The cult of culling

It is difficult to understand why culling (reduction of litter size) has become such a widely used procedure in reproductive toxicity studies since there appear to have been no prior investigations to ascertain that it would improve the efficiency of studies with respect to detecting adverse effects. Perhaps the only provable advantage of culling is with respect to economics and convenience. Post hoc rationalizations for culling lack conviction because many of the claims made for culling are erroneous, inconsistent, vague, and contradictory. Mostly, they are based on part truths derived from minimal studies, conducted for totally different purposes. That experimental animals have to be killed sooner or later is unquestioned, but for ethical and scientific reasons, it is imperative that the maximum amount of information is obtained from them. Currently, the most common practice is to cull litters to four per sex (total eight) on Day 4 postpartum. This is totally divorced from natural values for most rat strains and involves elimination, usually without adequate examination, of between 30 and 45% of offspring. Without culling most of these would survive, unless there was a treatment effect. Intuitively, it would seem that removal of such a proportion of offspring would severely limit the possibility of detecting the postnatal equivalent of fetal malformations. Culling totally nullifies litter size as an indicator of toxicity. Indirectly, it also nullifies the value of mean pup weight as an indicator of toxicity because it greatly increases the variation in mean pup weight. This is quite contrary to the claim that culling reduces variance. Further, the increased growth of offspring in culled lifters can have long-term consequences of a shorter overall and reproductive life span. (C) 1997 Society of Toxicology.