Principal component regression and linear mixed model in association analysis of structured samples: competitors or complements?

Principal component regression and linear mixed model in association analysis of structured samples: competitors or complements?