Syntax and Schizophrenia: A meta-analysis of comprehension and production
Bibliographic record
Abstract
Summary Background People with schizophrenia exhibit notable difficulties in the use of everyday language. This directly impacts one’s ability to complete education and secure employment. An impairment in the ability to understand and generate the correct grammatical structures (syntax) has been suggested as a key contributor; but studies have been underpowered, often with conflicting findings. It is also unclear if syntactic deficits are restricted to a subgroup of patients, or generalized across the broad spectrum of patients irrespective of symptom profiles, age, sex, and illness severity. Methods We conducted a systematic review and meta-analysis, registered on OSF, adhering to PRISMA guidelines, searching multiple databases up to May 1, 2024. We extracted effect sizes (Cohen’s d) and variance differences (log coefficient of variation ratio) across 6 domains: 2 in comprehension (understanding complex syntax, detection of syntactic errors) and 4 in production (global complexity, phrasal/clausal complexity, utterance length, and integrity) in patient-control comparisons. Study quality/bias was assessed using a modified Newcastle–Ottawa Scale. Bayesian meta-analysis was used to estimate domain-specific effects and variance differences. We tested for potential moderators with sufficient data (age, sex, study quality, language spoken) using conventional meta-regression to estimate the sources of heterogeneity between studies. Findings Overall, 45 studies (n=2960 unique participants, 64·4% English, 79 case-control contrasts, weighted mean age(sd)=32·3(5·6)) were included. Of the patient samples, only 29·2% were women. Bayesian meta-analysis revealed extreme evidence for all syntactic domains to be affected in schizophrenia with a large-sized effect (model-averaged d =0·65 to 1·01, with overall random effects d =0·86, 95% CrI [0·67-1·03]). Syntactic comprehension was the most affected domain. There was notable heterogeneity between studies in global complexity (moderated by the age), production integrity (moderated by study quality), and production length. Robust BMA revealed weak evidence for publication bias. Patients had a small-to-medium-sized excess of inter-individual variability than healthy controls in understanding complex syntax, and in producing long utterances and complex phrases (overall random effects lnCVR =0·21, 95% CrI [0·07-0·36]), hinting at the possible presence of subgroups with diverging syntactic performance. Interpretation There is robust evidence for the presence of grammatical impairment in comprehension and production in schizophrenia. This knowledge will improve the measurement of communication disturbances in schizophrenia and aid in developing distinct interventions focussed on syntax - a rule-based feature that is potentially amenable to cognitive, educational, and linguistic interventions. Research in Context Evidence before this study Prior studies have documented significant language deficits among individuals with psychosis across multiple levels. However, syntactic divergence—those affecting sentence structure and grammar—have not been consistently quantified or systematically reviewed. An initial review of the literature indicated that the specific nature and severity of syntactic divergence, as well as their impact on narrative speech production, symptom burden, and daily functioning, remain poorly defined. We conducted a comprehensive search of the literature up to May 1, 2024, using databases such as PubMed, PsycINFO, Scopus, Google Scholar, and Web of Science. Our search terms combined psychosis, schizophrenia, language production, comprehension, syntax, and grammar, and we identified a scarcity of meta-analytic studies focusing specifically on syntactic comprehension and production divergence in psychosis. Added value of this study This systematic review and meta-analysis is the first to quantitatively assess syntactic comprehension and production divergence in individuals with psychosis. This study provides estimated effect sizes associated with syntactic impairments as well as a quantification of the variance within patient groups for each domain of impairment. Besides a detailed examination of this under-researched domain, we also identify critical research gaps that need to be addressed to derive benefits for patients from knowledge generated in this domain. Implications of all the available evidence This study provides robust evidence of grammatical impairments in individuals with schizophrenia, particularly in syntactic comprehension and production. These findings can enhance early detection approaches via speech/text readouts and lead to the development of targeted cognitive, educational, and linguistic interventions. By highlighting the variability in linguistic deficits, the study offers valuable insights for future therapeutic trials. It also supports the creation of personalized formats of information and educational plans aimed at improving the effectiveness of any therapeutic intervention offered to patients with schizophrenia via verbal medium.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.021 | 0.040 |
| Meta-epidemiology (narrow) | 0.003 | 0.001 |
| Meta-epidemiology (broad) | 0.018 | 0.058 |
| Bibliometrics | 0.007 | 0.007 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.004 | 0.002 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.004 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".