Castrating the respiratory controller

That obstructive sleep apnoea (OSA) has a negative impact on health is well accepted, with diseased individuals having greater rates of cardiovascular and metabolic morbidity, as well as psychosocial and neurological symptoms. The incidence of OSA is greater in middle-aged to older males compared to age-matched females (Young et al. 1993). Although the mechanism is unknown, obvious differences in the hormonal milieu have long been implicated. Androgen replacement has been shown to induce OSA in some men (Sandblom et al. 1983). Consistent with these observations are recent results (Liu et al. 2003) demonstrating a reduction in total sleep time, longer hypoxic episodes and increases in the respiratory disturbance index (the number of apnoeas and hypopnoeas per hour) in older men exposed to high doses of testosterone. There were no concomitant changes in metabolic function or upper airway dimensions, suggesting that the effects of testosterone on nocturnal breathing are due to effects on the central respiratory controller. These are remarkable findings, which emphasize the need for more research into the influence of androgens on ventilatory control in wakefulness and sleep, and underscore potential problems associated with androgen replacement therapy to enhance bone and muscle mass.

In this issue of The Journal of Physiology, Mateika et al. (2004) examined the effects of androgens on the ventilatory response to chemoreceptor stimulation during wakefulness and sleep in young, healthy males. Rather than testosterone administration, they chose to suppress androgens with the long-acting gonadotropin-releasing hormone (GnRH) agonist, leuprolide acetate. Although short-term administration of the drug has the same stimulatory action as GnRH, long-term therapy suppresses the production of follicle stimulating hormone and luteinizing hormone by down-regulating GnRH receptors. As expected, 2 months of treatment with leuprolide resulted in reductions in testosterone, oestrone and oestradiol, to levels observed in castrated males. During wakefulness, chemoreceptor responses were studied in 11 subjects with CO₂ rebreathing trials initiated from a low end-tidal partial pressure of CO₂ (P_ET,CO2) that was induced by prior, volitional hyperventilation. From these studies they defined a ‘ventilatory recruitment threshold’, which corresponds to the P_ET,CO2 associated with an abrupt rise in ventilation during the CO₂ rebreathing trials. The ventilatory recruitment threshold was significantly less after chemical castration, whether the background gas was hyperoxic or hypoxic. The sensitivity (slope of the ventilation–P_ET,CO2 response) of the response to CO₂ was also increased, but only when the test was performed in hyperoxia. These data suggest that androgens influence both the recruitment threshold (more commonly referred to as the ‘CO₂ set-point’) and the sensitivity of the respiratory controller to changes in P_CO2. The change in the set point proved to be more robust than the change in sensitivity because the former was altered in both hyperoxia and hypoxia. The failure of the CO₂ sensitivity to change with leuprolide when the rebreathing trials were done with the hypoxic background suggests that the peripheral chemoreflex was attenuated rather than enhanced. However, more detailed studies of the hypoxic ventilatory response before and after leuprolide treatment will aid in the interpretation of these findings.

Five of the 11 subjects were also re-studied during NREM sleep on a separate occasion. In this case, the subjects were mechanically hyperventilated to induce central apnoea, and the P_ET,CO2 at the apnoea point was determined. Consistent with the data recorded in wakefulness, castration resulted in a lower apnoeic threshold during NREM sleep, and this effect could not be attributed to leuprolide-induced differences in baseline P_ET,CO2 or upper airway resistance. These observations show, for the first time, that chemical castration reduces the ventilatory recruitment threshold and CO₂ chemosensitivity in wakefulness, as well as the apnoeic threshold in NREM sleep. These data are consistent with earlier work showing that females have lower apnoeic thresholds than males, and that women have an increased apnoeic threshold following testosterone administration. On the other hand, testosterone replacement in hypogonadal men had little effect on CO₂ or hypoxic sensitivity (White et al. 1985).

Like most interesting experiments, the experiments conducted by Mateika and colleagues raise as many questions as they answer. First, if androgens do increase the apnoeic threshold during sleep, it is possible that an additional respiratory perturbation such as a sigh or a period of hyperpnoea following an obstructive apnoea would be more likely to drive the arterial CO₂ below threshold, leading to a central apnoea, and initiating the apnoea–hyperventilation–apnoea cycle commonly seen in patients with OSA. However, Mateika et al. found that androgen suppression increased CO₂ sensitivity; this suggests that the relatively low CO₂ sensitivity in males with a normal hormonal profile would attenuate the effects of the elevated apnoeic threshold, perhaps offering a degree of protection as lower CO₂ sensitivity is associated with a more stable ventilatory control system. Second, androgens down-regulate oestrogen and progesterone receptors in tumour cell lines (Poulin et al. 1989), and increase oestrone levels with no effect on the level of progesterone or oestradiol. However, in the present study the oestrogen levels declined markedly, probably reflecting the sustained stimulation of gonadotrophs by leuprolide, as opposed to the normally pulsatile release of GnRH. This points to the need for more studies on the effects of androgens (and oestrogens) on the control of breathing, in particular the potential for androgen replacement therapies to effect alterations in receptor densities on central respiratory neurones. Third, and most important from a clinical perspective, the studies by Mateika and colleagues emphasize the need for more careful consideration of nocturnal ventilatory control before androgen replacement therapy is instituted, particularly in older men, obese people of either sex, and postmenopausal women.