Prediction of long-term outcome by duration of the psychosis prodrome: Mixed effects models reveal continuity across 7 years but variation across quartile splits

Abstract

While associations between duration of untreated psychosis (DUP) and outcome are robust and have been influential in early intervention, whether similar associations and potential utility exist for the preceding duration of the psychosis prodrome (DPP) remain to be clarified. This study investigated prospectively across seven years: (i) whether DPP is associated with outcome; (ii) whether any prediction by DPP varies across long-term follow-up after initiating treatment; and (iii) whether these relationships vary across gradations of DPP values. Sixty subjects were evaluated for DPP at first episode psychosis (FEP) and for psychopathology and quality of life at both FEP and 7-year follow-up; functionality and service engagement were assessed at follow-up. Data were analysed using mixed-effects models with DPP and DUP included by quartile split. Longer DPP, like DUP, predicted greater severity of negative symptoms and lower quality of life but not positive symptoms; these predictions by longer DPP were stable between FEP and 7-year follow-up and derived from the longest DPP values. For assessments made only at follow-up, longer DPP values did not predict either functionality or service engagement, while the longest DUP values predicted each of these outcomes. DPP and DUP appear to be in part, but not in whole, two sequential moieties of a common process that has been dichotomised at an imprecise threshold along its continuum. Interventions to remediate early features of DPP may particularly benefit the negative symptom-quality of life domain that is a particularly pernicious feature of psychotic illness yet least responsive to antipsychotic drugs.