Principal investigator: Viviane Pires Ribeiro
Article title: EFFECT OF A CONDITIONAL CASH TRANSFER PROGRAM ON LENGTH-FOR-AGE AND WEIGHT-FOR-AGE IN BRAZILIAN INFANTS AT 24 MONTHS USING DOUBLY-ROBUST, TARGETED ESTIMATION
Article authors: Jeremy A. Labrecque, Jay S. Kaufman, Laura B. Balzer, Richard F. Maclehose, Erin C. Strumpf, Alicia Matijasevich, Iná S. Santos, Kelen H Schmidt and Aluísio JD Barros
Location of the intervention: Brazil
Sample size: 1703 children
Main theme: Health Insurance
Primary variable of interest: Children's health
Evaluation method: Maximum likelihood estimate
Evaluation Context
Conditional Cash Transfers are government programs that provide cash assistance to low-income families, provided they meet specific requirements. Most requirements relate to improving child health and education, such as health check-ups, vaccinations, prenatal care for the mother, school enrollment, or a specific level of school attendance. The Bolsa Família (BF) is a Brazilian monthly transfer program that began in 2004. Beneficiary families who met the requirements (children's health check-ups, up-to-date vaccinations, and school attendance of 85% or higher) received between R$15 and R$95 per month, depending on per capita family income and the number of children. In 2015, Bolsa Família covered more than 27 million low-income households, paying out a total of R$28,5 billion in 2016, according to data from the Brazilian Comptroller General's Office.
Lebrecque et al. (2018) argue that although a variety of anthropometric measures are used to assess children's nutritional status, height-for-age or length-for-age is of particular interest for two reasons: they are good proxys The cumulative nutritional status allows for the study of the long-term nutritional consequences of Bolsa Família; and poor childhood growth, particularly in the first 24 months, is an important predictor of future health. Weight-for-age is also used to assess nutritional status and can indicate changes in nutritional status in the short term. In this context, Lebrecque et al. (2018) estimate the effect of the Brazilian conditional cash transfer program, Bolsa Família, on child nutritional status, measured by length-for-age and weight-for-age, from birth to 24 months.
Intervention Details
The data used by the authors were obtained from the Pelotas Birth Cohort, which recruited 99% of hospital births in all five hospitals within the limits of the Brazilian city of Pelotas in 2004. Length-for-age and weight-for-age were calculated using the 2006 World Health Organization growth standards. Field researchers conducted home visits to measure the children and administer questionnaires to the parents. Information on prenatal, parental, and socioeconomic variables was thus obtained. Lebrecque et al. (2018) used data from birth, 12, and 24 months; twins and infants who died before 24 months were excluded from the analysis. The authors also limited the study to families who reported a per capita income of R$100 or less, as families with higher per capita income were not eligible to receive the Bolsa Família program.
The data on the amount of money received by BF beneficiaries was obtained from the Transparency Portal website (http://www.portaltransparencia.gov.brThe relationship between the money received from the Bolsa Família program and length-for-age and weight-for-age was calculated probabilistically, using the child's name, date of birth, parents' names, and age. A total of 1703 children were identified as eligible to receive the benefit.
Methodology Details
Lebrecque et al. (2018) divided children into three groups based on the total amount of BF received in the first 24 months: no BF (received R$ 0), low BF (received R$ 1000 or less), and high BF (received more than R$ 1000). The authors used a maximum likelihood estimation method, a doubly robust, semiparametric estimator, efficient replacement of statistical parameters, to estimate the effect of receiving low and high levels of BF on infant nutritional status, measured by length-for-age and weight-for-age, from birth to 24 months.
Results
Of the 1703 families studied, 110 belonged to the low BF group and 319 were in the high BF group in the first 12 months. At 24 months, 291 families were in the low BF group and 355 in the high BF group. On average, participating families received BF for 16 of the 24 months after the child's birth (with a standard deviation of 8). Of the families that started BF, 96% were still receiving it 24 months after the child's birth.
The pattern among mothers who received BF and participated in the analysis was: older age, shorter stature, lower education level, fewer prenatal visits, less weight gain during pregnancy, more likely to smoke, and less likely to work during pregnancy compared to mothers who did not receive BF. Birth length, birth weight, gestational age, and reported health problems at birth were all similar by BF level. All socioeconomic variables indicated that exposed families were notably poorer and had larger families.
The authors found a negative association between body fat (BF) and height-for-age and weight-for-age at almost all BF levels. The association was strongest for high BF at height-for-age and weight-for-age at 24 months, with a reduction of around 0,19 z-scores in both cases. Comparisons between low and high BF were all negative, but confidence intervals were also consistent with positive values. Estimates were lower at 12 months than at 24 months.
Lessons in Public Policy
What is the effect of Bolsa Família on child health? The results obtained by Lebrecque et al. (2018) indicate that in the first two years after a child's birth, families receiving Bolsa Família were negatively associated with child nutritional status. At 24 months, this negative association is equivalent to approximately three weeks of delayed growth in height-for-age and five weeks of delayed growth in weight-for-age.
The authors emphasize that, despite their efforts, the possibility of residual confounding cannot be eliminated. But if the results are not due to residual confounding of the estimates, there are possible causal mechanisms that explain why Bolsa Família has a negative effect on children's height-for-age and weight-for-age. Some studies suggest that receiving Bolsa Família may be associated with the purchase of less nutritious foods. Another study in northeastern Brazil found that families receiving Bolsa Família were three times more likely to consume nutrient-poor foods than families that did not receive Bolsa Família. A poorer diet can lead to an increase in infections or micronutrient deficiencies, which can, in turn, reduce height-for-age. However, there is also some research demonstrating a positive relationship between Bolsa Família and diet and nutrition. Therefore, Lebrecque et al. (2018) argue that further studies with more recent data are needed to determine whether this negative relationship between Bolsa Família and child health holds true today, given that the authors used data from the first years of Bolsa Família payments.
References
LABRECQUE, Jeremy A. et al. Effect of a conditional cash transfer program on length-for-age and weight-for-age in Brazilian infants at 24 months using doubly-robust, targeted estimation. Social Science & Medicine, v. 211, p. 9-15, 2018.
