# Analyzed 75 longevity papers. Most of your stack is a nothing-burger. Here's what actually moves mortality
Canonical: https://social-archive.org/gergan/PR7Fthoa6v
Original URL: https://www.reddit.com/r/Biohackers/comments/1sv0cyu/analyzed_75_longevity_papers_most_of_your_stack/
Author: toadlyBroodle
Platform: reddit
## Content
I compiled an open-source wiki of 75 peer-reviewed primary papers across longevity, rejuvenation, and preventive medicine. Trying to separate the signal from the multi-billion-dollar supplement industry. Sharing the verdict because if you're running a "longevity stack," most of it probably isn't doing what you think. **What the RCTs say doesn't work:** * **NMN / NR.** Blood NAD+ rises, clinical endpoints don't. The latest large RCT (NR in long-COVID, 2025) is mixed at best across cognition and recovery markers. * **Vitamin D supplementation in non-deficient adults.** VITAL trial 2019 NEJM, n=25,871, 5.3-year follow-up: null on cancer (HR 0.96), CVD (HR 0.97), and all-cause mortality (HR 0.99). Test before treating; supplement only documented deficiencies. * **"Young plasma" without a defined active fraction.** Plasmapheresis-without-IVIG was *negative* in a 2025 RCT (Horvath co-authored); some clocks accelerated. The TPE+IVIG protocol that actually worked (-2.6 yr biological age) suggests the IVIG is the active ingredient, not the plasma removal. * **Telomerase as a pill.** The 2022 PNAS paper claiming +41% lifespan was retracted in August 2025. * **Most "longevity stacks"** (resveratrol, anti-aging peptides, exotic herbal blends): no RCT support at endpoint level. **What actually moves mortality (free, no prescription):** * **VO2max.** Mandsager 2018 *JAMA Network Open*, n=122,007. Low vs elite cardiorespiratory fitness: **HR 5.04 for all-cause mortality**, no upper limit of benefit. Bigger effect than smoking, diabetes, or prior CAD in the same cohort. Norwegian 4×4 protocol raises VO2max \~13% in 8 weeks (Helgerud 2007 MSSE). * **Grip strength + resistance training.** PURE study (n=139,691, 17 countries): grip strength predicts CV mortality *more strongly than systolic BP*. Resistance training ≈21% lower mortality alone, ≈40% combined with cardio (Saeidifard 2019 meta-analysis). * **Sleep 7-9 hr.** Cappuccio 2010 meta-analysis (\~1.4M adults): U-shaped curve, short-sleep RR 1.12, long-sleep RR 1.30. * **Waist circumference, not BMI.** Pischon 2008 NEJM EPIC (n=359,387): waist + waist-to-hip ratio independently predict mortality at every BMI stratum. * **Don't smoke. ≤7 drinks/week.** Jha 2013 NEJM: smoking costs >10 years of life. Wood 2018 *Lancet* (n=599,912 drinkers): lowest-mortality threshold \~100 g ethanol/week. The "moderate drinking is protective" finding mostly vanished after correcting for sick quitters. **What actually works with a prescription:** * **Statins for primary prevention.** CTT 2012 *Lancet* IPD meta-analysis of 27 RCTs: each 1 mmol/L LDL-C reduction yields ≈21% lower vascular events per year, including in low-risk adults. Systematically underprescribed. * **BP target <120.** SPRINT 2015 NEJM (n=9,361 non-diabetic hypertensives): intensive control cut all-cause mortality 27%. Trial halted early for benefit. * **Measure apoB, not LDL-C.** Sniderman 2011 head-to-head meta-analysis: apoB beats LDL-C by 12% on relative-risk-reduction prediction. ESC/EAS now name it the preferred lipid metric. * **Rapamycin** (off-label). PEARL trial 2025 showed safe + healthspan markers improved at 1 year. Only emerging-class drug with real human RCT data. **What's actually exciting in the lab (not yet available to buy):** * Anti-IL-11 antibody: +25% mouse lifespan late-life. Already in human trials for fibrotic lung disease. * Trametinib + rapamycin combo: +27-29% additive mouse lifespan, both drugs FDA-approved separately (toxicity profile means wait for the combo trial). * AAV-Klotho gene therapy: +20% mouse lifespan with one shot, multi-organ rejuvenation. * CAR-T senolytics: single infusion persists >12 months in mice, now generalising from metabolism into gut aging. * Partial reprogramming (NewLimit, Retro Bio, Altos) heading toward first-in-human trials. If any of those worked the way the supplement industry claims their products work, the people selling you NMN would already be selling you those. [AI-maintained longevity research wiki](https://github.com/toadlyBroodle/science/tree/main/biology/longevity): 75 papers, 131 wiki pages, full citation graph, every effect size traces to a PMC link. All summaries written by hand from the primary sources, not generated. Built for myself; sharing because it's open-science. Top-level reader version with effect sizes per recommendation: `biology/longevity/recommendations.md`. Reverse-aging research-frontier analysis (partial reprogramming, foundation-model aging clocks, etc.): `biology/longevity/wiki/analysis/promising-reverse-aging.md`. Disagree? Feel free to send PR to improve the wiki.