Shah Mohammad Fahim

Malnutrition continues to be a major threat to health, particularly maternal and child health in low resource settings, resulting in impairments in cognitive function, growth, and development, and metabolic diseases later in life. Nutritional assessment is a cornerstone of any successful nutrition intervention or program whether in the community or at the clinic. Improved computational power and advances in technology may enable precision nutrition-based approaches for maternal and child health, which can complement current methods for nutritional assessment to identify clinical, biochemical, microbiome-related, social, and environmental characteristics to predict responses to nutritional interventions or programs. Precision nutrition has the potential to complement program monitoring, efficacy evaluation, and ultimately to inform design of interventions to improve maternal and child health.

Introduction: Treatment in the intensive care unit (ICU) generates complex data where machine learning (ML) modelling could be beneficial. Using routine hospital data, we evaluated the ability of multiple ML models to predict inpatient mortality in a paediatric population in a low/middle-income country.Method: We retrospectively analysed hospital record data from 0-59 months old children admitted to the ICU of Dhaka hospital of International Centre for Diarrhoeal Disease Research, Bangladesh. Five commonly used ML models- logistic regression, least absolute shrinkage and selection operator, elastic net, gradient boosting trees (GBT) and random forest (RF), were evaluated using the area under the receiver operating characteristic curve (AUROC). Top predictors were selected using RF mean decrease Gini scores as the feature importance values.Results: Data from 5669 children was used and was reduced to 3505 patients (10% death, 90% survived) following missing data removal. The mean patient age was 10.8 months (SD=10.5). The top performing models based on the validation performance measured by mean 10-fold cross-validation AUROC on the training data set were RF and GBT. Hyperparameters were selected using cross-validation and then tested in an unseen test set. The models developed used demographic, anthropometric, clinical, biochemistry and haematological data for mortality prediction. We found RF consistently outperformed GBT and predicted the mortality with AUROC of ≥0.87 in the test set when three or more laboratory measurements were included. However, after the inclusion of a fourth laboratory measurement, very minor predictive gains (AUROC 0.87 vs 0.88) resulted. The best predictors were the biochemistry and haematological measurements, with the top predictors being total CO₂, potassium, creatinine and total calcium.Conclusions: Mortality in children admitted to ICU can be predicted with high accuracy using RF ML models in a real-life data set using multiple laboratory measurements with the most important features primarily coming from patient biochemistry and haematology.

Background: A randomized controlled trial in Bangladeshi children aged 12 to 18 months with moderate acute malnutrition found that dietary supplementation with the microbiota-directed complementary food (MDCF-2) significantly improved weight gain and repaired gut microbiota compared to the ready-to-use supplementary food. However, the MDCF-2 formulation was made daily from locally available ingredients and the need for a packaged, nutritionally compliant, and organoleptically acceptable MDCF-2 prototype was essential for future large-scale clinical studies. Objective: The study aimed to develop and assess the acceptability of 3 alternative foil-packaged formulations of MDCF-2 in comparison to current MDCF-2. Methods: Of the 3 packaged formulations, the Jumpstart version was provided in 2 sachets, the other 2 formulations were provided in a retort-stable foil pouch extended by sterilization, and microbiological growth was monitored over 10 months. The acceptability study included 40 children aged 8 to 12 months living in an urban slum in Dhaka, and the organoleptic properties were assessed using a 7-point hedonic scale. Results: In the 100 g distributed over the 2 sessions, children consumed 82.5 ± 7.84 g (mean ± SD) of kitchen-prepared MDCF-2, 85.4 ± 7.15 g of the “Jumpstart” MDCF-2 formulation, 85.4 ± 8.70 g of the MDCF-2 with green banana powder, and 86.2 ± 4.26 g of the MDCF-2 with sweet potato formulation. The “Jumpstart” MDCF-2 and MDCF-2 with sweet potato achieved the highest overall acceptability scores on the hedonic scale; although none of the shelf-stable formulations were significantly different from the kitchen-prepared MDCF-2. Conclusions: Packaged, shelf-stable prototypes of MDCF-2 exhibited comparable acceptability among Bangladeshi children aged 8 to 12 months to the original freshly prepared formulation. Plain language title: Development and Acceptability of Shelf-Stable Microbiota-Directed Complementary Foods

Nutritional impairment during adolescence may result in adverse physical and reproductive health outcomes. We investigated the prevalence and determined the factors associated with underweight and overweight/obesity among ever-married adolescent girls in Bangladesh. We used Bangladesh Demographic and Health Surveys data conducted in 2004, 2007, 2011, 2014, and 2017. A total of 7040 ever-married adolescent girls aged 15–19 years were included in this analysis. Prevalence of underweight (body mass index [BMI]2) significantly decreased from 39.53% (95% CI = 36.71, 42.43) to 23.62% (95% CI = 21.35, 26.05) during 2004–2017 (p< 0.001). However, prevalence of overweight/obesity (BMI ≥ 23 kg/m²) significantly increased from 5.9% (95% CI = 4.67, 7.43) to 22.71% (95% CI = 20.39, 25.20) during the same period (p< 0.001). The girls with higher age (OR = 0.94, 95% CI = 0.90, 0.99,p= 0.023), higher level of education (OR = 0.60, 95% CI = 0.43, 0.83,p= 0.002), and richest wealth quintile (OR = 0.78, 95% CI = 0.62, 0.98,p= 0.035) had significantly lower risk of being underweight. Adolescent girls having more than one child (OR = 1.41, 95% CI = 1.15, 1.73,p= 0.001) were more likely to be underweight. Elderly adolescents with better economic status were more at risk of being overweight/obese (OR = 2.57, 95% CI = 1.86, 3.55,p< 0.001). Girls married to skilled/unskilled workers (OR = 0.58, 95% CI = 0.44, 0.77,p< 0.001) and persons involved in small businesses (OR = 0.66, 95% CI = 0.49, 0.89,p= 0.007) had lower risk of having a high BMI. Using contraceptive (OR = 0.8, 95% CI = 0.69, 0.94,p= 0.006) was negatively associated with overweight/obese. Although prevalence of undernutrition among ever-married adolescent girls is declining, the proportion of being overweight/obese is increasing in Bangladesh warranting effective strategies to improve adolescent nutrition.

Objectives There remain hesitations and miscommunication regarding appropriate public health behaviours and conceptions related to COVID-19. We tested the effectiveness of the community-based health cell approach in improving knowledge, attitude and practice (KAP) related to COVID-19. Setting Households of the Bauniabadh slum area in Mirpur, Dhaka, Bangladesh. Participants Household heads (HHs) and homemakers (HMs) of intervention (n=211) and comparison households (n=209). Interventions Behaviour change communication delivered at the community level in a quasi-experimental manner through small-scale community meetings and home visits. Outcome variables and methods The outcomes of interest were before–after mean and per cent changes in KAP scores. Data were collected from HHs and HMs before and after the intervention and difference-in-differences (DID) analysis technique was applied. Results We found statistically significant (pConclusion Scientifically valid information disseminated at the community level using the health cell approach could bring positive changes in KAP related to COVID-19.