Abstract
This study assessed whether enrollment in a national conditional cash transfer program was associated with wasting and stunting among children experiencing extreme poverty in the Philippines. Data were drawn from cross-sectional surveys collected from 10 regional areas in the Philippines between April 2018 and May 2019. A total of 2945 children aged between six months and 12 years comprised the analytical sample. Multilevel logistic regression was conducted to estimate the association between enrollment in Pantawid Pamilyang Pilipino Program (4Ps) and stunting and wasting, controlling for sociodemographic factors and clustering by region. There was no meaningful association between household enrollment in 4Ps and the wasting status of children, but enrollment in 4Ps was associated with lower odds of stunting and differed by geography type. Findings suggest that the current design of 4Ps may not address sudden shocks that contribute to wasting, but may address the underlying socioeconomic risk factors associated with stunting.
Keywords: child malnutrition, conditional cash transfers, social protection, poverty, stunting, wasting, Southeast Asia
What We Already Know
Conditional cash transfers have the potential to address the underlying socioeconomic risk factors associated with child malnutrition through providing financial incentives and health-related and education-related conditionalities.
Previous research has shown mixed results when investigating the association between conditional cash transfer programs and acute (wasting) and chronic (stunting) child malnutrition.
The Pantawid Pamilyang Pilipino Program (4Ps) is a national conditional cash transfer in the Philippines that targets populations experiencing poverty and aims to improve health and nutrition outcomes of children.
What This Article Adds
Among children experiencing extreme poverty, enrollment in the national conditional cash transfer program was not associated with wasting, but was associated with lower odds of stunting.
Findings suggest that the current program structure may not address needs arising from sudden shocks that contribute to wasting in children and may point to opportunities for leveraging existing programs for child malnutrition.
While enrollment in the conditional cash transfer program was associated with lower odds of stunting, the difference in odds according to geography type may suggest that the underlying risk of stunting is unevenly distributed or that program operations may vary across contexts.
Introduction
Child malnutrition continues to be a global health challenge in low-income and middle-income countries, with only one quarter of countries on track to meet nutritional targets set by the 2030 Sustainable Development Goals. 1 The United Nations Children’s Fund (UNICEF) Conceptual Framework of child malnutrition demonstrates how various macro-level and microlevel social and economic factors contribute to wasting and stunting in children. 2 In response to the complex factors that influence child malnutrition, conditional cash transfers are viewed as a nutrition-sensitive approach to support the underlying determinants of nutrition in children. 3 Although previous studies have pointed to the association between positive nutritional outcomes in children and participation in conditional cash transfer programs, results vary between wasting and stunting, and across different countries. 4
In the Philippines, 33.0% and 5.6% of children remain stunted and wasted, respectively. 5 The most recent data available from the 2018 Expanded National Nutrition Survey indicated that the prevalence of wasting and stunting in children across all age groups was consistently highest among households in rural areas and in the poorest income quintiles. 6 To support poverty alleviation among populations in the lowest income levels in the Philippines, the Pantawid Pamilyang Pilipino Program (4Ps) was launched in 2007 and has become the main social protection and welfare program in the country. 7 Led by the Department of Social Welfare and Development (DSWD), 4Ps is a conditional cash transfer program that aims to improve health and educational outcomes within low-income households.7,8 Upon meeting eligibility criteria and completing program requirements, households enrolled in 4Ps are entitled to receive cash transfers, food assistance, and health and educational grants (see Supplemental Appendix Figure A1).7,8
The theory of change for 4Ps demonstrates how program conditionalities and entitlements could potentially support the nutritional outcomes of children within recipient households (see Supplemental Appendix Figure A2). 9 The 4Ps conditionalities include attending maternal and child health care services, as well as participation in monthly Family Development Sessions that cover topics on health, well-being, family dynamics, and parenting.9 -11 Upon completion, participants are entitled to cash transfers and food subsidies that could support food security and improvements to housing conditions.9,10 Overall, a combination of both the conditionalities and entitlements are expected to provide families with the resources to create healthy environments that are supportive of children’s nutritional status. Although previous studies have examined the associations between enrollment in 4Ps and wasting and stunting among children, these evaluations were unable to capture the effect of 4Ps on the poorest households because of limitations in evaluation and study design, leaving gaps in understanding about the relationship between 4Ps and child malnutrition in populations that may benefit the most from its programming.9,12
The objectives of this study were (1) to describe and compare the sociodemographic characteristics of children (6 months to 12 years of age) according to household enrollment in 4Ps, and (2) to examine whether household enrollment in 4Ps was associated with wasting status and stunting status among children experiencing extreme poverty in the Philippines.
Methods
Study Context, Design, and Data Collection
This study was conducted in partnership with International Care Ministries (ICM), a nongovernmental organization based in the Philippines. ICM uses an asset-based poverty score card to determine a household’s eligibility to participate in its poverty alleviation programs, targeting similar populations as those enrolled in 4Ps. Data were drawn from cross-sectional household surveys collected between May 2018 and April 2019. To collect data, trained surveyors interviewed a representative of the household and informed verbal consent was acquired from respondents prior to survey administration. This study was approved by the University of Waterloo Office of Research Ethics (ORE no. 43368).
Variables of Interest
Dependent variables
Anthropometric measures of children aged between 6 months and 12 years were collected by ICM health trainers who measured children’s weight using a weighing scale; children’s length/height were measured using an Allen stick. 13 Recumbent length was measured among children below 2 years of age and standing height was measured among children aged 2 years and above.
Wasting status was determined using the weight-for-height z score (WHZ) of children aged below 5 years and the body mass index (BMI)-for-age z score (BAZ) of children aged 5 to 12 years. Stunting status was determined using height-for-age z score (HAZ). The calculation of WHZ, BAZ, and HAZ was based on the World Health Organization (WHO) Reference Growth Standards for children, using the z scorer package in R.14,15 Binary variables of wasting and stunting status were created by categorizing children with a WHZ of < –2 and a BAZ of < –2 as affected by wasting, and children with a HAZ of < –2 as affected by stunting.
Independent variable
Household enrollment in 4Ps was self-reported. The survey respondent was asked the question, “Is [name of household head] a recipient of 4Ps program?” and were provided response options of “Yes,” “No,” or “I don’t know.” Respondents who indicated “Yes,” were categorized as enrolled in 4Ps, and respondents who indicated “No” or “I don’t know” were categorized as not enrolled in 4Ps.
Covariates
The identification of covariates was informed by the UNICEF Conceptual Framework on the determinants of child malnutrition as well as the implementation of 4Ps.2,9 The sex and age of children were included to control for individual-level and biological factors that may confound the relationship between 4Ps enrollment and the nutritional status of children. 15 A higher order term for the age of the child (age of the child 2 ) was also included to account for the relationship between the anthropometric growth of children and age. 15 Self-reported measures of the education level of the household head (years), the age of the household head (years), and the number of household members were included to control for confounding effects of household sociodemographic characteristics.16,17 The wealth of the household was measured using a Wealth Index score, which was developed using principal component analysis (see Supplemental Appendix B). 18 Experiences of hunger and food insecurity were assessed using the Household Hunger Scale (see Supplemental Appendix C). 19 Finally, geography type was included in the model as the distribution of wasting and stunting varies across different locations in the Philippines.6,20 An interaction term between geography type and household enrollment in 4Ps was included because of the heterogeneous implementation of the program, as well as a recognition that varying experiences of 4Ps participants across different regions may moderate the association of 4Ps with stunting and wasting among children.11,21,22
Statistical Analyses
Descriptive statistics
Descriptive statistics were calculated to determine differences in sociodemographic characteristics of children in households enrolled and not enrolled in 4Ps to understand patterns in sociodemographic factors through which 4Ps may potentially be associated with the nutritional status of children. To characterize the sociodemographic characteristics of children, frequencies and proportions were calculated for categorical variables and tested using Pearson’s chi-square test. Means and standard deviations were calculated for continuous variables and were tested using Welch’s two-sample t test.
Regression modeling
Multilevel mixed-effects regression modeling was implemented to account for clustering by regional areas. 23 Two separate multilevel logistic regression models were conducted to model the association between household enrollment in 4Ps (ref: not enrolled in 4Ps) and each of wasting status (ref: not wasted) and stunting status (ref: not stunted), respectively, while controlling for sociodemographic factors that were theorized to confound or moderate the relationship between 4Ps and the nutritional status of children. Linear multilevel mixed-effects regression models were also analyzed to assess the association of BAZ/WHZ and HAZ with enrollment in 4Ps (see Supplemental Appendix F). Models were fitted using maximum likelihood estimation and analyses were conducted using the lme4 package in R Version 4.1.0.24,25
In the multilevel models, children formed level 1, whereas specific ICM regional areas formed level 2. Regression modeling was conducted by first examining an intercept-only model, followed by models that included main effects and interaction effects (see Supplemental Appendix D). Model 1 was an intercept-only model with no explanatory variables, and the intraclass correlation coefficient (ICC) was calculated to determine the proportion of the variance in wasting status and stunting status that was attributable to differences between regional bases. Because preliminary modeling indicated that rescaling variables improved model fit and convergence, subsequent models were run with standardized continuous variables. Model 2 was a random intercept model with main effects, whereas Model 3 was a random intercept model including main effects and interaction effects between household enrollment in 4Ps and geography type, both with standardized continuous variables. Model selection was first informed by the likelihood ratio test to determine the model with lowest deviance. Subsequently, the Akaike information criterion (AIC) of the models was reviewed to confirm selection of the model with the smallest AIC value. 23
Statistical significance was determined using a P value of .05, and 95% confidence intervals (CIs) for the adjusted log-odds of the model probabilities were calculated, with no adjustment for multiple testing. The log-odds estimates and CIs from the models selected for interpretation were converted to the unstandardized values before exponentiating to interpret the corresponding odds ratios (ORs).
Results
Data were available from 3401 children. Following participation criteria for ICM’s nutritional screening program, children who were aged below 6 months and above 12 years of age (n=204), had incomplete sociodemographic information (n=89), or had extreme outliers (WHZ, BAZ, and HAZ >6 or < –6; age of the household head ≥66 years; number of household members ≥11; n=183) were excluded from the final data set. The analytic sample consisted of 2945 children across 10 regions where ICM operated at the time of data collection.
Statistically significant differences were observed between children in households enrolled in 4Ps and children in households not enrolled in 4Ps (see Table 1). Children in households enrolled in 4Ps also had household heads with fewer years of education (7.7 years; SD=3.5) compared with those not enrolled in 4Ps (8.4 years; SD=3.7). Children in households enrolled in 4Ps also belonged to larger households (5.9 members; SD=1.8) compared with households that were not enrolled in 4Ps (4.7 members; SD=1.6). No statistically significant differences were observed between children in households enrolled and not enrolled in 4Ps in relation to the Wealth Index score and the Household Hunger Scale.
Table 1.
Descriptive Statistics (Frequency, Proportion, Mean, Standard Deviation) of Household Sociodemographic Characteristics and Geographic Distribution of Children (Ages 6 Months-12 Years) in Households Enrolled and Not Enrolled in the Pantawid Pamilyang Pilipino Program (N = 2945).
| Variable | Enrolled in 4Ps, n = 1163 (39.49%) | Not enrolled in 4Ps, n = 1782 (60.51%) | P value | ||
|---|---|---|---|---|---|
| n; mean | %/SD | n; mean | %/SD | ||
| Sex of child | .2 | ||||
| Female | 574 | (49.36%) | 833 | (46.75%) | |
| Male | 589 | (50.64%) | 949 | (53.25%) | |
| Age of child (months) | 69.79 | (36.50) | 53.87 | (31.15) | <.001 |
| Age of household head (years) | 40.50 | (8.80) | 35.37 | (9.97) | <.001 |
| Education of the household head (years) | 7.69 | (3.48) | 8.39 | (3.70) | <.001 |
| Number of household members | 5.87 | (1.77) | 4.71 | (1.59) | <.001 |
| Wealth Index score | −1.04 | (1.73) | −1.01 | (1.83) | .6 |
| Household Hunger Scale category | .9 | ||||
| Little to no hunger in household | 1043 | (89.68%) | 1,587 | (89.06%) | |
| Moderate hunger in household | 108 | (9.29%) | 175 | (9.82%) | |
| Severe hunger in household | 12 | (1.03%) | 20 | (1.12%) | |
| Geography type of area of residence | <.001 | ||||
| Coastal | 133 | (11.44%) | 191 | (10.72%) | |
| Urban mountain | 161 | (13.84%) | 180 | (10.10%) | |
| Rural plain | 284 | (24.42%) | 571 | (32.04%) | |
| Rural mountain | 547 | (47.03%) | 757 | (42.48%) | |
| Urban plain | 38 | (3.27%) | 83 | (4.66%) | |
| Region of residence | <.001 | ||||
| Bacolod | 47 | (4.04%) | 141 | (7.91%) | |
| Bohol | 90 | (7.74%) | 241 | (13.52%) | |
| Cebu | 49 | (4.21%) | 124 | (6.96%) | |
| Dipolog | 288 | (24.76%) | 210 | (11.78%) | |
| Dumaguete | 76 | (6.53%) | 160 | (8.98%) | |
| General Santos | 153 | (13.16%) | 210 | (11.78%) | |
| Iloilo | 105 | (9.03%) | 168 | (9.43%) | |
| Kalibo | 23 | (1.98%) | 70 | (3.93%) | |
| Koronadal | 112 | (9.63%) | 278 | (15.60%) | |
| Palawan | 220 | (18.92%) | 180 | (10.10%) | |
Abbreviation: 4Ps, Pantawid Pamilyang Pilipino Program; SD, standard deviation.
The ICC for the intercept-only model (Model 1) for both wasting (ICC=1.34%) and stunting (ICC=0.56%) indicated that a small proportion of the variances in stunting and wasting status were attributable to differences between regional bases (Tables 2 and 3).
Table 2.
Fixed- and Random-Effects Estimates and Model Information Criteria From Multilevel Logistic Regression Model of Household Enrollment in 4Ps on Wasting Status of Children Controlling for Sociodemographic and Geographic Variables (N = 2945).
| Predictors | Model 1 (intercept only) | Model 2 a b | Model 3 b | ||||||
|---|---|---|---|---|---|---|---|---|---|
| OR | SE | 95% CI | OR | SE | 95% CI | OR | SE | 95% CI | |
| Fixed effects | |||||||||
| Level 1 | |||||||||
| Household enrollment in 4Ps | |||||||||
| Not enrolled in 4Ps (ref) | |||||||||
| Enrolled in 4Ps | 1.02 | 0.15 | [0.77, 1.36] | 1.21 | 0.42 | [0.62, 2.37] | |||
| Sex of child | |||||||||
| Female [ref) | |||||||||
| Male | 0.72* | 0.09 | [0.56, 0.92] | 0.72* | 0.09 | [0.56, 0.93] | |||
| Age of child (months) | 0.99* | 0.24 | [0.97, 0.99] | 0.99* | 0.14 | [0.97, 0.99] | |||
| Age of child2 (months2) | 1.00** | 0.23 | [1.00, 1.01] | 1.00** | 0.46 | [1.00, 1.01] | |||
| Age of household head (years) | 1.00 | 0.07 | [0.99, 1.01] | 1.00 | 0.07 | [0.98, 1.01] | |||
| Education of household head (years) | 0.97 | 0.07 | [0.93, 1.01] | 0.97 | 0.07 | [0.93, 1.01] | |||
| Number of household members | 1.02 | 0.07 | [0.94, 1.01] | 1.02 | 0.07 | [0.94, 1.10] | |||
| Wealth Index score | 0.96 | 0.07 | [0.89, 1.04] | 0.98 | 0.04 | [0.89, 1.05] | |||
| Household Hunger Scale category | |||||||||
| Little to no hunger (ref) | |||||||||
| Moderate hunger | 2.04* | 0.65 | [1.10, 3.80] | 2.07* | 0.66 | [1.11, 3.87] | |||
| Severe hunger | 2.18** | 0.59 | [1.28, 3.71] | 2.19** | 0.60 | [1.29, 3.74] | |||
| Level 2 | |||||||||
| Geography type | |||||||||
| Coastal (ref) | |||||||||
| Urban mountain | 0.59* | 0.16 | [0.36, 0.99] | 0.68 | 0.24 | [0.34, 1.36] | |||
| Rural plain | 0.64* | 0.14 | [0.42, 0.97] | 0.64 | 0.18 | [0.37, 1.11] | |||
| Rural mountain | 0.61* | 0.12 | [0.41, 0.89] | 0.72 | 0.19 | [0.42, 1.21] | |||
| Urban plain | 0.63 | 0.23 | [0.31, 1.29] | 0.69 | 0.32 | [0.27, 1.73] | |||
| Interaction effects | |||||||||
| Urban mountain × enrolled in 4Ps | 0.77 | 0.39 | [0.28, 2.09] | ||||||
| Rural plain × enrolled in 4Ps | 1.07 | 0.45 | [0.47, 2.43] | ||||||
| Rural mountain × enrolled in 4Ps | 0.69 | 0.27 | [0.32, 1.49] | ||||||
| Urban plain × enrolled in 4Ps | 0.83 | 0.62 | [0.19, 3.63] | ||||||
| Intercept | 0.10*** | 0.01 | [0.08, 0.12] | 0.22*** | 0.06 | [0.13, 0.36] | 0.19*** | 0.06 | [0.11, 0.35] |
| Est. | SE | 95% CI | Est. | SE | 95% CI | Est. | SE | 95% CI | |
| Random effects | |||||||||
| Variance of random intercepts (regional base) | 0.04 | 0.09 | [−0.14, 0.23] | 0.07 | 0.15 | [−0.23, 0.36] | 0.06 | 0.76 | [−1.42, 1.54] |
| ICC | 0.01 | ||||||||
| Model information criteria | |||||||||
| Deviance | 1820.02 | 1772.66 | 1770.52 | ||||||
| AIC | 1824.02 | 1804.66 | 1810.52 | ||||||
Ni, 2945; where i = number of children; Nj, 10; where j = number of regional bases.
Abbreviations: 4Ps, Pantawid Pamilyang Pilipino Program; AIC, Akaike information criterion; CI, confidence interval; ICC, intraclass correlation coefficient; OR, odds ratio.
Continuous variables in the regression model are standardized and scaled to support model fit and convergence.
Model selected for interpretation.
P < .05. **P < .01. ***P < .001.
Table 3.
Fixed- and Random-Effects Estimates and Model Information Criteria From Multilevel Logistic Regression Model of Household Enrollment in 4Ps on Stunting Status of Children Controlling for Sociodemographic and Geographic Variables (N = 2945).
| Predictors | Model 1 | Model 2 a | Model 3 a b | ||||||
|---|---|---|---|---|---|---|---|---|---|
| OR | SE | 95% CI | OR | SE | 95% CI | OR | SE | 95% CI | |
| Fixed effects | |||||||||
| Level 1 | |||||||||
| Household enrollment in 4Ps | |||||||||
| Not enrolled in 4Ps (ref) | |||||||||
| Enrolled in 4Ps | 0.85 | 0.07 | [0.72, 1.01] | 0.55* | 0.13 | [0.35, 0.88] | |||
| Sex of child | |||||||||
| Female (ref) | |||||||||
| Male | 1.41*** | 0.11 | [1.21, 1.63] | 1.41*** | 0.11 | [1.22, 1.64] | |||
| Age of child (months) | 1.00 | 0.14 | [0.99, 1.00] | 0.99 | 0.14 | [0.99, 1.00] | |||
| Age of child2 (months2) | 1.00 | 0.14 | [0.99, 1.00] | 1.00 | 0.15 | [0.99, 1.00] | |||
| Age of household head (years) | 0.99 | 0.04 | [0.99, 1.00] | 1.00 | 0.04 | [0.99, 1.00] | |||
| Education of household head (years) | 0.96*** | 0.04 | [0.94, 0.98] | 0.96*** | 0.04 | [0.94, 0.98] | |||
| Number of household members | 1.13*** | 0.05 | [1.07, 1.18] | 1.13*** | 0.05 | [1.07, 1.18] | |||
| Wealth Index score | 0.90*** | 0.03 | [0.86, 0.94] | 0.90*** | 0.03 | [0.86, 0.94] | |||
| Household Hunger Scale category | |||||||||
| Little to no hunger (ref) | |||||||||
| Moderate hunger | 1.71 | 0.48 | [0.99, 2.95] | 1.71 | 0.48 | [0.99, 2.95] | |||
| Severe hunger | 1.35 | 0.25 | [0.93, 1.95] | 1.36 | 0.26 | [0.94, 1.97] | |||
| Level 2 | |||||||||
| Geography type | |||||||||
| Coastal (ref) | |||||||||
| Urban mountain | 0.88 | 0.14 | [0.64, 1.21] | 0.62* | 0.13 | [0.40, 0.94] | |||
| Rural plain | 0.93 | 0.13 | [0.71, 1.21] | 0.71* | 0.12 | [0.50, 1.00] | |||
| Rural mountain | 1.13 | 0.15 | [0.88, 1.46] | 1.02 | 0.17 | [0.73, 1.42] | |||
| Urban plain | 0.93 | 0.20 | [0.60, 1.43] | 0.79 | 0.22 | [0.46, 1.35] | |||
| Interaction effects | |||||||||
| Urban mountain × enrolled in 4Ps | 2.26* | 0.73 | [1.20, 4.24] | ||||||
| Rural plain × enrolled in 4Ps | 2.01* | 0.56 | [1.17, 3.46] | ||||||
| Rural mountain × enrolled in 4Ps | 1.29 | 0.34 | [0.78, 2.15] | ||||||
| Urban plain × enrolled in 4Ps | 1.46 | 0.68 | [0.59, 3.64] | ||||||
| Intercept | 0.98 | 0.06 | [0.87, 1.10] | 1.11 | 0.20 | [0.78, 1.59] | 1.32 | 0.27 | [0.89, 1.97] |
| Est. | SE | 95% CI | Est. | SE | 95% CI | Est. | SE | 95% CI | |
| Random effects | |||||||||
| Variance of random intercepts (regional base) | 0.02 | 0.04 | [−0.06, 0.01] | 0.01 | 0.04 | [−0.08, 0.26] | 0.01 | 0.47 | [−0.90, 0.92] |
| ICC | 0.01 | ||||||||
| Model information criteria | |||||||||
| Deviance | 4079.02 | 3963.90 | 3951.94 | ||||||
| AIC | 4083.02 | 3995.90 | 3991.94 | ||||||
Ni, 2945; where i = number of children; Nj, 10; where j = number of regional bases.
Abbreviations: 4Ps, Pantawid Pamilyang Pilipino Program; AIC, Akaike information criterion; CI, confidence interval; ICC, intraclass correlation coefficient; OR, odds ratio.
Continuous variables in the regression model are standardized and scaled to support model fit and convergence.
Model selected for interpretation.
P < .05. **P < .01. ***P < .001.
The standardized main effects model (Model 2) was determined as the most appropriate model for estimating the association between enrollment in 4Ps and wasting status of children based on deviance tests and evaluating the AIC (see Supplemental Appendix E). Overall, the sex and age of the child, the Household Hunger Scale, and geography type were associated with wasting status of children. While holding all other variables constant, the regression model indicated that there was no association between household enrollment in 4Ps and wasting status among children (adjusted OR=1.02; 95% CI [0.77, 1.36]).
The standardized interaction effects model (Model 3) was determined as the most appropriate model for estimating the association between enrollment in 4Ps and stunting status of children based on the deviance test statistic and the AIC (see Supplemental Appendix E). The sex of the child, the education level of the household head, the number of household members, and the Wealth Index score were associated with the stunting status of children. While holding all other variables constant, the odds of stunting were lower (adjusted OR=0.55; 95% CI [0.35, 0.88]) among children from households enrolled in 4Ps compared with children from households that were not enrolled in 4Ps. However, the odds of stunting differed among children in households enrolled in 4Ps when the interaction between geography type and household enrollment in 4Ps was included. Among children from households enrolled in 4Ps living in urban mountains, the odds of stunting were higher by a factor of 2.26 (95% CI [1.20, 4.24]) compared with children from households enrolled in 4Ps living in coastal areas. Similarly, children from households enrolled in 4Ps living in rural plains had higher odds of stunting by a factor of 2.01 (95% CI [1.17, 3.46]) compared with children from households enrolled in 4Ps living in coastal areas. While the odds of stunting were also higher for children from households enrolled in 4Ps living in rural mountains and urban plains, there was not a significant difference between these children compared to children from households enrolled in 4Ps living in coastal areas.
Discussion
This study investigated the associations between 4Ps and wasting status and stunting status among children experiencing extreme poverty in the Philippines. Findings from the descriptive statistics indicated that children in households enrolled in 4Ps belonged to larger households where the household head had fewer years of education. Moreover, the wealth status and experiences of food insecurity between children in households enrolled and not enrolled in 4Ps were similar. The multilevel logistic regression modeling indicated that household enrollment in 4Ps was significantly associated with stunting, but not with wasting. However, interaction effects between household enrollment in 4Ps and geography type indicated that geography type may be moderating the association between 4Ps and stunting status.
Differences were observed in the characteristics of children in households enrolled and not enrolled in 4Ps. Children in households enrolled in 4Ps may have a higher susceptibility to a poor nutritional status as they belonged to households that had household heads with fewer years of education and had more household members. These factors may be associated with less understanding of proper nutrition and limited resources directed to children to support healthy nutrition and development.16,17 In addition, similarities were observed in the Wealth Index score and the Household Hunger Scale between children in households enrolled and not enrolled in 4Ps. Given the cross-sectional design of this study, this may indicate that the cash transfers through 4Ps did not create significant differences in the wealth and food insecurity between children in households enrolled and not enrolled in 4Ps. Alternatively, it could also be hypothesized that 4Ps contributed to some improvements in the living conditions of recipient households, which could contribute to a similar wealth and food insecurity status between children in households enrolled and not enrolled in 4Ps. In addition, there may be similarities between the Wealth Index score of households enrolled and not enrolled in 4Ps because the sample was specifically targeted using an asset-based poverty score card, which may share common indicators used for constructing the Wealth Index.
The absence of an association between household enrollment in 4Ps and wasting status in children may suggest that the conditionalities and entitlements of the program are insufficient to provide socioeconomic resources that can address the underlying drivers of wasting. Alternatively, this finding may also suggest that 4Ps targeting does not reach households with children who are wasted. Our findings are consistent with previous evaluations of 4Ps using randomized controlled trial 12 and regression discontinuity study designs,26,27 which did not find significant associations between 4Ps enrollment and wasting in children. Theoretically, there is potential for cash transfer programs to direct more resources toward households experiencing poverty and enhance adaptive capacity prior to and in the midst of shocks. However, the present structure of conditional cash transfer programs (i.e., having fixed conditionalities and entitlements) may limit the ability of such programs to quickly respond to rapid shocks and emergencies that are associated with wasting.
Results from this study suggest that 4Ps may alleviate the underlying factors associated with chronic conditions of malnutrition. A previous randomized controlled trial found similar results wherein the prevalence of stunting was lower in areas where 4Ps was implemented compared to households in communities where 4Ps was not implemented. 12 Notably, the effect moderation from the interaction terms between geography type and household enrollment in 4Ps may point to the possible influence of geographic distance, infrastructure limitations, and gaps in health and social services on the capacity of 4Ps to support the nutrition of children across different contexts. Alternatively, the effect moderation observed in this study could also suggest that there are differences in underlying risk factors for stunting or that there is variability in the targeting mechanisms and outcomes for 4Ps enrollment and participation across regions.
The different relationships observed between household enrollment in 4Ps and wasting and stunting in children may be attributable to the varied combination of prevention and treatment approaches needed to effectively address these two conditions. Furthermore, 4Ps could theoretically contribute to reductions in wasting among children through the provision of cash and subsidies that can assist recipient households to cope with shocks that negatively affect food security and exacerbate illness. However, the conditionalities and entitlements provided through 4Ps may be inadequate to protect households from shocks that contribute to wasting among children. Indeed, a previous study in six Southeast Asian countries showed that higher household wealth was associated with lower odds of wasting among children and poor sanitation was associated with higher odds of wasting among children. 28 These findings highlight the importance of leveraging opportunities to strengthen the coordination and collaboration between 4Ps and existing nutrition-specific interventions to more effectively address acute malnutrition if the intended mechanisms of 4Ps to prevent wasting are insufficient. Conversely, as stunting is associated with persistent poverty and deprivation, enrollment in 4Ps may be able to support improvements to the underlying social determinants that are correlated with stunting in children. In addition, the use of a proxy means test to target 4Ps beneficiaries may be an effective approach to reach households that experience various intersecting conditions that are associated with stunting. Similar findings have been observed in Indonesia, where a cluster randomized study showed that conditional cash transfers was associated with increases in food consumption of milk and fish. 29 However, as indicated by the significance of the interaction term between household enrollment in 4Ps and geography type, it may be that these associations are not equally experienced across different areas in the Philippines.
The findings from this study contribute to the limited literature on the potential effects of cash transfer programs on child health in the Southeast Asia and Pacific region. Whereas a previous review has shown that enrollment in cash transfer programs was associated with higher HAZ and lower prevalence of stunting, only 20% of the studies were from Asia, and only Indonesia and the Philippines were included from the Southeast Asia subregion. 4 Moreover, research has shown that child health in Southeast Asia is associated with socioeconomic inequality within countries, and that poor health outcomes are disproportionately experienced by children in households experiencing poverty. 30 As the implementation and evaluation of conditional cash transfers expand in Southeast Asia, it is critical to shape and strengthen the conditionalities and entitlements of such programs to support and improve the underlying social determinants of child health and nutrition.
Although this study applied multilevel regression modeling and included a large sample size from a population that has been excluded from previous evaluations of 4Ps, there are also several limitations that should be noted. First, the sample in this study was not randomly selected and is not representative of all households enrolled in 4Ps in the Philippines. In addition, the number of groups for regional areas in our data set was limited, which may have affected the precision of our estimates. However, this study sample provides insight on health and nutritional outcomes among households experiencing extreme poverty, which is a population that is often underrepresented in national level surveys and evaluations. Second, the use of a cross-sectional study design does not allow for temporal and causal interpretation of the associations between household enrollment in 4Ps and the nutritional outcomes of children. In addition, the length of time that participants were enrolled in the 4Ps program was not available, further requiring longitudinal data to confirm the findings from this study. Third, as the Household Hunger Scale assesses severe experiences of hunger and food insecurity, applying it within a context where populations experience chronic poverty may lead to underestimated measures. However, the Household Hunger Scale was included in the analysis as previous research has indicated that cash transfers are associated with nutritional outcomes among children through reducing food insecurity. 3 Fourth, several covariates (e.g., sex of household head, anthropometry of mother) were not available for the study, which could have provided additional insight into the nature of wasting and stunting in the study context. Future research could collect data on additional determinants of child nutrition and sociocultural factors that shape dietary intake. Finally, as household enrollment in 4Ps was a self-reported variable in our study, there may have been misclassification bias in the analysis.
Conclusion
This study investigated the association between household enrollment in 4Ps with wasting and stunting among children experiencing poverty in the Philippines. Importantly, this study included a population experiencing extreme poverty that has been excluded from previous evaluations of the relationship between 4Ps and child malnutrition. Household enrollment in 4Ps was not associated with wasting, but was associated with stunting and was moderated by geography type. Future research using longitudinal data is necessary to confirm the findings regarding wasting and to further investigate the factors underlying the differences in stunting by geography type to inform more effective programming of 4Ps.
Supplemental Material
Supplemental material, sj-doc-1-aph-10.1177_10105395231189570 for Examining the Association Between Household Enrollment in the Pantawid Pamilyang Pilipino Program (4Ps) and Wasting and Stunting Status Among Children Experiencing Poverty in the Philippines: A Cross-Sectional Study by Monica Bustos, Lincoln L. Lau, Sharon I. Kirkpatrick, Joel A. Dubin, Helena Manguerra and Warren Dodd in Asia Pacific Journal of Public Health
Acknowledgments
The authors are grateful to International Care Ministries (ICM) for providing institutional support for this study, in addition to Kendall Wilson, Mia Choi, Joy Kimmel, and Daryn Go for their support in information management and data access. The authors would also like to acknowledge Nikki Abrera and Jesce dela Cruz for their insight on the ICM child nutrition program. Additionally, the authors acknowledge the enumerators, surveyors, and ICM staff who collected the data for this project. Finally, the authors are grateful for the participation of the children and their household members who provided information for this study.
Footnotes
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Authors L.L.L. and H.M. receive remuneration from International Care Ministries (ICM). The authors have been provided academic freedom by ICM to publish both negative and positive results.
Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.
ORCID iD: Warren Dodd 
https://orcid.org/0000-0003-0774-7644
Supplemental Material: Supplemental material for this article is available online.
References
- 1. World Health Organization. Levels and trends in child malnutrition. Published May 2021. Accessed October 30, 2022. https://www.who.int/publications/i/item/9789240025257
- 2. United Nations Children’s Fund. Improving Child Nutrition: The achievable imperative for global progress. New York, NY: UNICEF; 2013. Accessed October 30, 2022. https://data.unicef.org/resources/improving-child-nutrition-the-achievable-imperative-for-global-progress/ [Google Scholar]
- 3. Leroy JL, Ruel M, Verhofstadt E. The impact of conditional cash transfer programmes on child nutrition: a review of evidence using a programme theory framework. J Dev Eff. 2009;1(2):103-129. doi: 10.1080/19439340902924043 [DOI] [Google Scholar]
- 4. Manley J, Alderman H, Gentilini U. More evidence on cash transfers and child nutritional outcomes: a systematic review and meta-analysis. BMJ Glob Health. 2022;7(4):e008233. doi: 10.1136/bmjgh-2021-008233 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Global Nutrition Report. The burden of malnutrition at a glance. Philippines: Global Nutrition Report. Date unknown. Accessed September 15, 2021. https://globalnutritionreport.org/resources/nutrition-profiles/asia/south-eastern-asia/philippines/ [Google Scholar]
- 6. Department of Science and Technology—Food and Nutrition Research Institute. Philippine Nutrition Facts and Figures 2018 Expanded National Nutrition Survey (ENNS). Manila, Philippines: DOST-FNRI; 2020. Accessed October 30, 2021. http://enutrition.fnri.dost.gov.ph/site/uploads/2018_ENNS_Facts_and_Figures.pdf [Google Scholar]
- 7. Department of Social Welfare and Development. Guidelines on the Implementation of Pantawid Pamilyang Filipino Program (4Ps). Philippines: DSWD; 2008. Accessed September 15, 2021. https://www.dswd.gov.ph/issuances/AOs/AO_2008-016.pdf [Google Scholar]
- 8. Department of Social Welfare and Development. Listahanan Info Kit. Manila: DSWD; 2019. Accessed September 15, 2021. https://listahanan.dswd.gov.ph/wp-content/uploads/2019/11/listahanan_info_kit_7.pdf [Google Scholar]
- 9. Cho Y, Avalos J, Kawasoe Y, Rodriguez R. Optimizing Pantawid for Nutrition. Washington, DC: World Bank; 2020. Accessed September 15, 2021. https://openknowledge.worldbank.org/handle/10986/34784 [Google Scholar]
- 10. Acosta PA, Velarde R. “Sa Pantawid, Malapit Nang Makatawid!” (With Pantawid, We Are Closer to Getting Out of Poverty!): An Update of the Philippine Conditional Cash Transfer’s Implementation Performance. Washington, DC: World Bank; 2015. [Google Scholar]
- 11. Peñalba E. Exploring the health outcomes of conditional cash transfer program in rural Philippines. J Soc Work Educ Pract. 2019;4(3):37-51. doi: 10.31235/osf.io/4wzx7 [DOI] [Google Scholar]
- 12. Kandpal E, Alderman H, Friedman J, Filmer D, Onishi J, Avalos J. A conditional cash transfer program in the Philippines reduces severe stunting. J Nutr. 2016;146(9):1793-1800. doi: 10.3945/jn.116.233684 [DOI] [PubMed] [Google Scholar]
- 13. Talavera M, Gordoncillo N, Tandang N, Domingo D. Acceptability of height measuring equipment of different materials among community nutrition and health workers and parents in Laguna province, Philippines. Acta Med Philipp. 2021;56(3):30-36. doi: 10.47895/amp.vi0.3100 [DOI] [Google Scholar]
- 14. Myatt M, Guevarra E. z scorer: Child Anthropometry z-Score Calculator. Vienna, Austria: R Foundation; 2019. Accessed September 1, 2021. https://cran.r-project.org/web/packages/zscorer/zscorer.pdf [Google Scholar]
- 15. World Health Organization. WHO child growth standards: length/height-for-age, weight-for-age, weight-for-length, weight-for-height and body mass index-for-age: methods and development. Published November 2006. Accessed September 1, 2021. https://www.who.int/publications/i/item/924154693X
- 16. Ieiri MCA, Kosaka S, Tomitsuka E, Umezaki M. Factors affecting undernutrition among school children in Cebu, Philippines. Ecol Food Nutr. 2021;60(2):182-197. doi: 10.1080/03670244.2020.1813733 [DOI] [PubMed] [Google Scholar]
- 17. Wolde M, Berhan Y, Chala A. Determinants of underweight, stunting and wasting among schoolchildren. BMC Public Health. 2015;15:8. doi: 10.1186/s12889-014-1337-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Vyas S, Kumaranayake L. Constructing socio-economic status indices: how to use principal components analysis. Health Policy Plan. 2006;21(6):459-468. doi: 10.1093/heapol/czl029 [DOI] [PubMed] [Google Scholar]
- 19. Ballard T, Coates J, Swindale A, Deitchler M. Household Hunger Scale: Indicator Definition and Measurement Guide. Washington, DC: Food and Nutrition Technical Assistance; 2011. Accessed September 1, 2021. https://www.fantaproject.org/sites/default/files/resources/HHS-Indicator-Guide-Aug2011.pdf [Google Scholar]
- 20. Capanzana MV, Aguila DV, Gironella GMP, Montecillo KV. Nutritional status of children ages 0-5 and 5-10 years old in households headed by fisherfolks in the Philippines. Arch Public Health. 2018;76:24-28. doi: 10.1186/s13690-018-0267-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Conchada MIP, Tiongco M. Optimal allocation of school and health resources for effective delivery of the conditional cash transfer program in Bagac, Bataan. DLSU Res Congr La Salle Univ Manila. 2014;6:1-10. [Google Scholar]
- 22. Dodd W, Kipp A, Lau L, et al. Limits to transformational potential: analysing entitlement and agency within a conditional cash transfer program in the Philippines. Soc Policy Soc. 2022:1-18. doi: 10.1017/S1474746422000215 [DOI] [Google Scholar]
- 23. Raudenbush SW, Bryk AS. Hierarchical Linear Models: Applications and Data Analysis Methods. California: Sage Publications; 2002. doi: 10.2307/2290750 [DOI] [Google Scholar]
- 24. R Core Team. R: A language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing; 2019. Accessed September 1, 2021. https://www.r-project.org/ [Google Scholar]
- 25. Bates D, Mächler M, Bolker BM, Walker SC. Fitting linear mixed-effects models using lme4. J Stat Softw. 2015;67(1):1-48. doi: 10.18637/jss.v067.i01 [DOI] [Google Scholar]
- 26. Orbeta A, Abdon A, del Mundo M, Tutor M, Valera M, Yarcia D. Evaluating the Pantawid Pamilya Using Regression Discontinuity Design Second Wave Impact Evaluation Results. Metro Manila: DSWD; 2014. Accessed July 11, 2023. https://pdpb.dswd.gov.ph/download/evaluating-the-pantawid-pamilya-using-regression-discontinuity-design-second-wave-impact-evaluation-results/?wpdmdl=1687&refresh=64acf0efb822a1689055471 [Google Scholar]
- 27. Orbeta A, Melad K, Araos N. Reassessing the Impact of the Pantawid Pamilyang Pilipino Program: Results of the Third Wave Impact Evaluation (IE Wave 3). Quezon City: Philippine Institute for Development Studies; 2021. Accesssed July 1, 2022. https://www.pids.gov.ph/publication/discussion-papers/reassessing-the-impact-of-the-pantawid-pamilyang-pilipino-program-results-of-the-third-wave-impact-evaluation [Google Scholar]
- 28. Mutunga M, Frison S, Rava M, Bahwere P. The forgotten agenda of wasting in Southeast Asia: burden, determinants and overlap with stunting: a review of nationally representative cross-sectional demographic and health surveys in six countries. Nutrients. 2020;12(2):559. doi: 10.3390/nu12020559 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Kusuma D, McConnell M, Berman P, Cohen J. The impact of household and community cash transfers on children’s food consumption in Indonesia. Prev Med. 2017;100:152-158. doi: 10.1016/j.ypmed.2017.04.020 [DOI] [PubMed] [Google Scholar]
- 30. Acuin CS, Khor GL, Liabsuetrakul T, et al. Maternal, neonatal, and child health in Southeast Asia: towards greater regional collaboration. Lancet. 2011;377(9764):516-525. doi: 10.1016/S0140-6736(10)62049-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplemental material, sj-doc-1-aph-10.1177_10105395231189570 for Examining the Association Between Household Enrollment in the Pantawid Pamilyang Pilipino Program (4Ps) and Wasting and Stunting Status Among Children Experiencing Poverty in the Philippines: A Cross-Sectional Study by Monica Bustos, Lincoln L. Lau, Sharon I. Kirkpatrick, Joel A. Dubin, Helena Manguerra and Warren Dodd in Asia Pacific Journal of Public Health