Measurement error in polygenic indices (PGIs) attenuates the estimation of their effects in regression models. We analyze and compare two approaches addressing this attenuation bias: Obviously Related Instrumental Variables (ORIV) and the PGI Repository Correction (PGI-RC). Through simulations, we show that the PGI-RC performs slightly better than ORIV, unless the prediction sample is very small (N < 1000) or when there is considerable assortative mating. Within families, ORIV is the best choice since the PGI-RC correction factor is generally not available. We verify the empirical validity of the simulations by predicting educational attainment and height in a sample of siblings from the UK Biobank. We show that applying ORIV between families increases the standardized effect of the PGI by 12% (height) and by 22% (educational attainment) compared to a meta-analysis-based PGI, yet estimates remain slightly below the PGI-RC estimates. Furthermore, within-family ORIV regression provides the tightest lower bound for the direct genetic effect, increasing the lower bound for the standardized direct genetic effect on educational attainment from 0.14 to 0.18 (+29%), and for height from 0.54 to 0.61 (+13%) compared to a meta-analysis-based PGI.
Bibliographical noteFunding Information:
The authors gratefully acknowledge participants of the 23andMe, Inc. cohort for sharing GWAS summary statistics for educational attainment. This work made use of the Dutch national e-infrastructure with the support of the SURF Cooperative using grant no. EINF-2327. The authors also acknowledge funding from NORFACE through the Dynamic of Inequality across the Life Course (DIAL) program (462-16-100; H.v.K., P.B., S.v.H., C.A.R., S.F.W.M., D.M., R.D.P.), from the European Research Council (DONNI 851725 to S.v.H. and GEPSI 946647 to C.A.R.), from the National Institute on Aging of the National Institutes of Health (RF1055654, R56AG058726 to T.J.G., H.v.K. and R01AG078522 to T.J.G.), and from the Dutch Research Council (016.VIDI.185.044 to T.J.G.). H.v.K., P.B., T.J.G., S.v.H., and C.A.R. also acknowledge the European Union’s Horizon 2021 research and innovation program under the Marie Skłodowska-Curie grant agreement (ESSGN 101073237). This research was supported by the National Institute for Health Research (NIHR Cambridge BRC-1215-20014 for E.A.W.S.). The views expressed are those of the authors and not necessarily those of the National Institutes of Health, NIHR, or the Department of Health and Social Care. We would like to thank Sjoerd van Alten, Dan Belsky, Neil Davies, Ben Domingue, Michel Nivard, and Elliot Tucker-Drob for valuable comments.
© 2023, The Author(s).