MMPI correlates of sleep and respiratory disturbance in obstructive sleep apnea.

STUDY
OBJECTIVE: To evaluate associations between polysomnographic variables in obstructive sleep apnea (OSA) and a variety of psychological responses (including depressive symptoms) as assessed by the Minnesota Multiphasic Personality Inventory (MMPI).
DESIGN: Cross-sectional.
SETTING: University sleep disorders center. PATIENTS: One-hundred seventy eight consecutive clinical OSA patients.
INTERVENTIONS: Not applicable. MEASUREMENTS AND
RESULTS: Patients completed the MPI prior to overnight diagnostic polysomnography. Fifty-eight percent demonstrated at least one MMPI elevation (mean = 1.8 elevations), with Depression (D) elevated for 32%, Hypochondriasis (Hs) for 30%, and Hysteria (Hy) for 21%. Thirty-eight percent demonstrated two or more elevations, with several variations of Hs-D and Hs-D-Hy configurations evident. "Conversion V" profiles were fairly rare, and a large number of miscellaneous configurations occurred once. Significant correlations were detected between several MMPI scale scores and total sleep time, the apnea-hypopnea index (AHI) during REM, and particularly arterial oxygen saturation, even when partialling out variance related to body mass index (BMI). In contract, D scores were not correlated with any polysomnographic parameters. Based upon MMPI configuration, the sample was subdivided into the following seven profile groups: Nonelevated (n = 74); Single D (n = 11); Single non-D (n = 25); Combined D plus (a) HS or HY (n = 7), (b) Hs and Hy (n = 10), or (c) other (n = 29); and Multiple non-D (n = 22). Multivariate analysis controlling for age and gender indicated higher AHI in the Single non-D, Combined D plus other, and Multiple non-D groups, compared to the Single D group. Also, there was lower average oxygen saturation in the Multiple non-D group, compared to Single D, Single non-D, and Nonelevated groups. The Combined D plus HS and/or Hy groups did not differ from each other or from other groups, even when merged. The Multiple non-D findings were unattributable to any specific scale or overall number of elevations.
CONCLUSIONS: OSA patients who have core depressive symptoms (as measured by MMPI scale D) without significant psychological symptoms in other areas tend to have less severe OSA, whereas those with a diverse set of other psychological symptoms overshadowing depressive symptoms (e.g., somatic focus, emotional reactivity, family/marital problems, cognitive problems, etc.) tend to have greater AHI and lower oxygen saturation. Although it seems probable that these MMPI differences primarily reflect OSA effects, prospective research is needed to confirm this hypothesis.