BackgroundGesture deficits in patients with schizophrenia are highly pronounced, and often linked to poor social functioning, motor abnormalities, and frontal lobe dysfunction. Although gesture performance has been associated to both negative and positive symptoms, its relationship to the severity of these symptoms is still unclear. Here, we examine how gesture performance varies as symptoms change. Furthermore, we aimed to compare gesture performance at two time points to healthy controls and first-degree relatives of schizophrenia patients. Gesture performance in relatives may indicate whether the deficits are associated with genetic liability to schizophrenia. We hypothesize that gesture performance in controls and relatives would be stable; while we expect improvement in patients when symptom severity declines.MethodsThe present study included 36 patients with schizophrenia (DSM-5 criteria; mean age 35.5 years), 28 unaffected first-degree relatives of schizophrenia patients (mean age 49.9 years) and 38 healthy controls (mean age 39.9 years). All three groups performed the Test for Upper-Limp Apraxia (TULIA), which includes pantomime (performance on verbal command) and imitative (performance upon demonstration) gestures, at two different time points, baseline and re-test (between 1–4 weeks). TULIA performance is recorded on videos and rated blind to diagnosis and stage. In addition, 22 of the 36 patients performed the TULIA at a follow-up session 6-months after baseline. Symptom severity was assessed with the Positive And Negative Syndrome Scale (PANSS). Analysis between the three groups and within patients across the different time points was done using repeated measures ANOVA in R.ResultsSymptom severity in patients declined between baseline and week 4 (T = 6.7, p<0.001, PANSS total). A 3x2x2 repeated measures ANOVA revealed significant main effects of Group, Type of Gestures, as well as, a significant interaction between Group and Time Point (all F > 3.8; p<0.5) Post hoc analysis, bonferroni corrected, revealed that patients underperformed in both pantomime and imitative gestures compared to healthy controls (p<0.0001) and relatives of schizophrenia patients, although this effect did not reach significance (p=0.26). Performance of pantomime gestures was poorer compared to imitative gestures. Interestingly, this pattern was also observed during the re-test time point (p<0.0001), though gesture performance for imitative gestures significantly declined compared to baseline in patients (p<0.05). In contrast, healthy controls performed better than both patients (p<0.0001) and relatives (p-0.09) and remained stable during the re-test. Likewise, relatives performed intermediate between patients and healthy controls at both time points with scores reaching significance only at re-test (p<0.001). At baseline, imitation was better than pantomime in relatives, but with re-test imitation scores declined while pantomime scores remained stable. Finally, at the 6-month follow-up patients still exhibited lower gesture performance compared to baseline (F=22.25; p<0.05).DiscussionGesture performance in schizophrenia patients remained significantly impaired across time-points, suggesting an extended effect on poor social functioning despite symptom change. In addition, schizophrenia patients and their relatives showed a significant impairment when performing imitative gestures during the re-test compared to baseline. These results call for interventions specifically targeting gesture and social cognition, which would greatly improve patients’ quality of life. Finally, our findings suggest a trait component to gesture behavior that might be linked to genetic liability to psychosis.