplementation, suggesting that the impact of maternal unemployment adversely affects nutritional recovery. This result points to healthcare limits in controlling malnutrition, showing the need to change public policies directed to societal structure, for example, increasing the offer of jobs for women with malnourished children.
For children at nutritional risk, frequency of attendance to ICCN, maternal breastfeeding and family income >3 minimum wages stand out as significantly associated to recovery; negative outcomes are associated to absence of maternal work.
Likewise, for well-nourished children, program participation was associated to nutritional evolution, in that, for these children, enrolling in the ICCN between 12 and 24 months of age was also positively associated. It is important to highlight that well-nourished children had lower response than children with nutritional deficits-a trend observed in other studies conducted with supplemental feeding program populations that included children with light or nonexistent nutritional deficits 30, 12.
Based on result analysis, we can verify that frequency of attendance to the program, which implicated in receiving supplemental feeding, and participation in educational activities, was associated to the nutritional improvement of all children, regardless of other factors.
Historically, supplemental feeding programs have been developed in the basic healthcare network, where other preventive actions are developed too. When supplemental feeding is regarded as a mere distribution of food, without the development of any educational action amongst the population, results can be inexpressive34. However, when accompanied by other health and education actions, supplemental feeding has shown to be effective to recover malnutrition 35,36.
In the case of the ICCN in Mogi das Cruzes, when mothers went to the supplement distribution site, they took part in nutritional education actions. Perhaps this differential can explain part of the recovery attained by children.
A study conducted in Vietnam with malnourished children that used nutritional education interventions, found significant association of non-deterioration of child growth in the families that received educational actions, when compared to the children in control families 37. Income has been related to children's health and nutrition conditions, which, in turn, is related to mother's level of education and maternal work. In this study, income did not show any association, probably because it was a homogenous population, where 85.2% of children's families received up to 3 minimum wages.
Possible selection, measurement and confounding biases sometimes present in studies like this seem not to have occurred or, if so, to have been small. In relation to selection, the number of children that enrolled in the ICCN in Mogi das Cruzes in the period assessed coincides with the estimate of malnourished children and children at nutritional risk of NUPENS (Center of Epidemiological Research in Nutrition and Health) for the municipality.
With regard to measurement bias, before program start, 4th year interns of Nutrition from a private university in the same city were trained to measure children's height, with direct supervision along the program. Additionally, children's height measurements were verified by sampling, which provided greater reliability to results.
With regard to confounding biases, a multivariate model was used and the independence of effects was measured for the variables that remained in the final model. We could not identify seasonal influence on child nutritional evolution, since admission was along all months of the year and evolution was followed monthly for twelve months. In relation to economic changes, we tried to identify them by asking the child's legal representative at the end of the 12 months' period, if during intervention someone in the family had got a job and, therefore, increased family income. Any occurrence of child hospitalization was also verified.
A problem found in studies with repeated samples is data self-correlation. Effects of self-correlation were controlled by using the model applied in data analysis. Considering the validity of this study due to the little presence of biases and suitability of the statistical tool employed, we can say that the children the benefited from the ICCN in the city of Mogi das Cruzes showed improvement in nutritional status, assessed by the average variation of Z score along 12 months. The best results were attained by children with the highest nutritional deficits, corroborating studies that assessed supplementation programs developed in different parts of the world.
In a positive manner, higher birth weight, maternal breastfeeding and age at enrollment in the ICCN (12 to 24 months) were associated to children recovery. Special attention should be given, therefore, to pre-natal care with regard to encouraging maternal breastfeeding, following pregnant women's nutritional status and including children in supplemental feeding programs at the age there is better possibility of response.
It was also possible to verify that absence of maternal employment was associated to worse child recovery, as well as non attainment of work by some family member configured as a possible association to worse nutritional recovery. These factors should be studied in greater samples since, in this study, they showed borderline values.
Also, it is important to make some considerations about the limitations described in literature with regard to assessments of supplemental feeding programs without a control group. According to Taddei27, associations found between supplementation and evolution in nutritional status in the group studied cannot be inferred as causal relations, since such modifications may occur due to other factors not addressed in this study. In such case, possible associations found may be undue.
However, this study sought, by means of the statistical model employed, to adjust the variables described in literature as associated to malnutrition, which made the results attained more powerful.
Another limitation refers to possible presence of prematurity, which was not assessed in the present study. It is possible that recovery is overestimated among malnourished children and that some factors identified could be adjusted by the variable of prematurity.
It is also pertinent to discuss the methodological option made in this study, separating children according to baseline nutritional status. Another alternative would be to consider the program's global effectiveness and consider the initial nutritional status as another exposure. Our option may have reduced the statistic power to identify a greater number of variables by expanding confidence intervals; however, it enabled understanding that children's age at enrollment, birth weight and mother's age are specific factors of malnourished children.
The ICCN experience in Mogi das Cruzes suggests that supplemental feeding has a relevant role in nutritional recovery, particularly for more severe cases. It should be started as soon as possible, and reinforcing the protective role of maternal breastfeeding.
In addition to offering food, our results showed absence of maternal employment as a risk factor for worse child nutritional recovery, which points to the need of increasing the offer of employment and education to women, particularly to those in more vulnerable families.
We can also conclude that the policies that positively affect pre-natal care aiming to reduce low birth weight may equally accelerate nutritional recovery.
1. Benício MHA; Monteiro CA. Desnutrição infantil nos municípios brasileiros: risco de ocorrência. Brasília: NUPENS/USP/UNICEF; 1987. [ Links ]
3. CEBRAP/Ministério da Saúde. Departamento de Ciência e Tecnologia. Pesquisa Nacional de Demografia e Saúde da Criança e da Mulher: PNDS, 2006. Brasília, DF; 2008. [ Links ]
4. Monteiro CA, Benício MHD'A, Iunes RF, Gouveia NC, Cardoso MAA. Evolução da Desnutrição Infantil. In: Monteiro CA. Velhos e novos males da saúde no Brasil: a evolução do país e suas doenças. (2ª ed. rev. e aumentada) São Paulo: Hucitec; 2000. pp. 93-114. [ Links ]
5. UNICEF - Fundo das Nações Unidas para Infância e Adolescência. Estratégia para melhorar a nutrição de crianças e mulheres nos países em desenvolvimento. New York: UNICEF; 1990. [ Links ]
6. Monte CMG. Desnutrição: um desafio secular à nutrição infantil. J Pediatr 2000; 76(S3): 285-97. [ Links ]
7. Chaves SP, Stefanini MLR, Lei DLM, Lerner BR. Evolução do estado nutricional de desnutridos matriculados em programas de intervenção. Rev Nutr PUCCAMP 1989; 2(2): 161-77. [ Links ]
8. Taddei JAA. Avaliação antropométrica do impacto nutricional de um programa de suplementação alimentar [tese de doutorado] São Paulo: Faculdade de Saúde Pública da USP; 1987. [ Links ]
9. Castro IRR. Efetividade da suplementação alimentar na recuperação nutricional de crianças: avaliação do programa "Leite é saúde" no município do Rio de Janeiro [tese de doutorado]. São Paulo: Faculdade de Saúde Pública da USP; 1999. [ Links ]
10. Goulart RMM. Evolução do estado nutricional de crianças submetidas a programa de suplementação alimentar e ações educativas à família no município de Itaquaquecetuba - SP [dissertação de mestrado]. São Paulo: Faculdade de Saúde Pública da USP; 1998. [ Links ]
11. Puccini RF, Goihman S, Nóbrega FJ. Avaliação do programa de recuperação de desnutridos do município de Embú, na região metropolitana de São Paulo. J Pediatr 1996; 72(2): 72-9. [ Links ]
12. Soares BT, Parente WG. Desnutrição e resultados de reabilitação em Fortaleza. Rev Nutr Campinas 2001; 14(2): 103-10. [ Links ]
13. D'Innocenzo S. Avaliação da efetividade de programas de suplementação alimentar sobre incremento ponderal de crianças em município semiárido baiano [dissertação de mestrado]. Bahia: Instituto de Saúde Coletiva da Universidade Federal da Bahia; 2006. [ Links ]
14. Morris SS, Olinto P, Flores R, Nilson EAF, Figueiró AC. Conditional cash transfers are associated with a small reduction in the rate of weight gain of preschool children in Northeast Brazil. J Nutr 2004; 134(9): 2336-41. [ Links ]
15. Barreto ML O conhecimento científico e tecnológico como evidência para políticas e atividades regulatórias em saúde. Ciênc Saúde Coletiva 2004; 9(2): 329-38. [ Links ]
16. Victora CG Avaliando o impacto de intervenções em saúde. Rev Bras Epidemiol 2002; 5(1): 2-4. [ Links ]
17. Ministério da Saúde. Incentivo ao combate às carências nutricionais. Manual de Orientação. Brasília; 2000. [ Links ]
19. Secretaria da Saúde. Fundação Sistema Estadual de Análise de Dados - SEADE. Taxa de mortalidade infantil: Municípios Paulistas 2005. Disponível em: http://portal.saude.sp.gov.br/resources/profissional/documentos_tecnicos/estudos_anali ses/mortalidade_infantil.pdf. [Acessado em 20 de Janeiro de 2009] [ Links ].
20. World Health Organization Working Group. An evaluation of infant growth: the use and interpretation of anthropometric in infants. Bull WHO 1995; 73: 165-74. [ Links ]
21. Goulart RMM, França Junior I, Souza MFM. Recuperação nutricional de crianças desnutridas e em risco nutricional em programa de suplementação alimentar no município de Mogi das Cruzes, São Paulo. Brasil. Cad Saúde Pública 2007; 23(1): 147-56. [ Links ]
22. Ministério da Saúde. Organização Pan-Americana da Saúde. Guia alimentar para crianças menores de dois anos. Brasília; 2002. [ Links ]
23. Dean AG, Dean JA, Coulombier D, Brendel KA, Smith DC, Burton AH et al. Epi Info, Version 6: A word processing, database, and statistics program for epidemiology on microcomputers. Atlanta (Ge): Centers for Disease Control and Prevention; 1994. [ Links ]
24. Stata Corp. (2003). Stata statistical software: realease 8.1. College Station, TX, Stata Corporation. [ Links ]
25. Horton NJ, Lipsitz SR. Review of software to fit generalized estimating equation regression models. Am Stat Assoc 1999; 53: 160-9. [ Links ]
26. Hosmer DW, Lemeshow S. Applied logistic regression. USA: John Wiley & Sons; 1989. [ Links ]
27. Taddei JAAC. Avaliação Antropométrica do impacto nutricional de um programa de suplementação alimentar [tese de doutorado]. São Paulo: Faculdade de Saúde Pública da USP; 1987. [ Links ]
28. Aitchison TC, Durnin JVGA, Beckett C, Pollitt E. Effects of an energy and micronutrient supplement on growth and activity, correcting for non-supplemental sources of energy input in undernourished children in Indonesia. Eur J Clin Nutr 2000; 54(S2): 69-73. [ Links ]
29. Obatolu VA. Growth pattern of infants fed with a mixture of extruded malted maize and cowpea. Nutr Africa 2003; 19: 174-8. [ Links ]
30. Rivera JA, Habicht JP. Effect of supplementary feeding on the prevention of mild-to-moderate wasting in conditions of endemic malnutrition in Guatemala. Bull WHO 2002; 80(12): 926-32. [ Links ]
31. Schroeder DG, Martorell R, Rivera JA, Ruel MT, Habicht JP. Age differences in the impact of nutritional supplementation on growth. J Nutr 1995; 125(S):151-9. [ Links ]
32. Santos H. Recuperação nutricional como atividade do programa de assistência integral à saúde da criança: avaliação dos aspectos operacionais e de impacto nutricional em uma unidade de saúde da Secretaria de Estado da Saúde de São Paulo [dissertação de mestrado]. São Paulo: Faculdade de Saúde Pública da Universidade de São Paulo; 1990. [ Links ]
33. Carvalhaes MABL, Benício MHD´A. Capacidade materna de cuidar e desnutrição infantil. Rev Saúde Pública 2002; 36(2): 188-97. [ Links ]
34. Monte CMG, Ashworth A, As MLB, Diniz RLP. Effectiveness of nutrition centers in Ceará state, northeastern Brazil. Rev Pan Salud Pública 1998; 4(6): 375-82. [ Links ]
35. Paknawin-Mock J, Jarvis L, Jahari AB, Husaini MA, Pollitt E. Community-level determinants of child growth in an Indonesian tea plantation. Eur J Clin Nutr 2000; 54(S2): 28-42. [ Links ]
36. Roy SK et al. Intensive nutrition education with or without supplementary feeding improves the nutritional status of moderately-malnourished children in Bangladesh. J Health Popul Nutr 2005; 23(4): 320-30. [ Links ]
37. Schroeder DG, Pachón H, Dearden KA, Ha TT, Lang TT, Marsh DR. An integrated child nutrition intervention improved growth of younger, more malnourished children in northern Viet Nam. Food and Nutr Bull 2002; 23(4): 50-7. [ Links ]
Correspondência: Tese de doutorado
apresentada na Faculdade de Saúde Pública da Universidade de São
Paulo em 2003. Título: Evolução do estado nutricional de
crianças desnutridas e em risco nutricional em programa de suplementação
Rita Maria Monteiro Goulart
Rua Piraquara, 450
São Paulo - SP CEP 03688-000
Tese de doutorado apresentada na Faculdade de Saúde Pública da Universidade de São Paulo em 2003. Título: Evolução do estado nutricional de crianças desnutridas e em risco nutricional em programa de suplementação alimentar.