A method making fewer assumptions was recommended for estimating exposure–outcome associations in stratified case–cohort studies.

EJ, MJS and SGT were supported in part by the UK Medical Research Council and the British Heart Foundation. SJS was supported by the UK Medical Research Council [Unit Programme number MC_UU_12015/1]. The EPIC-InterAct study was funded by the EU FP6 programme (LSHM_CT_2006_037197).