Well-designed population-based twin study with large sample and validated methods quantifying heritability of postpartum depression; rigorous design and clear effect size, but a genetic parameter estimate rather than a clinical trial outcome.
A well-designed cross-sectional mediation analysis in a large clinical sample identifies plausible pathways between childhood trauma and suicide outcomes, but cross-sectional design prevents causal inference and limits actionability.
Secondary analysis of an FDA-regulated trial showing meaningful reductions in suicidal ideation with both TMS protocols, with accelerated dosing achieving faster onset, but limited by exploratory design and lack of a non-TMS control.
Exploratory post-hoc analysis of pooled data from two small RCTs examining a personality trait association that failed multiplicity correction and showed no active-versus-sham difference, requiring independent replication before clinical utility.
Cross-sectional network analysis of brain structure and self-reported symptoms in adolescents, identifying associations without establishing causation, clinical outcomes, or intervention efficacy.
A rigorous observational study with propensity-score adjustment showing a clear short-term protective effect (one-year reattempt reduction) in a clinically relevant population, but limited by observational design, lack of long-term benefit, and age-specific effects that require confirmation.
A well-designed observational cohort study with adequate sample size and rigorous statistical method (linear mixed modelling) reporting clear differences in cognitive trajectories between TRD and non-TRD patients, but limited to 12 weeks and requiring confirmation in longer-term, interventional designs.
Large observational cohort study with propensity-score weighting comparing antipsychotics to antidepressants for a hard clinical outcome (psychiatric emergency visits/admissions) in adolescents with difficult-to-treat depression, showing increased risk with antipsychotics and consistent results across sensitivity analyses.