Only if the deaths (or more correctly: any effects harsh enough to impede mating) happen before the typical age of becoming a parent. If it's just "live healthy to very old age" vs "cancer at fifty", then there is no selection bias since our anecdata here is based on the next generation.