A self-directed weight-loss programme missed its mark in sleep apnea

A low-cost, self-directed weight-loss programme did not meaningfully help patients with obstructive sleep apnea lose weight, according to the POWER trial published in the American Journal of Respiratory and Critical Care Medicine [s1]. It is a useful negative result: weight loss is standard advice for sleep apnea, and this is a real-world test of whether a scalable, hands-off programme can deliver it [s1].

Why the question matters

Obstructive sleep apnea, in which the airway repeatedly collapses during sleep, is closely tied to excess weight, and losing weight can ease it. But intensive, in-person weight-loss programmes are hard to deliver at scale. A self-directed programme — materials and remote coaching a patient works through largely on their own — could in principle reach far more people, if it works in routine practice [s1]. POWER was built to test exactly that pragmatic question rather than efficacy under ideal conditions.

What the trial did

POWER enrolled patients with obstructive sleep apnea and a body-mass index of 30 to 45 kg/m², recruited from Veterans Affairs sites across the United States [s1]. Participants were randomised to either a self-directed lifestyle intervention — based on the Diabetes Prevention Program, with video-based sessions, readings and regular coaching reminders — or usual care [s1]. The trial had two co-primary outcomes measured from baseline to 12 months: the change in the Functional Outcomes of Sleep Questionnaire (FOSQ), which captures how sleepiness affects daily life, and clinical weight in kilograms [s1]. Because there were two primary outcomes, the researchers set the bar for statistical significance at p=0·025 rather than the usual 0·05 [s1]. The trial is registered as NCT05104450 [s2].

In all, 696 participants were randomised [s1]. Their mean age was 56·2 years (SD 12·2), and 19·7% identified as female [s1] — a heavily male population reflecting the VA setting, which limits how far the findings extend to women.

What it found

Twelve-month weights were available for 81·9% of the intervention group and 83·6% of the control group [s1]. Unadjusted weight fell by 2·1 kg (SD 6·8) in the intervention group and 0·9 kg (SD 5·9) in usual care, for an adjusted between-group difference of −1·1 kg (95% CI −2·0 to −0·1; p=0·037) [s1]. That p-value clears the conventional 0·05 line but not the trial's own prespecified 0·025 threshold — so by the standard the researchers set in advance, the weight-loss endpoint was not met [s1].

On the sleep-related quality-of-life endpoint, FOSQ scores at 12 months were available for 78·2% of the intervention group and 90·5% of controls [s1]. Scores improved by 1·4 points (SD 2·7) with the intervention and 0·9 points (SD 2·9) with usual care, an adjusted difference of 0·6 points (95% CI 0·2 to 1·0; p=0·001) [s1]. This endpoint did reach significance — but the authors themselves judge the size of the improvement to be of limited clinical importance [s1].

What it means

The researchers' own conclusion is blunt: among people with sleep apnea and obesity, the self-directed programme did not improve weight loss, and the gains in sleep-related quality of life were small enough to be of limited clinical value [s1]. The differential dropout on the FOSQ measure — more complete data in the control group than the intervention group — is also worth noting, because participants who disengage from a programme may differ from those who stay, which can colour patient-reported results.

The pragmatic design is both the study's strength and the key to reading it. Pragmatic trials test an intervention under everyday conditions rather than the tightly controlled setup of an efficacy study, so a weak result here reflects how the programme is likely to perform if rolled out — not a failure to give it the best possible chance [s1]. By the same token, it says little about whether a more intensive, in-person or clinician-led programme would do better, because that is not what was tested. The recruitment from Veterans Affairs sites, with a population that was 19·7% female and had a mean age of 56·2 years, also means the results speak most directly to older men and should be generalised to women and younger patients only with caution [s1].

The value of a trial like this is in what it rules out. A hands-off, mail-and-video weight-loss programme is attractive precisely because it is cheap and scalable, but POWER shows that scalability is not the same as effectiveness: the programme moved the scale by about a kilogram more than doing nothing, and that was not enough to meet its own success criteria [s1]. For patients with sleep apnea hoping weight loss will ease their symptoms, the implication is that a self-guided programme is unlikely to be sufficient on its own, and more intensive or medically supported approaches may be needed.