0764 The Diagnostic Conundrum Posed By Antidepressant Medications While Conducting The Pediatric Mslt

Abstract

Introduction: Research examining the influence of antidepressants on multiple sleep latency test (MSLT) parameters in pediatric populations is limited. We examined the impact of rapid eye movement (REM)-suppressant antidepressant medications and other clinical, actigraphy and polysomnography (PSG) characteristics on mean sleep latency (MSL) and sleep-onset REM episodes (SOREMs) in a large pediatric clinical sample. Method(s): This was a retrospective chart review. We identified 164 MSLTs performed in patients aged <18 years. All of the data were manually abstracted. Correlations between clinical, actigraphy and PSG characteristics and MSL as well as SOREMs were examined using univariate and regression analyses. Result(s): Mean age of the sample was 11.9 years (SD 4.19); 62% were female, 28 (17%) were on REM-suppressant antidepressants (48% of whom were able to discontinue these prior to MSLT) and mean pediatric daytime sleepiness score was 21.7 (SD 6.1). MSL was 11.27 min (SD=5.77) and mean number of SOREMs was 0.55 (1.04). Twelve patients met criteria for narcolepsy and 40 for idiopathic hypersomnia. In the overall sample, MSL positively correlated with average time in bed on actigraphy, sleep-onset latency on PSG, and negatively correlated with age, use of REM-suppressant antidepressants and number of SOREMs (all p<0.05). The number of SOREMs positively correlated with age, BMI and arousal index and negatively correlated with use of REM-suppressant antidepressants, time in bed and total sleep time on actigraphy as well as self-reported sleep duration (all p<0.05). Similar findings were noted in the narcolepsy and idiopathic hypersomnia groups. In regression analyses accounting for factors significant in univariate analyses and sex, MSL and SOREMs continued to be associated with the use of REM-suppressant antidepressants (p=0.008 and 0.003 respectively). Further analyses examining the impact of tapering these medications in the weeks prior to MSLT did not change these associations. Conclusion(s): Clinicians should account for REM-suppressant antidepressant use while interpreting the results of MSLT. Future studies examining the impact, if any, and the precise timing of the discontinuation of these medications to reduce the likelihood of spurious results are required.