Methods: This study is a secondary analysis of baseline data from participants in a randomized clinical trial (1R01DK096028). The parent study evaluates if persons with treatment of OSA + diabetes education have better glucose control and self-management behavior compared to those on a sham (placebo) device + diabetes education. Inclusion criteria for the baseline assessment include age > 18 years; self-report of T2DM, CPAP naïve, and willing to be randomized to CPAP if found to have OSA (apnea + hypopnea index [AHI] ≥ 10). Measures included demographics (sex, age, race, education), sleep quality (Pittsburgh Sleep Quality Index [PSQI] Buysse, Reynolds, Monk, Berman, Kupfer, 1989), (scores>5=poor sleep quality), insomnia (Insomnia Severity Index [ISI] Morin, Belleville, Belanger, Ivers, 2011); scores 0-14 = “no insomnia” to “mild insomnia” and scores ≥ 15 =“moderate insomnia” to “severe insomnia”), diabetes related distress (Problem Areas in Diabetes [PAID], Welch, Jacobson, Polonsky, 1997) with higher scores = more distress, and mental and physical HRQoL component scores (SF-12v2 MCS, PCS, Ware, Kosinski, Keller, Ware, Kosinski, Keller, 1996).) with higher scores = better mental or physical HRQoL. The clinical evaluation included A1C for glycemic control; height and weight were measured to calculate BMI. Descriptive statistics include mean (SD) for continuous variables and percent and frequency for categorical variables. Inferential statistics included independent Students t-tests, and Pearson correlations. Linear multiple regression models were conducted to see if age, race, college education, BMI, diabetes related distress, or insomnia predicted mental and physical HRQoL (MCS, PCS). The level of statistical significance was set at p<.05
Results: The sample (N=194) was primarily middle age (mean age=56.8 years ±10.7 [range 26-88 years], overweight (mean BMI= 34.7 ±6.8), had suboptimal glucose control (mean A1c= 7.9% ±1.8), and moderate-to-severe insomnia (41%). Participants were well distributed by sex (male 46%; n=90), race (white 54%, n=105), and college graduate (31%, n=60). No differences were observed in mental or physical HRQoL (MCS, PCS) by sex, race, marital status; participants with a college education had significantly (p<.01) increased physical HRQoL. Age was significantly associated with improved mental HRQoL (r=.25, p<.01). Participants with moderate-severe insomnia had significantly worse diabetes related distress [PAID], lower mental and physical HRQoL (MCS, PCS) than those who reported no insomnia/mild insomnia (all p values <0.5). Using the enter method it was found that age, BMI, diabetes related distress and insomnia explain a significant amount of the variance in mental HRQoL (F=12.564, p<.001, R2=.296, R2 adjusted =.273). The regression model for physical HRQoL found that age, college graduate, BMI, and insomnia explained a significant amount (F=7.575,p<.001, R2=.202, R2adjusted =.176). Limitations to this study include the cross-sectional sample, secondary analysis design, relatively modest sample size, and that the sample might not be reflective of all persons with T2DM.
Conclusions: Insomnia was highly prevalent in the recruited sample; insomnia was found to have significant negative impact on mental and physical HRQoL after controlling for age, race, education, BMI and diabetes-related distress. Future research is needed to determine if insomnia is associated with worse glucose control or diabetes self-management behavior.