Authors :
Presenting Author: Jana Jones, PhD – University of Wisconsin School of Medicine & Public Health
Aaron Struck, MD – University of Wisconsin-Madison
Dace Almane, MS – University of Wisconsin School of Medicine & Public Health
Bruce Hermann, PhD – University of Wisconsin
Rationale:
From the initial descriptions of JME, alterations in behavior have been considered a characteristic feature of the disorder. Over the subsequent decades a diversity of behavioral tests and inquiries into rates of DSM-defined Axis I and II disorders in JME have characterized the degree of behavioral risk associated with JME. Recently, the application of unsupervised machine techniques to behavioral and psychiatric data in other epilepsies (e.g., TLE) has demonstrated the existence of distinct behavioral phenotypes, each associated with unique sociodemographic, clinical and neuroimaging features. This conceptual move, from a focus on tests to a precise grouping of patients, has yet to be undertaken in JME and represents the purpose of this report. In addition, unaffected siblings are included to address the issue of familial aggregation of behavioral risk compared to unrelated controls.Methods:
Inclusion criteria included chronological age between12-25 years, English speaking, with a diagnosis of JME supported by at least two of the three following criteria: 1) early morning myoclonic jerks 2) generalized tonic-clonic seizures, 3) an EEG with bursts of 3.5-5Hz generalized spike-wave and/or polyspike wave discharges. The control group was composed of siblings of JME participants (n=76) and healthy controls (n=43) and sibling controls (n=19). All participants underwent a structured psychiatric interview based on age using the Kiddie Schedule for Affective Disorders and Schizophrenia (KSADS) for school age children or Structured Clinical Interview (SCID) for adults to evaluate the presence of current and lifetime-to-date depression, anxiety, and ADHD. Latent class analysis was performed to identify current/LTD psychiatric groups. Classes from 2 to 6 were examined and the lowest BIC was used to determine the optimal number of classes. Results:
Based on the entire cohort, 4 latent classes were found, including no psychiatric disorders, ADHD, Anxiety, and Depression, with a significantly different distribution of latent classes across groups (Table 1). Inspection of the distribution of latent psychiatric groups across the JME, unaffected siblings and unrelated control participants revealed no significant differences between the JME and sibling participants (p=0.259). Significant differences existed between JME and unrelated controls (p=.006) with a higher proportion of no current/LTD psychiatric disorders in unrelated controls compared to JME participants (60.5% vs 31.5%) and lower proportion of current/LTD ADHD in unrelated controls compared to JME subjects (0% vs 11.8%).
Conclusions:
In this first application of cluster analysis to the current/LTD psychiatric history of participants with JME, siblings and unrelated controls, specific behavioral phenotypes are found to characterize the spectrum of psychiatric comorbidities. JME is underrepresented in the no current/LTD psychiatric disorder cluster and overrepresented in the ADHD cluster. Siblings occupy an intermediate position with no differences between JME and unrelated controls. JME is uniquely linked to ADHD, with shared susceptibility to the other behavioral categories.
Funding: NIH 1R01NS111022-01A1