Quantifying the impact of selective attrition on estimates of direct and indirect genetic effects in MoBa cohort under various conditions: Using observed and simulated data

Résumé fourni par la source

Selective attrition remains a major challenge in cohort studies like MoBa. Given that MoBa, along with similar cohorts, is a central resource for the estimation of direct and indirect genetic effects via genotyped trios, understanding the impact of selective attrition in such analyses is a crucial step to allow researchers to avoid biased estimates. In this study, we aim to quantify the impact of attrition by conducting trio polygenic score analyses under different conditions. We will use maternal and paternal depressive symptoms (SCL-5) measured at gestational week 15, and child length measured at birth as our outcomes, and take MoBa families with complete data on these outcomes as our reference samples. We will analyse these samples with trio polygenic score prediction of outcomes, and then repeat analyses while progressively restricting the analytic sample to participants who stayed active in MoBa at different waves. In addition, we will simulate alternative attrition patterns based on different hypothetical causal scenarios for drop-out. These simulated attrition patterns will again be used to progressively restrict the reference sample for repeated analysis of maternal and paternal SCL at gestational week 15 and child length measured at birth. We will compare direct and indirect genetic effect estimates from analyses of the various selected samples (based on both observed and simulated patterns of attrition) against estimates from the full reference samples. In all cases, we will assess differences in effect sizes and standard errors in fully observed vs attritioned samples.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.