Salivary cortisol in early psychosis: New findings and meta-analysis

Schizophrenia is a multifactorial disorder and environmental risk factors for it might contribute to hypothalamo-pituitary-adrenal axis (HPA) dysregulation. While increased cortisol levels have been reported in schizophrenia, as well as in early psychosis (compared to healthy controls), a crucial unresolved issue is whether elevated cortisol levels could be related to the distress of an emerging illness, rather than being specific to psychosis. Here, we report new findings from the first French cohort of young help-seekers (ICAAR) including ultra-high risk subjects (UHR), first-episode of psychosis (FEP) and non at-risk help seekers controls (HSC), followed by a meta-analysis of all available reports on salivary basal cortisol levels in early psychosis (UHR and FEP).

In the ICAAR study, 169 individuals (15-30 years old) had their basal cortisol levels sampled and they were categorized (at baseline) as either UHR, FEP, or HSC using the criteria of the Comprehensive Assessment of At-Risk Mental States (CAARMS). The three groups were compared at baseline, and the UHR and HSC individuals were also included in a one-year longitudinal follow-up. UHRs who converted to psychosis at the follow up (UHR-P) were compared to non-converters (UHR-NP). We also performed a meta-analysis from case-control studies with basal salivary measures of cortisol, drawing from a systematic bibliographic search using the keywords 'cortisol', 'glucocorticoid', 'HPA' with 'UHR', 'CHR', 'at-risk mental state', 'schizotypal ', 'prodromal schizophrenia', 'first-episode psychosis', 'first episode schizophrenia', 'newly diagnosed schizophrenia', 'recent onset schizophrenia' [in Medline, Web of Knowledge (WOS), EBSCO], followed by a systematic screening of the resulting articles.

Basal cortisol levels were not significantly different between UHR, FEP, and HSC controls in the ICAAR cohort. Interestingly, initial cortisol levels were correlated with positive symptoms at the one year follow-up in the ICAAR cohort. The meta-analysis revealed a significant elevation of the salivary basal cortisol levels in UHR individuals compared to controls (8 studies--1060 individuals), but not between FEP and controls (6 studies--441 individuals). Indirect comparison of salivary basal cortisol levels between UHR and FEP did not yield significant differences. Finally, no differences were detected between the baseline cortisol of UHR-P and UHR-NP (4 studies--301 individuals).

The meta-analysis (including new data) indicates that basal cortisol levels were increased in UHR compared to controls, but FEP levels were not different from UHR or controls. Many confounding factors could decrease the effect size in FEP especially medication intake. Taken together with our new results (which made use of help-seeker controls, and not merely healthy controls), the findings indicate that basal cortisol levels may not be a reliable biomarker for early psychosis. Further studies are needed to clarify the precise role of the HPA axis in psychotic conversion.