© 2003 American Thoracic Society
The Hormone Replacement Dilemma for the PulmonologistBrigham and Women's Hospital Boston, Massachusetts Although it has been recognized for years that obstructive sleep apnea (OSA) is more common in men than in women, the impact of menopause on the prevalence of sleep apnea has been less clear (1–3). A number of studies have suggested an increased prevalence in postmenopausal women, but confounding due to age and body mass index have left this issue unresolved (1–3). It has also been proposed that the female hormones, estrogen and progesterone, may modulate apnea pathophysiology and protect premenopausal women, thus explaining the male predominance in this disorder. Although this could also explain the increased prevalence of OSA in postmenopausal women, the administration of estrogen and progesterone to postmenopausal apnea patients has yielded mixed results in terms of improvement in apnea frequency (4–6).
Two papers in this issue of AJRCCM (pp. 1181–1185 [7] and 1186–1192 [8]) have brought near closure to several of these issues. First, Young and coworkers (7), utilizing the Wisconsin Sleep Cohort (n = 589), including both longitudinal and cross-sectional data, have reported a gradual increase in the prevalence of OSA through the perimenopausal period with a substantial increase in postmenopausal women. Their fully adjusted model yielded an odds ratio of 2.59 for OSA (apnea–hypopnea index These observations raise two fundamental questions. By what mechanism does estrogen and progesterone modulate apnea pathogenesis? Should HRT be utilized routinely in postmenopausal women, either to prevent apnea development or as therapy for sleep apnea? Neither study provides any particularly mechanistic information regarding the effect of estrogen and/or progesterone on apnea pathophysiology other than the observation that this effect is not modulated purely through body weight or body mass index. Numerous investigators have attempted to delineate the mechanisms explaining the strong male predominance in OSA with no clear single answer emerging. Subtle sex differences however, in virtually all of the component processes that could lead to sleep apnea, have been reported. These include sex differences in ventilatory control (hypoxic/hypercapnic ventilatory responses (9) and sleep apnea PCO2 thresholds [10]), pharyngeal anatomy (surrounding fat distribution (11) and airway length [12]), upper airway muscle activation (13), and pharyngeal mechanics (collapsibility while asleep [14]). One has to suspect that estrogen and/or progesterone could affect some or all of these processes in a manner that reduces the probability of developing sleep apnea (less collapsible airway, improved ventilatory stability, and increased dilator muscle activation). Sex effects, however, are also likely mediated by other mechanisms as well (androgens, nonhormone effects, and so on). In addition, the effects of these hormones (estrogen and progesterone) are not likely so large that they can overcome major anatomic or physiologic predispositions to OSA. For instance, if an individual's pharyngeal airway is remarkably small in size secondary to obesity or bony structures, female hormones will not likely have a large enough effect to overcome this anatomic predisposition. On the other hand, a woman with an anatomically quite large upper airway will not likely develop sleep apnea when she becomes postmenopausal. However, when these predispositions are more subtle, the influence of these hormones may be strong enough to have an important impact on apnea frequency. Thus the loss of these hormones (menopause) and a failure to replace them (no HRT) increases the prevalence of sleep apnea importantly. The larger question, however, is how should we manage postmenopausal women in whom there is 11–12% prevalence of clinically relevant sleep apnea? It is probably not reasonable to place all postmenopausal women on HRT to prevent sleep apnea because 88–89% will not have clinically important apnea and thus do not require treatment. In addition, although sleep apnea is believed to contribute to the development of cardiovascular disease (15), most current evidence does not suggest that HRT improves cardiovascular outcomes (16). Thus, if HRT does reduce sleep apnea, the resultant effect on cardiovascular disease is either negligible or is offset by some other effect of the hormones. There is also no current precedent for apnea screening in high-risk populations (as postmenopausal women seem to be) in that inexpensive, sensitive/specific screening tools are not available and current treatment approaches are not likely acceptable to asymptomatic or minimally symptomatic patients. As a result, we are left with physician-based case identification, which in men is relatively easily accomplished if primary care physicians ask the appropriate questions (snoring, witnessed apnea/gasping, waking hypersomnolence, and so on). In women, symptom-based identification may be more difficult because classic symptoms are less commonly reported. Thus careful questioning may be required, which does not often occur. This speaks to the substantial education task ahead. The two studies in this issue of AJRCCM suggest that such education is broadly needed if we are to address this growing problem. Finally, if apnea is identified in a postmenopausal woman, how should it be treated—hormones or continuous positive airway pressure? Based on the current literature (4–6) and the pathophysiologic principles described above, I suspect that patients with an apnea–hypopnea index greater than 25 or 30 will not respond to HRT and will ultimately require continuous positive airway pressure. In women with less severe disease, HRT may be a viable therapeutic approach and should be considered. However, randomized clinical trials of HRT in postmenopausal apnea patients are needed before a rationale approach can be defined. REFERENCES
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
15) in postmenopausal versus premenopausal women. Second, Shahar and coworkers (
