Abstract
OBJECTIVE: Non-adherence is an important factor in clinical trials, which has not been investigated in people at ultra-high risk (UHR) of developing a first episode of psychosis. METHODS: Exploratory analysis of data from NEURAPRO, a multicenter, placebo-controlled trial of long-chain omega-3 polyunsaturated fatty acids (omega-3 PUFAs) in 304 individuals at UHR. We examined correlates of non-adherence with study medication (omega-3 PUFAs or placebo), including patient, illness and treatment factors, plus transition to psychosis. Non-adherence was defined as <75% study medication intake over 6 months and, post hoc, by the number of returned pills. RESULTS: Of 285 randomized participants with baseline fatty acid data, 163 (57.2%) were non-adherent. In univariate analyses, non-adherence was associated with baseline omega-3 index, pre-baseline duration of untreated symptoms, smoking, cannabis use, lower baseline Social and Occupational Functioning Assessment Scale, Global Functioning: Social and Role Scale scores and transition to psychosis. Transition to psychosis risk was significantly lower in the adherent than non-adherent group (4.2%, 95% CI = 0.7-7.7% vs 17.3%, 95% CI = 10.4-24.2%), Kaplan-Meier Log-rank test, chi-square = 10.675, p = 0.001), independent of omega-3 PUFA treatment status. Similarly, Cox regression analysis, covarying for the aforementioned factors significantly associated with non-adherence, also revealed non-adherence as an independent predictor of transition to psychosis (B = 1.452, p = 0.005). Finally, non-adherence was also significantly associated with transition to psychosis, even when defining non-adherence by number of returned pills. CONCLUSION: Non-adherence predicted a higher risk of progressing to psychosis in UHR individuals. Further studies are needed to better understand factors contributing to non-adherence and how non-adherence is related to transition to psychosis.