Covariates didn't change GWAS too much?

Question

[okiedokie1208]

I used different sets of covariates (i.e. Sex+age VS. Sex+age+Sex^2) to run GWAS based on a mixed linear model( software: Regenie) However , I don't see a huge difference from GWAS summary statistics and I got a similar Manhanttan plot. Does that mean my genotype has strong association with phenotype so the covariates will not affect too much?

[Sam]

Maybe more like your covariates ain't correlated with the genotype. E.g. for traits affected by Sex, if you don't adjust for Sex, you will get significant results for SNPs related to sex but not necessarily the trait of interest. By adjusting for sex, you retain only those significance that are independent of sex.