Abstracts

Rationale: Previous studies of persons with psychogenic non-epileptic spells (PNES) indicate female predominance, a history of physical or sexual abuse, a later onset of events and a 7-10 year delay in official diagnosis by video-EEG. The literature also indicates persons with PNES have significantly higher rates of comorbid psychiatric conditions when compared with populations with or without epilepsy. Previous research also suggests headaches, chronic pain and obesity are more common in persons with PNES. However, investigations examining other common somatic conditions have not been conducted. The purpose of this study was to examine demographic, clinical and comorbid health conditions (psychiatric and somatic) predictors of persons with diagnosed PNES versus persons diagnosed with epilepsy from our epilepsy monitoring unit (EMU). Methods: Data was retrospectively gathered from existing medical records of persons who were admitted to the Ohio State University Medical Center’s EMU between 2002 and 2007. Statistical analyses were conducted to examine differences between persons diagnosed with PNES versus those a diagnosis of epilepsy. Persons who were diagnosed with both conditions were excluded. Results: A total of 324 (53%) participants had their typical events diagnosed as PNES and 282 (47%) were diagnosed with epilepsy by video-EEG. Seventy-one percent of those diagnosed with PNES were female compared to 54% that were diagnosed with epilepsy. Mean age of onset of spells (or seizures) for persons with PNES was 28 compared to 18 for epilepsy (p < 0.001). Persons with PNES had experienced their events for an average of 9 years before diagnosis in the EMU. Both populations report similar rates of disability, 38% for PNES vs. 41% for epilepsy. Both populations also reported high rates of smoking, 45% for PNES vs. 38% for epilepsy (p = 0.07) and high rates of obesity 49% vs. 40% (p = 0.02). Univariate logistic regression analyses revealed 23 individual factors that were significant predictors of PNES vs. epilepsy. Once these factors were entered into a multivariable logistic regression model, 9 predictors remained significant. Numbers of years with events, Odds Ratio (OR 1.1), history of physical/sexual abuse (OR 3.6), history of head injury (OR 1.9), anxiety (OR 1.7), asthma/COPD (OR 3.0), GERD (OR 1.6) and pain (OR 2.4) were significant predictors of PNES. Marital status was the only protective factor against PNES (OR 0.6).