Sep 28, 2026 · Sports, Health & Nutrition Daily Digest
A roundup of sports, fitness, wellness, and nutrition news for Sep 28, 2026, with summaries, links, and commentary.
I. Nutrition & Diet
1. King's College London: A Mediterranean-style diet tracks with lower odds of adult acne, and the pattern differs by sex (Nutrition)
Summary:
King's College London reported on Sep 28, 2026, that Bochen Li, Wendy Hall, and colleagues analyzed UK Biobank data in the Journal of the European Academy of Dermatology and Venereology. Among 228,757 adults, 1,603 were classed as having current acne from recent diagnoses or self-report. A Mediterranean-style pattern (more vegetables, fruit, whole grains, and healthy fats) was linked with lower odds of adult acne, especially in women. In men, lower-carbohydrate, higher animal-food patterns were linked with higher odds. Cooked vegetables, cheese, and yogurt pointed toward lower odds; skimmed milk and milk-based desserts pointed toward higher odds. The authors state that the cross-sectional analysis cannot show that diet causes or prevents acne, and that dairy should not be treated as one category.
Links:
Commentary:
For adult acne, start with overall diet quality rather than cutting every dairy food; yogurt, cheese, and skimmed milk do not belong in the same rule, and causation still needs a trial.
2. EASD: Highest intake of meat-derived Neu5Gc linked with about 63% higher type 2 diabetes risk, no longer significant after red-meat adjustment (Nutrition)
Summary:
Leopold Fezeu of Université Sorbonne Paris Nord and colleagues presented preliminary NutriNet-Santé results at the European Association for the Study of Diabetes meeting in Milan (Sep 28–Oct 2, 2026); the EASD release is dated Sep 27. In 89,953 adults without diabetes or cancer at baseline (mean age about 42; about 79% women), three 24-hour recalls estimated N-glycolylneuraminic acid (Neu5Gc) from red meat and dairy. Over a median 7.2 years, 966 incident type 2 diabetes cases occurred. Versus the lowest quintile of meat-derived Neu5Gc, the highest had a fully adjusted HR of 1.63 (95% CI 1.29–2.06) and the second-highest 1.41 (1.11–1.77). After further adjustment for total red-meat intake, the association weakened and was no longer significant. Dairy-derived Neu5Gc was not associated with diabetes; the highest quintile of total dietary Neu5Gc had about 37% higher risk, largely from the meat component. Results are not yet finalized, the design is observational, and intake estimates can be mismeasured. Median red meat in the top quintile was about 97 g/day in men and 73 g/day in women, versus about 5 g and 0 g in the bottom quintile; those grams describe the groups, not a safety cutoff.
Links:
Commentary:
If the goal is lower diabetes-related risk, swap beef and lamb toward poultry or plant protein; the 63% figure does not survive adjustment for red meat itself, and it is not a reason to buy a sugar-molecule supplement.
3. EASD: Metagenomes from more than 220,000 adults show a diabetes-like microbiome before blood glucose leaves the normal range (Nutrition)
Summary:
Gabriel Baldanzi and Nicola Segata of the University of Trento and colleagues are presenting at the EASD meeting opening Sep 28, 2026; the EASD release is the same day. Metagenomes and health data covered 229,025 adults in the UK and US (about 73.8% women; mean age about 50), including 3,627 with type 2 diabetes, 14,022 with prediabetes, and 211,376 with normal glucose. After age, sex, and BMI, 789 species were associated with diabetes (168 higher, 621 lower); 587 of them were also associated with unmedicated prediabetes. Among 135,093 normoglycemic participants, people whose glucose stayed high longer in the two hours after a meal were more likely to carry the diabetes-like pattern, and the shift grew as insulin resistance increased. The authors say microbiome profiles do not replace HbA1c, fasting glucose, or an oral glucose tolerance test. The sample size depends largely on the ZOE cohort; Tim Spector, a co-author, is ZOE's scientific co-founder.
Links:
Commentary:
There is no reason yet to pay for a gut test to "prevent diabetes"; diet and activity remain the levers that change risk, and the microbiome here is a research signal, not a clinical assay.
II. Exercise & Training Science
4. BMJ network meta-analysis: About 2–3 hours a week of brisk walking or jogging is near a clinically meaningful bone-density dose after age 40 (Training)
Summary:
A hierarchical Bayesian network meta-analysis led by Changbo Lu and Xiaojiang Yang (Jinling Hospital, Nanjing University, and the 940th Hospital) was published online in The BMJ on Sep 9, 2026 (PubMed 42716561; DOI: 10.1136/bmj-2026-100561) and was widely summarized in Chinese science coverage on Sep 28. It pooled 124 randomized trials and 18,429 adults aged 40 or older. For lumbar-spine bone mineral density, brisk walking or jogging differed from control by 0.013 g/cm² (95% CrI 0.006–0.021; low certainty) and combined aerobic-resistance training by 0.013 (0.010–0.016; low to moderate certainty). At the femoral neck, brisk walking or jogging (0.009, 0.001–0.018) and mind-body exercise (0.007, 0.002–0.013) ranked highest; at the total hip, only brisk walking or jogging stood out (0.021, 0.005–0.038; low certainty). Dose-response curves were generally an inverted U. A minimal clinically important difference was reached at about 400 MET-minutes/week for the lumbar spine and about 600 for the femoral neck and total hip, roughly 2–3 hours a week of brisk walking or jogging. In fracture analyses (26 trials, 11,132 people), mixed aerobic exercise had an OR of 0.29 (0.11–0.78) and mind-body exercise 0.58 (0.36–0.93), both low certainty; other modalities were near neutral for fractures.
Links:
Commentary:
After 40, build toward about 2–3 hours a week of brisk walking or jogging and keep some strength or tai chi/yoga; more volume is not automatically more bone, and balance work still matters for falls.
5. University of Tsukuba acute trial: One 5.5 mg/kg caffeine dose kept a repeated-sprint benefit for at least 3.5 hours (Training)
Summary:
Mizuki Nishikawa and colleagues at the University of Tsukuba report a randomized, double-blind crossover in Medicine & Science in Sports & Exercise (2026; DOI: 10.1249/mss.0000000000004129), with the university write-up on Sep 28. Twelve healthy young adults (4 women) took caffeine 5.5 mg/kg or placebo 60 minutes before exercise, then completed three repeated-sprint cycling sets across 3.5 hours. Versus placebo, mean sprint power was about 0.5%–2.4% higher through 3.5 hours, and peak power in the second and third sets rose by up to about 3.7%. Noradrenaline was higher, ratings of exertion and leg-muscle pain were lower, heart rate rose, and end-tidal CO2 and middle-cerebral-artery blood velocity fell. The trial measured one dose, not weeks of training adaptation or sleep.
Links:
- Medicine & Science in Sports & Exercise — Pre-exercise caffeine and repeated sprints
- Medical Xpress — Caffeine sprint effects lasting 3.5 hours
Commentary:
A pre-event caffeine dose about an hour before a long intermittent session can be rational, but 5.5 mg/kg is about 385 mg at 70 kg; do not push the last dose into the evening if you still need to sleep.
III. Sleep & Mental Health
6. Sleep Medicine Reviews umbrella review: Insomnia linked with about 26% higher stroke risk and about 28% higher odds of any hospitalization (Sleep)
Summary:
Luca Vignatelli (IRCCS Institute of Neurological Sciences of Bologna), Stefan Seidel (Medical University of Vienna), and colleagues published an umbrella review in Sleep Medicine Reviews (2026; 90:102372; DOI: 10.1016/j.smrv.2026.102372). The European Academy of Neurology release is dated Sep 27 and ABC News reported it on Sep 28. MEDLINE and EMBASE were searched through August 2025. Existing meta-analyses were combined with three new reviews on stroke, occupational injury, and hospitalization. For stroke, 6 of 9 studies entered the meta-analysis (more than 1.3 million participants): HR 1.26 (95% CI 1.05–1.53), with wide variation between studies. For hospitalization for any reason, 3 of 5 studies were pooled: OR 1.28 (1.14–1.43). Prior reviews more consistently linked insomnia with depression, suicidal behavior, and Alzheimer's disease; all-cause dementia was less consistent, vascular dementia was mixed, and death was not associated. Insomnia definitions differed, and sleep medicines were rarely separated. The European Academy of Neurology funded the work; Idorsia, Jazz Pharmaceuticals, and Takeda gave independent scientific grants that the authors say did not affect design or analysis.
Links:
- EAN / EurekAlert — Insomnia, stroke and hospital admission
- Sleep Medicine Reviews — DOI 10.1016/j.smrv.2026.102372
Commentary:
Persistent insomnia is a reason to get evaluated, not a personal forecast of stroke or dementia; look for accessible cognitive behavioral therapy for insomnia before turning this association into long-term sleeping pills.
IV. Public Health & Disease Prevention
7. EASD: Starting continuous glucose monitoring in type 2 diabetes on basal insulin was linked with about 44% lower one-year all-cause mortality (Public health)
Summary:
Jochen Seufert (University Hospital of Freiburg), with Abbott and collaborating institutions, presented a target-trial emulation at the EASD meeting opening Sep 28, 2026. Using de-identified Truveta records (more than 140 million people in the United States), they matched 13,935 adults who filled a first continuous glucose monitor (CGM) prescription with 13,935 non-users among people on basal insulin, with or without non-insulin diabetes drugs, from Jan 1, 2017, to Mar 18, 2024. Matching used 29 baseline characteristics. Mean age was about 58 and about 47% were women. All-cause death was 1.16% versus 2.09% at 12 months (about 44% lower risk) and 2.12% versus 3.26% at 24 months (about 35% lower). A first macrovascular event was about 26% less common at 12 months (1.03% vs 1.39%) and about 24% less common at 24 months; recurrent macrovascular events were about 35% and 36% lower. Heart-failure hospitalization was about 54% lower at 12 months and 47% lower at 24 months. Directions held after excluding the COVID-19 period and deaths in the first 90 days. The authors note that unmeasured confounding remains possible and that later medication changes were not modeled.
Links:
Commentary:
People already using basal insulin can ask whether a CGM is available to them; this is a matched observational association with a device maker in the analysis, not a promise that wearing a sensor cuts personal death risk by 44%.
8. The Lancet Respiratory Medicine: Tuberculosis preventive treatment fades faster in high-burden settings and can last years where burden is lower (Public health)
Summary:
Lauren R. Linde and colleagues report a systematic review and individual-participant meta-analysis in The Lancet Respiratory Medicine (DOI: 10.1016/S2213-2600(26)00219-5); The Hindu covered it on Sep 28, 2026. From 466,523 participants in 47 studies, they included 84,128 close contacts with tuberculosis infection from 44 cohorts in 24 countries. Preventive treatment was strongly protective in the first year, then declined, with more durable effects in medium- and low-burden settings. Adjusted HRs at one year were 0.57 (95% CI 0.31–1.07; the interval crosses 1) in high-burden settings, 0.14 (0.09–0.21) in medium-burden settings, and 0.21 (0.17–0.26) in low-burden settings. Protection lasted at least about eight years where burden was medium (adjusted HR 0.15, 0.07–0.33) and about 13 years where it was low (0.65, 0.43–1.00). In medium- and low-burden settings, people who completed treatment had lower early risk, but by about two years they were no longer statistically different from non-completers. The authors read the high-burden fade as a reason to pair treatment with measures that cut transmission.
Links:
- The Lancet Respiratory Medicine — Duration of tuberculosis preventive treatment
- PubMed — TPT effectiveness across burden settings
Commentary:
Finish a prescribed course of tuberculosis preventive treatment; in high-burden settings, do not stop early because protection may wane, and household case-finding still matters to limit reinfection.
9. PNAS model: Florida's current measles-vaccine coverage is already below the level that stops sustained spread, and further drops would enlarge outbreaks (Public health)
Summary:
A transmission model by Abba B. Gumel and colleagues at the University of Maryland is published in Proceedings of the National Academy of Sciences on Sep 28, 2026 (DOI: 10.1073/pnas.2610455123), with Medical Xpress the same day. The model was calibrated to the 2025 Gaines County, Texas, outbreak and used Florida demographics, MMR coverage, and age and geographic mixing. Even without removing school-entry requirements, coverage in the model is already below the level needed to prevent sustained transmission. Simulated further declines after removal of school-entry rules increased the potential for large outbreaks, including pediatric deaths on a scale associated with the prevaccine era. The news report characterizes that policy scenario as possibly the largest U.S. measles epidemic in nearly 500 years, but it does not provide a checkable death count from the paper. The same report says 3,471 cases have been confirmed nationally this year across 45 states plus Washington, D.C., and New York City, and a co-author cited about a 650% rise in cases from 2024 to 2025. This is a model, not an observed Florida death toll.
Links:
Commentary:
Two on-schedule doses of MMR remain the main way to stop measles; the model is about what happens when coverage falls, not a lifestyle substitute for vaccination.
V. Evidence & Clarification
10. Mouse study: A phosphocholine sugar from the human gut bacterium Clostridium immunis reduced visceral fat — not a weight-loss probiotic (Evidence clarification)
Summary:
Neeraj Surana's group at Duke University School of Medicine reports in Cell Host & Microbe (online Sep 23, 2026; DOI: 10.1016/j.chom.2026.08.018), with a Duke release on Sep 24 and Sci.News on Sep 28. The newly identified human commensal Clostridium immunis secretes a phosphocholine-modified exopolysaccharide. In obese mice the molecule raised energy expenditure and reduced visceral fat without making the animals eat less or lose muscle. Loss and gain of the phosphocholine biosynthesis locus licABC showed that the chemical tag is required. The sugar lowers IL-22 in the small intestine and visceral fat, increasing metabolic activity specifically in visceral adipose tissue; the effect disappears without IL-22 or group 3 innate lymphoid cells. Sci.News describes a single oral dose cutting visceral fat by roughly one-third within a week, with little change in subcutaneous fat, plus blunted weight gain and better glucose tolerance across 12 weeks of a high-fat diet. Phosphocholine-biosynthesis genes were less abundant in human metagenomes from people with obesity or high triglycerides, which is an association only. The authors say the work does not yet show that the bacterium or its sugar treats obesity in people.
Links:
- Duke University School of Medicine — Commensal sugar and visceral fat in mice
- Cell Host & Microbe — DOI 10.1016/j.chom.2026.08.018
Commentary:
Do not buy a "visceral-fat probiotic" from this paper; it is a mouse mechanism, the human gene finding is only correlational, and visceral fat still moves with diet, activity, and prescribed treatment.
Today's Summary
- Diet evidence today is about patterns, not single products: a Mediterranean-style diet and adult acne is a cross-sectional clue, the red-meat diabetes signal weakens after the meat itself is accounted for, and the microbiome is not a diagnostic test.
- Training doses got more concrete: after 40, bone-density benefit clusters around 2–3 hours a week of brisk walking or jogging plus some strength or balance, while caffeine can stretch a small sprint gain for hours at a dose that also changes brain blood flow and can cost sleep.
- Sleep and chronic-disease care both argue for acting earlier: persistent insomnia tracks with stroke and hospitalization without predicting an individual's fate, and people on basal insulin can discuss continuous glucose monitoring while treating the mortality figure as an industry-involved association.
- Infection prevention is about coverage and time: finish tuberculosis preventive treatment and still reduce reinfection where burden is high; the measles model says school-entry coverage, once it slips, enlarges outbreaks out of proportion.
Daily Framing:
This was a dose-and-boundary day: walking time, red meat, caffeine, and vaccine coverage all have clearer magnitudes, while percentage headlines, microbiome tests, and mouse fat loss are not personal prescriptions.
This digest is compiled from real-time search results and is for reference only.