Abstract
Introduction
Sleep depth decreases in the transition from childhood to adolescence, even in typically developing (TD) youth. However, it remains unknown whether this developmental trajectory in NREM sleep depth differs across adolescents with psychiatric/behavioral disorders.
Methods
We analyzed the sleep EEG of 392 subjects aged 5–12 at baseline and 12–22 at follow-up (45.2% female, 23.2% racial/ethnic minority), of whom 246 were TD adolescents (controls), 62 were diagnosed with a psychiatric/behavioral disorder and were taking stimulant, anti-depressant, anxiolytic, sedative and/or anti-psychotic medications, and 84 were un-medicated. NREM sleep depth was measured at both time points using the odds ratio product (ORP), which provides a standardized continuous EEG measure of NREM sleep depth/arousability (higher ORP reflects lighter NREM sleep). General linear models examined mean differences between groups on the percent change in ORP between baseline and follow-up (ΔORP) while adjusting for sex, race/ethnicity, age, BMI and AHI at follow-up, and PSG system, psychiatric/behavioral disorders, psychoactive medications and ORP at baseline as well as time-to-follow-up.
Results
Overall, medicated (80.4%, 95%CI=66.2–94.6) and un-medicated (66.1%, 95%CI=53.0–79.1) subjects showed a higher ΔORP compared to controls (52.2%, 95%CI=40.0–64.5, p<0.01 and p<0.05, respectively) but did not differ between each other (p=0.134). Specifically, un-medicated subjects with ADHD (n=56) showed a higher ΔORP (77.3%, 95%CI=62.4–92.1) compared to controls (p<0.01), while subjects with ADHD on stimulant medication (n=36) did not differ (66.1%, 95%CI=48.9–93.2) from controls (p=0.268) or from un-medicated ADHD subjects (p=0.303). Subjects with internalizing disorders on psychoactive medications (n=29) showed a higher ΔORP (104.9%, 95%CI=82.8–127.0) compared to controls (p<0.01) and to un-medicated subjects (n=27) with internalizing disorders (60.1%, 95%CI=36.8–83.3, p<0.01), who did not differ from controls (p=0.772).
Conclusion
The greater increase in ORP in the transition to adolescence in un-medicated youth with ADHD suggests that decreased NREM sleep depth may be a biomarker of the disorder. In contrast, the greater increase in ORP in medicated youth with internalizing disorders suggests that psychoactive medications impact NREM sleep depth in these children as they transition to adolescence. These data have important implications for sleep EEG studies that include medicated and un-medicated youth with comorbid psychiatric disorders.
Support (if any)
NIH Awards Number R01MH118308, R01HL136587, R01HL97165, R01HL63772, UL1TR000127
Long-, intermediate- and short-acting benzodiazepine effects on human sleep EEG spectra
