Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2026 Aug 5;20(8):e70244. doi: 10.1111/eip.70244

Psychotic Vulnerability in Adolescents With Obsessive‐Compulsive Disorder: First Longitudinal Evidence From a Prospective Pilot Study

Davide Fausto Borrelli 1,2,✉, Luca Gambolò 1,2, Anna Di Donna 1,2, Carlo Marchesi 1,3, Matteo Tonna 1,3
PMCID: PMC13440141  PMID: 42554468

ABSTRACT

Aim

Childhood‐onset Obsessive‐Compulsive Disorder (OCD) has been associated with an increased vulnerability to psychosis, yet no prospective studies have examined this risk in paediatric populations.

Methods

In this pilot prospective study, 59 adolescents with a primary diagnosis of OCD were assessed at baseline for psychotic vulnerability, measured through Cognitive‐Perceptual basic symptoms (COPER) and the high‐risk criterion Cognitive Disturbances (COGDIS), as well as for OCD severity and psychosocial functioning. The number of subjects who developed a first episode of psychosis (FEP) during the 18‐month follow‐up period was evaluated. Logistic regression and moderation analyses were performed.

Results

During the follow‐up period, 18.6% of participants presented with a FEP. These patients were older and displayed poorer psychosocial functioning compared with those who did not develop a FEP. Logistic regression identified higher COPER symptoms and lower psychosocial functioning as the only significant predictors of FEP. Although moderation effects were not statistically significant, exploratory simple slopes suggested that both COPER and COGDIS symptoms were more strongly associated with psychosis onset in adolescents with lower psychosocial functioning.

Conclusions

This study highlights a notable rate of progression to psychosis among adolescents with OCD over an 18‐month period. Despite methodological limitations, our findings offer preliminary evidence that COPER symptoms and impaired psychosocial functioning are key risk factors in adolescents with OCD for progression to psychosis.

Keywords: adolescents, at‐risk mental state, basic symptoms, early onset, OCD, pilot study, psychosis

1. Introduction

In the last decades, the comorbidity between obsessive‐compulsive disorder (OCD) and schizophrenia has attracted growing research attention. This interest is supported by epidemiological evidence indicating that the prevalence of OCD and obsessive‐compulsive symptoms (OCS) is, respectively, 7.9% and 21.4%, among individuals at risk for psychosis; 10.5% and 34.0% among those with first‐episode psychosis (Martinho et al. 2023); and 12%–14% and 30%, among individuals with schizophrenia (Swets et al. 2014). Moreover, in around half individuals with schizophrenia and co‐occurring OCD, OCS emerge more than 3 years prior to the first psychotic symptoms (Faragian et al. 2012).

Despite these epidemiological findings, little is known about psychotic vulnerability in individuals with OCD/OCS, particularly in childhood and adolescence. Considering non‐clinical samples, a data‐driven machine learning approach was applied to a large sample of adolescents to investigate predictors of psychotic experiences arising from subclinical OCS (Sawai et al. 2025). The authors found that close‐knit family bonds and limited social connections outside the family predicted the emergence of psychotic experiences among adolescents with OCS. Considering clinical samples, two parallel and independent studies, one conducted by our group (Averna et al. 2018; Borrelli, Ottoni, et al. 2024; Borrelli, Tonna, and Dar 2024), have shown that approximately 40% of youth with OCD meet criteria for a high‐risk state for psychosis and, in those subjects, psychotic vulnerability is associated with poor psychosocial functioning. The two studies offered a comprehensive assessment of clinical high‐risk states in paediatric OCD population through different, yet complementary approaches: the ‘ultra‐high risk’ (UHR) and ‘basic symptoms’ models (Fusar‐Poli et al. 2013; Schultze‐Lutter et al. 2015). In addition, a recent study (Borrelli et al. 2025) adopted a first‐level screening tool (the Thought Problems scale of the Child Behaviour Checklist) (Achenbach 1991) to examine the prevalence and clinical correlates of psychotic vulnerability in youth with OCD across two large cohorts of Scandinavian (n = 215) and North American (n = 125) patients. The authors found that one‐third of the participants in both samples scored high enough to warrant an in‐depth second‐level diagnostic investigation of psychosis risk and extended follow‐up.

Taken together, these findings emphasise the need for an exhausting evaluation of psychotic vulnerability in adolescents with OCD. Moreover, little is known about the pathophysiological mechanisms underlying the development of psychosis in OCD adolescent population. Given the substantial attention devoted to individuals at ultra‐high risk for psychosis, this gap represents a significant limitation in the literature, with not negligible clinical implications in the search of early predictors of psychotic vulnerability.

Therefore, the present study aims to fill the gap by examining, over an 18‐month follow‐up, the onset of psychosis in a cohort of OCD adolescents assessed at baseline using COPER and COGDIS symptom criteria. These criteria arise from basic symptoms, which represent subclinical disturbances in mental processes, such as thinking, speech, (body) perception, motor action, drive, affect, and stress tolerance, and represent the earliest subjectively experienced symptoms of the underlying neurobiological impairment of psychosis (Hartmann et al. 2024). Among basic symptoms, those encompassed within the COPER and COGDIS criteria have been identified as significant predictors of subsequent psychosis onset (Schultze‐Lutter and Theodoridou 2017). According to the current developmental models of psychotic disorders (Fusar‐Poli et al. 2013; Schultze‐Lutter et al. 2010), the trajectory leading to psychotic onset is as follows: (1) predictive basic symptoms (Cognitive–Perceptual basic symptoms—COPER); (2) high‐risk criterion Cognitive Disturbances (COGDIS); (3) ultra‐high risk symptoms (including attenuated psychotic symptoms and brief limited intermitted psychosis); (4) full‐blown psychotic symptoms. Although their use in routine clinical practice is limited, the assessment of COPER and COGDIS basic symptoms allows for the identification of early subjective experiences that reflect initial neurobiological alterations preceding psychosis onset (Schultze‐Lutter and Theodoridou 2017).

The primary objective of the present pilot prospective study was to evaluate the transition rate to psychosis in a cohort of OCD adolescents. The secondary aim was to identify possible reliable predictors of psychotic onset. To this end, all patients were assessed at baseline using the COPER and COGDIS criteria. We expected that a significant proportion of patients meeting COPER and COGDIS criteria would develop psychosis at follow‐up. In addition, we hypothesised that specific clinical variables (i.e., earlier age at onset and poorer psychosocial functioning) might play an important role in the development of psychosis. These hypotheses are supported by the available evidence from previous cross‐sectional studies, that psychotic vulnerability in youth with OCD is associated with an earlier onset of OCS (Borrelli, Ottoni, et al. 2024; Borrelli, Tonna, and Dar 2024) as well as with markedly impaired psychosocial functioning (Borrelli et al. 2023, 2025). In a similar vein, the literature on UHR population has found that level of functioning is strongly correlated with the risk of transition to psychosis (Hartmann et al. 2024).

2. Material and Methods

2.1. Participants

Fifty‐nine adolescents with OCD were recruited from January 2021 to December 2022. Participants were outpatients referring to the Childhood and Adolescence Service of the Department of Mental Health of Parma, Italy, with a diagnosis of Obsessive‐Compulsive Disorder. Inclusion criteria were: (1) a principal diagnosis of OCD according to DSM‐5 criteria (APA 2013) (American Psychiatric Association 2013); (2) being between 12 and 18 years old; (3) providing an informed consent signed by parents/legal guardians. Exclusion criteria were: (1) a current mental disorder related to a general medical condition (e.g., central nervous system tumours and infections and endocrine disorders); (2) OCD arising after the introduction of pharmacological treatment with antipsychotics; (3) a current affective disorder (e.g., bipolar disorder) comorbid to the OCD diagnosis; (4) a current psychotic disorder (e.g., schizophrenia spectrum disorders) comorbid to the OCD diagnosis; (5) a current condition related to drug or alcohol abuse or dependence; (6) a cognitive disorder (defined as a Mini‐Mental State Examination score lower than 25; although MMSE has limited psychometric validation in adolescent populations, it was used as a broad screening tool for significant cognitive impairment that would preclude study participation); (7) a clinical condition too severe to allow for testing. All participants were provided with a complete and exhaustive description of the study before their assent/consent to participate. All participants and their parents/legal guardians provided written informed assent and consent.

The study was approved by the Local Institutional Ethics Committee of ‘Area Vasta Emilia Nord’ (Emilia‐Romagna region, approved on October 06, 2020, n.35, 564).

2.2. Procedure

Assessment procedure at the baseline has been described in detail elsewhere (Borrelli et al. 2023). At the 18‐month follow‐up, we contacted each patient's treating specialist to determine whether a diagnosis of first‐episode psychosis (FEP) had been made. The primary outcome measure was binary FEP status. Repeated assessments of baseline symptomatology (COPER, COGDIS) were not performed at follow‐up, as these early prodromal markers typically precede psychotic onset and are known to fluctuate over time (Schultze‐Lutter et al. 2020). To meet criteria for FEP, the individual must have experienced full threshold psychotic symptoms (e.g., hallucinations, delusions, or disorganised behaviour) for at least 7 days (Gannon et al. 2024).

2.3. Measures

2.3.1. Children's Yale‐Brown Obsessive Compulsive Scale (CY‐BOCS)

The severity of OCD during the last week was assessed using the Children's Yale‐Brown Obsessive Compulsive Scale (CY‐BOCS) (Scahill et al. 1997). It consists of 10 items rated on a scale ranging from 0 (no symptoms) to 4 (extreme symptoms).

2.3.2. Children Global Assessment Scale (CGAS)

Psychosocial functioning was evaluated using the clinician rated Children Global Assessment Scale (CGAS), which is a one‐item scale ranging from 0 to 100, with lower scores indicating poorer functioning (Shaffer 1983).

2.3.3. Schizophrenia Proneness Instrument‐Child and Youth Version (SPI‐CY)

Psychotic vulnerability was assessed using the Schizophrenia Proneness Instrument‐Child and Youth version (SPI‐CY) (Fux et al. 2013; Schultze‐Lutter, Marshall, and Koch 2012; Schultze‐Lutter, Ruhrmann, et al. 2012). COPER criteria are met when at least 1 of the COPER symptoms is present with at least weekly occurrence during the last 3 months, and the first occurrence of symptoms was more than 12 months prior to the evaluation. COGDIS criteria are met when at least two of the COGDIS symptoms are present with at least weekly occurrence in the last 3 months; the symptoms must have not been present during the premorbid phase of the illness and are not related to drug use. We examined the number of items each patient reported and classified each participant according to the COPER and COGDIS criteria.

2.4. Statistics

R Studio (version 4.4.0) was used for all statistical analyses.

First, we compared participants who developed psychosis (Psy+) with those who did not (Psy−), examining differences in sex, age, age at OCD onset, OCD severity, and psychosocial functioning.

Second, we investigated the proportion of participants who, having met COPER, COGDIS, or both criteria at baseline, transitioned to psychosis during the follow‐up period. We examined the predictive value of baseline COPER and COGDIS criteria by computing odds ratios (OR), Fisher's exact tests, absolute risk differences (RD), and the number needed to treat (NNT).

Third, we performed a weighted logistic regression to identify predictors of psychotic onset. Assuming a possible low baseline rate of transition to psychosis, we decided to apply double weighting to positive cases (individuals who transitioned to psychosis) to address class imbalance and reduce bias in parameter estimation. Because of the small sample size and in order to limit the risk of overfitting, only variables deemed clinically and theoretically relevant were included in the model. Following previous studies (Borrelli et al. 2023; Borrelli, Ottoni, et al. 2024; Borrelli, Tonna, and Dar 2024), we decided to include as independent variables the following ones: age of onset of OCD, OCD symptom severity, COPER and COGDIS criteria and global functioning level. To avoid multicollinearity, the COPER and COGDIS criteria were assessed in two separate models. For each model, calibration was checked with the Hosmer‐Lemeshow test, while discrimination was assessed by ROC curve and Area Under the Curve (AUC). The optimal cut‐off according to Youden's index was also calculated, with subsequent evaluation of sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV).

Finally, we examined whether baseline psychosocial functioning moderated the relationship between COPER/COGDIS criteria and psychotic onset. Model fit was assessed using likelihood tests, and odds ratios (OR). To explore potential interactions, we conducted a simple slopes analysis, which examines the effect of X on Y at specific values of the moderator M (e.g., at the mean, ±1 SD, or selected percentiles).

3. Results

3.1. Socio‐Demographic and Clinical Features in Psy+ and Psy− Groups

In the overall sample (n = 59; Mage = 15.44 ± 2.48, range = 8–18; females = 35), 11 participants (18.6%) presented with a FEP during the follow‐up period. Compared to non‐converters (Psy−), those who developed psychosis (Psy+) were significantly older at baseline (Psy + = 16.7 ± 2.0, Psy− = 15.1 ± 2.5; t = −2.24, p = 0.038, Cohen's d = 0.70) and exhibited lower psychosocial functioning (CGAS: Psy + = 52.9 ± 12.2; Psy− = 69.5 ± 12.8; t = 4.03, p = 0.001, Cohen's d = 1.35). No significant differences were found in terms of sex, age of OCD onset and OCD severity. All comparisons between groups are reported in Table 1.

TABLE 1.

Socio‐demographic and clinical features in the Psy‐ and Psy+ groups.

Psy− Psy+
N 50 9
Gender (female) 30 5

χ2 = 0.48,

p = 0.48
Mean (SD) Mean (SD) p t Cohen's d
Age at baseline (years) 15.1 (± 2.5) 16.7 (± 2.0) 0.038* −2.24 0.65
Age at OCD onset (years) 12.1 (± 2.8) 10.9 (± 2.6) 0.212 1.35 −0.43
CY‐BOCS total 18.4 (± 9.1) 22.5 (± 6.0) 0.081 −1.84 0.48
CY‐BOCS obs. 9.1 (± 5.2) 11.2 (± 3.0) 0.097 −1.72 0.41
CY‐BOCS comp. 9.3 (± 4.5) 11.4 ± 3.6 0.117 −1.65 0.47
CGAS 69.5 (± 12.8) 52.9 ± 12.2 0.001** 4.01 −1.31
Open in a new tab

Note: CGAS = Children Global Assessment Scale; CY‐BOCS = Children's Yale‐Brown Obsessive Compulsive Scale; Psy− = participants who did not develop psychosis; Psy+ = participants who developed psychosis.

**p ≤ 0.01; *p ≤ 0.05.

3.2. Transition Rate According to COPER and COGDIS Criteria

Twenty‐six participants met COPER criteria at baseline, while 33 did not. Among those positive at baseline, nine (34.6%) presented with a FEP during the follow‐up period, compared with two (6.1%) among those negative (OR = 8.21, CI = 0.59–42.41; Fisher's exact test, p = 0.007; RD = 0.29, NNT = 3.5).

Twenty‐one participants met COGDIS criteria at baseline, while 38 did not. Among those positive at baseline, nine (42.9%) presented with a FEP during the follow‐up period, compared with two (5.3%) among those negative (OR = 13.5, CI = 2.55–71.41; Fisher's exact test, p < 0.001; RD = 0.38, NNT = 2.6).

Considering both COPER and COGDIS criteria, 27 participants met COPER or COGDIS criteria at baseline, while 32 did not. Among those positive at baseline, nine (33.3%) presented with a FEP during the follow‐up period, compared with two (6.25%) among those negative (OR = 7.5, CI = 1.46–38.6; Fisher's exact test, p = 0.02; RD = 0.27, NNT = 3.7).

3.3. Logistic Regressions Analyses

In the logistic regression model with COPER symptoms (COPER model), COPER symptoms (OR = 1.80, 95% CI = 1.12–3.45, p = 0.027) and baseline functioning (OR = 0.93, 95% CI = 0.86–9.96, p = 0.04) were the only significant predictors of psychotic onset, suggesting that higher scores on COPER symptoms and lower lever of functioning were associated with an increased risk of developing psychosis (see Table 2). The Hosmer‐Lemeshow test indicated a good calibration (X 2 = 5.64, p = 0.13). The AUC of the model was 0.864, indicating high discriminative ability. The optimal cut‐off calculated by Youden index showed a sensitivity of 0.8, specificity of 0.91, PPV of 0.67 and NPV of 0.95. The Receiver Operating Characteristic (ROC) curve is shown in Figure 1.

TABLE 2.

Logistic regression—COPER model.

Independent variable OR 95% CI p
Age of onset 1.00 0.76, 1.31 0.97
OCD severity 0.94 0.84, 1.04 0.23
COPER symptoms 1.80 1.19, 3.07 0.01**
Baseline functioning 0.93 0.86, 9.96 0.04*
Open in a new tab

Abbreviations: CI = confidence interval, OR = odds ratio. **p ≤ 0.01; *p ≤ 0.05.

FIGURE 1.

FIGURE 1

Open in a new tab

ROC curve—COPER model. AUC = area under the ROC curve; FPR = false positive rate; ROC = receiver operating characteristic curve; TPR = true positive rate.

In the logistic regression model with COGDIS symptoms (CODGIS model), baseline functioning (OR = 0.91, 95% CI = 0.83–0.98, p = 0.026) was the only significant predictors of psychotic onset, suggesting that lower levels of functioning were associated with an increased risk of developing psychosis (see Table 3). The Hosmer‐Lemeshow test indicated a good calibration (X 2 = 5.17, p = 0.17). The AUC of the model was 0.864, indicating high discriminative ability. The optimal cut‐off calculated by Youden index showed a sensitivity of 0.8, specificity of 0.89, PPV of 0.61 and NPV of 0.95. The Receiver Operating Characteristic (ROC) curve is shown in Figure 2.

TABLE 3.

Logistic regression—COGDIS model.

Independent variable OR 95% CI p
Age of onset 0.98 0.69, 1.38 > 0.9
OCD severity 0.95 0.82, 1.08 0.4
COGDIS symptoms 1.44 0.86, 2.62 0.2
Baseline functioning 0.91 0.83, 0.98 0.026*
Open in a new tab

Abbreviations: CI = confidence interval, OR = odds ratio. **p <= 0.01; *p <= 0.05.

FIGURE 2.

FIGURE 2

Open in a new tab

ROC curve—COGDIS model. AUC = area under the ROC curve; FPR = false positive rate; ROC = receiver operating characteristic curve; TPR = true positive rate.

To address the potential concern regarding class imbalance adjustment, we conducted a sensitivity analysis comparing weighted and unweighted logistic regression models. In the unweighted COPER model, COPER symptoms remained significantly associated with FEP onset (OR = 1.85, 95% CI = 1.12–3.45, p = 0.027) with negligible change in effect size compared to the weighted model (OR = 1.80, p = 0.0120). Similarly, baseline functioning remained significant (OR = 0.92, p = 0.0743) though the p‐value increased slightly, consistent with the loss of information resulting from unweighting. The AUC remained identical in both approaches (0.864), indicating no loss of discriminative ability. Results were similar when examining the COGDIS model: baseline functioning remained the sole significant predictor (OR = 0.91, 95% CI = 0.83–0.98, p = 0.0259), replicating the weighted finding (p = 0.0102) and confirming that COGDIS criteria do not independently predict psychotic onset regardless of weighting approach. The AUC was again identical (0.851). These results demonstrate that our main findings are robust to the choice of weighting strategy and not artefacts of our handling of class imbalance (see Table S1).

3.4. Moderation Analyses

In the COPER model, the main effect of COPER symptoms was significant (LR χ 2(1) = 7.40, p = 0.006), while CGAS showed a trend (LR χ 2(1) = 2.39, p = 0.122). Notably, the COPER × CGAS interaction was not statistically significant (LR χ 2(1) = 2.53, p = 0.112), indicating that moderation effects were not established. In the COGDIS model, the main effect of COGDIS symptoms was not significant (LR χ 2(1) = 3.69, p = 0.055), while the effect of CGAS was significant (LR χ 2(1) = 4.84, p = 0.028). The COGDIS × CGAS interaction was not significant (LR χ 2(1) = 2.25, p = 0.133).

However, given the pilot nature of this study, we conducted exploratory simple slopes analyses to characterise the conditional associations at different functional levels, which may generate hypotheses for future larger studies.

At low CGAS (−1 SD), higher COPER scores were associated with greater odds of transition (OR = 5.25, 95% CI = 1.69–26.3, p = 0.016), whereas the effect was markedly attenuated or absent at mean CGAS (OR = 1.34, 95% CI = 0.27–5.62, p = 0.681) or high CGAS (+1 SD; OR = 0.34, 95% CI = 0.01–5.86, p = 0.494). Similarly to the COPER scores, simple slopes analyses showed that at low CGAS (−1 SD), higher COGDIS scores were associated with increased odds of transition (OR = 3.31, 95% CI = 1.22–13.1, p = 0.040), with non‐significant effects at mean CGAS (OR = 0.97, 95% CI = 0.19–3.79, p = 0.965) and high CGAS (+1 SD; OR = 0.28, 95% CI = 0.01–4.31, p = 0.417). These findings should be interpreted cautiously as exploratory patterns pending replication in larger samples rather than confirmatory evidence of moderation.

Figures 3 and 4 show the interaction plot of the simple slopes, depicting how the association between COPER and COGDIS symptoms and transition to psychosis varies across levels of psychosocial functioning (CGAS).

FIGURE 3.

FIGURE 3

Open in a new tab

Interaction plot of the simple slopes—COPER model.

FIGURE 4.

FIGURE 4

Open in a new tab

Interaction plot of the simple slopes—COGDIS model.

4. Discussion

The present pilot prospective study was aimed to evaluate the progression rate to psychosis and its potential predictors in a cohort of Italian adolescents with OCD. At baseline, all participants were assessed using COPER and COGDIS criteria, which identify basic symptoms as prodromal markers of psychotic onset. Age at OCD onset, OCD symptom severity, and baseline functioning were also examined as potential predictors.

As to progression rate to psychosis, we found that 11 out of 59 patients (18.6%) presented with a FEP during the 18‐month follow‐up. This finding is consistent with the evidence that youth with OCD are at increased risk for psychotic disorders (Cheng et al. 2019; Meier et al. 2014) and in line with the finding that 8% of adolescents and adults with a primary diagnosis of OCD develop schizophrenia during an 11‐year follow‐up period (Chen et al. 2023). Moreover, this result is coherent with the epidemiological evidence that individuals first diagnosed with OCD have a three‐fold higher risk of receiving a later diagnosis of schizophrenia and a five‐fold higher risk of receiving a later diagnosis of schizoaffective disorder compared to individuals without OCD (Cederlöf et al. 2015). Notably, the 18‐month psychotic transition rate observed in our OCD adolescent sample (18.6%) approximates the 3‐year UHR transition rate (25%) reported in a recent meta‐analytic study (Salazar de Pablo et al. 2021). This correspondence suggests that a diagnosis of OCD in adolescence constitutes a significant risk factor for psychotic onset. This finding underscores the importance of routine psychotic risk screening in paediatric OCD populations.

As to potential predictors of psychosis transition, significant group differences (Psy + vs. Psy‐) emerged for baseline age, and psychosocial functioning, whereas no significant differences were observed for age of OCD onset or OCD symptom severity.

Participants who developed psychosis (Psy+) were significantly older at baseline than Psy‐ subjects. This finding may suggest that Psy+ participants have a longer history of illness, and it cannot be ruled out that they may also have experienced a longer duration of untreated illness than Psy− patients. Considering that in UHR individuals, the duration of symptoms prior to referral to specialised services has been shown to predict the transition to psychosis (Hartmann et al. 2024), it may be hypothesised that in young patients with OCD the duration of their OCS could have predictive validity for developing psychosis, perhaps masking underlying pre‐psychotic symptoms.

Participants who developed psychosis (Psy+) had significantly lower levels of functioning at the baseline. This finding was further corroborated by the significant effect of baseline functioning in both the regression models. Therefore, in our sample of adolescents with OCD low functional level at baseline appears a strong predictor of the development of psychosis. These results are in line with the evidence that high‐risk states are inherently associated with functional impairment (Fusar‐Poli et al. 2013) and that baseline functioning represents a strong predictor for transition in people at risk for psychosis(Andreou et al. 2023).

In our study both COPER and COGDIS criteria demonstrated robust effect sizes (OR, risk difference, NNT) with respect to transition rates, thus corroborating the evidence of their potential clinical utility in screening for psychotic vulnerability. However, when examined within multivariate models, only COPER criteria retained predictive significance, whereas COGDIS criteria did not independently predict transition to psychosis. In fact, after performing regression analyses, baseline functioning and COPER symptoms were the only significant predictors of developing psychosis at a18‐month follow‐up. In fact, the other variables considered (i.e., age of onset, OCD severity, COGDIS symptoms) did not reach statistical significance. The divergent findings regarding COPER and COGDIS symptoms call for a more nuanced interpretation. According to Frauke Schultze‐Lutter et al. (Schultze‐Lutter, Ruhrmann, et al. 2012), both criteria occur with adequate frequency prior to the onset of a first psychotic episode, while remaining rare in the general population, thereby supporting their utility in the early identification of individuals at risk for psychosis. Moreover, even though COPER and COGDIS criteria share a subset of symptoms (e.g., interference from emotionally neutral thoughts; thought blocking; crowding, pressure of thoughts; flight of ideas), COPER criteria are characterised by more prominent perceptual symptoms whereas COGDIS criteria include specific cognitive symptoms (e.g., disturbances in symbolic understanding; receptive language disturbances). It may be hypothesised that basic perceptual symptoms, more represented in the COPER criteria, could capture peculiar sensory features of OCD, underlying a distinct form of psychotic vulnerability. In this vein, it has been suggested that specific abnormal psycho‐sensorial experiences in OCD, also referred to as ‘sensory phenomena’ (Ferrão et al. 2012; Chen et al. 2023), may represent a pathophysiological bridge between OCD and psychosis (Poletti et al. 2023). Moreover, perceptual symptoms, as captured by both COPER and COGDIS criteria in various ways reflect a loss of control over one's own stream of consciousness, and thus an altered sense of agency, a domain that is known to be impaired in both OCD (Borrelli, Ottoni, et al. 2024; Borrelli, Tonna, and Dar 2024) and psychosis (Rossetti et al. 2022, 2024). However, perceptual basic symptoms, which are specific to COPER criteria (e.g., difficulty distinguishing between imagination and perception, between memories and fantasies), appear towards the end of adolescence (i.e., around the age of 18), while cognitive disturbances appear later, in the early twenties (Schultze‐Lutter et al. 2020). Therefore, in our adolescent sample, the higher predictive value of perceptive basic symptoms may be due to their earlier appearance, underlying altered developmental trajectories preceding the onset of psychosis. Altogether, the greater predictive value of perceptive basic symptoms relative to cognitive symptoms may point to an early neuro‐sensory vulnerability in youth with OCD, potentially intercepting different developmental pathways. In this connection, childhood and adolescence OCD may cover a bunch of heterogeneous vulnerabilities, which interacting with environmental and epigenetic factors, developmentally unfold a variety of clinical outcomes (Tonna et al. 2025).

We did not find a significant effect of OCD severity as potential predictor of later psychosis. This result is consistent with other studies showing that in OCD patients with pre‐psychotic symptoms and in psychotic patients with OCD symptoms, the severity of OCD and psychotic symptoms are largely independent from each other (Borrelli et al. 2025; Tonna et al. 2016, 2024).

Similarly, age of OCD onset did not emerge as a significant predictor of psychotic onset. This finding stands in contrast to our previous study conducted in a sample of adolescent and adult patients, where early‐onset OCD was associated with a higher risk of developing pre‐psychotic symptoms (Borrelli, Ottoni, et al. 2024; Borrelli, Tonna, and Dar 2024). Nonetheless, the limited duration of follow‐up and the narrow age range of our sample may preclude the identification of distinct trajectories in individuals with an earlier onset of the disorder.

Finally, moderation analyses failed to find a moderating effect of psychosocial functioning on COPER/COGDIS criteria in predicting psychotic onset. However, the predictive value of COPER and, to a lesser extent, COGDIS was influenced by lower levels of psychosocial functioning. In other words, both COPER and COGDIS symptoms were associated with transition, especially among adolescents with lower CGAS scores. Although this result should be considered with caution, it is consistent with previous evidence that psychotic vulnerability should be hypothesised in cases of adolescents with OCD and low psychosocial functioning (Averna et al. 2018; Borrelli et al. 2023, 2025).

The present study is subject to several limitations. First, the relatively small sample size precluded the inclusion of additional variables (e.g., OCD symptoms dimensions, OCD dysfunctional cognitive beliefs, depressive symptoms) that might play a decisive role in the transition to psychosis. Second, the limited follow‐up duration (18 months) may not have been sufficient to adequately characterise the temporal trajectory of each case. As argued by Salazar et al. (Salazar de Pablo et al. 2021), cohorts with a short follow‐up period (i.e., 1–2 years) limit the ability to determine longer‐term clinical outcomes and to examine temporal changes in psychopathological trajectories. Third, baseline assessments were not repeated. Nevertheless, it is worth noting that repeating the evaluation of COPER/COGDIS symptoms might have yielded limited additional clinical information, given that these symptoms usually precede psychosis onset and present a fluctuating course over time. Fourth, the assessment of psychosis risk was limited to the COPER and COGDIS criteria, while the UHR criteria were not applied. The combined use of these two approaches has been shown to improve the accuracy of identifying young people most vulnerable to the onset of psychotic disorders, at least over a medium‐term follow‐up period (Youn et al. 2020).

Fifth, the exclusion of participants with comorbid affective disorders may limit the generalisability of these findings, as mood‐related symptoms and their management could influence the trajectory towards psychosis in real‐world clinical populations.

With the caveat of these limitations, the findings of the present study may have important clinical implications. To the best of our knowledge, no previous study has prospectively examined the development of psychosis in adolescents with OCD. It is important to frame the present work as a pilot prospective study, providing the first longitudinal evidence that adolescents with OCD may face psychotic transition rates comparable to those observed in ultra‐high risk cohorts. Our results suggest that a substantial proportion (18.6%) of adolescents with OCD may be at risk of developing psychosis, with lower baseline functioning and more pronounced COPER symptoms as putative risk factors of this transition. These findings may have a relevance for the prevention of different psychopathological trajectories in young people with OCD and, in line with recent studies (Borrelli et al. 2025; Preti et al. 2022), support the importance of careful screening for the risk of transition to psychosis in young people with OCD and low levels of functioning. In order to validate these preliminary results, larger studies with repeated assessments are needed.

In conclusion, this pilot study yields the first prospective evidence that a significant proportion of adolescents with OCD may progress to psychosis in a relatively short follow‐up period.

Furthermore, our findings support the systematic assessment of psychotic vulnerability in young people seeking help for OCD, to better identify those at risk for psychosis and guide early intervention strategies.

Funding

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Table S1: Sensitivity Analysis: Logistic Regression Models with Double Weighting versus Standard Weighting.

EIP-20-0-s001.docx (13KB, docx)

Acknowledgements

We wish to acknowledge all young people and their families for study participation.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

References

  1. Achenbach, T. M. 1991. Integrative Guide for the 1991 CBCL/4‐18, YSR, and TRF Profiles. Department of Psychiatry, University of Vermont. [Google Scholar]
  2. American Psychiatric Association . 2013. Diagnostic and Statistical Manual of Mental Disorders. 5th ed. American Psychiatric Publishing. [Google Scholar]
  3. Andreou, C. , Eickhoff S., Heide M., de Bock R., Obleser J., and Borgwardt S.. 2023. “Predictors of Transition in Patients With Clinical High Risk for Psychosis: An Umbrella Review.” Translational Psychiatry 13, no. 1: 286. 10.1038/s41398-023-02586-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Averna, R. , Pontillo M., Demaria F., et al. 2018. “Prevalence and Clinical Significance of Symptoms at Ultra High Risk for Psychosis in Children and Adolescents With Obsessive–Compulsive Disorder: Is There an Association With Global, Role, and Social Functioning?” Brain Sciences 8, no. 10: 181. 10.3390/brainsci8100181. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Borrelli, D. F. , Cervin M., Ottoni R., Marchesi C., and Tonna M.. 2023. “Psychotic Vulnerability and Its Associations With Clinical Characteristics in Adolescents With Obsessive‐Compulsive Disorder.” Research on Child and Adolescent Psychopathology 51, no. 10: 1535–1548. 10.1007/s10802-023-01089-2. [DOI] [PubMed] [Google Scholar]
  6. Borrelli, D. F. , Højgaard D. R. M. A., Storch E. A., et al. 2025. “Psychotic Vulnerability in Pediatric Obsessive‐Compulsive Disorder: Associations With Clinical Features and Treatment Outcome in Two Clinical Samples.” European Child & Adolescent Psychiatry 34: 3449–3462. 10.1007/s00787-025-02757-y. [DOI] [PubMed] [Google Scholar]
  7. Borrelli, D. F. , Ottoni R., Provettini A., et al. 2024. “A Clinical Investigation of Psychotic Vulnerability in Early‐Onset Obsessive–Compulsive Disorder Through Cognitive–Perceptive Basic Symptoms.” European Archives of Psychiatry and Clinical Neuroscience 274, no. 1: 195–205. 10.1007/s00406-022-01543-0. [DOI] [PubMed] [Google Scholar]
  8. Borrelli, D. F. , Tonna M., and Dar R.. 2024. “An Investigation of the Experience of Control Through the Sense of Agency in People With Obsessive–Compulsive Disorder: A Review and Meta‐Analysis.” CNS Spectrums 29, no. 4: 224–232. 10.1017/S1092852924000117. [DOI] [PubMed] [Google Scholar]
  9. Cederlöf, M. , Lichtenstein P., Larsson H., et al. 2015. “Obsessive‐Compulsive Disorder, Psychosis, and Bipolarity: A Longitudinal Cohort and Multigenerational Family Study.” Schizophrenia Bulletin 41, no. 5: 1076–1083. 10.1093/schbul/sbu169. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Chen, M.‐H. , Tsai S.‐J., Liang C.‐S., et al. 2023. “Diagnostic Progression to Schizophrenia in 35,255 Patients With Obsessive–Compulsive Disorder: A Longitudinal Follow‐Up Study.” European Archives of Psychiatry and Clinical Neuroscience 273, no. 3: 541–551. 10.1007/s00406-021-01361-w. [DOI] [PubMed] [Google Scholar]
  11. Cheng, Y.‐F. , Chen V. C.‐H., Yang Y.‐H., Chen K.‐J., Lee Y.‐C., and Lu M.‐L.. 2019. “Risk of Schizophrenia Among People With Obsessive‐Compulsive Disorder: A Nationwide Population‐Based Cohort Study.” Schizophrenia Research 209: 58–63. 10.1016/j.schres.2019.05.024. [DOI] [PubMed] [Google Scholar]
  12. Faragian, S. , Fuchs C., Pashinian A., Weizman R., Weizman A., and Poyurovsky M.. 2012. “Age‐Of‐Onset of Schizophrenic and Obsessive‐Compulsive Symptoms in Patients With Schizo‐Obsessive Disorder.” Psychiatry Research 197, no. 1–2: 19–22. 10.1016/j.psychres.2012.02.024. [DOI] [PubMed] [Google Scholar]
  13. Ferrão, Y. A. , Shavitt R. G., Prado H., et al. 2012. “Sensory Phenomena Associated With Repetitive Behaviors in Obsessive‐Compulsive Disorder: An Exploratory Study of 1001 Patients.” Psychiatry Research 197, no. 3: 253–258. 10.1016/j.psychres.2011.09.017. [DOI] [PubMed] [Google Scholar]
  14. Fusar‐Poli, P. , Borgwardt S., Bechdolf A., et al. 2013. “The Psychosis High‐Risk State.” JAMA Psychiatry 70, no. 1: 107. 10.1001/jamapsychiatry.2013.269. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Fux, L. , Walger P., Schimmelmann B. G., and Schultze‐Lutter F.. 2013. “The Schizophrenia Proneness Instrument, Child and Youth Version (SPI‐CY): Practicability and Discriminative Validity.” Schizophrenia Research 146, no. 1–3: 69–78. 10.1016/j.schres.2013.02.014. [DOI] [PubMed] [Google Scholar]
  16. Gannon, L. , Mullen E., McGorry P., and O'Donoghue B.. 2024. “Prevalence and Predictors of Admission at the Time of Presentation in First Episode Psychosis.” Social Psychiatry and Psychiatric Epidemiology 59, no. 7: 1143–1151. 10.1007/s00127-023-02552-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Hartmann, S. , Dwyer D., Cavve B., et al. 2024. “Development and Temporal Validation of a Clinical Prediction Model of Transition to Psychosis in Individuals at Ultra‐High Risk in the UHR 1000+ Cohort.” World Psychiatry 23, no. 3: 400–410. 10.1002/wps.21240. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Martinho, F. P. , Magalhães D., Felício R., Ferreira T. F., and Jorge S.. 2023. “Obsessive‐Compulsive Symptoms in First Episode Psychosis and Risk States: Systematic Review With Meta‐Analysis.” Schizophrenia Research 255: 41–51. 10.1016/j.schres.2023.03.024. [DOI] [PubMed] [Google Scholar]
  19. Meier, S. M. , Petersen L., Pedersen M. G., et al. 2014. “Obsessive‐Compulsive Disorder as a Risk Factor for Schizophrenia.” JAMA Psychiatry 71, no. 11: 1215. 10.1001/jamapsychiatry.2014.1011. [DOI] [PubMed] [Google Scholar]
  20. Poletti, M. , Gebhardt E., Pelizza L., Preti A., and Raballo A.. 2023. “Neurodevelopmental Antecedents and Sensory Phenomena in Obsessive Compulsive Disorder: A Systematic Review Supporting a Phenomenological‐Developmental Model.” Psychopathology 56, no. 4: 295–305. 10.1159/000526708. [DOI] [PubMed] [Google Scholar]
  21. Preti, A. , Meneghelli A., Poletti M., and Raballo A.. 2022. “Through the Prism of Comorbidity: A Strategic Rethinking of Early Intervention in Obsessive‐Compulsive Disorder.” Schizophrenia Research 239: 128–133. 10.1016/j.schres.2021.11.038. [DOI] [PubMed] [Google Scholar]
  22. Rossetti, I. , Mariano M., Maravita A., Paulesu E., and Zapparoli L.. 2024. “Sense of Agency in Schizophrenia: A Reconciliation of Conflicting Findings Through a Theory‐Driven Literature Review.” Neuroscience & Biobehavioral Reviews 163: 105781. 10.1016/j.neubiorev.2024.105781. [DOI] [PubMed] [Google Scholar]
  23. Rossetti, I. , Repossi M., Florio V., et al. 2022. “Sense of Body Ownership and Body Agency in Schizophrenia.” Scientific Reports 12, no. 1: 19369. 10.1038/s41598-022-23988-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Salazar de Pablo, G. , Radua J., Pereira J., et al. 2021. “Probability of Transition to Psychosis in Individuals at Clinical High Risk.” JAMA Psychiatry 78, no. 9: 970–978. 10.1001/jamapsychiatry.2021.0830. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Sawai, Y. , Tanaka R., Minami R., et al. 2025. “Predictors of Psychotic Experiences Among Adolescents With Obsessive–Compulsive Symptoms: A Data‐Driven Machine Learning Approach.” Psychiatry and Clinical Neurosciences Reports 4, no. 2: e70e103. 10.1002/pcn5.70103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Scahill, L. , Riddle M. A., Mcswiggin‐Hardin M., et al. 1997. “Children's Yale‐Brown Obsessive Compulsive Scale: Reliability and Validity.” Journal of the American Academy of Child & Adolescent Psychiatry 36, no. 6: 844–852. 10.1097/00004583-199706000-00023. [DOI] [PubMed] [Google Scholar]
  27. Schultze‐Lutter, F. , Marshall M., and Koch E.. 2012. Schizophrenia Proneness Instrument, Child and Youth Version (SPI‐CY). Giovanni Fioriti Editore SRl. [DOI] [PubMed] [Google Scholar]
  28. Schultze‐Lutter, F. , Michel C., Schmidt S. J., et al. 2015. “EPA Guidance on the Early Detection of Clinical High Risk States of Psychoses.” European Psychiatry 30, no. 3: 405–416. 10.1016/j.eurpsy.2015.01.010. [DOI] [PubMed] [Google Scholar]
  29. Schultze‐Lutter, F. , Ruhrmann S., Berning J., Maier W., and Klosterkotter J.. 2010. “Basic Symptoms and Ultrahigh Risk Criteria: Symptom Development in the Initial Prodromal State.” Schizophrenia Bulletin 36, no. 1: 182–191. 10.1093/schbul/sbn072. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Schultze‐Lutter, F. , Ruhrmann S., Fusar‐Poli P., Bechdolf A. G., Schimmelmann B., and Klosterkotter J.. 2012. “Basic Symptoms and the Prediction of First‐Episode Psychosis.” Current Pharmaceutical Design 18, no. 4: 351–357. 10.2174/138161212799316064. [DOI] [PubMed] [Google Scholar]
  31. Schultze‐Lutter, F. , Ruhrmann S., Michel C., Kindler J., Schimmelmann B. G., and Schmidt S. J.. 2020. “Age Effects on Basic Symptoms in the Community: A Route to Gain New Insight Into the Neurodevelopment of Psychosis?” European Archives of Psychiatry and Clinical Neuroscience 270, no. 3: 311–324. 10.1007/s00406-018-0949-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Schultze‐Lutter, F. , and Theodoridou A.. 2017. “The Concept of Basic Symptoms: Its Scientific and Clinical Relevance.” World Psychiatry 16, no. 1: 104–105. 10.1002/wps.20404. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Shaffer, D. 1983. “A Children's Global Assessment Scale (CGAS).” Archives of General Psychiatry 40, no. 11: 1228. 10.1001/archpsyc.1983.01790100074010. [DOI] [PubMed] [Google Scholar]
  34. Swets, M. , Dekker J., van Emmerik‐van Oortmerssen K., et al. 2014. “The Obsessive Compulsive Spectrum in Schizophrenia, a Meta‐Analysis and Meta‐Regression Exploring Prevalence Rates.” Schizophrenia Research 152, no. 2–3: 458–468. 10.1016/j.schres.2013.10.033. [DOI] [PubMed] [Google Scholar]
  35. Tonna, M. , Borrelli D. F., Aguglia E., et al. 2024. “The Relationship Between Obsessive‐Compulsive Symptoms and Real‐Life Functioning in Schizophrenia: New Insights From the Multicenter Study of the Italian Network for Research on Psychoses.” European Psychiatry: The Journal of the Association of European Psychiatrists 67, no. 1: e37. 10.1192/j.eurpsy.2024.1747. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Tonna, M. , Borrelli D. F., Marchesi C., Gerra M. C., and Dallabona C.. 2025. “Childhood Obsessive‐Compulsive Disorder, Epigenetics, and Heterochrony: An Evolutionary and Developmental Approach.” Development and Psychopathology: 1–15. Advance online publication. 10.1017/S0954579425000124. [DOI] [PubMed] [Google Scholar]
  37. Tonna, M. , Ottoni R., Paglia F., Ossola P., De Panfilis C., and Marchesi C.. 2016. “Obsessive–Compulsive Symptoms Interact With Disorganization in Influencing Social Functioning in Schizophrenia.” Schizophrenia Research 171, no. 1–3: 35–41. 10.1016/j.schres.2016.01.002. [DOI] [PubMed] [Google Scholar]
  38. Youn, S. , Phillips L. J., Amminger G. P., et al. 2020. “Basic Symptoms in Young People at Ultra‐High Risk of Psychosis: Association With Clinical Characteristics and Outcomes.” Schizophrenia Research 216: 255–261. 10.1016/j.schres.2019.11.047. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Table S1: Sensitivity Analysis: Logistic Regression Models with Double Weighting versus Standard Weighting.

EIP-20-0-s001.docx (13KB, docx)

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Articles from Early Intervention in Psychiatry are provided here courtesy of Wiley

RESOURCES