Thesis title: I Disturbi Della Forma del Pensiero: Dimensioni Psicopatologiche e Valore Prognostico in Acuto
Introduction. The dimensional structure of formal thought disorder (FTD) remains poorly understood. The Thought and Language Disorder Scale (TALD) is a recent rating scale for FTD, but its factor structure has been investigated only in samples including healthy controls, which limits its reliability in clinical populations. Moreover, the discriminant value of FTD between diagnostic groups in acute conditions, as well as its prognostic value for relevant clinical outcomes, has been insufficiently investigated.
Aims. To investigate the structure of FTD in an acute inpatient sample; to compare FTD dimensions across diagnostic groups; and to assess their prognostic value for key outcomes in acute settings.
Methods. A total of 226 acute psychiatric inpatients diagnosed with psychotic, bipolar, or depressive disorder were assessed with the Thought and Language Disorder Scale (TALD) at admission. Exploratory factor analysis (EFA) with minimum residuals extraction and promax rotation was conducted on TALD items to examine the structure of FTD. Confirmatory factor analysis (CFA) was performed on aggregated reproduced correlation matrices from previous TALD factor-analytic studies. A multivariate analysis of variance (MANOVA) with post hoc comparisons was used to test the effect of diagnosis on TALD dimensions derived from the EFA. Analysis of covariance (ANCOVA) was added to control for verbosity between psychotic and bipolar subjects. Regression and negative binomial analyses were conducted to examine the prognostic value of TALD dimensions for use of restraint, compulsory admission, length of stay, and risk of rehospitalization.
Results. The EFA identified four independent dimensions: Thought Disorganization, Schizophasia, Poverty, and Subjective Language Impairment. The CFA indicated an adequate model fit (χ² = 212.0; SRMR = 0.06; RMSEA = 0.02; GFI = 0.96; AGFI = 0.95; CFI = 0.99; NNFI = 0.99). Diagnostic category had a significant effect on TALD dimensions (V = 0.371, F(2, 8) = 12.6, p < 0.001, partial η² = 0.185). Compared to the bipolar subgroup, the psychotic group showed higher levels of Poverty (F = 6.3; p < 0.01) and Subjective Language Impairment (F = 9.7; p < 0.01), with no significant differences in Thought Disorganization and Schizophasia. Compared to the depressive group, both psychotic and bipolar groups showed higher Thought Disorganization (F = 34.1; p < 0.01) and Schizophasia (F = 9.3; p < 0.01), and lower Subjective Language Impairment (F = 9.7; p < 0.01). TALD dimensions globally predicted use of restraint (χ² = 20.8; p < 0.001; R² = 0.16), compulsory admission (χ² = 16.0; p < 0.01; R² = 0.11), length of stay (F = 3.6; p < 0.001; R² = 0.06), and risk of rehospitalization (χ² = 10.52; p = 0.03; pseudo-R² = 0.05). When other clinical and demographic predictors were included, Thought Disorganization remained an independent predictor of use of restraint (B = 0.07; p = 0.05; OR = 1.07) and risk of rehospitalization (β = 0.06; p = 0.01; Exp(B) = 1.05).
Discussion and conclusions. This is the first study to examine the TALD factor structure exclusively in an acute psychiatric inpatient sample. The dimensional structure of the TALD is largely consistent with previous findings; however, the distinction between Thought Disorganization and Schizophasia represents a meaningful contribution, at least to some extent, although it has so far been only weakly supported by empirical evidence.
FTD demonstrated limited discriminant value between diagnostic groups in acute settings. Schizophasia may differentiate psychosis from mania when controlling for verbosity, but this finding requires further investigation. Thought Disorganization significantly enhances the prognostic value of models predicting use of restraint and risk of rehospitalization and may be a useful addition to predictive models of clinical outcomes in acute psychiatry. The main limitation concerns the sample size relative to the EFA, warranting cautious interpretation and further data collection to strengthen the robustness of these findings.