Associate Professor & Associate Chair University of Toledo Toledo, Ohio
It is unclear whether self-report measures of schizotypy identify only individuals who are truly at higher risk for developing a schizophrenia-spectrum disorder, or if individuals with symptoms of non-schizophrenia-spectrum (NSS) disorders are being identified as having high levels of schizotypy and therefore incorrectly deemed at-risk. An earlier study carried out by the authors found that there is significant overlap among the response patterns of individuals with symptoms of schizophrenia (SCZ) and individuals with symptoms of NSS disorders on commonly used measures of schizotypy. The current study aimed to build upon that study by examining the latent profiles that emerge from the same sample using only the dimensional scores from the Multidimensional Schizotypy Scale - Brief (MSS-B), a measure that was developed as an improvement over older measures of schizotypy. We hypothesized that there would be at least three latent profiles showing clear separation between the response patterns of individuals with SCZ symptoms, those with NSS disorder symptoms, and those without any symptoms. Further, we hypothesized these profiles would be characterized by high, mixed, and low scores across dimensions of schizotypy, respectively. A sample of 676 participants (Mage = 42.4, SD = 15.7; 83% female; 82% White) was recruited via ResearchMatch. The sample included individuals endorsing significant symptoms of SCZ (N = 73), depression (DEP; N = 124), autism spectrum disorder (ASD; N = 111), or obsessive-compulsive disorder (OCD; N = 149), as well as individuals who were non-symptomatic (N = 219). Participants completed an online survey via Qualtrics that included measures assessing symptoms of the disorders listed above, four self-report measures of schizotypy, and demographic questions. A latent profile analysis (LPA) was conducted using Mplus. The best-fitting LPA model contained five profiles and had an entropy value of 0.935, indicating highly accurate classification. One profile characterized by high schizotypy was mostly comprised of individuals from the group endorsing SCZ symptoms (40%) and the group endorsing OCD symptoms (35%). Another profile characterized by low schizotypy was mostly comprised of individuals from the non-symptomatic group (47%). The other three profiles, characterized by mixed scores across all dimensions of schizotypy, were each comprised primarily of individuals from the groups endorsing symptoms of NSS disorders (i.e., DEP, ASD, and OCD; 63% - 73%). Thus, we obtained partial support for our hypotheses. However, the composition of the latent profiles included more overlap among the groups than expected. These results further confirm previous results, that individuals who are not at risk for developing a schizophrenia-spectrum disorder obtain similarly high scores as individuals who are exhibiting symptoms of SCZ. This suggests that the way schizotypy is currently measured results in false positives, a limitation that future researchers need to be aware of. The MSS-B could benefit from modifications that ensure only those who are truly at higher risk for schizophrenia-spectrum disorders obtain elevated scores.
