Research Radar is being rebuilt. On September 30, 2026 we checked every study summarized on this page against PubMed. None of the 12 could be matched to a published paper, so all of them, and the five issue summaries built on them, have been removed. New issues will list only studies with a PubMed record, linked by PMID and DOI. Details are in our corrections log.
The longevity field produces hundreds of studies per month. Most don't matter. We read them so you don't have to — and tell you honestly what each finding means, what the methodology allows us to conclude, and whether you should change anything about your approach.
📅 Latest: October 2026
🔬 12 studies reviewed this issue
📚 47 total studies reviewed
✍️ By Derek Giordano, IQ Healthspan
Browse Issues
Impact Signal🟢 High ImpactChanges what we recommend🔵 ModerateStrengthens or refines current guidance🟡 PreliminaryInteresting; watch for replication🟣 CautionaryChallenges a popular assumption
🔍
9 studies
Issue 1 · October 2026
Lifestyle versus metformin, GLP-1 outcomes, and two supplement trials that moved biomarkers but not results
Nine human studies published July to September 2026, each checked against its PubMed record: a 25-year follow-up from the Diabetes Prevention Program, pooled GLP-1 drug outcomes, two supplement trials that raised blood or brain levels without improving cognition, and new meta-analyses on exercise, bone and diet. Issue compiled 2026-10-03.
9
Studies Reviewed
3
Strong Evidence
2
Null Results
Oct 3, 2026
Published
Editorial Standards
What earns a spot in Research Radar
We review human studies only for actionability grades — animal data is noted where mechanistically important but not treated as practice-changing. We prioritize RCTs, large prospective cohorts, and pre-registered trials. We evaluate every study for sample size, follow-up duration, control quality, and funding source. Studies funded exclusively by supplement manufacturers without independent replication receive a flag. We tell you when a finding is exciting but fragile — and when a "groundbreaking" press release is outrunning the actual data.
Strong evidenceDrugs & InterventionsJAMA
A lifestyle program in prediabetes was linked to fewer chronic conditions over two decades; metformin was not
In the Diabetes Prevention Program follow-up, people originally assigned to the lifestyle program had a lower risk of developing two or more chronic conditions (HR 0.79); metformin showed no difference from placebo (HR 0.91).
Study Design
Long-term follow-up of a randomized trial
Sample Size
n = 1,173 with Medicare data (of 3,234 enrolled)
Duration
About 25 years
Funding
Not stated in the abstract
What They Did
Adults with prediabetes were randomly assigned to an intensive lifestyle program, metformin or placebo in 1996-1999 and followed through 2021 using Medicare records for the 1,173 who consented.
What They Found
By the end of follow-up, most participants had two or more chronic conditions. The lifestyle group had a lower risk of multimorbidity than placebo (HR 0.79, 95% CI 0.68-0.93); metformin did not differ from placebo (HR 0.91, 0.78-1.07). Results held when diabetes was excluded.
⚠️ Methodological Caveats
Only about a third of the original participants had Medicare data, the placebo was stopped after the original trial, and later analyses are observational, so the findings are associations.
What To Do With This
Evidence that lifestyle change in prediabetes pays off for decades remains strong; this does not support taking metformin to prevent chronic disease.
Strong evidenceDrugs & InterventionsDiabetes Res Clin Pract
Tirzepatide lowered major cardiovascular events and deaths across 22 trials
Spiazzi et al. Diabetes Res Clin Pract. 2026. PMID 42442557 · DOI
⚡ Key Takeaway
Pooling 22 randomized trials (29,023 people with type 2 diabetes or obesity), tirzepatide lowered the odds of major cardiovascular events by 13% and of death from any cause by 16%.
Study Design
Systematic review and meta-analysis of RCTs
Sample Size
n = 29,023
Duration
At least 24 weeks per trial
Funding
Not stated in the abstract
What They Did
Researchers pooled randomized trials of tirzepatide in adults with type 2 diabetes or overweight or obesity, assessed risk of bias and used trial sequential analysis.
What They Found
Major adverse cardiovascular events were reduced (odds ratio 0.87, 95% CI 0.79-0.94; high certainty), as was all-cause mortality (OR 0.84, 0.75-0.93). No association was seen for individual components of MACE or heart failure hospitalizations.
⚠️ Methodological Caveats
Most included trials were not designed to measure cardiovascular outcomes, and the pooled effect comes mostly from a few large trials.
What To Do With This
Adds to evidence that GLP-1-based drugs reduce cardiovascular risk in people with diabetes or obesity; it says nothing about use in healthy people.
Semaglutide reduced serious kidney outcomes across three large trials
Mann et al. Lancet Diabetes Endocrinol. 2026. PMID 42567173 · DOI
⚡ Key Takeaway
In 30,787 participants from the SELECT, FLOW and SOUL trials, semaglutide lowered the risk of a combined kidney outcome by 16% compared with placebo.
Study Design
Prespecified pooled analysis of three RCTs
Sample Size
n = 30,787
Duration
Mean 39.5-47.5 months
Funding
Novo Nordisk
What They Did
Participant-level data from three placebo-controlled trials in people with chronic kidney disease or cardiovascular disease were pooled.
What They Found
The primary kidney composite (major eGFR decline, kidney failure, or kidney or cardiovascular death) was reduced (HR 0.84, 95% CI 0.77-0.91). Safety was in line with other GLP-1 drug trials.
⚠️ Methodological Caveats
Doses and forms differed between the trials (weekly injections and a daily pill), the analysis was funded by the manufacturer, and participants already had cardiovascular or kidney disease.
What To Do With This
Relevant for people with cardiovascular or kidney disease discussing treatment with their doctor; not evidence for preventive use in healthy adults.
A post hoc analysis of the MIND trial found that participants who improved their diet scores most showed small gains in global cognition, although the trial itself found no difference between diets.
Study Design
Post hoc analysis of an RCT
Sample Size
n = 590
Duration
3 years
Funding
Not stated in the abstract
What They Did
Researchers grouped MIND trial participants by how much their diet scores improved, checked with blood nutrient levels, and compared changes in cognition.
What They Found
Those with substantial improvement showed a small annual increase in global cognition; modest improvement was not linked to change. The association did not differ by trial arm.
⚠️ Methodological Caveats
Post hoc and observational within a trial whose randomized comparison was null; people who change diet most may differ in other ways.
What To Do With This
Not proof that the MIND diet protects memory, but a reason to keep studying adherence.
A BMJ network meta-analysis of 124 trials found that brisk walking or jogging and combined aerobic-resistance exercise probably improve bone density, with meaningful effects at about 2 to 3 hours of brisk walking a week.
Study Design
Network meta-analysis of RCTs
Sample Size
n = 18,429
Duration
Varied by trial
Funding
Not stated in the abstract
What They Did
Researchers compared exercise types and doses across randomized trials in adults aged 40 and older, rating evidence certainty.
What They Found
Brisk walking or jogging and combined aerobic-resistance exercise probably improved spine bone density; dose-response looked inverted-U shaped. Mixed aerobic exercise (OR 0.29) and mind-body exercise may reduce fractures, on low-certainty evidence.
⚠️ Methodological Caveats
Much of the evidence was low to moderate certainty, and the fracture findings rest on fewer trials.
What To Do With This
Supports regular brisk walking or jogging plus strength work for bone health in midlife and beyond.
Across 126 studies of 4,019 women, resistance training improved strength and body composition similarly before and after menopause, with no sign that age limited the benefit.
Study Design
Systematic review and meta-analysis
Sample Size
n = 4,019
Duration
Varied by study
Funding
Not stated in the abstract
What They Did
Researchers pooled studies of resistance training in healthy women aged 18 and over, comparing pre- and postmenopausal women.
What They Found
Strength improved substantially in both groups with no subgroup difference; lean mass rose and fat mass fell. Age, training duration and frequency did not predict the size of the benefit.
⚠️ Methodological Caveats
Included studies varied in design and quality, and many were small.
What To Do With This
It is never too late to start strength training; general guidelines appear to apply to women of any age.
A meta-analysis of 33 randomized trials (3,476 people) found a Mediterranean diet lowered hs-CRP, IL-6 and IL-17, but not several other inflammation or antioxidant markers.
Study Design
Systematic review and meta-analysis of RCTs
Sample Size
n = 3,476
Duration
Varied by trial
Funding
Not stated in the abstract
What They Did
Researchers pooled randomized trials comparing a Mediterranean diet with a control diet and measured blood markers of inflammation and oxidative stress.
What They Found
hs-CRP, IL-6 and IL-17 fell significantly; standard CRP, IL-10, TNF-alpha and total antioxidant capacity did not change.
⚠️ Methodological Caveats
These are biomarkers, not health outcomes, and the authors call for further high-quality studies.
What To Do With This
Consistent with the Mediterranean diet's cardiovascular trial evidence; the inflammation changes alone do not prove benefit.
In a 12-week pilot trial in adults with amnestic mild cognitive impairment, NR doubled blood NAD+ but did not improve cognition, total brain blood flow or blood pressure.
Study Design
Randomized, double-blind, placebo-controlled pilot
Sample Size
n = 42 completed
Duration
12 weeks
Funding
Not stated in the abstract
What They Did
Adults with amnestic mild cognitive impairment took nicotinamide riboside or placebo for 12 weeks.
What They Found
Blood NAD+ increased twofold with NR, without serious side effects. There were no improvements in cognitive function (the primary outcome), total cerebral blood flow or blood pressure; exploratory analyses hinted at regional blood flow changes.
⚠️ Methodological Caveats
Small and short, and the hippocampal finding was exploratory.
What To Do With This
Raising NAD+ with supplements is easy; showing a benefit is not. This adds to the run of null clinical results for NAD+ precursors.
In a 24-month trial in 365 older adults at risk for dementia, 2 g a day of DHA raised DHA in spinal fluid but did not change brain volumes or cognitive performance.
National Institute on Aging and the Alzheimer's Drug Discovery Foundation
What They Did
Adults aged 55 to 80 without dementia, with low dietary DHA and at least one dementia risk factor, took 2 g of DHA a day or placebo, stratified by APOE4 status.
What They Found
DHA increased the spinal-fluid DHA-to-arachidonic acid ratio at 6 months regardless of APOE4 status, but there were no treatment differences in brain volumes or cognition over 24 months.
⚠️ Methodological Caveats
Dropout was 38%, largely during COVID-19, and the trial was powered for a biochemical outcome rather than cognition.
What To Do With This
Another well-run trial in which an omega-3 supplement changed a biomarker without changing outcomes.
Try clearing the search or selecting a different category.
How to Read Longevity Science
The critical literacy skills that separate signal from noise
The longevity field produces hundreds of studies per month. The majority are interesting but not practice-changing. A small fraction are genuinely important. Developing the ability to distinguish between them is a skill that Research Radar aims to help you build.
Correlation vs. Causation
Most published longevity research is observational — it shows that people who do X tend to live longer, not that X causes longer life. People who eat more vegetables also tend to exercise more, smoke less, and have higher incomes. Separating the effect of the vegetable from the healthy user bias requires randomized controlled trials, which are expensive, slow, and ethically complex for longevity endpoints. When Research Radar rates a study as "Preliminary," it typically reflects this causation gap — promising association data that hasn't yet been tested in a controlled experiment.
Effect Size vs. Statistical Significance
A study of 100,000 people can find a statistically significant result that represents a trivially small absolute effect. A "20% reduction in risk" sounds dramatic until you learn the baseline risk was 1% — meaning the absolute reduction is 0.2 percentage points. Research Radar always reports absolute risk changes, not just relative risk ratios. The distinction matters enormously: relative risk language is how supplement companies market products with modest absolute effects as life-changing interventions.
The Replication Standard
A single impressive trial — even in Nature or the New England Journal — is not sufficient reason to change behavior. The reproducibility crisis in science is real: a 2015 replication project found only 36–39% of psychology and social science findings replicated with the same effect size. Biomedical research has a similar problem. Resveratrol, DHEA at supraphysiological doses, and antioxidant megadosing all had impressive single-study data before failing replication. Research Radar requires independent replication before awarding a High Impact signal to any finding.
Publication Bias
Studies with positive results are approximately three times more likely to be published than studies with null results. This means the published literature systematically overstates the effectiveness of interventions. For every published paper showing a supplement works, there may be three unpublished studies showing it doesn't — a phenomenon called the "file drawer problem." Pre-registered trials (where the study design is locked before data collection) are a partial solution, and we note pre-registration status in our study reviews. When we see a cluster of positive findings from one research group without independent replication, this is a red flag we call out explicitly.
Every study in Research Radar links to its PubMed record and DOI. Summaries describe only what the published abstract reports, including null results and the authors' stated limitations.
Review Methodology
How we select and evaluate research
Most science journalism covers research by summarizing abstracts. We read full papers — methodology, supplemental data, and funding disclosures — and evaluate them against these criteria before including them in Research Radar.
🧪
Study Design Hierarchy
RCTs rank highest, followed by pre-registered prospective cohorts, then retrospective studies. Cross-sectional and correlation studies are noted but cannot justify behavior change recommendations. Animal studies are included only when mechanistically illuminating.
📊
Effect Size, Not P-Value
A statistically significant result in a 10,000-person study can represent a trivially small effect. We always report absolute risk changes, not just relative risk ratios. A "40% reduction in risk" means nothing without the baseline rate.
💰
Funding Transparency
Industry-funded studies are included but flagged. The majority of supplement research is manufacturer-funded — this doesn't make findings wrong, but it does require independent replication before we upgrade a finding's impact signal.
🔁
Replication Standard
A single impressive trial earns Preliminary status at most. High Impact designations require replication from independent research groups. This is why we downgrade many widely-covered studies — including some from major journals.
Get new research by email
Every study that matters, and every one that doesn't, explained honestly. Unsubscribe anytime.