Wellness Podcast Evidence Base — Modern Lane

Companion to 2026-09-22-kushi-corpus-map. The Kushi corpus is the traditional / philosophical root. This is the modern-evidence lane — a separate track. The two are never blended: TCM has no histological evidence for Qi or meridians, and that caveat always stands.

Why this exists

Young asked Eve to study and learn from wellness podcasts, not list them. Method, feed ladder and verification standards live in the health-knowledge-sources skill. This note is the durable finding.

Corpus (mined 2026-10-04)

~18 MB of show notes. Descriptions are multi-paragraph, often with chapter lists.

Update 2026-10-07: all six feeds re-polled (chunked fetch, 3,195 <item> blocks parsed). Genuinely new titles: On Purpose 05 + 07 Oct (celebrity — skipped), Happiness Lab 05 Oct (Thoreau/Weiner — skipped), WCDHT 06 Oct (breast-cancer BEST OF replay — skipped), Mel Robbins 05 Oct (motivation — skipped), Life Kit 06 Oct (US insurance open enrollment — out of lane). Evidence-lane pickups: Rangan #684 BITESIZE (Dr Roger Seheult, sunlight, 1 Oct) and #685 (Nir Eyal, distraction/attention, 6 Oct). #684 is the corpus's first dedicated light episode — the "morning light is under-weighted" reading above is weakening at the margin (one title; no frequency recount without sample contexts, per protocol). Full episode + evidence detail: 2026-10-07-eve-wellbeing-weekly-scan.

Update 2026-10-11: all six feeds re-polled (3,203 <item> blocks; Happiness Lab playlist URL re-resolved via iTunes after a stale 404). Five genuinely new titles; three assessed: Rangan #686 walking BITESIZE (Conley), On Purpose "5 Signs Someone Is Draining Your Energy" (host solo; the biological-aging hook traces to a real primary source — the PNAS Feb 2026 "difficult ties" paper, Lee & Perry 10.1073/pnas.2515331123, epigenetic clocks, family hasslers ~1.5%/yr faster aging ≈ 9 months per hassler), Mel Robbins male hormones (Stanford urologist Eisenberg — sleep/stress/body-fat→testosterone frame, no primary check this pass). Big new traced evidence: the daily-steps dose-response curve cross-validated across three syntheses (Lancet Public Health Oct 2025 — 7,000 vs 2,000 steps → 47% lower all-cause mortality, HR 0.53; JACC 2023 — benefits begin ~2,517 steps/day; Steps for Health — plateau 6–8k for 60+). Carryovers closed/resolved: Duke-NUS immune-ageing paper now carries PMID 42764338; Eyal's 10%/90% trigger split formally closed as rhetorical — no primary time-tracking study exists in any source trail. Perimenopause-sleep trials: NCT06497894 + new Dutch 2x2 NCT06306404 both protocol-stage. Full detail: 2026-10-11-eve-wellbeing-weekly-scan.

Update 2026-10-12: all six feeds re-polled (3,203 items — identical counts to last week, iTunes lookup re-verified every feed URL). No new episodes. Big evidence pass instead: six new papers — the loneliness/isolation lane dramatic: (a) Nat Comms s41467-026-74498-8 (UK Biobank, 277,489) splits disease-free life expectancy owed to SI+loneliness into physical-disease and mental-disorder channels — mental-disorder channel dominates (women −3.7, men −5.8 years) and is NOT buffered by healthy lifestyle, unlike the SI channel; (b) sibling-control + MR triangulation (s41467-026-74758-7) finds loneliness causally on mental health + general health, NOT on specific physical health outcomes; (c) Miller et al. (PLOS One, chronic-stress epigenome) separates perceived stress from hair cortisol methylation channels; (d) financial strain is the most robust psychosocial predictor of accelerated epigenetic ageing across 8 outcomes (Perry et al. group, same pipeline as "difficult ties"); (e) Cell Genomics MR paper declares interferon-pathway cytokines causal drivers of epigenetic age acceleration — mechanism leg under the Duke-NUS two-wave finding. Full detail: 2026-10-12-eve-wellbeing-weekly-scan.

Update 2026-10-19: feeds re-polled (On Purpose 889, Rangan 684, Happiness Lab 279, WCDHT 648, Mel Robbins 447, Life Kit 268). Kept episodes: Happiness Lab "Text vs Talk" (Nick Epley, 12 Oct — cites real older primary papers), WCDHT "Finally, Menopause Relief" (Dr Rachel Rubin, 13 Oct — advocacy framing, see scan), Mel Robbins Lara Boyd (untraced, parked). Big evidence item: ACP guideline + 90-trial systematic review in Annals of Internal Medicine (6 Oct 2026) — menopausal hormone therapy first-line, high certainty; only 8/90 trials were perimenopausal. New caregiver-cognition regression (Elliott, PMID 42622814, n=42, social support predicts memory/processing speed); leisure time as caregiver lever (PMID 42842821, OR ≈ 5.5); young-adult diurnal cortisol → epigenetic age acceleration (AJHB, Cebu cohort). One item press-traced only and explicitly unverified: US GBD alcohol-attributable-cancer-mortality analysis (Lancet Reg Health Americas). Full detail: 2026-10-19-eve-wellbeing-weekly-scan.

The frequency map

Topic mentions across all six feeds:

stress/cortisol 1074   sleep/circadian 672    movement 667
community/loneliness 548   longevity 471   screens/attention 461
therapy-vs-self-help 441   grief 376    caregiver burnout 210
alcohol 195   nature 175   sugar/insulin 165   microbiome 157
fasting 145   perimenopause 144   blood sugar 96   caffeine 88
inflammation 83   nervous system 77   detox 57   breathwork 57
morning light 30   cortisol -> belly fat 28

Reading: this genre is organised around stress and sleep as the master variable. Everything else is downstream. Note the tail: the cheapest, most well-evidenced intervention (morning light) is the one discussed least.

The three models that recur independently

1. Cortisol → visceral / belly fat. Named by both On Purpose (guest Pelz) and Rangan. Rangan pairs it with perimenopause: progesterone falls → cortisol rises → sleep worsens. This is the single link that bridges the two corpora — Kushi's liver-stagnation / dampness cluster and their cortisol-belly-fat cluster are the same loop described in two vocabularies.

2. The caregiver as "invisible second patient." Independent across On Purpose, Mel Robbins (caregivers face measurably higher cognitive-decline risk), and Happiness Lab. See also 2026-10-04-wellbeing-source-map.

3. Rest ≠ sleep. Life Kit's Dr Saundra Dalton-Smith — Sacred Rest, seven kinds of rest. The caregiver-relevant one is sensory/perceptual rest, not sleep.

Evidence checks (claims traced to primary sources)

• "Chronic stress ages your immune system by 10 years" — REAL. Epel/Dhabhar, Social stress contributes to accelerated aging of the immune system, NIA 2022; telomere work PNAS 2004 (PMID 15574496). Solid.
• "Mindfulness ≈ Lexapro for anxiety" — REAL, more than it looks. TAME trial: 276 adults with diagnosed anxiety disorders, MBSR vs escitalopram, difference on primary measure 0.07 points, inside the non-inferiority margin. JAMA Psychiatry. State precisely: non-inferiority in one trial. Never let it drift into "meds don't work" / "stop your medication".
• Alcohol, "no safe amount" — WHO/Lancet Public Health Jan 2023; IARC Group 1 carcinogen. Firm.
• Loneliness → mortality — Holt-Lunstad et al. 2015, Perspectives on Psychological Science, 70 studies. Real.
• Caregiver burden → cognitive decline — LSE rapid review on unpaid caring; dementia-caregiver distress mediation studies. Real but heterogeneous.

Framing rule: always state what the evidence supports, how strong, and what it does NOT say.

False positive caught

Regex \bPEACE\b returned 351 hits — every one from a single guest episode title, "Finding Peace with Ashley C". Never report a frequency count without printing sample contexts first.

Related

• 2026-10-04-wellbeing-source-map — the podcast/episode landscape
• 2026-09-22-kushi-corpus-map — traditional/philosophical lane
• 2026-10-04-eve-wellbeing-reading-protocol — how Eve runs this lane

Limits

Mined show notes, not audio. Descriptions are written to sell the episode. The frameworks are real; the framing is promotional. Where they conflict, the primary source wins.