RESUMEN
BACKGROUND: Neurocognitive deficits have been widely reported in clinical high-risk for psychosis (CHR) populations. Additionally, rates of cannabis use are high among CHR youth and are associated with greater symptom severity. Cannabis use has been sometimes shown to be associated with better neurocognition in more progressed psychosis cohorts, therefore in this study we aimed to determine whether a similar pattern was present in CHR. METHODS: CHR participants ages 12-30 from the North American Prodromal Longitudinal Study (NAPLS-3) (N = 698) were grouped according to: "minimal to no cannabis use" (n = 406), "occasional use" (n = 127), or "frequent use" (n = 165). At baseline, cannabis use groups were compared on neurocognitive tests, clinical, and functional measures. Follow-up analyses were used to model relationships between cannabis use frequency, neurocognition, premorbid, and social functioning. RESULTS: Occasional cannabis users performed significantly better than other use-groups on measures of IQ, with similar trend-level patterns observed across neurocognitive domains. Occasional cannabis users demonstrated better social, global, and premorbid functioning compared to the other use-groups and less severe symptoms compared to the frequent use group. Follow-up structural equation modeling/path analyses found significant positive associations between premorbid functioning, social functioning, and IQ, which in turn was associated with occasional cannabis use frequency. DISCUSSION: Better premorbid functioning positively predicts both better social functioning and higher IQ which in turn is associated with a moderate cannabis use pattern in CHR, similar to reports in first-episode and chronic psychosis samples. Better premorbid functioning likely represents a protective factor in the CHR population and predicts a better functional outcome.