Matsumoto · Clinical endocrinology 1985 · prospective before-and-after study · n=5

Testosterone replacement in hypogonadal men: effects on obstructive sleep apnoea, respiratory drives, and sleep.

Cited 249 times in the scientific literature.

Level 4 - case-series / case-control

Prospective pre-post interventional case series without a control group

PubMed 4017261 · doi:10.1111/j.1365-2265.1985.tb00161.x · record verified 2026-08-31

What was done

Five androgen-deficient hypogonadal men underwent waking ventilatory drive studies and overnight polysomnography while off treatment and again after six weeks of testosterone replacement therapy (testosterone oenanthate 200 mg intramuscularly every two weeks). Waking hypoxic and hypercapnic ventilatory drives, sleep architecture, and obstructive sleep apnoea events were assessed.

What was found

Hypoxic ventilatory drive decreased significantly on testosterone therapy, falling from 158 ± 39 (mean ± SEM) off treatment to 88 ± 19 on treatment (P < 0.05). Hypercapnic ventilatory drive did not change significantly. The proportion of sleep time in REM sleep increased from 14 ± 3% off treatment to 22 ± 2% on treatment (P < 0.01). Obstructive sleep apnoea developed de novo in one participant and markedly worsened in a second, with both men experiencing marked oxygen desaturations, nocturnal cardiac dysrhythmias, and large increases in haematocrit; apnoea events occurred predominantly during non-REM sleep. The other three participants showed no significant sleep apnoea either on or off testosterone.

Why it matters

This study provides early clinical evidence that physiological testosterone replacement can depress hypoxic ventilatory drive and induce or exacerbate obstructive sleep apnoea in susceptible hypogonadal men.

Limits

The study is limited by an extremely small sample size (n = 5) and an uncontrolled before-and-after design lacking a placebo group. The follow-up duration was short (six weeks), and clinical characteristics distinguishing responders from non-responders could not be reliably determined.

Cited by