The authors, and the originator of the “staying alive” hypothesis (Campbell, Reference Campbell1999), identify that sex differences in the timing of male and female contributions to fitness can result in the evolution of sex differences in survival and the mechanisms that prolong life. The hypothesis seeks to explain women's greater longevity as the result of female-specific longevity-promoting adaptations. Those adaptations are thought to have evolved because, relative to men, women have historically gained greater marginal fitness from post-partum contributions to offspring and grand-offspring fitness. This proposal complements the more commonly articulated argument that stronger selection among men to succeed in pre-copulatory sexual competition leads them to take more risks, be more violent, and thus die younger than women.

Our commentary points to some areas where “staying alive” might be more thoroughly integrated with theory and empirical research on ageing and life histories. Such integration would allow both critical testing of “staying alive” and other human-centric hypotheses and enrich the evolutionary study of sex-dependent human ageing. Many of the background references in the target article are from research on ageing, and many of the traits discussed – including cancers, dementia, cardiovascular diseases, autoimmune diseases, and age-dependent immune responses – are related to ageing. Placing “staying alive” into the theoretic context of ageing biology would constitute an important next step for the hypothesis.

Ageing research has found value in distinguishing extrinsic mortality due to age-independent hazards such as accidents and general predation from intrinsic mortality due to age-dependent decreases in performance (Kirkwood & Austad, Reference Kirkwood and Austad2000; Williams, Reference Williams1957). While the distinction can often be artificial, it remains useful in thinking about the evolution of mechanisms that slow or prevent ageing, including immune and behavioural self-protection. The fact that some extrinsic mortality is unavoidable means that mechanisms of somatic repair are always going to be under weaker selection in older cohorts than in younger ones.

This distinction is important because the reasons presented for greater female lifespan in humans and in other animals are often tied to greater extrinsic mortality in males. Higher predation due to male ornamentation, accidents due to risk-taking, and male–male aggression all have the potential to weaken the selection against both diseases of ageing and self-preservation behaviours in males relative to females (Austad & Bartke, Reference Austad and Bartke2016; Bonduriansky, Maklakov, Zajitschek, & Brooks, Reference Bonduriansky, Maklakov, Zajitschek and Brooks2008). High male extrinsic mortality can thus lead on to higher intrinsic mortality.

Observations across a variety of taxa suggest that such a pattern may commonly pertain (see Bonduriansky et al., Reference Bonduriansky, Maklakov, Zajitschek and Brooks2008 for a review; Promislow & Harvey, Reference Promislow and Harvey1990). In many taxa, males are more likely than females to “live fast and die young” due to differences in the intensity of sexual competition (Promislow, Montgomerie, & Martin, Reference Promislow, Montgomerie and Martin1992). In humans, evidence suggests that young men are particularly prone to discount the future and risk their lives in the pursuit of status, wealth, and, ultimately, matings (Wilson & Daly, Reference Wilson and Daly1985), and that women's fitness is more likely to benefit from long-term investment in children and grandchildren (Hawkes, Reference Hawkes2004).

Humans are, by any comparison, a long-lived species with extended contributions from mothers, fathers, and alloparents to offspring. Viewed in the context of other species, both women and men have extraordinary self-preservation and anti-ageing traits that drive very long mean and maximum lifespans. Nonetheless, sex-differences in extrinsic mortality may be responsible for sex-dependent patterns of intrinsic mortality, including many of the traits reviewed in the target article. These traits may be over-represented in women both because staying alive enhances female fitness and because higher male extrinsic mortality has weakened selection in favour of these traits among men.

The claim that women's fitness benefits more from self-preservation than men's fitness, while consistent with the majority of evidence presented in the target article, is thus an empirical claim that needs critical testing. Theoreticians and comparative biologists have pointed out that in animal species in which small numbers of males live long enough to grow large, develop large weapons, or achieve high social rank, selection may promote male lifespan and suppress male ageing more strongly than female (Clinton & Le Boeuf, Reference Clinton and Le Boeuf1993; Graves, Reference Graves2007). In humans, the importance of social status to men's mating success may have a similar effect, and has been suggested to be at least partly responsible for the evolution of extended human lifespans (Tuljapurkar, Puleston, & Gurven, Reference Tuljapurkar, Puleston and Gurven2007).

Further, although women's life expectancy currently exceeds men's in most countries, that has not always been the case. Throughout history, extraordinary numbers of women have died in childbirth, and higher parities and costs of reproduction are, in some places and at some times, associated with men living longer than women (Bolund, Lummaa, Smith, Hanson, & Maklakov, Reference Bolund, Lummaa, Smith, Hanson and Maklakov2016; Maklakov, Reference Maklakov2008).

Sex-differences in lifespan, age-dependent reproduction, and thus selection are variegated, complex, and often environmentally contingent. To add further complication, changes in the timing of reproductive effort and in ageing-related traits can both precede and follow changes in survival (Austad & Bartke, Reference Austad and Bartke2016; Bonduriansky et al., Reference Bonduriansky, Maklakov, Zajitschek and Brooks2008). As a result, relationships between selection on lifespan and the traits that result are often more complex and confusing than the target article concedes. This is not a criticism of a target article and the choices made under publishing constraints, but rather a call for further theoretic development and hypothesis testing.

In conclusion, the authors of the target article are to be congratulated on assembling such a varied and thorough set of published observations which, together, appear to weigh heavily in favour of the “staying alive” hypothesis. Greater integration with life history theories of ageing, and with the genomics of sex-dependent variation in fitness traits (Bonduriansky et al., Reference Bonduriansky, Maklakov, Zajitschek and Brooks2008), together with a commitment to critically testing the adaptive intuition behind the hypothesis will likely broaden its relevance and spur a flurry of exciting research.