PURPOSE: Introduction: Studies have shown strong association between insulin resistance, diabetes mellitus(DM) and sleep apnea (OSA). Ethnic variance in the prevalence of DM in patients with OSA documented by full polysomnography (PSG) has not been well described. The objective of this study was to describe the prevalence of DM in a cohort of Hispanic and Caucasian patients with OSA and to assess PSG related differences, particularly in REM sleep, in that cohort if any.
METHODS: A retrospective chart review of 521 selected patients (255 Hispanic, 266 Caucasian) with OSA was carried out. All patients had undergone full night baseline PSG. We excluded charts of patient who underwent split-night studies and those who did not have any REM sleep. DM was assessed based upon patient self-identification on our standard sleep and medical history questionnaire.
RESULTS: Of 521 patients 32.43% had self-reported DM –there was a significantly higher proportion of DM amongst Hispanics vs. Caucasian (36.86% vs. 28.19% respectively; p = 0.008). Patients with DM, as a group were older, in both ethnicities. No differences were seen in any of the PSG variables such as AHI, REM AHI, O2 nadir, and minutes spent with O2 saturation <85% between patients with DM and those without, in either ethnic cohort.(table 1).
CONCLUSION: There is a high prevalence of DM in a unselected cohort of patients with PSG-proven OSA. The prevalence is significantly more in the Hispanic population.Differences in PSG parameters do not account for presence of DM nor are they predictive of the observed ethnic differences.
CLINICAL IMPLICATIONS: Patients with OSA should routinely be screened for presence of glucose intolerance. Routine PSG parameters do not reliably differentiate those patients with DM from those without - further studies are needed to define what other PSG findings, if any, may prove to be helpful in predicting presence of glycemic intolerance.
DISCLOSURE: Shyam Subramanian, None.