Sex Hormones and Asthma.

A potential causal association between changes in sex hormones and asthma incidence or severity has long been supported by several well-known epidemiological observations; for example, asthma is more prevalent and severe in male children, yet after childhood, women are more commonly affected throughout adulthood (1). Perimenstrual worsening has been reported to occur in 20–40% of patients with severe or difficult-to-control asthma, particularly among obese females (2). On the basis of this information, suppressing sex hormone fluctuations through hormonal replacement therapy (HRT) would indeed make logical sense to reduce asthma incidence or potentially improve asthma outcomes; however, this has not been consistently shown, and in fact, some longitudinal studies have shown that HRT actually increased the odds of developing this disease (3, 4). Several factors such as the type and timing of HRT, the type study design, and individual characteristics (obese vs. nonobese) may potentially explain the variability in response to this treatment. It may also imply, however, that we do not fully understand how variations in sex hormone levels affect asthma, and therefore, we do not know in whom and when HRT could be useful.

To this effect, in this issue of the Journal, Han and colleagues (pp. 158–166) present the first large population-representative study evaluating the association between serum estradiol and serum free testosterone levels with self-referred current asthma diagnosis in men and women participating in the National Health and Nutrition Examination Survey (5). This cross-sectional analysis of 7,615 (3,953 men and 3,662 women) adults shows that women with higher serum testosterone levels have between 30% and 44% lower odds of current asthma when compared with the women with the lowest testosterone levels. Given the significant interaction found between obesity and sex hormones on asthma, the authors showed in the fully adjusted model that in obese women, higher serum estradiol and serum free testosterone were, respectively, associated with a 40% and a 40–50% reduction in the odds of current asthma. The associations in men were less impressive, as only the highest serum estradiol quartile, relative to the lowest, was associated with reduced asthma odds in nonobese subjects.

Whereas prior large-scale epidemiological studies have attributed changes in asthma-related outcomes to sex hormones only by proxy (i.e., puberty or menstrual period), the results from this study, by directly measuring serum levels, significantly strengthen causality. Further, the association with testosterone, which had been largely overlooked in many prior asthma studies, potentially adds new insights into the pathophysiology of sex hormones and airway diseases. A mechanistic link between testosterone and asthma is supported by studies showing that it is positively associated with lung function in both male and female children with asthma, and by the fact that inhaled dehydroepiandrosterone-3-sulfate (an androgen derivative) had been shown to improve short-term control in patients with moderate to severe asthma (6, 7). Also, in experimental models, testosterone has been shown to lessen airway smooth muscle contraction and to decrease lung innate type 2 cell numbers and lessen allergen-induced IL-5 and IL-13 expression in innate type 2 cells (8–10).

The interaction between serum testosterone levels and body mass index on asthma is certainly intriguing and could potentially play a role in understanding late-onset asthma in women, which after the age of 45 years becomes the predominant phenotype, particularly among those who are also obese (10). One could easily speculate that losing the protective androgenic effect with aging (11) could potentially explain why late-onset asthma is more common in older women, particularly in conjunction with other risk factors such as obesity and/or metabolic syndrome. The potential causal role of estradiol levels in asthma is more difficult to understand, given the inconsistencies in epidemiological and experimental studies. It is also possible that the estradiol associations with asthma found in this study are explained by unmeasured confounding.

In summary, this study represents a major advancement in the epidemiology of asthma and sex hormones and should spearhead further longitudinal and intervention studies related to primary asthma prevention. However, given its cross-sectional nature and the use of a self-referred asthma case definition, these findings should be interpreted with some caution.