Search

Custer randomized trials are often used to study large-scale public health interventions. In large trials, even small improvements in statistical efficiency can have profound impacts on the required sample size and cost. Pair matched randomization is one strategy with potential to increase trial efficiency, but to our knowledge there have been no empirical evaluations of pair-matching in large-scale, epidemiologic field trials. Location integrates many socio-demographic and environmental characteristics into a single feature. Here, we show that geographic pair-matching leads to substantial gains in statistical efficiency for 14 child health outcomes that span growth, development, and infectious disease through a re-analysis of two large-scale trials of nutritional and environmental interventions in Bangladesh and Kenya. We estimate relative efficiencies ≥ 1.1 for all outcomes assessed and relative efficiencies regularly exceed 2.0, meaning an unmatched trial would have needed to enroll at least twice as many clusters to achieve the same level of precision as the geographically pair-matched design. We also show that geographically pair-matched designs enable estimation of fine-scale, spatially varying effect heterogeneity under minimal assumptions. Our results demonstrate broad, substantial benefits of geographic pair-matching in large-scale, cluster randomized trials.

BackgroundUnderstanding pathogen-specific relationships with climate is crucial to informing interventions under climate change.MethodsWe matched spatiotemporal temperature, precipitation, surface water, and humidity data to data from a trial in rural Bangladesh that measured diarrhea and enteropathogen prevalence in children 0-2 years from 2012-2016. We fit generalized additive models and estimated percent changes in prevalence using projected precipitation under Shared Socio-Economic pathways describing sustainable development (SSP1), middle of the road (SSP2), and fossil fuel development (SSP5) scenarios.FindingsAn increase from 15°C to 30°C in weekly average temperature was associated with 5.0% higher diarrhea, 6.4% higher Norovirus, and 13.0% higher STEC prevalence. Above-median precipitation was associated with 1.27-fold (95% CI 0.99, 1.61) higher diarrhea; higher Cryptosporidium, tEPEC, ST-ETEC, STEC, Shigella, EAEC, Campylobacter, Aeromonas, and Adenovirus 40/41; and lower aEPEC, Giardia, Sapovirus, and Norovirus prevalence. Other associations were weak or null. Compared to the study period, diarrhea prevalence was similar under SSP1 (7%), 3.4% (2.7%, 4.3%) higher under SSP2, and 5.7% (4.4%, 7.0%) higher under SSP5. Prevalence of pathogens responsible for a large share of moderate-to-severe diarrhea in this setting (Shigella, Aeromonas) were 13-20% higher under SSP2 and SSP5.InterpretationHigher temperatures and precipitation were associated with higher prevalence of diarrhea and multiple enteropathogens; higher precipitation was associated with lower prevalence of some enteric viruses. Under likely climate change scenarios, we projected increased prevalence of diarrhea and enteropathogens responsible for clinical illness. Our findings inform pathogen-specific adaptation and mitigation strategies and priorities for vaccine development.FundingBill & Melinda Gates Foundation, National Institute of Allergy and Infectious Diseases, National Heart, Lung, And Blood Institute, Stanford University School of Medicine, Chan Zuckerberg BiohubResearch in ContextEvidence before this studyWe searched Google Scholar and Scopus for studies published from January 1, 2000 to present using the following three queries: 1) child; and diarrhea OR “enteric infection”; and meteorological OR environmental OR “surface water” OR “standing water”; and risk AND/OR factors; 2) Climate AND change AND project* AND diarrhea OR diarrhoea; 3) climate AND change AND project* OR model AND enter* AND infect* AND E. coli. Studies generally focused on individual risk factors for diarrhea transmission or enteric infection, with an emphasis on temperature and precipitation. Studies found that higher temperatures were associated with higher incidence of bacterial diarrhea and lower incidence of viral diarrhea; few studies have investigated associations between temperature and parasitic diarrhea. Heavy rainfall, particularly after dry periods, was associated with higher diarrhea prevalence, though heavy rainfall during rainy seasons was found to protect against diarrhea incidence. Similarly, flooding of surface water and shallow wells was also associated with higher diarrhea prevalence. Very few studies investigated associations between diarrhea or enteropathogen carriage and surface water presence, or humidity. A recent individual participant meta-analysis of studies in 19 low- and middle-income countries found that higher precipitation was associated with a small decrease in enterotoxigenic E. coli (ETEC) and Campylobacter spp. prevalence and no difference in Shigella, Cryptosporidium, or Giardia, or enteric virus prevalence. Weekly average temperature increases of 10-40° C within the study period were associated with higher risk of Campylobacter, ETEC, Shigella, Cryptosporidium, Giardia, and adenovirus, and lower risk of sapovirus and rotavirus, and generally, associations were stronger. Higher humidity was associated with higher risk of enteric bacterial infections and lower risk of enteric virus infection. A small number of studies have projected diarrhea under climate change in low- and middle-income countries. Studies have estimated a 15-20% increase in global diarrhea risk in 2040-2069 relative to 1961-1990, up to 21% increase in diarrhea incidence in northern India from 2013 to the 2040s, 3-10% increase in diarrhea cases in the Gaza Strip associated with 1.5° to 2°C increases in temperature, and an 8% increase in diarrhea burden by 2050 in Nepal. One study estimated an additional 1,625,073 ETEC diarrhea cases in Bangladesh from 2046-2065 due to climate change. No studies estimated changes in the presence of multiple enteropathogens under possible climate change scenarios.Added value of this studyThis study matched remote sensing data on precipitation, temperature, surface water, and humidity to data on diarrhea and enteropathogen carriage in children under 2 years from a trial in rural Bangladesh. We fit flexible models to investigate potentially non-linear relationships between each climate and environmental risk factor and each health outcome. In addition, we predicted the prevalence of each outcome under three possible climate change scenarios. Our predictions leveraged the spatiotemporal distributions of both climate precipitation projections and diarrhea and enteropathogen carriage in the trial. To our knowledge, this is the first study to predict prevalence under climate change scenarios for both diarrhea and multiple enteropathogens in a rural, low-resource south Asian setting. Our study sheds light on how climate change may impact diarrhea and enteropathogen carriage in Bangladesh, which is highly vulnerable to climate change due to its low altitude and seasonal flooding.Implications of all the available evidenceWe found that both higher temperatures and precipitation were associated with higher diarrhea prevalence, consistent with prior studies. Compared to prior studies, which have primarily found associations between diarrhea and heavy rainfall following a dry period, we found that moderate levels of weekly precipitation were associated with higher diarrhea prevalence. Relationships between climatic and environmental variables and enteropathogen carriage varied by taxa, as other studies have found, but for some pathogens and risk factors, we observed associations in different directions than prior studies. Under middle of the road and fossil fuel-based development climate change scenarios, we projected increased diarrhea prevalence and taxa-specific changes in enteropathogen prevalence. Our findings suggest that climate change may exacerbate diarrhea burden in rural Bangladesh in the absence of sustainable development and inform prioritization of pathogen-specific mitigation and/or adaptation interventions (e.g., vaccines) for young children in rural, South Asian settings under climate change.

BackgroundClimate change may influence the effectiveness of environmental interventions. We investigated if climate and environment modified the effect of low-cost, point-of-use water, sanitation, and handwashing (WASH) interventions on diarrhea and predicted intervention effectiveness under climate change scenarios.MethodsWe analyzed data from a cluster-randomized trial in rural Bangladesh that measured diarrhea prevalence in children 0-2 years from 2012-2016. We matched remote sensing data on temperature, precipitation, humidity, and surface water to households by location and measurement date. We estimated prevalence ratios (PR) for WASH interventions vs. control stratified by environmental factors using generative additive models and targeted maximum likelihood estimation. We estimated intervention effects under predicted precipitation in the study region in 2050 for climate change scenarios from different Shared Socioeconomic Pathways (SSPs).FindingsWASH interventions more effectively prevented diarrhea under higher levels of total precipitation in the previous week and when there was heavy rain in the previous week (heavy rainfall PR = 0.38, 95% CI 0.23-0.62 vs. no heavy rainfall PR = 0.77, 0.60-0.98). We did not detect substantial effect modification by other environmental variables. WASH intervention effectiveness increased under most climate change scenarios; in a fossil-fueled development scenario (SSP5), the PR was 0.46 (0.44-0.48) compared to 0.67 (0.65-0.68) in the study.InterpretationWASH interventions had the strongest effect on diarrhea under higher precipitation, and effectiveness may increase under climate change without sustainable development. WASH interventions may improve population resilience to climate-related health risks.FundingBill & Melinda Gates Foundation, National Institute of Allergy and Infectious Diseases, National Heart, Lung, And Blood InstituteResearch in Context PanelEvidence before this studyWe searched Google Scholar using the search terms “sanitation” OR “hygiene” OR “WASH” OR “water quality”; and “heterogen*” or “effect modif*”; and “climate” OR “environmental”; and “diarrhea” OR “enteric infection”; and “risk” AND/OR “factors”. In general, the effect modification of sanitation and hygiene on diarrhea or enteropathogen infection by climate and environment is not well studied. One study in Ecuador investigated different relationships between rainfall, diarrhea, and unimproved sanitation and water sources. They found that poor sanitation was most strongly associated with elevated diarrhea after precipitation minimums, whereas poor water sources were most strongly associated with elevated diarrhea after precipitation maximums. One study in Rwanda also found that high levels of runoff were protective against diarrhea only in households with unimproved toilets. In Bangladesh, one study found that access to tubewells was most effective at reducing childhood diarrhea in non-flood controlled areas. No prior studies have projected changes in WASH effectiveness under different climate change scenarios.Added value of this studyTo our knowledge, this is the first study to assess differences in WASH intervention effectiveness for multiple environmental variables and the first to predict future intervention effectiveness under different climate change scenarios. We spatiotemporally matched individual-level data from a trial in rural Bangladesh to remote sensing data on temperature, precipitation, humidity, and surface water presence. We compared the effectiveness of WASH interventions to prevent diarrhea under different levels of potential environmental effect modifiers. We then used climate models with high spatiotemporal resolution to predict changes in intervention effectiveness under different climate change scenarios.Implications of all the available evidenceWe found that WASH interventions were more effective under increasing precipitation and following heavy rain (over 80th percentile) events. Prior studies have found that diarrhea tends to increase with rainfall. Several studies have documented that precipitation-associated increases in diarrhea occurred primarily after heavy rain that followed dry periods, but we found that WASH intervention effectiveness in preventing diarrhea was higher even at moderate precipitation levels that were not necessarily preceded by a dry period. Under climate change scenarios with higher precipitation, we projected that WASH interventions may become more effective. In settings similar to rural Bangladesh, our findings suggest that regions that are expected to experience greater rainfall should be prioritized for intervention installation, promotion, and maintenance and that these activities might be most impactful during the rainy season.

ImportanceA regulated stress response is essential for healthy trajectories, but the integrated effects of early childhood environmental and nutritional interventions on stress physiology are unknown.ObjectiveTo assess the effects of a combined nutritional, water, sanitation, and handwashing intervention on physiological stress response, oxidative stress, and DNA methylation.Design, Setting, and ParticipantsIn a trial in rural Bangladesh, we randomized geographical clusters of pregnant women and their in-utero children into either the combined nutritional, water, sanitation, and handwashing intervention or the control group. Physiological stress response, oxidative stress, and methylation levels of 757 children were measured at ages one and two years. Analysis was intention-to-treat.InterventionsThe intervention group received combined nutritional counseling and lipid-based nutrient supplements, chlorinated drinking water, upgraded sanitation, and handwashing with soap (N+WSH). The control group did not receive interventions.Main Outcomes and MeasuresWe measured four isomers of urinary F2-isoprostanes [iPF(2α)-III; 2,3-dinor-iPF(2α)-III; iPF(2α)-VI; 8,12-iso-iPF(2α)-VI] at year one. At year two, we measured pre- and post-stressor concentrations of salivary alpha-amylase and cortisol, overall methylation of the glucocorticoid receptor (NR3C1) exon 1F promoter including methylation levels at the nerve growth factor-inducible protein A (NGFI-A) binding site, mean arterial pressure, and resting heart rate.ResultsChildren in the N+WSH group had lower levels of F2-isoprostanes compared to controls (difference -0.16 to -0.19 log ng/mg of creatinine, PPP=0.023) in the N+WSH group. Children in the N+WSH group exhibited decreased logit-transformed methylation of the NGFI-A transcription factor binding site (−0.04; 95% CI, -0.08 to 0; P=0.037).Conclusions and RelevanceA nutritional, water, sanitation, and handwashing intervention reduced oxidative stress, enhanced hypothalamic-pituitary-adrenocortical axis activity, and reduced methylation levels in a transcription factor binding site of the glucocorticoid receptor gene. A targeted environmental and nutritional intervention affected the set point, reactivity, and regulation of the physiological stress system in early childhood, which may have implications for long-term health and developmental trajectories.Trial RegistrationClinicalTrials.govNCT01590095Key PointsQuestionDo intensive nutritional and environmental interventions alter physiological stress response, oxidative stress levels, and epigenetic programming during the first two years of life?FindingsIn this cluster randomized clinical trial of 757 children, a combined nutritional, drinking water, sanitation, and handwashing intervention significantly reduced oxidative stress, enhanced the cortisol response, and reduced methylation levels in a transcription factor binding site of the glucocorticoid receptor gene.MeaningThe integrated nutritional, drinking water, sanitation, and handwashing intervention enhanced adaptive responses of the physiological stress system in early childhood, which may have implications for long-term health and developmental trajectories.

While studies have speculated that immune function may play a role in how water, sanitation, hygiene, and nutritional (N+WSH) interventions may individually impact child growth and development, the combined effects of these interventions on immune system development are unknown. Here, we report on a trial in rural Bangladesh, where we cluster-randomized pregnant women into control and N+WSH arms. Among the birth cohort, we quantified plasma IL-1β, IL-6, TNF-α, IL-2, IL-12p70, IFN-γ, IL-4, IL-5, IL-13, IL-17A, IL-21, IL-10, and GM-CSF at ages 14 and 28 months. Cytokine ratios were included as prespecified outcomes to examine the net inflammatory environment. We assessed 704 children. After one year, TNF-α/IL-10, IL-12/IL-10, and IL-17A/IL-10 ratios were lower in the intervention group compared to the control group (mean difference: -0.12 to -0.19, p

Author(s): Benjamin-Chung, Jade; Mertens, Andrew; Colford, John M; Hubbard, Alan E; van der Laan, Mark J; Coyle, Jeremy; Sofrygin, Oleg; Cai, Wilson; Nguyen, Anna; Pokpongkiat, Nolan N; Djajadi, Stephanie; Seth, Anmol; Jilek, Wendy M; Jung, Esther; Chung, Esther O; Dayal, Sonali; Hejazi, Nima; Malenica, Ivana; Li, Haodong; Hafen, Ryan; Subramoney, Vishak; Haggstrom, Jonas; Norman, Thea; Brown, Kenneth H; Christian, Parul; Arnold, Benjamin F | Abstract: Globally 149 million children under five are estimated to be stunted (length more than 2 standard deviations below international growth standards). Stunting, a form of linear growth faltering, increases risk of illness, impaired cognitive development, and mortality. Global stunting estimates rely on cross-sectional surveys, which cannot provide direct information about the timing of onset or persistence of growth faltering- a key consideration for defining critical windows to deliver preventive interventions. We performed the largest pooled analysis of longitudinal studies in low- and middle-income countries to date (n=32 cohorts, 52,640 children, ages 0-24 months), allowing us to identify the typical age of linear growth faltering onset and to investigate recurrent faltering in early life. The highest incidence of stunting onset occurred from birth to age 3 months. From 0 to 15 months, less than 5% of children per month reversed their stunting status, and among those who did, stunting relapse was common. Early timing and low reversal rates emphasize the importance of preventive intervention delivery within the prenatal and early postnatal phases coupled with continued delivery of postnatal interventions through the first 1000 days of life.

Author(s): Mertens, Andrew; Benjamin-Chung, Jade; Colford, John M; Hubbard, Alan E; van der Laan, Mark J; Coyle, Jeremy; Sofrygin, Oleg; Cai, Wilson; Jilek, Wendy; Dayal, Sonali; Nguyen, Anna; Pokpongkiat, Nolan N; Djajadi, Stephanie; Seth, Anmol; Chung, Esther O; Malenica, Ivana; Hejazi, Nima; Li, Haodong; Hafen, Ryan; Subramoney, Vishak; Haggstrom, Jonas; Norman, Thea; Christian, Parul; Brown, Kenneth H; Arnold, Benjamin F | Abstract: AbstractAcute malnutrition accounts for an immense disease burden and is implicated as a key, underlying cause of child mortality in low resource settings. Child wasting, defined as weight-for-length more than 2 standard deviations below international standards, is a leading indicator to measure the Sustainable Development Goal target to end malnutrition by 2030. Prevailing methods to measure wasting rely on cross-sectional surveys that are unable to measure onset, recovery, and persistence — key features of wasting epidemiology that could inform preventive interventions and disease burden estimates. Here, we show through an analysis of 18 longitudinal cohorts that child wasting is a highly dynamic process of incident onset and recovery, and that peak incidence is between birth and 3 months — far earlier than peak prevalence at 12-15 months. By age 24 months the proportion of children who had ever experienced a wasting episode (33%) was more than 5-fold higher than prevalence (6%), suggesting that the wasting burden is likely far higher than cross-sectional surveys suggest. Seasonally driven changes in population mean weight-for-length were large (g0.5 z in some cohorts) and were synchronous with rainfall across diverse settings, creating potential for seasonally targeted interventions. Our results motivate a new focus on extending preventive interventions for child wasting to pregnant and lactating mothers, and for preventive and therapeutic interventions to include children below age 6 months in addition to current targets of ages 6-59 months.