Intergenerational Risk and Resilience of Rohingya in Displacement

Background and rationale. By 2024, an estimated 117.8 million people globally had been forcibly displaced by persecution, armed conflict, or other human rights violations (UNHCR, 2026). The vast majority reside in low- and middle-income countries (LMICs). The Rohingya refugees in Cox's Bazar, Bangladesh - numbering over one million - represent one of the world's largest and longest-standing refugee populations. A substantial body of research has established that exposure to trauma and adversity before and during pregnancy can affect offspring development through neurobiological mechanisms including fetal programming and intergenerational epigenetic transmission, as well as through postnatal disruptions to caregiving. Despite the scale of this global challenge, large-scale longitudinal studies of early child development in humanitarian contexts that begin prenatally remain extremely rare.

The iRRRd study was designed to address this gap. Drawing on Bronfenbrenner and Morris' bioecological systems framework, it captures risk and protective factors spanning multiple levels - from molecular processes to household and family dynamics to broader contextual forces such as camp governance, humanitarian infrastructure, and environmental stressors - through a multi-method, interdisciplinary approach combining qualitative, survey, observational, physiological, and neurobiological data.

The 36-month wave. The 36-month time point is a critical developmental window: cognitive, language, social-emotional, and behavioral development can be more reliably measured using gold-standard instruments; executive function (EF) emerges as a distinct and predictive construct; and caregiver-child playful interactions become more observable and directly relevant to developmental outcomes. This wave extends the longitudinal scope from infancy into toddlerhood, enabling examination of developmental trajectories from the prenatal period through early childhood.

A central scientific focus of this wave is the role of play and playful caregiver-child interactions as a source of resilience. Preliminary data from older siblings of enrolled children suggest that playful interactions buffer EF development specifically when mothers are experiencing elevated psychological distress. The 36-month wave allows formal testing of this and related hypotheses with the target cohort.

Key hypotheses. Guided by the bioecological framework above, this wave is designed to test several related hypotheses: (H1) caregiver-child playful interaction quality will buffer the association between maternal psychological distress and child cognitive/executive function outcomes, such that the association is attenuated at higher observed play quality; (H2) maternal executive function and autonomic regulation (HRV) will each independently predict caregiver-child interaction quality, which will in turn mediate the association between maternal functioning and child developmental outcomes; (H3) child psychosocial adversity (CPAS) will be associated with poorer child cognitive, executive function, and behavioral outcomes, and this association will be moderated by caregiver mental health and caregiving quality; (H4) physiological synchrony between mother and child during the caregiver-child interaction observation, indexed by concurrent HRV, will be positively associated with interaction quality and, secondarily, with child self-regulation outcomes. These hypotheses will be tested using the primary and secondary outcomes and the predictor, mediator, and moderator measures specified above.

Exploratory biological analyses. In parallel with the confirmatory hypotheses above, the study maintains a biobank of saliva and buccal swab specimens (described in Part 4) that will support exploratory, hypothesis-generating analyses as funding and laboratory capacity permit. Planned exploratory work includes untargeted and targeted proteomic and metabolomic profiling (e.g., inflammatory markers, steroid hormones including cortisol, oxylipins) and epigenetic analyses (e.g., DNA methylation, telomere length). Because the analytic pipelines, sample subsets, and quality-control criteria for these assays are still being finalized, this work is not registered as formal outcome measures in this record. Findings will be reported as exploratory, and specific analyses will be preregistered separately on a per-study basis once hypotheses and analytic plans are finalized.

Note on timing of registration. The iRRRd study began in 2023 and was not preregistered at inception due to logistical and timing constraints inherent to launching large-scale research in a complex humanitarian context. This registration is filed prospectively with respect to the 36-month data collection wave. See Brief Summary for details on the timing of registration relative to fieldwork. Data from Waves 1-5 were collected without registration; analyses of those data are not governed by this registration, though individual papers are being preregistered on an individual basis.

Inclusion criteria

• Mother/primary female caregiver was enrolled in the iRRRd prenatal cohort (recruited February 2023 - March 2024 when pregnant)
• Target child is alive and 36 months of age (+/- 2 months)
• Household can be located and reached within the study area
• Caregiver provides informed consent for continued participation

Exclusion criteria

• Target child has died prior to the 36-month visit
• Household has permanently migrated out of the study area and cannot be located
• Caregiver is in severe acute distress and unable to participate safely
• Caregiver refuses continued participation