Metabolic syndrome is present in over half of adults with confirmed sleep apnoea

Design

Systematic review and three-level random-effects meta-analysis with meta-regression, PROSPERO registered

Population

34,013 adults with polysomnography-confirmed obstructive sleep apnoea across 102 studies from 28 countries

Primary outcome

Prevalence of metabolic syndrome

Effect

Pooled prevalence 55.4 per cent (95 per cent CI 51.0 to 59.8), I-squared 97.8 per cent; prevalence rose with cohort mean body mass index (beta 0.0772, P < 0.0001)

One hundred and two studies covering 34,013 adults from 28 countries, all with obstructive sleep apnoea confirmed on standard polysomnography rather than by questionnaire, were pooled for metabolic syndrome prevalence. The combined estimate was 55.4 per cent (95 per cent CI 51.0 to 59.8).

Heterogeneity was 97.8 per cent, and the subgroup analyses show why: prevalence differed significantly by geographic region, study design, which definition of metabolic syndrome was used, and what apnoea-hypopnoea index threshold defined the sleep apnoea. Meta-regression found prevalence rose with the mean body mass index of the cohort (beta 0.0772, P < 0.0001), which is the expected and slightly deflating explanation - much of what links the two conditions is the obesity underlying both.

The number is nonetheless usable, because the direction of the clinic visit is one-way. A patient reaches a sleep laboratory because of snoring and daytime sleepiness; nobody arrives asking about their waist circumference and triglycerides. If more than half of them meet criteria for metabolic syndrome, then blood pressure, waist circumference, fasting glucose or HbA1c and a lipid profile belong in the same episode of care as the polysomnography report, not in a referral to somebody else. That is a low-cost addition in Indian practice, where the metabolic phenotype appears at lower body mass index than these largely Western cohorts would suggest.

• Measure waist circumference, blood pressure, glucose or HbA1c and lipids when sleep apnoea is confirmed.
• Do it in the same episode as the sleep study rather than referring onward.
• Expect the association to be driven substantially by shared obesity, not by apnoea alone.
• Use Asian waist circumference thresholds; the pooled cohorts are mostly not South Asian.
• Treat the 55 per cent figure as a reason to screen, not as an individual's risk.

The statistics, in plain English

An I-squared of 97.8 per cent means the studies are estimating genuinely different quantities, so the pooled 55.4 per cent is an average across incompatible populations rather than a figure to quote for any one clinic. The confidence interval of 51.0 to 59.8 describes precision of that average, not its applicability. The meta-regression on body mass index is the most informative part: it identifies obesity as a substantial driver, which means the prevalence in a leaner cohort will be lower.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.