Details of the association statistics of the covariates and global ancestries of BMI.

Sun, Hanxiao; Lin, Meng; Russell, Emily M.; Minster, Ryan L.; Chan, Tsz Fung; Dinh, Bryan L.; Naseri, Take; Reupena, Muagututi‘a Sefuiva; Lum-Jones, Annette; Cheng, Iona; Wilkens, Lynne R.; Le Marchand, Loïc; Haiman, Christopher A.; Chiang, Charleston W. K.

doi:10.1371/journal.pgen.1009273.s011

pgen.1009273.s011.docx (18.5 kB)

Details of the association statistics of the covariates and global ancestries of BMI.

journal contribution

posted on 2021-02-11, 18:30 authored by Hanxiao Sun, Meng Lin, Emily M. Russell, Ryan L. Minster, Tsz Fung Chan, Bryan L. Dinh, Take Naseri, Muagututi‘a Sefuiva Reupena, Annette Lum-Jones, Iona Cheng, Lynne R. Wilkens, Loïc Le Marchand, Christopher A. Haiman, Charleston W. K. Chiang

Model 1 models the non-genetic covariates according to the heuristic described in the Methods. The residual from model 1 is then inverse normalized and tested in model 2. Models 1A and 2A repeats the procedure but included quintiles of nSES levels in a mixed effect model (Methods); in this case, the R² in Model 1A reported include both the fixed and the random effect. * edu4 was a binary variable created from the original categorical variable of education status by grouping levels 1,2,3 and coded 0, while education status level 4 was coded as 1. This was done because there were no significant associations between education levels 1 through 3 and BMI. See S21 Table for description of these education levels.

(DOCX)