IQ predictors in pediatric opsoclonus myoclonus syndrome: a large international cohort study

AIM To determine predictors of full-scale IQ (FSIQ) in an international pediatric opsoclonus myoclonus syndrome (OMS) cohort. METHOD In this retrospective and prospective cohort study at three academic medical centers (2006-2013), the primary outcome measure, FSIQ, was categorized based on z-score: above average (≥+1), average (+1 to -1), mildly impaired (-1 to -2), and impaired (<-2). Univariate analysis and multivariable linear regression modeling using stepwise selection with Akaike's information criterion was performed to understand the relationship between exposures and FSIQ. RESULTS Of 81 participants, 37 with sufficient data had mean FSIQ 84.38 (SD 20.55) and median 90 (40-114) at latest available evaluation (mean age 8y 5mo). Twenty (54%), nine (24.3%), and eight (21.6%) had normal, mildly impaired, and impaired FSIQ respectively. The final multivariable linear regression model included 34 participants with evaluable data: number of relapses occurring before neuropsychological testing (p<0.001) and OMS severity score at last follow-up (p<0.001) predicted FSIQ (adjusted R2 =0.64). There was a mean decrease of 2.4 FSIQ points per OMS relapse. INTERPRETATION Number of relapses negatively correlates with FSIQ in pediatric OMS. Demographic and clinical measures available at OMS onset did not predict FSIQ. Strategies to reduce OMS relapses may improve intellectual outcomes.