SEX AND GENDER DIFFERENCES IN HEALTH
Despite the rather modest attention paid to sex and gender in recent publications, as discussed in the preceding section, men and women continue to have different patterns of illness and life expectancies. Although sex and gender differentials have narrowed in recent years, they remain substantial.30 Questions about the origins and mechanisms for these disparities have long intrigued scientists from many disciplines. In particular, how and why sex and gender differentials unfold in aging over the life course are not well understood.21
Overall, sex and gender differences in health and longevity are pervasive and persistent. Lower survival rates and life expectancy for males than for females have been found across time, place, and even species.4,31 Among humans, the “sex health paradox,” or the observation that females have higher rates of morbidity but lower rates of mortality, motivated a series of analyses with national health statistics in the United States.32–34 This research found the following patterns: (1) sex differences in the distribution of illness conditions; namely, women have higher rates of acute illnesses and most nonfatal daily symptoms and chronic conditions, whereas men have higher rates of the leading fatal conditions; (2) sex differences in age-related patterns of disease, such that men die from more life-threatening chronic diseases (cardiovascular disease and cancer) at younger ages; and (3) that controlling for a broad array of social behavioral factors reduces female excess in morbidity and reveals consistent greater male vulnerability to illness and death. These findings may have reconciled the discrepancy in trends in sex differences in morbidity and mortality, but they beg a further question: why are there sex gaps in health at all?
Because of sex-specific variation in biology, several theories propose a biological basis for sex differences in health and survival. First, such differences are consistent with marked sexual dimorphism across the species in major physiological functions such as immune competence related to reproductive biology and regulated by sex gonadal hormones. Second, aging research suggests sex differences in insulin-like growth factor 1, signaling, and oxidative stress production. A review of related hypotheses is available in Austad.4 Third, sex determination is asserted to be the major genetic determinant of longevity differences within human populations, fueling the idea that the biological differences between males and females may stem ultimately from the genes on the X and Y chromosomes themselves and in X inactivation.25
One line of reasoning is that the female advantage can be explained by the heterogametic sex hypothesis, which posits that the sex with 2 different sex chromosomes (in mammals, that would be the XY male) is shorter lived.4,36 The logic is that a single copy of genes on the X chromosome exposes males to more risk because any deleterious alleles on the X chromosome will have no compensatory allele. By contrast, cell mosaicism, uniquely owned by females, is advantageous because it ameliorates the deleterious effects of X-linked mutations and contributes to physiological diversity and robustness.37
Because one or the other female X chromosome is inactivated randomly throughout the tissues, cells with one of the 2 X chromosomes that have fewer deleterious alleles—the better one—will survive better and gradually predominate in all the cells in a specific tissue as aging progresses. About 15% of the genes on the “inactivated” X chromosome are not fully inactivated, which may also provide a survival advantage.38
In addition, it has been suggested that telomere length in humans may be longer in women than men.4 Although telomere length was first thought to be negatively related to cell aging, a recent review of the evidence on the telomere-aging link calls this assumption into question.39
In a parallel way, because of gender differences in society, social and epidemiological research has advanced explanations that emphasize social structural, behavioral, or psychosocial factors, such as social class, smoking, and stress, as factors that contribute to sex and gender differences in health and longevity. For instance, women’s lower socioeconomic status (in terms of educational attainment, participation in paid employment, income) and fewer job hours have been shown to compromise their health.34 By contrast, men are disproportionately disadvantaged by occupational hazards, gender role socialization, and health behaviors, including higher levels of cigarette smoking, alcohol consumption, meat and fat consumption, and aggressive behavior and violence.40–43 Nonetheless, men benefit relative to women from more physical activity and healthier body weight, as well as beneficial psychosocial risk factors, such as a higher sense of mastery and self-esteem.34,44 Furthermore, psychosocial stress, which has been positively associated with cardiovascular disease, may result from gendered processes, such as uneven family responsibilities, gender-specific harassment or discrimination, and unequal levels of poverty.23 In addition, women’s poorer health may reflect constrained choices.3,19 In sum, social behavioral models have demonstrated considerable power in explaining gender differences in morbidity and mortality. The discovery of a male health disadvantage after controlling for social factors is in line with biological theories about persistent male excess in mortality, and offers some resolution to the sex health paradox. Nonetheless, our knowledge is far from complete about why such a sex difference exists in the first place.
One major deficiency in extant research is that few studies simultaneously consider biological and social forces as explanations of sex and gender differentials in health and longevity. Thus, we know little about the multifaceted interconnections between social and biological processes or how social conditions can shape sex- and gender-specific patterns of health. Furthermore, research to date tends to ignore age and changing social contexts in the investigation of health. Gender inequalities in health over the life course vary.45,46 The gender gap can grow with age, consistent with cumulative advantage theory, or converge with age, consistent with an age-as-leveler or selective survival process. In addition, these age-related changes can be further conditioned or modified by contextual factors specific to birth cohort, factors that represent social historical changes exogenous to individual trajectories of physical and mental states. For example, the longitudinal study by Yang and Lee, based on a nationally representative sample of close to 4000 US respondents, found that the male-female gap in depressive symptoms converges with age, but with less convergence in more recent cohorts, and hence, a potentially weakening age-as-leveler process over time.46 This pattern not only affirms the need for a developmental approach to understanding the interplay between social and biological factors, but also an approach that features changing contextual variation.
Age variations in sex differences in health contribute to the complexity of explanations for sex differences in health but have not been systematically examined from a biological and developmental perspective nor rigorously modeled. One reason is that prospective longitudinal data are less common in the biological sciences than in the social sciences. Analysis of change over the life course may offer considerable leverage in understanding the nature of differences in health by sex. Specific patterns of age variation may provide important clues to the underlying biological mechanisms that contribute to sex and gender differences, once these variations are put in the context of development and aging. Recent population research using explicit measures of biological robustness shows strong evidence for post-reproductive change in sex differentials in physiological functions such as systemic inflammation, metabolic syndrome, and the allostatic load that are consistent with and partly account for the reduction of sex differences in overall and cardiovascular disease mortality in old age.47,48
In terms of genetic research on health, sex and gender differences, when they are considered, are often implicitly represented as biologically based, as discussed previously. An alternative approach may start with gender as an environment and explore the ways in which gendered experiences are embodied and reflected in physiological processes, and how and when genomic factors enable, constrain, or shape this process, or are, in turn, reshaped by it. An integrated research agenda could begin to tease out the nature of these interactions, and in so doing, elucidate pathways that move us toward explanation. Such an approach would not focus on sex-linked genes as genetically relevant variants. Instead, it would recognize that genetic variants similar in men and women may have different health implications when expressed in “contexts” or “environments” that are gendered.
SOCIAL AND BIOLOGICAL FACTORS OVER THE LIFE COURSE
We present 2 examples of social and biological factors affecting health over the life course with applications to gender differences in health. These examples extend previous studies by Yang and Kozloski and illustrate the importance of addressing some of the underexplored issues we have outlined above.47,48 Based on data from about 38 000 adults in the National Health and Nutrition Examination Survey (NHANES) conducted between 1988–1994 (III) and 1999–2006, we estimated multivariate regression models to assess the parametric relationships of sex and age with various biological functions and assess how social behavioral factors may account for such relationships.
Figure 2 shows the age curves of the inflammation burden by sex and smoking status predicted from the best fitting model that adjusts for other covariates. The curves indicate that the female excess in inflammation decreases with age (P < .001 for the sex by age interaction). Interestingly, the patterns of age variations in sex differentials depend on smoking behaviors. Cigarette smoking is associated with elevated risks of inflammation for both sexes, but the association is significantly larger for men than women (P < .001 for the sex by smoking interaction), such that the sex gap before old age is much smaller, converges much earlier, and even reverses later in life for current smokers compared with never or past smokers.