Reports on the association between obstructive sleep apnea (OSA) and risk of death, cardiovascular (CV) events, diabetes and depression have been inconsistent.

We conducted a systematic review of the prognostic value of clinical and polysomnographic (PSG) characteristics of OSA for adverse long-term outcomes of untreated OSA in adult patients. A comprehensive search strategy for prognosis studies, OSA, CV events, mortality, depression and diabetes was developed in collaboration with a medical information specialist. All English language studies, from Jan 1999 to Dec 2011, with longitudinal design in adults with OSA diagnosed by PSG recording, found through Medline, Embase and bibliographies of identified articles, were considered eligible. Quality was assessed using published guidelines.

Among 26 articles, ten evaluated the association of OSA with mortality, 9 with a composite CV outcome, 4 with stroke, 2 with diabetes and 1 with depression. Significant relationships between the apnea–hypopnea index (AHI) and outcomes of interest were reported in 18 studies: seven for all-cause mortality, six for composite CV events, three for stroke, one for diabetes and one for depression. The effect of AHI was attenuated by female gender, older age, absence of daytime sleepiness and higher body mass index. Due to clinical heterogeneity between studies, meta-analyses were not performed.

Evidence exists in men for a relationship between OSA and all-cause mortality and a composite CV outcome. Associations between OSA and other outcomes remain uncertain. Among OSA-specific markers, only AHI was a consistent predictor. Other consistent predictors were traditional CV risk factors. Research is required to identify effect modifiers and the predictive ability of various AHI threshold values and hypopnea definitions. An enhanced set of OSA-specific predictors will allow better risk stratification to guide OSA treatment.