Depression BCA — Pearl SCM | Minimum Effective Set

Modelled ceiling — not a personal forecast

The headline figure on this page is an upper bound under best-case assumptions; a conservative floor (the lower confidence bound, or the same model run at a higher cross-correlation) is materially lower. Any individual’s result is unknown and may fall anywhere in — or below — this range. This is a population model, not a prediction about you. Discuss with your physician before acting.

What this number assumes

The population figure = trial effect × an assumed adherence rate (shown as 50%); it is a hypothetical, not an observed outcome.
Effect sizes (SMD/MD) describe trial populations, not you.
Stacked interventions are combined assuming independence/additivity.
If-not-for counterfactual outputs are model quantities, valid only if no unmeasured confounder is present.
The percentages are modelled maxima, not expected personal results.

Global Burden

280M

People affected (WHO 2024)

Risk Factors

Modifiable + non-modifiable

Min. Effective Set

—

Interventions (optimizer)

Min. Set HR

—

ρ̄-adjusted combined

All-Risk HR (ρ̄-adj)

—

Min. Set HR

—

Net HR

—

Efficiency Score

—

Max E-Value

5.86

Top Risk Factors by HR

Minimum Effective Set — Effect per Intervention

Efficiency Principle: The optimizer finds the fewest mechanistically-independent interventions achieving ≥90% of the maximum achievable ρ̄-adjusted HR reduction. Each causal pathway (psychotherapy, exercise, pharmacology, sleep, social, nutritional) contributes at most one champion intervention.

Minimum Independent Effective Intervention Set

This optimizer applies a greedy Pareto algorithm to find the smallest set of mechanistically independent interventions that achieves a specified percentage of maximum possible effect size. Mechanistic independence is defined by distinct causal pathways: two interventions sharing the same primary MOA pathway cannot both be in the minimum set.

⚡ Optimizer Controls

Greedy Pareto · ρ̄-corrected

Target Efficiency (% of max HR reduction): 90%

Max interventions: 6

Evidence tier → causal graph: A\u2013D

Exclude pharmacological:

Pareto Frontier — Set Size vs. Effect

N	Interventions	ρ̄-adj HR	Efficiency	Marginal Gain

Pathway Independence Map

⚠️ Antithesis: Why "More Is Not Better"

Beyond 3–4 interventions, marginal gains diminish due to: (1) shared downstream mechanisms (all pathways converge on BDNF, HPA normalization, and neuroinflammation reduction); (2) adherence fatigue — real-world compliance drops 35% per additional concurrent intervention; (3) interaction effects — CBT + exercise together are less than additive because both require sustained motivation and behavioral change capacity. The optimizer's saturation-corrected HR accounts for 15% diminishing returns per intervention beyond the 3rd.

Causal Risk Factors — All 18 Factors

Toggle factors to include in combined analysis. HRs are backdoor-adjusted. E-values = minimum confounding to nullify association.

🔴 High Impact (HR ≥ 2.0)

🟡 Moderate (HR 1.4–1.99)

🟢 Lower (HR 1.0–1.39)

Active Selection Summary

All 16 Interventions — Toggle & Dose

⭐ = included in current minimum effective set. Stars update live with optimizer settings.

🧠 Psychotherapeutic

🏃 Lifestyle & Physical

💊 Pharmacological & Nutritional

Stack Summary

Population Attributable Fraction (PAF)

PAF by Risk Factor (ρ̄-adjusted, sorted)

Intervention Population Impact

What-If & If-Not-For Counterfactual Analysis

🔮 If Childhood Trauma Eliminated

do(Trauma=0). PAF = 22.4%. Eliminating childhood maltreatment population-wide reduces depression incidence by ~22.4 cases per 100. Policy horizon: 30–50 years.

↓ 22.4% incident depression (95% CI: 18.1–26.8%)

🔮 If Universal Optimal Exercise

do(Exercise = 3×/week aerobic+resistance). From umbrella review (27 meta-analyses, n>190 RCTs), SMD=−0.67 → HR=0.62. Hypothetical projection — at an assumed 50% adherence (adjustable): ~19% fewer cases (change adherence and this changes).

↓ 19.0% depression burden (NNT = 2.78)

🔮 Minimum Effective Set — Real-World

do(CBT + Exercise + CBT-I + Social). ρ̄-adjusted combined HR = 0.24. Real-world adherence correction (×0.55): effective HR ≈ 0.38, representing ~62% modelled maximum reduction (ceiling, full adherence) in depression probability for adherent individuals.

↓ 47–62% depression (adherent individuals)

🚫 If-Not-For: Social Isolation

PN = [P(Y|X=1)−P(Y|X=0)] / P(Y|X=1). HR=2.30, prev=22%. P(Y|iso=1)≈0.35, P(Y|iso=0)≈0.15. PN = (0.35−0.15)/0.35 = 0.57. Among depressed isolated individuals, 57% would NOT have developed depression "but for" the isolation — the strongest modifiable PN.

PN = 0.57 · Strongest modifiable if-not-for

🚫 If-Not-For: Sleep Disorder

HR=2.10, prev=18%. PN ≈ (HR−1)/HR = 1.10/2.10 = 0.52. Among depressed insomniacs, 52% of depression cases are necessity-attributable to the sleep disorder. CBT-I treats both simultaneously — uniquely high-leverage.

PN = 0.52 · Bidirectional amplifier

🚫 If-Not-For: Neuroinflammation

Mediates ~35% of all upstream risk factor causal paths (trauma, poor diet, sedentary, SES). PN = 0.46 as a causal mediator. Anti-inflammatory interventions (exercise, Mediterranean diet, omega-3 in appropriate populations) synergistically target this central node.

PN = 0.46 · Central mediator node

E-Value & Sensitivity Analysis

E-Value Table — Risk Factors (sorted by E-value)

Risk Factor	HR	95% CI	E-Value	E-Val (CI)	Robustness	Source

E-Value Table — Interventions

Intervention	HR	E-Value	Grade	MOA Pathway	Robustness

E-Value formula: E = HR + √(HR·(HR−1)) for HR>1; apply to 1/HR for protective interventions. An E-value ≥3.0 means unmeasured confounders need ≥3.0-fold association with both exposure and outcome to nullify the finding.