Select age group, sex, and anxiety type to adjust Bayesian priors and disable age-inappropriate interventions. 41 interventions including 17 supplements with proven effect sizes. Threshold-controlled Pareto finds the minimum effective set for any benefit target, evidence quality, and caution level.
Interventions sharing mechanisms are not additive. Eigendecomposition penalty applied automatically.
| #. Intervention | Sleep g | Marginal % | Cumul. % | Status |
|---|
✓ = Selected. ★ = In Pareto set (not selected). — = Below threshold or not in Pareto set. Marginal % = contribution to total achievable benefit. Cumul. % = cumulative total as interventions are added in rank order.
Directed Acyclic Graph: arrows = causal direction. Dashed = bidirectional feedback. Gold node = primary modifiable. PN = Probability of Necessity (Pearl counterfactual).
Pearl's counterfactual: do(X=x) forces a variable to a new value — "what if the story had unfolded differently?"
This analysis challenges its own conclusions. Genuine counter-arguments — not straw men.
E-value = minimum confounder strength to nullify result (>2.0 = causally credible). a: anxiety g; s: sleep g.
| Intervention | g (a/s) | 95% CI | E-Val | Stability |
|---|
| # | Pathway | Causal Chain | PN | Primary Targets |
|---|---|---|---|---|
| 1 | Cortisol Loop (HPA) | Anxiety→Cortisol↑→Fragmented Sleep→Fatigue→Withdrawal→Anxiety | 0.44 | Tai Chi, Yoga, PMR, Ashwagandha, Rhodiola, Exercise |
| 2 | Racing Mind ★ Strongest | Anxiety→Rumination→Long Sleep Onset→Low Sleep→Irritability | 0.71 | CBT-I, MBCT, ACT, PMR, Biofeedback, Lavender |
| 3 | Circadian Fracture | Anxiety→SNS Activation→Melatonin Delay→Phase Shift→Poor Sleep | 0.35 | BLT, Melatonin, Sleep Hygiene, Morning Walk, Magnesium |
| 4 | Social Contraction | Anxiety→Withdrawal→Loneliness→Amplified Anxiety→Worse Sleep | 0.35 | Social Engagement, Pet Therapy, Volunteering, Dance, Group Classes |
| 5 | Pharmacological Hazard | Sleep Problems→Benzo Rx→Architecture↓→Falls→More Anxiety | Harm pathway | CBT-I (prevents); Sleep Hygiene; Buspirone (safer alt.) |
PN = Probability of Necessity. HPA = Hypothalamic-Pituitary-Adrenal. SNS = Sympathetic Nervous System. ★ = strongest causal node. Rx = prescription.
All 17 supplements with proven anxiety effect sizes. Ranked by anxiety g. Evidence from peer-reviewed meta-analyses.
| Supplement | Anxiety g | Sleep g | Mechanism | Dose | Min Age | Evidence | Key Caution |
|---|---|---|---|---|---|---|---|
| CBD / Cannabidiol | 0.92* | 0.50 | Endocannabinoid system; 5-HT1A agonism | 300-600mg/day | 18 | MODERATE (N=316, wide CI) | Wide CI; legal varies; drug interactions |
| PMR (mind-body) | 1.11 | 1.74 | Sympathetic NS reduction; somatic relaxation | 20-30 min/day | 10 | HIGH — 31 RCTs, N=2,277 | None |
| Lavender / Silexan | 0.73 | 0.55 | VOCC inhibition; serotonin modulation | 80-160mg/day oral | 12 | HIGH — N=1,213; licensed medicine | None significant |
| Ashwagandha KSM-66 | 0.65 | 0.58 | HPA axis / cortisol reduction | 600mg/day | 18 | MODERATE — N=873 (15 RCTs) | Drug interactions; banned in Denmark/France |
| Kava | 0.63 | 0.35 | GABA-A modulation; kavalactones | 120-240mg kavalactones | 18 ⚠ | MODERATE — 13 RCTs | ⚠ Hepatotoxicity — liver monitoring essential |
| Passionflower | 0.50 | 0.40 | GABA-A partial agonist | 300-400mg/day | 12 | MODERATE — 2 RCTs vs oxazepam | Avoid in pregnancy |
| Saffron | 0.45 | 0.35 | Serotonergic; antioxidant; crocin | 28-30mg/day | 14 | MODERATE — 3/3 trials positive | Serotonin overlap w/ antidepressants |
| Ginkgo Biloba EGb-761 | 0.42 | 0.25 | GABA modulation; antioxidant; circulation | 240mg/day standardised | 50 | LOW-MOD — small trials; elderly | ⚠ Anticoagulant interaction |
| Omega-3 EPA+DHA | 0.38 | 0.28 | Anti-inflammatory; neuronal membrane | 2g+/day EPA emphasis | All | MODERATE — 19 RCTs, N=2,240 | Blood thinning at high dose |
| Lemon Balm | 0.38 | 0.35 | GABA transaminase inhibition (rosmarinic acid) | 300-1800mg/day | 6 | MODERATE — 2022 clinical recommendation | Potentiates sedatives |
| Rhodiola rosea | 0.35 | 0.25 | HPA axis; adaptogenic; AMPK | 300-600mg/day | 16 | LOW-MOD — 11 RCTs; stress-anxiety | Antidepressant interactions |
| Magnesium Glycinate | 0.30 | 0.45 | GABA-R modulation; cortisol regulation; NMDA | 200-400mg evening | All | MODERATE — 7 RCTs | Diarrhea at high doses (glycinate minimises) |
| Chamomile Extract | 0.30 | 0.30 | GABA-A partial agonist (apigenin) | 220-1100mg/day | All | LOW-MOD — positive signal | Warfarin interaction (bleeding risk) |
| Probiotics (multi-strain) | 0.30 | 0.22 | Gut-brain axis; vagus nerve; SCFA | Multi-strain 4-8wk course | All | LOW-MOD — 34 RCTs; SMD=0.34 | None significant at therapeutic doses |
| L-Theanine | 0.28 | 0.35 | Alpha-wave enhancement; NMDA modulation | 200mg (acute or daily) | All | LOW-MOD — 9 RCTs | None significant |
| B-Vitamin Complex | 0.28 | 0.18 | GABA synthesis (B6); methylation (B12/folate) | B6 50mg; B12 1000mcg | All | LOW-MOD — strongest in deficiency | High B6 long-term: peripheral neuropathy |
| Valerian Root | 0.25 | 0.40 | GABA-A; valerenic acid; sleep latency | 450-600mg evening | All | LOW-MOD — insufficient HAM-A; sleep better | Morning drowsiness; sedative potentiation |
| Zinc | 0.25 | 0.18 | NMDA receptor modulation; antioxidant | 25-50mg/day | All | LOW — primarily deficiency correction | Nausea; copper depletion at high doses |
g = Hedges' g vs placebo. *CBD CI very wide (-1.80 to -0.04). VOCC = Voltage-Operated Calcium Channel. GABA = Gamma-Aminobutyric Acid. HPA = Hypothalamic-Pituitary-Adrenal. NMDA = N-Methyl-D-Aspartate. SCFA = Short-Chain Fatty Acids. AMPK = AMP-activated protein kinase. All supplements require physician consultation before use.