Intralocus sexual conflict can resolve the male-female health-survival paradox

Abstract

At any given age, men are more likely to die than women, but women have poorer health at older ages. This is referred to as the "male-female, health-survival paradox", which is not fully understood. Here, we provide a general solution to the paradox that relies on intralocus sexual conflict, where alleles segregating in the population have late-acting positive effects on male fitness, but negative effects on female health. Using an evolutionary modelling framework, we show that male-benefit, female-detriment alleles can spread if they are expressed after female reproduction stops. We provide support for our conflict based solution using experimental Drosophila data. Our results show that selecting for increased late-life male reproductive effort can increase male fitness but have a detrimental effect on female fitness. Furthermore, we show that late-life male fertility is negatively genetically correlated with female health. Our study suggests that intralocus sexual conflict could resolve the health-survival paradox.

Fig. 1

Population-level allele frequencies are determined by sex-specific fitness costs and benefits. The graphs show equilibrium allele frequency from simulating our two-sex models over a wide range of male-fitness benefits (relative reproductive success of males with the allele compared to males lacking the allele) and female-fitness costs (relative reproductive success in females with the allele compared to females lacking the allele). The red regions show the conditions where the sexually antagonistic allele goes to fixation, the blue regions indicate where the allele will be lost and the white/colour mixed regions show where both allelic variants coexist. Each panel assumes a different mode of inheritance: the allele is dominant in both sexes (a), dominant in males but recessive in females (b), recessive in males but dominant in females (c) or recessive in both sexes (d)

Fig. 2

Detrimental effect of selection for late-life male fertility on female longevity (=net health). Changes in female lifespan were regressed against improvements in male fertility in populations subject to artificial selection for male late-life fertility relative to the stock population from which experimental flies were derived (correlation coefficient = −0.93; P = 0.022). Thus the stock population acts as a baseline against which evolution was assessed. If a value is equal to zero, the trait average in the selected population is identical to the trait average in the control population. As the value increases, the experimental population has an increasingly higher trait value relative to the control population. As the value declines, the experimental population has a lower trait value than the control. Line averages and 95% confidence intervals are shown

Fig. 3

Associations between late-life male fitness and female health measures. a Male fertility at old age (35 days old) plotted against female performance in negative geotaxis assays (where a high number indicates a long distance travelled and high performance) (correlation coefficient: −0.715, P = 0.046) and b female recovery time from anaesthesia (where a large number is a slow recovery and hence a negative indicator of performance) (correlation coefficient: 0.257, P = 0.577) in a sample of iso-female lines (standardised genotypes). Line averages and 95% confidence intervals are shown