Sleep and gestation duration are correlated, but not with cancer risk or prevalence across vertebrates
Cancer prevalence varies across species, with traits such as litter/clutch size, gestation duration, carnivory, and adult mass, partly explaining this variation. Yet no coherent explanation exists for why shorter gestation and carnivory both correlate with cancer prevalence or risk. Given that carnivores sleep more than herbivores, sleep reportedly being a compensation for brain immaturity after gestation, longer sleep appearing in species with shorter gestation, and shorter gestation in species with higher cancer prevalence, we hypothesize that sleep may mediate the gestation-cancer association. We obtained neoplasia prevalence, cancer prevalence, cancer mortality risk (n [≥] 20 individuals/species), gestation duration, and sleep duration data across vertebrates. We tested whether sleep duration mediates the known gestation-cancer prevalence association, whether sleep duration is directly correlated with neoplasia/cancer prevalence or risk, and tested the known gestation-sleep duration association using more vertebrate species and phylogenetic generalized least squares analyses. Gestation and sleep duration were not correlated with neoplasia prevalence, cancer prevalence, or cancer mortality risk. Gestation and sleep duration were negatively correlated. These results highlight the complexity of understanding how multiple physiology variables explain the variation in cancer prevalence or risk across vertebrates, and the need to verify previously known associations with more powerful and robust statistical tools.