Baseline/Base-Rate Transition-to-Psychosis Figures for CHR Cohorts

What the evidence says

Across clinical high-risk (CHR/ARMS/UHR) cohorts, the reported rate of transition to a full psychotic disorder rises with length of follow-up. At roughly one year, reported rates cluster in the 15-30% range: a meta-analysis by Fusar-Poli et al. (2012), cited in Barnes (2019), puts the 12-month transition rate at 15-30%, and Perkins (2014) reports "about 20%-25% developing psychosis within a year" citing prior work establishing CHR clinical criteria. By two to three years, reported rates move higher, into the 30-36% range: Perkins (2014) reports 30-35% at two years and, separately, "~30% risk of psychosis in the next 2 years" for persons meeting COPS criteria; Barnes (2019), again citing Fusar-Poli et al. (2012), reports "over 36% after 3 years." No source in this set establishes a causal mechanism for these figures — they are observational baseline rates drawn from meta-analyses and cohort follow-up, not effects of an intervention or policy lever, so no mechanism is stated.

Not every cohort matches this trajectory. Morrison et al. (2012), cited in Barnes (2019), reports an 8% transition rate at 1-2 years in a sample of 288 ARMS subjects — well below the 15-30% one-year range from the Fusar-Poli meta-analysis. A separate and more recent report, Barajas (2026), found a 3.3% transition rate at 6 months in UHR individuals receiving care within an early intervention program, based on a single transition event in a small sample.

For prodromal presentations more broadly, Sahoo (2026) gives a wider range: "10%-50% of individuals" presenting with prodromal symptoms convert to schizophrenia specifically, rather than to psychotic disorder broadly. The source frames this spread as evidence that prodromal diagnosis "lacks specificity" in cross-sectional or prospective evaluations.

Barnes (2019) also flags, without quantifying, that the transition rate within ARMS cohorts "appears to be declining" over time, citing Yung et al. (2007) and Hartmann et al. (2016). No number accompanies this trend in the source, so it can be noted only as a described but unquantified pattern, not a measured magnitude.

No source in this set reports transition rates beyond three years of follow-up, and none reports a rate for CHR/ARMS populations outside clinical or early-intervention program settings.

Where the evidence splits

The 8% transition rate at 1-2 years (Morrison et al. 2012, via Barnes 2019) sits well below the 15-30% one-year rate and the >36% three-year rate drawn from the Fusar-Poli et al. (2012) meta-analysis, also via Barnes (2019). Barnes's own text frames the 8% figure as one of the "lower rates" reported in the literature, but the source does not state what in Morrison's study design, sample, or era produced the lower figure — the difference is unexplained by the sources on this page.

The 3.3% transition rate at 6 months (Barajas 2026) is also markedly lower than the 15-30% one-year rate from the Fusar-Poli meta-analysis. Here the claim itself names one difference: the Barajas cohort was "receiving care within an early intervention program," a treatment setting not specified for the meta-analytic or Perkins cohorts. The sources do not say whether that setting explains the gap, and the Barajas figure rests on a single transition event in a small sample.

The 10-50% range for prodromal-to-schizophrenia conversion (Sahoo 2026) differs from the ARMS/CHR ranges above in what it measures: conversion to schizophrenia specifically, versus the broader "psychotic disorder" outcome used in the Fusar-Poli, Morrison, Perkins, and Barajas figures. The sources do not state whether this outcome difference alone accounts for the wider range.

The declining-rate claim (Yung et al. 2007; Hartmann et al. 2016, via Barnes 2019) is offered in the source as a possible explanation for cross-study heterogeneity in transition rates generally, but it is not quantified and not tied to any specific pair of figures above, so it cannot be used here to reconcile any one split.

Thresholds and dose-response

No threshold established in the current sources.

Open empirical questions

• Whether the described decline in ARMS transition rates over time (Yung 2007, Hartmann 2016) is quantifiable, and what accounts for it.
• What in Morrison et al. (2012)'s cohort or methodology produced an 8% transition rate versus the 15-36% range reported elsewhere.
• Whether the early intervention program setting in Barajas (2026) accounts for its lower 6-month rate, or whether longer follow-up in that cohort would converge with other reported rates.
• No source here reports transition rates beyond three years of follow-up.