Animal models have shown a quantal slowing of respiratory pattern when exposed to opioid agonist, in a pattern similar to that observed in central sleep apnea. We postulated that opioid-induced hypoventilation is more likely to be associated with sleep apnea rather than hypoventilation alone. Since we did not have a direct measure of hypoventilation we used hypoxemia as an indirect measure reasoning that significant hypoventilation would not occur in the absence of hypoxemia.

We conducted a retrospective analysis of 98 consecutive patients on chronic opioid medications who were referred for overnight polysomnography. All patients on chronic opioids seen in the chronic pain clinic were referred for a sleep study regardless of whether they had sleep symptoms or not. Sleep-related hypoxemia was defined as arterial oxyhemoglobin saturation of less than 90% for more than 5 min with a nadir of ≤85%, or greater than 30% of total sleep time at an oxyhemoglobin saturation of less than 90%.

Of the 98 patients, 36% (95% CI 26–46%) had obstructive sleep apnea, 24%, (95% CI 16–33%) had central sleep apnea, 21% (95% CI 14–31%) had combined obstructive and central sleep apnea, in 4% (95% CI 0–10%) sleep apnea was classified as indeterminate, and 15% (95% CI 9–24%) had no sleep apnea. Opioids were potentially responsible for hypoxemia during wakefulness in 10% of patients (95% CI 5–18%) and for hypoxemia during sleep not clearly associated with apneas/hypopneas in 8% of patients (95% CI 4–15%). Two patients (2%, 95% CI 0–7%) had sleep-related hypoxemia in the absence of sleep apnea or hypoxemia during wakefulness.

Patients on chronic opiate therapy for chronic pain have an extremely high prevalence of sleep apnea and nocturnal hypoxemia. Hypoxemia can occur during quiet wakefulness in patients on chronic opioid medications with and without sleep apnea. In patients on chronic opioid therapy, isolated nocturnal hypoxemia without coexisting sleep apnea or daytime hypoxemia is very uncommon.