Eve Wellbeing Weekly Scan — 2026-10-07

Standing policy: 2026-10-04-eve-wellbeing-reading-protocol. No diagnosis, no prescription, no medication advice. Alcohol: Young's agreed 1-2 weekend beers as wind-down stands — nothing here relitigates it. Modern lane only; the TCM/Kushi lane (2026-09-22-kushi-corpus-map) is never blended with this material.

1. New episodes since the 2026-10-04 corpus build

All six feeds polled live (chunked fetch; 3,195 <item> blocks parsed). Genuinely new titles vs the evidence-base episode table:

None of the Happiness Lab, WCDHT, On Purpose, Mel Robbins, or Life Kit entries are evidence-lane material this week (celebratory interview, best-of replay, US insurance navigation, confidence/motivation content) — skipped per the standing rule.

2. Episodes read and assessed

#685 Nir Eyal (Rangan, 6 Oct) — screens/attention thread. Show-note claim: "only around 10 per cent of phone checks are triggered by an external notification; the other 90 per cent come from within." This is a guest-author statistic, untraced to a primary source this pass — treat as unverified until it is. Framework overlap is real, though: attention as internally triggered fits the corpus's dominant stress→screens→sleep loop, and the "psychological nutrients" framing for children's screen time is the carer-relevant angle for two teens. Duplicate-ish content (Eyal's fourth appearance across these feeds) — the echo chamber filter applies: same voice again, not new confirmation.

#684 Seheult sunlight BITESIZE (Rangan, 1 Oct) — morning-light thread. This is the first time this feeds-assembled corpus has dedicated a whole episode (bitesize) to light: the evidence-base's claim that morning/monthly light content was under-weighted relative to its evidence strength is weakening at the margin — the topic just earned a headline slot from the GP-voice show. Content is sensible (20 min outside, even cloudy; mood/energy/sleep framing) but is promotional repackaging of episode #586 — evidence checks still to primary sources.

3. New primary evidence (each traced this pass)

3.1 Immune ageing in waves around 40 and 60 — Nature Communications

Duke-NUS, Park/Behmoaras et al., Nature Communications 2026, article DOI 10.1038/s41467-026-76737-4 (press push 1 Oct via EurekAlert; article dated Aug 2026). Multi-ethnic single-cell transcriptome of Peripheral Blood Mononuclear Cells (PBMCs): 3.8 million cells from 1,828 healthy adults aged 19-97, 35% Asian. Findings: immune ageing is not linear — two peaks of differential gene expression, around 40 (mainly CD4 T cells) and after 60 (mainly CD8 T cells), alongside declining RNA/protein homeostasis and inflammatory polarization, with sex-dependent kinetics. Sex-stratified deep-learning biological-age clocks outperformed sex-combined models.

• Supports: the standing Epel/Dhabhar stress→immune-ageing thread is now population-level, single-cell, both sexes — and worth noting that it puts Young (mid-40s) squarely inside the first documented acceleration wave.
• Strength: strong for description/correlation — observational, healthy adults, no intervention tested.
• Does NOT say: that any specific exposure (or practice) moves these waves; no stress, caregiving, or lifestyle causal claims.

3.2 Living with a person with dementia → impaired daily function from cognitive decline — n=17,596

Annals of Geriatric Medicine and Research (PMC13054570), cross-sectional analysis of the 2024 Korea Community Health Survey, adults 50+. Adjusted OR 1.86 (95% CI 1.63–2.12) for self-reported cognitive-decline impairment in daily activities among those cohabiting with a family member with dementia vs not (23.3% vs 7.8% prevalence).

• Supports: the caregiver-cognitive-decline thread (2026-10-04-wellbeing-source-map) with a very large, nationally representative population — beyond the LSE rapid review's heavier-than- opportune heterogeneity note.
• Strength: moderate-to-strong for association; cross-sectional, so direction of causation is not established and cohabitation is not caregiving intensity.
• Does NOT say: anything direct about parents caring for teenage children. The sample is 50+, cohabiting with older/ill family members — Young's solo-parent-of-two-teens situation is structurally adjacent (the "invisible second patient" model) but not personally equivalent, and this should not be overstated onto his exact case.

3.3 Alcohol-attributable cancer mortality doubled 1990→2023 — Lancet Regional Health – Americas

Jani et al., secondary Global Burden of Disease analysis, published online Oct 2026 (10.1016/j.lana.2026.101599). Alcohol-attributable US cancer deaths rose from 11,361 (1990) to 23,126 (2023); age-standardized mortality rate rose in males (+14.3%) and fell in females (−16%), with 55+ carrying most of the burden (liver, esophageal, colorectal; breast highest in females). Two companion 2026 papers specify the policy picture: Lancet Reg Health – Europe modelling (146,194 EU cancer cases in 2022 attributable to alcohol; taxation/availability/marketing interventions all reduce incidence) and Lancet Public Health (Canada's minimum unit pricing plus cancer warning labels, Mar 2026, modelling; refines the intervention side rather than the risk mechanism).

• Supports: the "no safe amount," IARC Group 1 carcinogen, and 7-cancer links already in 2026-10-04-wellness-podcast-evidence-base — the magnitude and trend claim now has a 2026 primary paper behind it.
• Strength: very strong for description; this is mortality-trend analysis based on GBD's well-established methodology.
• Does NOT say: anything that applies to deciding what one adult's moderate, agreed, wind-down intake should be. This is population level. Young's 1-2 weekend beers agreement stands untouched, as the protocol requires.

3.4 Perimenopause sleep — real-world fezolinetant signal (preliminary, industry)

Astellas PRN release, presented 2 Oct 2026 at the International Menopause Society congress (OPTION-VMS, Phase 4, real-world). Preliminary claim: with fezolinetant 45mg, mean reduction of 2.6 nighttime awakenings by week 12; patient-reported sleep disturbance moved from mild to no-disturbance band by week 4; objective actigraphy measures improved at weeks 4, 8, 12. US label retains a boxed warning for liver injury.

• Supports: the Rangan perimenopause progesterone→cortisol→sleep chain (2026-10-04-wellbeing-source-map) gets a new pharmacological data point — neurokinin-B antagonism affecting sleep in a real-world cohort. Familial relevance is indirect; Young is sole carer of a teenage daughter — sleep-disruption mechanism literacy, not personal applicability.
• Strength: weak-to-moderate for inference — industry-sponsored, non- randomized, real-world, preliminary, no between-group comparison, press-release staged rather than peer-review staged.
• Does NOT say: anything about non-medication paths to sleep improvement, or that this drug is preferable to alternatives; it explicitly was not designed to compare treatments; not medical advice in either direction.
• Watch item: an unrelated open-label randomized trial protocol (Trials, Springer, n=54, late perimenopause) will directly compare hormone replacement therapy vs cognitive behavioral therapy for insomnia (CBT-I) vs sleep hygiene on subjective and objective sleep — worth tracking for a later scan, as CBT-I-vs- hormonal results may change the field's default answer.

3.5 (Mechanism, mouse-only) Psychological stress accelerates immune ageing via the gut

Cell Stem Cell, 2 July 2026 (Sun Yat-sen University, Guangzhou). In mouse models, chronic psychological stress → reduced activity in medial prefrontal cortex and periaqueductal gray → gut microbiota shift (loss of Lactobacillus reuteri, lower spermidine) → hematopoietic stem-cell ageing signatures (fewer lymphocytes, loss of regenerative capacity).

• Supports: an actual mechanism bridging "chronic stress ages the immune system" (Epel/Dhabhar/NIA 2022, already in 2026-10-04-wellness-podcast-evidence-base) to the gut-brain-bone-marrow axis — three independent groups (Duke-NUS, Sun Yat-sen, Nature Aging News & Views, Aug 2026) now circling the same stress→immune-ageing territory from different angles, which is stronger signal than one paper.
• Strength: mechanism-level, mouse only. Real for mechanism, not yet for humans.
• Does NOT say: that any human supplement, diet, or practice (or probiotic) reproduces this in people. No human replication this pass.

4. Contradictions with existing notes (flagged prominently)

1. Nothing falsified. No note in 2026-10-04-wellness-podcast-evidence-base or 2026-10-04-wellbeing-source-map was contradicted this week.
2. Refinement (morning light is losing its "under-discussed" status). The evidence-base records morning light at 30 mentions — the lowest of any high-evidence topic. Rangan #684 (Seheult, sunlight) is now a dedicated whole bitesize. The claim "most under-discussed high-value intervention" is directionally correct in aggregate but weakening; a single new episode does not change the frequency count (no recount without sample contexts, per protocol) but does change the qualitative read.
3. Refinement (loneliness → mortality numbers are smaller and more specific than the Holt-Lunstad headline). Two newer meta-analyses refine rather than refute the standing thread: Schutter et al. (older adults; loneliness HR 1.10, 95% CI 1.06–1.14) and Aging Clinical and Experimental Research (Jan 2025; 86 studies; loneliness HR 1.14, social isolation HR 1.35, living alone HR 1.21; heterogeneity I²=84%). Objective isolation outperforms subjective loneliness as a mortality predictor. This makes the practical edge sharper, not softer: structural contact (showing up, sharing space) carries more weight than the feeling of loneliness alone.

5. Unverified / unresolved / not-yet-checkable

• Nir Eyal's "10% external trigger / 90% internal" phone-check statistic — guest-book-sourced; no primary trace this pass; will not be repeated as a number until traced.
• Duke-NUS Nature Communications paper — title confirmed via Nature PDF and multiple secondary sources; exact PMID/epub date not captured this pass. Effect claims are associative only.
• OPTION-VMS / GRACE / ADELE-AU fezolinetant findings — industry preliminary, no peer-reviewed publication located; treat as a signal, not a finding.
• The CBT-I-vs-HRT-vs-sleep-hygiene perimenopause trial — protocol only (n=54); results pending; track next week.
• No new WHO / IARC / NIA / Epel quantitative updates landed this week beyond what the Lancet GBD-piece already embodies.
• Life Kit open-enrollment episode — US health-insurance navigation, out of scope for an Australian lane this week; noted not assessed.

6. Implication for the carer-under-load lane (structural, not prescriptive)

The shape of this week is: immune ageing is being demystified at the population scale (waves at 40 and 60), and the caregiver thread — already marked real but heterogeneous — gained a large cohort data point with explicitly bounded scope. Nothing here changes the standing applied read: chronic non-negotiable stress with no recovery window is still the master variable, morning light is still cheap and now GP-presented, and social contact remains the intervention with the least promo and the most evidence behind it. No new action request is implied; no technique is being pushed from a magazine scale into his life.

Related

• 2026-10-04-wellness-podcast-evidence-base — corpus, frequency map, checks
• 2026-10-04-wellbeing-source-map — carer-under-load applied read
• 2026-10-04-eve-wellbeing-reading-protocol — standing policy
• 2026-09-22-kushi-corpus-map — traditional lane (kept separate)