Program Impact Pathways and Contexts: A Commentary on Theoretical Issues and Research Applications to Support the EsIAN Component of Mexico's Conditional Cash Transfer Program.

This commentary on the Integrated Strategy for Attention to Nutrition (EsIAN) journal supplement begins with a discussion about the challenges that implementation researchers confront with respect to analyzing complex impact pathways. We note that the research on the implementation of the EsIAN component of Mexico's conditional cash transfer program was based implicitly or explicitly on a program impact pathway approach, which used both quantitative and qualitative methods to examine bottlenecks in program implementation. We then identify 5 categories of contexts that affect the impact, implementation, and survival of intervention programs: 1) biological, 2) social-cultural, 3) delivery modalities and platforms, 4) bureaucratic, and 5) political. Each of these contexts presents theoretical and methodological challenges for investigators. In this commentary, we focus primarily on biological and social-cultural contexts, discussing the theoretical and methodological challenges the investigators faced and the research strategies they used to address them, which have produced a unique compilation of "learning by doing" studies. We also touch briefly on the political context in which the Prospera program research was conducted. We conclude with statements that highlight the exceptional value of the journal supplement, not only with respect to the analysis of the interventions the studies cover and the sustained examination of a long-term program but also as a major contribution to the literature in implementation science in nutrition.

A Fortified Food Can Be Replaced by Micronutrient Supplements for Distribution in a Mexican Social Protection Program Based on Results of a Cluster-Randomized Trial and Costing Analysis.

BACKGROUND: Despite positive nutrition impacts, the prevalence of malnutrition among beneficiaries of Mexico's conditional cash transfer (CCT) program remains high. Greater nutrition impact may have been constrained by the type of nutritional supplements provided. OBJECTIVE: The objective of this study was to inform a potential modification to the supplements distributed to pregnant and lactating women and children. METHODS: Impact was assessed using 2 cluster-randomized trials (pregnant women, children) run simultaneously. Communities (n = 54) were randomly assigned to the fortified foods provided by the program (Nutrivida women, Nutrisano children) or alternatives: tablets (women), syrup (children), or micronutrient powders for women (MNP-W) and children (MNP-C). Each supplement for women/children contained the same micronutrients based on the formulations of Nutrivida and Nutrisano, respectively. Pregnant women (aged >18 y) were recruited before 25 weeks of gestation and followed to 3 mo postpartum. Children aged 6-12 mo were recruited and followed to age 24 mo. Primary outcomes were anemia for women and length growth for children. Statistical analyses appropriate for cluster-randomized designs were used, and structural equation modeling to estimate dose-response effects. Supplement costs per beneficiary (daily dose for 18 mo) were estimated for production and distribution. RESULTS: There was no significant difference in change of anemia prevalence between supplement groups in women, or in length growth between groups in children. One daily dose of any supplement was associated with 0.8 cm greater length growth. From baseline to age 24 mo, the prevalence of anemia in the Nutrisano, syrup, and MNP-C groups decreased by 36.7, 40.8, and 37.9 percentage points, respectively (within-group, P < 0.05; between groups, P > 0.05). Costs per beneficiary ranged from $12.1 (MNP-C) to $94.8 (Nutrivida). CONCLUSIONS: The CCT program could distribute alternative supplements at lower cost per beneficiary without compromising potential for impact. Acceptance among beneficiaries should also be considered in choice of alternatives. This trial was registered at www.clinicaltrials.gov as NCT00531674.

Bottlenecks and predictors of coverage and adherence outcomes for a micronutrient powder program in Ethiopia.

A theory-driven evaluation was conducted to assess performance of a trial to deliver micronutrient powder (MNP) through the Ethiopian Ministry of Health. We adapted an approach to coverage assessment, originally developed to identify bottlenecks in health service delivery, to examine sequential program outcomes and their correlates using cross-sectional survey data of caregivers of children 6-23 months (N = 1915). Separate multivariable Poisson regression models were used to estimate adjusted risk ratios of conceptually relevant determinants of coverage and adherence. Caregivers of children >11 months were more likely to have received MNP than caregivers of younger infants, yet children 12-17 months were 32% (P < 0.001) and children 18-23 months 38% (P < 0.001) less likely to have been fed MNP in the 14 days preceding the survey than children 6-11 months. Among caregivers who initiated feeding MNP, the most frequently reported reasons for discontinuing use were not obtaining additional supply (36.1%) and perceived child rejection of food with MNP (22.9%). For each additional time a caregiver met with frontline workers in the 3 months preceding the survey, they were 13% more likely to have recently fed MNP (P < 0.001). Caregivers' perception that MNP produced positive changes in children was associated with a 14% increase in the likelihood of having recently fed it (P < 0.001). These results emphasize the importance of counselling for MNP and infant and young child feeding for initial use and the importance of multiple contacts with frontline workers for continued use.

Identifying Nutrition and Health-Relevant Behaviors, Beliefs, and Values of School-Going Adolescent Girls in Rural Bangladesh: Context for Interventions.

BACKGROUND: Focusing on adolescent schoolgirls in rural Bangladesh, this study aimed to describe the nutrition-relevant context required for effective intervention planning. It included attention to dietary behaviors, daily schedules and activities, and girls' beliefs and values. We placed a special emphasis on iron because anemia has been identified as a significant problem in adolescent girls in Bangladesh. OBJECTIVES: The study was undertaken to inform a larger project to develop a replicable model for integrating nutrition activities into the multiple social programs in BRAC, a large nongovernmental service delivery organization in Bangladesh. METHODS: Following an initial phase of exploratory key informant interviewing, data collection was conducted through the use of focused ethnographic methods in 2 additional phases. Phase II consisted of in-depth interviews with school-going adolescents in rural communities in Rangpur District (n = 23). Interview modules included sociodemographic information; a qualitative 24-h dietary recall; respondents' concepts of "health," "healthy foods," and "anemia"; exposure to nutrition messages and ratings of their importance and feasibility; and actions to maintain health. In phase III, ratings and other data were collected systematically from a separate sample of 20 adolescent girls. Analyses included thematic analysis of transcribed and translated text and quantitative analysis of numeric data. RESULTS: Key findings include evidence that school-going girls consume a variety of foods and have substantial knowledge about good health and nutrition practices. However, we also found beliefs and practices that challenge the development of nutritionally sound food practices, including iron nutrition. The study revealed the importance of teachers and mothers as information sources, and the dietary and cultural roles of purchased snacks, which occur in connection with school attendance. CONCLUSIONS: These findings, together with insights about girls' values, particularly conceptions related to "a healthy life," have implications for expanding BRAC activities to support the nutrition of adolescent school girls.

The Pathways from a Behavior Change Communication Intervention to Infant and Young Child Feeding in Bangladesh Are Mediated and Potentiated by Maternal Self-Efficacy.

Background: Although self-efficacy is a potential determinant of feeding and care behaviors, there is limited empirical analysis of the role of maternal self-efficacy in low- and middle-income countries. In the context of behavior change interventions (BCIs) addressing complementary feeding (CF), it is possible that maternal self-efficacy can mediate or enhance intervention impacts. Objective: In the context of a BCI in Bangladesh, we studied the role of maternal self-efficacy for CF (MSE-CF) for 2 CF behaviors with the use of a theoretically grounded empirical model of determinants to illustrate the potential roles of MSE-CF. Methods: We developed and tested a locally relevant scale for MSE-CF and included it in a survey (n = 457 mothers of children aged 6-24 mo) conducted as part of a cluster-randomized evaluation. Qualitative research was used to inform the selection of 2 intervention-targeted behaviors: feeding green leafy vegetables in the last 24 h (GLV) and on-time introduction of egg (EGG) between 6 and 8 mo of age. We then examined direct, mediated, and potentiated paths of MSE-CF in relation to the impacts of the BCI on these behaviors with the use of regression and structural equation modeling. Results: GLV and EGG were higher in the intensive group than in the nonintensive control group (16.0 percentage points for GLV; P < 0.001; 11.2 percentage points for EGG; P = 0.037). For GLV, MSE-CF mediated (ß = 0.345, P = 0.010) and potentiated (ß = 0.390, P = 0.038) the effect of the intensive group. In contrast, MSE-CF did not mediate or potentiate the effect of the intervention on EGG. Conclusions: MSE-CF was a significant mediator and potentiator for GLV but not for EGG. The divergent findings highlight the complex determinants of individual specific infant and young child feeding behaviors. The study shows the value of measuring behavioral determinants, such as MSE-CF, that affect a caregiver's capability to adopt intervention-targeted behaviors.

Addressing epidemiological and public health analytic challenges in outcome and impact research: a commentary on 'Prechewing Infant Food, Consumption of Sweets and Dairy and Not Breastfeeding are Associated with Increased Diarrhea Risk of Ten Month Old Infants'.

Based on a paper by Conkle et al 2016, in which the authors use a descriptive epidemiological design to examine the relationship of premastication and other dietary behavioral variables to childhood diarrhea in the US, we address larger issues of "plausible causality" and the challenges involved in moving from epidemiological studies to public health policy. Drawing on examples from breastfeeding research and water, sanitation and hygiene (WASH) research, we discuss the following propositions: 1. Effective outcome analyses require simultaneous investigation of different, even contradictory, pathways; 2. Outcome versus impact assessments require different analytic procedures including context analysis; 3. Impact analysis requires understanding the trade-offs between detrimental and beneficial outcomes in relation to potential interventions; 4. No estimates exist for the likely detrimental and beneficial impacts of banning premastication, much less for their trade-offs.

Using height-for-age differences (HAD) instead of height-for-age z-scores (HAZ) for the meaningful measurement of population-level catch-up in linear growth in children less than 5 years of age.

BACKGROUND: Evidence from studies conducted in nutritionally deprived children in low- and middle-income countries (LIMC) in past decades showed little or no population-level catch-up in linear growth (mostly defined as reductions in the absolute height deficit) after 2 years of age. Recent studies, however, have reported population-level catch-up growth in children, defined as positive changes in mean height-for-age z-scores (HAZ). The aim of this paper was to assess whether population-level catch-up in linear growth is found when height-for-age difference (HAD: child's height compared to standard, expressed in centimeters) is used instead of HAZ. Our premise is that HAZ is inappropriate to measure changes in linear growth over time because they are constructed using standard deviations from cross-sectional data. METHODS: We compare changes in growth in populations of children between 2 and 5 years using HAD vs. HAZ using cross-sectional data from 6 Demographic and Health Surveys (DHS) and longitudinal data from the Young Lives and the Consortium on Health-Orientated Research in Transitional Societies (COHORTS) studies. RESULTS: Using HAD, we find not only an absence of population-level catch-up in linear growth, but a continued deterioration reflected in a decrease in mean HAD between 2 and 5 years; by contrast, HAZ shows either no change (DHS surveys) or an improvement in mean HAZ (some of the longitudinal data). Population-level growth velocity was also lower than expected (based on standards) in all four Young Lives data sets, confirming the absence of catch-up growth in height. DISCUSSION: We show no evidence of population-level catch-up in linear growth in children between 2 to 5 years of age when using HAD (a measure more appropriate than HAZ to document changes as populations of children age), but a continued deterioration reflected in a decrease in mean HAD. CONCLUSIONS: The continued widening of the absolute height deficit after 2 years of age does not challenge the critical importance of investing in improving nutrition during the first 1000 days (i.e., from conception to 2 years of age), but raises a number of research questions including how to prevent continued deterioration and what is the potential of children to benefit from nutrition interventions after 2 years of age. Preventing, rather than reversing linear growth retardation remains the priority for reducing the global burden of malnutritionworldwide.

Gender perceptions predict sex differences in growth patterns of indigenous Guatemalan infants and young children.

BACKGROUND: Nearly one-half of Guatemalan children experience growth faltering, more so in indigenous than in nonindigenous children. OBJECTIVES: On the basis of ethnographic interviews in Totonicapán, Guatemala, which revealed differences in maternal perceptions about food needs in infant girls and boys, we predicted a cumulative sex difference in favor of girls that occurred at ∼6 mo of age and diminished markedly thereafter. We examined whether the predicted differences in age-sex patterns were observed in the village, replicated the examination nationally for indigenous children, and examined whether the pattern in nonindigenous children was different. DESIGN: Ethnographic interviews (n = 24) in an indigenous village were conducted. Anthropometric measurements of the village children aged 0-35 mo (n = 119) were obtained. National-level growth patterns were analyzed for indigenous (n = 969) and nonindigenous (n = 1374) children aged 0-35 mo with the use of Demographic and Health Survey (DHS) data. RESULTS: Mothers reported that, compared with female infants, male infants were hungrier, were not as satisfied with breastfeeding alone, and required earlier complementary feeding. An anthropometric analysis confirmed the prediction of healthier growth in indigenous girls than in indigenous boys throughout the first year of life, which resulted in a 2.98-cm height-for-age difference (HAD) between sexes in the village and a 1.61-cm HAD (P < 0.001) in the DHS data between 6 and 17 mo of age in favor of girls. In both data sets, the growth sex differences diminished in the second year of life (P < 0.05). No such pattern was seen in nonindigenous children. CONCLUSIONS: We propose that the differences in the HAD that first favor girls and then favor boys in the indigenous growth patterns are due to feeding patterns on the basis of gendered cultural perceptions. Circumstances that result in differential sex growth patterns need to be elucidated, in particular the favorable growth in girls in the first year of life.

Linear growth deficit continues to accumulate beyond the first 1000 days in low- and middle-income countries: global evidence from 51 national surveys.

Growth faltering is usually assessed using height-for-age Z-scores (HAZs), which have been used for comparisons of children of different age and sex composition across populations. Because the SD (denominator) for calculating HAZ increases with age, the usefulness of HAZs to assess changes in height over time (across ages) is uncertain. We posited that population-level changes in height as populations age should be assessed using absolute height-for-age differences (HADs) and not HAZs. We used data from 51 nationwide surveys from low- and middle-income countries and graphed mean HAZs and HADs by age. We also calculated annual changes in HAZs and HADs and percentage of total height deficit accumulated annually from birth to age 60 mo using both approaches. Mean HAZ started at -0.4 Z-scores and dropped dramatically up to 24 mo, after which it stabilized and had no additional deterioration. Mean HAD started at -0.8 cm, with the most pronounced faltering occurring between 6 and 18 mo, similar to HAZ. However, in sharp contrast to HAZ, HAD curves had continued increases in the deficit of linear growth from 18 to 60 mo, with no indication of a leveling off. Globally, 70% of the absolute deficit accumulated in height (HAD) at 60 mo was found to be due to faltering during the first "1000 days" (conception to 24 mo), but 30% was due to continued increases in deficit from age 2 to 5 y. The use of HAZ masks these changes because of age-related changes in SD. Therefore, HAD, rather than HAZ, should be used to describe and compare changes in height as children age because detecting any deficit compared with expected changes in height as children grow is important and only HAD does this accurately at all ages. Our findings support the current global programmatic momentum to focus on the first 1000 d. Research is needed to better understand the dynamics and timing of linear growth faltering using indices and indicators that accurately reflect changes over ages and to identify cost-effective ways to prevent growth faltering and its consequences throughout the lifecycle.

Maternal education mitigates the negative effects of higher income on the double burden of child stunting and maternal overweight in rural Mexico.

Globally, the rate at which maternal overweight and obesity increase with rising wealth is higher than the accompanying decline in the prevalence of child stunting, resulting in the double burden of malnutrition. The positive association between household wealth and child linear growth is larger in households with a more educated mother. However, whether a similar positive and synergistic association between maternal education and household wealth is observed for maternal body weight is unknown. Our objective was to assess the potential protective role of maternal education in the etiology of the double burden of malnutrition (stunted child with an overweight mother). We used data on 1547 nonpregnant mothers (aged 18-49 y) and their children (aged 0-5 y) collected in a cross-sectional survey in 235 rural communities in southern Mexico. Child height-for-age Z-score and maternal body weight were regressed on household wealth, women's schooling, and the interaction between both, controlling for relevant covariates. A similar model was used for the prevalence of double-burden pairs (stunted child with an overweight mother). In mothers with less than primary school, a doubling in wealth was not associated with improved child's height but was associated with an increase in mother's weight (3.7%, P < 0.01). In mothers who had completed primary school, the reverse was found: a doubling in wealth score was associated with improved child height-for-age Z-score (0.32 SD, P < 0.01) but not with mother's weight. As a result, a 100% increase in wealth among households with less schooled mothers was associated with a 4.5 percentage point increase (P < 0.05) in double-burden pairs; in households with mothers with primary schooling or more, it was not associated with the occurrence of double-burden pairs. Maternal schooling effectively mitigated the negative effects of household wealth on the prevalence of double-burden households in rural Mexico. Where maternal schooling is low, poverty reduction must be accompanied by effective behavior change communication to prevent child stunting and to protect women from unhealthy weight gain.

From biological to program efficacy: promoting dialogue among the research, policy, and program communities.

The biological efficacy of nutritional supplements to complement usual diets in poor populations is well established. This knowledge rests on decades of methodologic research development and, more recently, on codification of methods to compile and interpret results across studies. The challenge now is to develop implementation (delivery) science knowledge and achieve a similar consensus on efficacy criteria for the delivery of these nutrients by public health and other organizations. This requires analysis of the major policy instruments for delivery and well-designed program delivery studies that examine the flow of a nutrient through a program impact pathway. This article discusses the differences between biological and program efficacy, and why elucidating the fidelity of delivery along the program impact pathways is essential for implementing a program efficacy trial and for assessing its internal and external validity. Research on program efficacy is expanding, but there is a lack of adequate frameworks to facilitate the process of harmonizing concepts and vocabulary, which is essential for communication among scientists, policy planners, and program implementers. There is an urgent need to elaborate these frameworks at national and program levels not only for program efficacy studies but also for the broader research agenda to support and improve the science of delivering adequate nutrition to those who need it most.

Maternal knowledge after nutrition behavior change communication is conditional on both health workers' knowledge and knowledge-sharing efficacy in rural Haiti.

In the context of a food assistance program in rural Haiti, we developed measures of the effectiveness of community health worker (CHW)-delivered behavior change communication (BCC). We administered knowledge tests to 954 mothers and 38 CHWs to define 4 measures: CHW knowledge, maternal knowledge, knowledge-sharing efficacy (proportion of CHW knowledge shared), and shared correct knowledge between the CHWs and the mothers with whom they interacted. On the basis of the tests, CHWs had high knowledge (93% correct), mothers scored 72% on maternal knowledge, the proportion of CHW knowledge shared was 75%, and shared correct knowledge between CHWs and mothers was 70%. Factors affecting maternal knowledge included CHW characteristics (unmarried: ß = -0.070, P < 0.05), long duration (19-45 mo) of program participation (ß = 0.034, P < 0.05), and having multiple sources of health and nutrition information (ß = 0.072, P < 0.01). Shared correct knowledge and CHW knowledge-sharing efficacy were positively associated with CHW (age, education) and program participation characteristics. We parsed the relative contributions of CHW characteristics to total and proportion of shared CHW knowledge. We observed a positive association between CHW education and shared correct knowledge between the CHWs and mothers (ß = 0.328, P < 0.01), not because of more knowledge (ß = -0.012, P > 0.05) but rather because of greater knowledge-sharing efficacy (ß = 0.340, P < 0.01). These measures provide a means of examining features of program delivery and interpreting impact evaluation results. They show that BCC outcomes are conditional on both frontline workers' knowledge and knowledge-sharing efficacy. Whereas most programs focus on content training to improve CHWs' knowledge, it is also important to strengthen process training and support to foster knowledge-sharing efficacy.

The WHO evidence-informed guideline development process and implications for vitamin and mineral research priorities: symposium rationale and summary.

The WHO evidence-informed guidelines provide recommendations to Member States and their partners on interventions with vitamins and minerals. Evidence gathered and synthesized through systematic reviews contributes to the development of these guidelines, a process that is dependent on the availability and quality of evidence. Although the guideline development process is stringently governed and supervised to maintain clarity and transparency, the lack of adequacy and specificity of available evidence poses limitations to the formulation of recommendations that can be easily applied for policy and program decision making in diverse contexts. The symposium created a space for dialogue among scientists and public health practitioners to improve the understanding of how evidence fulfills the needs and reflect on mechanisms by which policy and program guidance and priorities for research could be better informed by policy and program needs. Ultimately, programmatic success depends not only on identifying efficacious agents but ensuring effective delivery to those with the potential to respond. To do this, we must understand the rationale for recommending interventions, the biological pathways by which interventions work, delivery systems required to make efficacious interventions work, and other contextual factors that might limit or facilitate successful implementation.

Prenatal micronutrient supplements cumulatively increase fetal growth.

Prenatal multiple micronutrients (UNIMMAP) improve fetal growth only moderately compared to iron and folic acid alone (IFA). Whether this is due to insufficient amounts of UNIMMAP or to IFA being in reality an active control is unknown. We assessed the association between cumulative micronutrient intake (CMI) and fetal growth by secondary analysis of a randomized controlled trial in Burkina Faso where tablet intake was directly observed. We applied 2-part residual regression models adjusted for main confounders. Among the 1056 single pregnancies included, the mean CMI (± SD) was 124 ± 54 tablets. The odds of delivering a small-for-gestational-age baby was reduced by 21% [(95%CI: 5, 35); P = 0.013] for each additional tertile of CMI. The association between CMI and birth weight was positively modified by gestational age at enrollment (P-interaction = 0.001). Each unit of CMI was associated with a 1.6-g [(95%CI: 0.3, 3.1); P = 0.019] higher birth weight at a mean-centered gestational age at enrollment, with a higher gradient observed later in pregnancy. Maternal BMI at enrollment was also a positive modifying factor (P-interaction = 0.02), with no association of CMI with birth weight for low BMI. There was no evidence of an effect modification by group allocation; i.e., we observed the same change in birth weight per unit of CMI with either IFA or UNIMMAP. Yet UNIMMAP increased birth weight by 69 g [(95%CI: 58, 81); P < 0.001] relative to IFA. We found similar results for thoracic and cephalic circumferences. In conclusion, for both IFA and UNIMMAP, the effect on fetal growth is cumulative. The supplementation should therefore begin as early as possible in pregnancy, even if the growth increment per CMI is higher in late than in early pregnancy. Women with a low BMI should also receive extra energy.

Multiple micronutrient interventions are efficacious, but research on adequacy, plausibility, and implementation needs attention.

Evidence for the efficacy of multiple micronutrient (MMN) supplementation has been established by state-of-the-art randomized controlled trials (RCT). These efficacy trials have also provided strong evidence of the widespread occurrence of deficiencies. Trials intended to demonstrate a public health benefit must show that the magnitude of benefit is adequate for policy considerations. In the MMN efficacy studies the magnitude of impact was generally inadequate. The extent to which this was due to various factors that affect individuals' potential to respond to MMN supplementation has not been examined, and trials have not been designed to provide insights into why impact is below expectation. For instance, when birth weight was the outcome of concern, impact was not related to presumed need, judged by baseline birth-weight values. Also, contrary to expectations, the impacts were greater among the heavier mothers. Our inability to examine issues of adequacy, plausibility, and implementation with efficacy trial approaches as they are currently conducted, calls into question the present standards for designing and interpreting community RCT in nutrition. Improving the capacity of efficacy studies to yield more meaningful data requires a number of modifications to current practices, such as including measures of the intermediary behavioral and biological steps between intervention and biological outcomes to assess the adequacy and plausibility of the findings. The progression from RCT to program interventions must also extend research to program delivery and uptake to ascertain the full program impact pathway. This in turn requires novel organizations of relationships between research and program development and implementation.

Enhancing the intrinsic work motivation of community nutrition educators: how supportive supervision and job design foster autonomy.

Mixed-methods research investigated the work motivation of paraprofessional community nutrition educators (CNEs) delivering a long-running public health nutrition program. In interviews, CNEs (n = 9) emphasized "freedom," supportive supervision, and "making a difference" as key sources of motivation. Community nutrition educator surveys (n = 115) confirmed high levels of autonomy, which was associated with supervisors' delegation and support, CNE decision-making on scheduling and curricula, and job satisfaction. Supervisors (n = 32) rated CNEs' job design as having inherently motivating characteristics comparable to professional jobs. Supervisory strategies can complement job design to create structured, supportive contexts that maintain fidelity, while granting autonomy to paraprofessionals to enhance intrinsic work motivation.