Quantifying and adjusting for selection biases in the Norwegian Mother, Father and Child Cohort Study using population-wide individual-level registry information

Background: Selective participation in research studies hampers researchers’ ability to draw valid and generalizable inferences from analyses. Quantifying and adjusting for selective participation is desirable, but can be challenging given the paucity of data for non-participants. Methods: We used individual-level information from population registers to predict initial and continued participation in the Norwegian Mother, Father and Child Cohort Study (MoBa). Inverse probability weights were computed from logistic regression models with elastic-net regularization. We predicted selective participation and attrition in 296,987 mothers, of whom 29% returned the first MoBa questionnaire, and 12,5% who returned a follow-up questionnaire eight years later. To quantify bias we computed sample characteristics and outcome-exposure effects in the reference population and stratified samples. To compare approaches for adjusting for bias, we computed weighted sample estimates using three sets of weights. Results: Unweighted sample estimates were systematically different from population values, indicating bias due to selective participation. Initial participation weights substantially reduced the impact of selection bias on mean values (93%) and effects (73%). Attrition weights performed similarly in reducing attrition bias (78% and 59%), but were less effective for reducing participation bias (22% and 11%). Discussion: Sample characteristics and effect estimates are substantially different from population values. Participation weights were relatively effective at reducing bias due to selective participation, but attrition weights were not. Estimates from studies that use attrition weights may still contain non-negligible selection bias - particularly if the baseline sample is not representative of the population. Future studies should prioritise opportunities for deriving sampling weights for participation as well as attrition.