Single-Item Anhedonia in Schizophrenia Marks Illness Severity but Does Not Independently Predict Outcomes: An Item-Level Reanalysis of the Clinical Antipsychotic Trials of Intervention Effectiveness Trial

Abstract Anhedonia has been proposed as a marker of poor functional outcomes in schizophrenia. A model that omits baseline functioning estimates a different quantity from one that includes it: the former assesses whether anhedonia identifies patients who function worse later, and the later assesses whether it adds prognostic information beyond baseline functioning. We estimated both associations using data from the Clinical Antipsychotic Trials of Intervention Effectiveness trial. We reanalysed the Clinical Antipsychotic Trials of Intervention Effectiveness Phase 1/1A Main Trial schizophrenia sample (759 rated and decisive model 672). We moved from the first quantity to the second quantity by removing the Heinrichs-Carpenter Quality of Life Scale anhedonia item from the outcome and then adding the baseline function. We repeated the test on the three Quality of Life Scale subscales without the item, screened all 21 items the same way, set anhedonia against 15 other baseline features, and tested independent outcomes, drug interactions and attrition. Baseline anhedonia predicted the 21-item Quality of Life Scale total at 12 months (standardised β=0.309, p=4.2×10−15). Removing the item from the outcome measure did not change the association (β=0.301). Adding the baseline function collapsed it (β=0.022, p=0.58). All 21 items behaved the same way: 21 were significant in models with no baseline term, 21 with the item also removed, and 0 once baseline function was adjusted. Anhedonia ranked 13th of 21 items, 13th of 16 features, and predicted no independent outcomes. Baseline single-item anhedonia in Clinical Antipsychotic Trials of Intervention Effectiveness marks concurrent clinical and functional burden. It carries no incremental prognostic value beyond the baseline function.