Sleep studies count the oxygen you lose. This one measured what comes back.

Sleep medicine has spent a decade trying to replace the apnea-hypopnea index with something that predicts outcomes better. Hypoxic burden — the depth, duration and frequency of oxygen desaturations — has been the leading candidate. A study published in the Journal of Clinical Sleep Medicine on 31 August proposes measuring the opposite half of the same curve: not how far oxygen saturation falls, but how far it overshoots on the way back [s1].

The measure

Oxygen overshoot burden is defined as the area under the oxygen saturation curve above a stable sleep baseline; a wakefulness baseline was also considered [s1]. The rationale is oxidative stress. Repeated cycles of desaturation followed by reoxygenation are a plausible source of oxidative injury, and the reoxygenation phase has been largely unmeasured. The authors note that the role of oxidative stress in central sleep apnea specifically remains unknown [s1].

The analysis drew on 7,530 participants from two long-running cohorts: the Sleep Heart Health Study (mean age 64 years, 47.0% male) and the Osteoporotic Fractures in Men Study (mean age 76, 100% male) [s1]. Overshoot burden was modelled continuously in the primary analysis. Secondary analyses classified participants as having central sleep apnea, obstructive sleep apnea, or no sleep apnea, and stratified each by high versus low overshoot burden at the median [s1]. Cox models adjusted for covariates estimated hazard ratios for major adverse cardiovascular events [s1].

What it found

Over a follow-up of 8.92 years, 2,258 major adverse cardiovascular events occurred [s1].

Continuous oxygen overshoot burden was associated with increased risk, with stronger associations among people with any central apneas (P = 0.003 for the sleep baseline and P = 0.02 for the wakefulness baseline) [s1].

The stratified analysis is where the pattern becomes specific. Participants with central sleep apnea and high overshoot burden had a cardiovascular event incidence of 54.6% against 25.6% in controls, a hazard ratio of 1.45 (95% CI, 1.16–1.82; P = 0.001) [s1]. Central sleep apnea with low overshoot burden showed risk comparable to controls. So did obstructive sleep apnea with either high or low overshoot burden [s1].

The results were robust to adjustment for the apnea-hypopnea index, the central apnea index, loop gain and hypoxic burden [s1] — meaning the overshoot measure was not simply restating severity already captured by the standard indices.

The scoring problem underneath it

A finding that turns on distinguishing central from obstructive events runs directly into a measurement problem the field has not solved. An expert panel discussion published in the same journal on 27 August argues that while there is general agreement on the criteria for scoring obstructive and central apneas, there is much less guidance and clarity on how and when to classify obstructive and central hypopneas [s2].

That gap is not academic. The panel reports that when hypopneas are systematically classified, a substantial number of patients turn out to have a high proportion of coexisting central disease [s2]. Obstructive and central sleep apnea share pathophysiological features and often coexist in the same patient [s2] — so a classification convention that defaults ambiguous hypopneas to obstructive will systematically undercount central disease.

The panel argues the distinction is growing in importance for evaluating patients for neurostimulation and for emerging pharmacological therapies, and calls for reproducible approaches to hypopnea classification [s2].

Read together, the two papers describe a circular problem. A phenotype defined by central events carries elevated cardiovascular risk [s1]; the scoring rules that identify central events are inconsistent between laboratories [s2].

What it does not establish

The authors are explicit that the findings require validation in larger and more diverse populations, citing sample limitations and potential pulse oximetry bias across skin pigmentation [s1]. That second caveat is a serious one for any measure derived from oximetry, and it applies to hypoxic burden as much as to overshoot.

The cohort composition compounds it. One of the two studies is entirely male and has a mean age of 76 [s1]; the combined sample is weighted toward older adults, in whom central sleep apnea is more prevalent and more often accompanied by cardiac disease. Whether the phenotype identifies risk in younger or more balanced populations is untested.

And this is an association measured over 8.92 years of follow-up, not an intervention [s1]. Nothing here shows that reducing overshoot burden reduces cardiovascular events. What it offers is a candidate for risk stratification — a way of separating the central sleep apnea that appears to matter from the central sleep apnea that, in this data, does not.

Sources

• [s1] "Oxygen overshoot burden of central sleep apnea and its association with cardiovascular outcomes," Journal of Clinical Sleep Medicine, 31 August 2026. https://doi.org/10.1007/s44470-026-00161-0
• [s2] "The importance of classifying central hypopneas: a call to action," Journal of Clinical Sleep Medicine, 27 August 2026. https://doi.org/10.1007/s44470-026-00177-6