M119. Linguistic Content in Schizophrenia and Bipolar Disorder: Relationships With Cognition and Social Functioning

Abstract

Background: Individuals with schizophrenia and bipolar disorder suffer from impairments in social functioning. People with schizophrenia tend to use fewer words overall when speaking compared healthy controls, and fewer words with positive emotional valence, yet it is unknown how this linguistic structure compares to the structure in bipolar disorder and how linguistic structure is related to social functioning. Method(s): Thirty-nine individuals with bipolar disorder and 42 individuals with schizophrenia were randomly selected from a larger study. All participants were audio recorded while they completed the Social Skills Performance Assessment (SSPA), which comprises role plays of making plans with a friend, greeting a new neighbor, and negotiating with a landlord. Recordings were transcribed and processed using the Linguistic Inquiry and Word Count (LIWC2015). Participants also completed a standard cognitive assessment battery, the Positive and Negative Syndrome Scale (PANSS), Beck Depression Inventory (BDI), and the Specifc Levels of Functioning rating scale (SLOF). Result(s): Compared to manualized norms, individuals with schizophrenia and bipolar disorder used signifcantly fewer complex words, d = 0.53, fewer positive emotion words, d = 0.66, and more nonfuencies and fller words, d = 1.58. There was no difference in the number of negative emotion words used. Participants with bipolar/schizophrenia used signifcantly fewer frst person words, d = 1.98, and more second person words, d = 1.90 than in normative data. Between diagnostic groups, people with schizophrenia used signifcantly fewer words, d = 0.86, and fewer negative emotion words, d = 0.61 than people with bipolar disorder. Total number of words used and number of negative emotion words were related to both the cognitive composite score and real world interpersonal functioning (rs =.33-.55), however linguistic variables contributed small, nonsignifcant variance to interpersonal functioning after accounting for clinical symptoms. Conclusion(s): Generally, people with schizophrenia and bipolar disorder communicated similarly; however, both groups exhibited substantial dif-ferences in linguistic structure compared to normative data. Treatments focused on helping people use more positively valenced words and normalizing the number of second person references compared to frst person references may help improve interpersonal functioning for people with severe mental illness.