Genetically heterogeneous mice exhibit a female survival advantage that is age- and site-specific: Results from a large multi-site study

Abstract

The female survival advantage is a robust characteristic of human longevity. However, underlying mechanisms are not understood, and rodent models exhibiting a female advantage are lacking. Here, we report that the genetically heterogeneous (UM-HET3) mice used by the National Institute on Aging Interventions Testing Program (ITP) are such a model. Analysis of age-specific survival of 3,690 control ITP mice revealed a female survival advantage paralleling that of humans. As in humans, the female advantage in mice was greatest in early adulthood, peaking around 350 days of age and diminishing progressively thereafter. This persistent finding was observed at three geographically distinct sites and in six separate cohorts over a 10-year period. Because males weigh more than females and bodyweight is often inversely related to lifespan, we examined sex differences in the relationship between bodyweight and survival. Although present in both sexes, the inverse relationship between bodyweight and longevity was much stronger in males, indicating that male mortality is more influenced by bodyweight than is female mortality. In addition, male survival varied more across site and cohort than female survival, suggesting greater resistance of females to environmental modulators of survival. Notably, at 24 months the relationship between bodyweight and longevity shifted from negative to positive in both sexes, similar to the human condition in advanced age. These results indicate that the UM-HET3 mouse models the human female survival advantage and provide evidence for greater resilience of females to modulators of survival.

Keywords: age-specific mortality; gender differences; lifespan; sex differences; smoothed hazard estimation; somatotrophic axis.

Figure 1

Sex‐specific survival dynamics of genetically heterogeneous UM‐HET3 mice. (a) Survival curves for male and female mice, showing the proportion surviving at each age. Survival curves are stratified by sex and study cohort (year). Survival for males (M) is shown in blue, females (F) in red. Average cohort size for males n = 332, females n = 282. Total number of animals n = 3,690. (b) Mortality hazard in male and female mice. Each line (males in blue, females in red) represents the smoothed estimated hazard as a function of age. Confidence intervals (95%) are shaded in gray. Vertical dotted line indicates age of weaning. (c) Ratio of male to female mortality hazard at each age in mice. Black line represents the hazard ratio of males to females as a function of age. Dashed blue lines demarcate the bootstrapped 95% confidence limits. Individual green lines represent hazard ratios estimated from resampling replicates. Horizontal dotted black line represents a hazard ratio of 1, indicating no difference in estimated hazard between males and females

Figure 2

Site differences in survival across study cohorts, showing greater variability in males. Mortality hazard in male and female mice is shown for each site (a). Each line (males in blue, females in red) represents the smoothed estimated hazard as a function of age. Confidence intervals (95%) are shaded in gray. Lower panel (b) shows median lifespan of males and females at each site, split by study cohort (year). TJL = The Jackson Laboratory (red), UM = University of Michigan (green), UT = UT Health San Antonio (blue)

Figure 3

Relationship between bodyweight and lifespan by sex and age. Each panel shows the scatterplot and linear regression line of lifespan by bodyweight in males (blue) and females (red) at four different ages of bodyweight measurement (age in months, righthand axis). A nonlinear fit is shown in gray to assess departure from linearity