Abstract
Individual differences in telomere length are associated with individual differences in behaviour in humans and birds. Within the human epidemiological literature this association is assumed to result from specific behaviour patterns causing changes in telomere dynamics. We argue that selective adoption-the hypothesis that individuals with short telomeres are more likely to adopt specific behaviours-is an alternative worthy of consideration. Selective adoption could occur either because telomere length directly affects behaviour or because behaviour and telomere length are both affected by a third variable, such as exposure to early-life adversity. We present differential predictions of the causation and selective adoption hypotheses and describe how these could be tested with longitudinal data on telomere length. Crucially, if behaviour is causal then it should be associated with differential rates of telomere attrition. Using smoking behaviour as an example, we show that the evidence that smoking accelerates the rate of telomere attrition within individuals is currently weak. We conclude that the selective adoption hypothesis for the association between behaviour and telomere length is both mechanistically plausible and, if anything, more compatible with existing empirical evidence than the hypothesis that behaviour is causal.This article is part of the theme issue 'Understanding diversity in telomere dynamics'.