(PTFS03-03-26) Maternal Risk Stratification During Pregnancy and Adverse Birth Outcomes in Rural Bangladesh: A Latent Class Analysis

Saturday, July 25, 2026

1:08 PM - 1:12 PM ET

Location: Prince George's Hall B-E (The Hub) - Poster Theater 2 / Poster Board Number: PT2-3

Presenting Author(s)

Mary E. de Boer, PhD, MSc (she/her/hers)

Postdoctoral Fellow
Stanford University
Stanford, California, United States

Disclosure(s): No relevant financial relationship(s) with ineligible companies to disclose.

Co-Author(s)

EZ

Eleonor Zavala, MSPH (she/her/hers)

PhD Student
Johns Hopkins Bloomberg School of Public Health
Baltimore, Maryland, United States
AK

Anna Kalbarczyk

Johns Hopkins University, United States
HA

Hasmot Ali, MBBS, MPH

Senior Research Physician
JiVitA Project Bangladesh
Gaibandha, Rajshahi, Bangladesh
RH

Rezwanul Haque, DCS, MA

JiVitA Project Bangladesh
Gaibandha, Rangpur, Bangladesh
LW

Lee S. Wu, MHS

Johns Hopkins University
Baltimore, Maryland, United States
TS

Towfida Siddiqua

JiVitA Project Bangladesh, Bangladesh
DM

Diwakar Mohan

Johns Hopkins University, United States
DE

Daniel J. Erchick, PhD, MPH

Johns Hopkins University, United States
PC

Parul Christian, PhD, MSc

Professor
Johns Hopkins University
Baltimore, Maryland, United States
AT

Andrew L. Thorne-Lyman, ScD MHS

Associate Professor
Johns Hopkins Bloomberg School of Public Health
Baltimore, Maryland, United States

Disclosure(s):

Mary E. de Boer, PhD, MSc: No relevant financial relationship(s) with ineligible companies to disclose.

Objectives: Maternal risk factors for adverse birth outcomes in low- and middle-income countries are diverse and often co-occur, yet screening and interventions typically target using single risk factors in isolation. Latent class analysis (LCA) was used to identify profiles of maternal risk for more precise risk stratification and their association with birth outcomes, for potential targeting of antenatal nutrition interventions.

Methods: Baseline data from 3,359 pregnant women from the TargetBEP cluster-randomized trial in northwest rural Bangladesh (2022–2025) were analyzed; the sample was limited to singleton live births, stillbirths, or miscarriages. Twenty nutritional, sociodemographic, and psychosocial indicators were used in LCA to identify maternal risk profiles. Associations were examined between class membership and adverse birth outcomes including low birth weight (LBW), small for gestational age (SGA), preterm birth (PTB) and neonatal mortality, using generalized estimating equations adjusted log-binomial regression models. Missing data were imputed via multiple imputation with chained equations.

Results: Fit statistics (AIC/BIC, likelihood ratio tests) and an entropy of 0.88 supported a four-class model with good class separation. The classes included: Class 1 "Anemic, short stature, high parity, low SES, depressed" (n=919, 27%); Class 2 "Adequate diet, high education and SES" (n=757, 23%); Class 3 "Low BMI, low agency, nulliparous, adolescent" (n=761, 23%); and Class 4 "Primiparous, low SES, low education" (n=922, 27%). Compared to Class 2, all other classes had significantly elevated LBW risk (RR 1·25-1·44). Class 3 exhibited highest SGA prevalence (48·7%) with elevated risk (RR 1·59, 95% CI 1·35-1·87), while Class 1 showed highest PTB prevalence (18·7%) and PTB risk (RR 1·08, 95% CI 0·81-1·39). Classes were not associated with the risk of neonatal mortality.

Conclusions: Maternal vulnerability in rural Bangladesh (defined using distinct, identifiable risk profiles) was associated with adverse birth outcomes. Screening to identify high-risk groups for targeted interventions could benefit from use of risk profiles based on readily measurable antenatal indicators.

Funding Sources: Gates Foundation, Edesia Nutrition