Research has consistently shown that language abilities represent a core dimension of psychosis; however, to date, very little is known about syntactic comprehension performance in the early stages of psychosis. This study aims to compare the linguistic abilities involved in syntactic comprehension in a large group of First Episode Psychosis (FEP) patients and healthy controls (HCs).

A multiple choice test of comprehension of syntax was administered to 218 FEP patients (166 non-affective FEP patients [FEP-NA] and 52 affective FEP patients [FEP-A]) and 106 HCs. All participants were asked to match a sentence they listen with one out of four vignettes on a pc screen. Only one vignette represents the stimulus target, while the others are grammatical or non-grammatical (visual) distractors. Both grammatical and non-grammatical errors and performance in different syntactic constructions were considered.

FEP committed greater number of errors in the majority of TCGB language domains compared to HCs. Moreover, FEP-NA patients committed significantly more non-grammatical (z = −3.2, p = 0.007), locative (z = −4.7, p < 0.001), passive-negative (z = −3.2, p = 0.02), and relative (z = −4.6, p < 0.001) errors compared to HCs as well as more passive-affirmative errors compared to both HCs (z = −4.3, p < 0.001) and FEP-A (z = 3.1, p = 0.04). Finally, we also found that both FEP-NA and FEP-A committed more grammatical (FEP-NA: z = −9.2, p < 0.001 and FEP-A: z = −4.4, p < 0.001), total (FEP-NA: z = −8.2, p < 0.001 and FEP-A: z = 3.9, p = 0.002), and active-negative (FEP-NA: z = −5.8, p < 0.001 and FEP-A: z = −3.5, p = 0.01) errors compared to HCs.

This study shows that the access to syntactic structures is already impaired in FEP patients, especially in those with FEP-NA, ultimately suggesting that language impairments represent a core and inner feature of psychosis even at early stages.

The GET UP multi-element psychosocial intervention proved to be superior to treatment as usual in improving outcomes in patients with first-episode psychosis (FEP). However, to guide treatment decisions, information on which patients may benefit more from the intervention is warranted.

To identify patients' characteristics associated with (a) a better treatment response regardless of treatment type (non-specific predictors), and (b) a better response to the specific treatment provided (moderators).

Some demographic and clinical variables were selected a priori as potential predictors/moderators of outcomes at 9 months. Outcomes were analysed in mixed-effects random regression models. (Trial registration: ClinicalTrials.gov, NCT01436331.)

Analyses were performed on 444 patients. Education, duration of untreated psychosis, premorbid adjustment and insight predicted outcomes regardless of treatment. Only age at first contact with the services proved to be a moderator of treatment outcome (patients aged ≥35 years had greater improvement in psychopathology), thus suggesting that the intervention is beneficial to a broad array of patients with FEP.

Except for patients aged over 35 years, no specific subgroups benefit more from the multi-element psychosocial intervention, suggesting that this intervention should be recommended to all those with FEP seeking treatment in mental health services.