{
  "study_id": "NCT06041152",
  "agent": "framework",
  "run": 1,
  "p_primary_hypothesis_supported": 0.12,
  "effect_direction": "favours intervention",
  "effect_size_90pct_interval": {"low": -6, "high": 12, "metric": "Percent difference in SynVesT-1 volume of distribution, psilocybin vs placebo (positive = higher density after psilocybin)"},
  "moderator_prediction": "aMCI participants may show a slightly larger relative rise (more lost synaptic density to recover, lower baseline), but with 6 vs 4 participants split across arms no moderator difference will be detectable; effectively none.",
  "claim_ids_used": ["PM-0028", "PM-0401", "PM-0062", "PM-0061", "PM-0372", "PM-0092", "PM-0086"],
  "rationale": "The framework treats psychedelics as relaxing belief precision (PM-0028, PM-0401) and holds that belief is written in tissue that changes slowly, not never (PM-0061, PM-0062). One saturated session can book heavily (PM-0092), so a small directional rise in synaptic density is plausible. But SV2A density is a level-4 biomarker, the outcome class the framework expects to move least (PM-0086, PM-0372), and the rodent spinogenesis signal is modest and may not carry to human PET a week later. With N=10 split across four arms (about 5 vs 5), even a real 5% effect is far below detectable given test-retest variability. As a pilot, the report may also be only descriptive. So significant support is unlikely."
}