23 mg/dL off LDL from a six-ingredient supplement tablet [Best Read]
A six-ingredient lipid tablet tested in 90 adults, broccoli extract against muscle damage, bitter melon only alongside metformin.
In my work as a Silicon Valley based startup executive and longevity researcher, I track the gap between what the labs are publishing and what’s actually worth adding to your protocol. Here’s what stood out this week — with the numbers that matter.
LDL fell 13.2% on the tablet and rose 1.6% on placebo over 12 weeks [Nutrients]
HUMAN randomized, double-blind, placebo-controlled, parallel-group trial. Ninety adults with low-to-moderate cardiovascular risk and elevated LDL-C were randomized to NUT2 (n = 45) or placebo (n = 45), one tablet daily for 12 weeks alongside standardized dietary advice. The formulation is worth reading closely, because it is the sum of six things biohackers already buy separately: artichoke extract 100 mg, bergamot extract 300 mg (equivalent to flavonoids 120 mg), Berberis aristata extract 588 mg, astaxanthin 0.50 mg, folic acid 0.20 mg and chromium 0.04 mg.
At week 12 the baseline-corrected between-group difference in LDL-C was −23 mg/dL (95% CI −29 to −16; p < 0.001), and total cholesterol fell by a corrected −26 mg/dL (95% CI −36 to −17; p < 0.001). Homocysteine was significantly reduced and fasting insulin concentrations were lower, and metabolic syndrome prevalence at week 12 was 13.3% with NUT2 versus 37.8% with placebo (p = 0.008) — which the authors themselves flag as exploratory and based on small participant numbers. There was no significant between-group difference in hsCRP. Three caveats belong next to that 23 mg/dL: the study was funded by Meda Pharma S.p.A. (the product sits in the Armolipid brand, now owned by Cooper Consumer Health) and two authors are company employees; baseline lipid values differed between the groups, which is why every headline number is baseline-corrected; and although ethics approval dates to 13 January 2020, the ClinicalTrials.gov registration (NCT07492264) is dated 23 March 2026.
HR 0.42 for cardiovascular death among patients who kept exercising after coronary disease [Nutrients]
HUMAN systematic review and meta-analysis, and by a wide margin the largest evidence base in this issue: 40 reports representing 34 data sources and at least 260,354 unique patients with coronary heart disease, comprising six randomized trials, 30 observational reports and four comparative nonrandomized studies, searched from January 2015 through 15 June 2026.
Maintained or continued physical activity relative to persistent inactivity carried a hazard ratio of 0.64 for all-cause mortality (95% CI 0.55 to 0.74; I² = 29%) and 0.42 for cardiovascular mortality (95% CI 0.28 to 0.64; I² = 72%). Closer Mediterranean diet adherence was inversely associated with all-cause mortality (HR 0.74, 95% CI 0.60 to 0.90; I² = 44%), while cardiovascular mortality for diet remained uncertain. The single randomized comparison of a Mediterranean versus a low-fat diet gave an HR of 0.72 for major adverse cardiovascular events (95% CI 0.54 to 0.96; moderate certainty). Two things to hold onto: 30 of the 40 reports are observational and only six are randomized, so the physical-activity hazard ratios rest largely on comparisons between people who kept exercising and people who stopped rather than on assignment; and the I² of 72% on that 0.42 signals substantial heterogeneity between studies. The authors’ own conclusion is that randomized evidence on exercise or combined dietary and activity programs was sparse.
Bitter melon protein tracked with a 0.23% HbA1c drop — only in metformin users [Nutrients]
Two studies in one paper. The ANIMAL arm used db/db mice: a bitter melon blood-sugar-modulating protein (BMP) significantly reduced fasting blood glucose and HbA1c, with greater glucose-lowering when combined with metformin than metformin alone (approximately 300 mg/dL versus 200 mg/dL reduction; p < 0.05), and skeletal muscle transcriptomics identified 27 shared pathways involved in glucose and lipid metabolism.
The HUMAN arm is an exploratory, retrospective, self-controlled observational cohort of 119 individuals with prediabetes or diabetes taking daily BMP for three months. Among metformin users, BMP supplementation was associated with an HbA1c reduction of 0.23% (p = 0.0028), total cholesterol −7.5 mg/dL and LDL cholesterol −6.1 mg/dL; the largest observed HbA1c reduction, 0.32% (p < 0.0001), occurred in participants on 1000 mg/day of metformin. In non-users, no significant changes were observed. The authors state plainly that given the retrospective self-controlled design and baseline clinical differences between treatment groups, these clinical findings reflect observational associations rather than causal effects. That interaction pattern — a botanical whose association with better numbers showed up only alongside a drug — is the part worth filing away.
Creatine kinase 121.52 on sulforaphane versus 291.84 on placebo, in 14 people [Nutrients]
HUMAN double-blind, placebo-controlled crossover trial, and the authors label it “Proof of Principle” for a reason. Fourteen healthy participants (7 male, 7 female; age 25.07 ± 1.04 years) took sulforaphane glucosinolate capsules at 200 µmol/day or placebo for two weeks, crossed over after a three-week washout, and performed a maximal exercise test after each period.
Sulforaphane did not alter endurance capacity or blood cell counts. It did attenuate serum creatine kinase (p = 0.034; sulforaphane 121.52 ± 16.89, placebo 291.84 ± 133.07) and myoglobin (p = 0.011; sulforaphane 15.82 ± 1.95, placebo 27.31 ± 5.71) after exercise; in the discussion the estimated marginal means for myoglobin were 17.81 (95% CI 13.35–22.27) on sulforaphane versus 26.47 (95% CI 22.01–30.94) on placebo. Absolute oxidative stress markers showed no significant change, and the significant supplement interactions appeared only in percent change from baseline for the d-ROMs (p = 0.038) and OXY-adsorbent (p = 0.032) tests. Read the dispersion before the means: the placebo creatine kinase standard deviation of 133.07 is close to half its own mean, n is 14, the abstract reports these values without units, and the authors state that the study received funding from Murakami Farm Co., Ltd.
Hesperetin restored a longevity gene in aged mouse liver — no effect sizes published [npj Aging]
ANIMAL study in naturally aged mice, plus human liver tissue for validation. Late-life administration of the citrus flavonoid hesperetin restored hepatic Cisd2 expression, improved liver pathology, and partially reversed aging-associated transcriptomic alterations. Mechanistically the authors identify Hmgcs2 as a candidate molecular target of hesperetin and show that Hmgcs2 interacts with Pparα to activate Cisd2 transcription through a Ppar response element in the Cisd2 promoter; the protective effects are largely Cisd2-dependent, being significantly attenuated in hepatocyte-specific Cisd2 knockout mice. PPARα and CISD2 expression decline with age in human liver tissues.
The reason this gets a mechanism in the headline instead of a number is that it does not publish one. This is an unedited early-access manuscript, and the abstract carries no effect sizes, no doses and no animal numbers — so there is nothing here to translate into a human quantity, and “late-life administration” in a mouse has no defined equivalent in a person. What it does establish is a specific, testable axis (Hmgcs2 → Pparα → Cisd2) that a citrus flavonoid appears to engage in an aged liver, which is a more useful thing to watch than another correlation.
Click here to share your thoughts
This newsletter is for informational purposes only and is not medical advice. Consult your physician before changing your protocol.