Highly active adults had 22% lower fatty liver risk, with gut microbe shifts

If you have been told there is fat in your liver, or you suspect it, the advice is nearly always the same: move more. What is less clear is whether your gut bacteria are part of how movement helps. A cohort study from Anhui Medical University in China, published in Annals of Hepatology, followed adults from 2020 to 2025 to connect physical activity, gut microbes, a blood metabolite and new cases of fatty liver.

Key Takeaways

• The most active third of adults had a 22% lower risk of new steatotic liver disease (fatty liver) than the least active third; the third who sat most had a 26% higher risk.
• In a 245-person subset, Proteus gut bacteria rose over a year in low-activity people and fell in high-activity people, and higher Proteus tracked with more fatty liver.
• Blood L-valine was lower in active people and higher in people with fatty liver.
• These are associations with borderline confidence intervals and self-reported activity, not advice on what to take, avoid or test.

Physical activity and steatotic liver disease: 22% lower risk in the most active

Adults in the top third of moderate-to-vigorous physical activity (MVPA, movement at a moderate pace or harder) had a 22% lower risk of developing steatotic liver disease (SLD, the umbrella term for fat building up in the liver) than adults in the bottom third (hazard ratio, HR, where below 1 means fewer new cases: 0.78, 95% CI 0.63 to 0.98). Sitting went the other way: the top third for sedentary time had a 26% higher risk (HR 1.26, 95% CI 1.00 to 1.58). When the researchers combined both habits, only people who sat less and moved more were clearly lower than the reference group of long sitters with low activity (HR 0.74, 95% CI 0.57 to 0.95). Moving more but still sitting a lot showed no clear difference (HR 0.99, 95% CI 0.77 to 1.27).

Fatty liver is common and often silent

Steatotic liver disease affects over 30% of adults worldwide, the authors cite, with projections of 55% by 2040. A 2023 consensus statement replaced the older term NAFLD with MASLD (metabolic dysfunction-associated steatotic liver disease); 460 of the 565 new cases here were MASLD. The most common form usually causes few or no symptoms, so many people learn of it from a scan or blood test. The authors note it raises risk of cardiovascular disease and of liver fibrosis, cirrhosis and liver cancer.

How the study tracked 1,417 adults free of fatty liver

The main analysis followed 1,417 adults in Anhui Province, judged free of fatty liver at the start (from 2,942 recruited in 2020), until January 2025, by which point 565 had been newly diagnosed. Participants reported their activity on a validated questionnaire, and new fatty liver was detected with a FibroScan (an ultrasound-based scan that estimates liver fat). For the gut work, 245 adults wore a wrist accelerometer (motion sensor) for three days and gave stool samples twice, a year apart, for 16S rRNA sequencing (which identifies bacterial genera). A further 179 had shotgun metagenomics (whole-DNA sequencing that resolves species), and blood was screened for 121 metabolites. Models adjusted for age, sex, diet quality, smoking, alcohol, diabetes and hypertension, among other factors.

Proteus bacteria and L-valine: the gut and blood clues

Of 74 gut bacterial genera analyzed, 12 changed over the year in relation to activity level, and Proteus stood out. Proteus and Synergistes rose over a year in low-activity people and fell in high-activity people, while Faecalibacterium and Oscillospira increased in the high-activity group. Higher Proteus tracked with 32% higher odds of fatty liver per standard-deviation increase (OR 1.32, 95% CI 1.01 to 1.72), and shotgun sequencing found one of six Proteus species, Proteus cibarius, clearly linked (OR 1.72, 95% CI 1.19 to 2.54). Synergistes showed no clear link (OR 0.85, 95% CI 0.65 to 1.11). The authors note most Proteus species make lipopolysaccharide, an inflammatory molecule, and suggest that could feed liver disease, a step this study did not test.

In blood, L-valine, a branched-chain amino acid, was the only one of 121 metabolites that stayed lower with more activity after multiple-testing correction (beta -0.23, FDR 0.031). Higher valine went with higher odds of fatty liver (OR 1.78, 95% CI 1.22 to 2.63) and correlated weakly with Synergistes (partial r 0.13, p = 0.048).

What this could mean if you are worried about fatty liver

The study tests no treatment, so it cannot say what to do about a gut bacterium or a blood amino acid. It adds another data point that steady moderate movement tracks with less fatty liver and long sitting tracks the other way, which fits lifestyle advice already in guidelines. One detail matters for how you think about exercise: on-the-job activity such as housework, farming and carrying loads was linked to lower risk (HR 0.94 per standard deviation, 95% CI 0.91 to 0.97), while leisure-time exercise was not (HR 1.04, 95% CI 0.99 to 1.08). The authors suggest this reflects how much of this population's activity was work-based. Fatty liver is often silent, so it is reasonable to ask a doctor whether a liver check makes sense for you. Nothing here shows that a Proteus stool test, a probiotic or changing valine intake helps the liver.

Limits: observational data, self-reported activity, a small microbiome subset

This is an observational study, so it cannot show that sitting causes fatty liver or that Proteus does. Activity came from a questionnaire whose validity coefficient against activity diaries was 0.66 for MVPA. The gut analysis rested on 245 people, and the bacteria and valine findings were odds of having fatty liver, which cannot show which came first. Several intervals are borderline: the MVPA interval reaches 0.98, the sitting interval starts at 1.00, and Proteus starts at 1.01. Participants came from central-eastern China, and only 31 MetALD (fatty liver with heavier alcohol use) and 21 alcohol-related cases occurred, so chance findings there cannot be ruled out. The authors call the multi-omics results hypothesis-generating.

What comes next for activity and gut microbes

The authors call for larger cohorts in more diverse populations, plus intervention studies that test whether changing activity shifts these microbes and metabolites. They also want deeper methods, such as metatranscriptomics (which microbial genes are switched on) and untargeted metabolomics. The paper gives no timeline.

Common Questions

Does sitting cause fatty liver? Not shown. Longer sitting tracked with a 26% higher risk, but the interval starts at 1.00 and observational data cannot prove cause.

Is Proteus the culprit? Not shown. Higher Proteus tracked with fatty liver in 245 people, and one species, Proteus cibarius, stood out, but the study cannot say the bacteria caused the disease.

Does exercise change gut bacteria? Other work points the same way. A cross-sectional study of 8,416 Swedish adults linked sedentary time and moderate or vigorous activity to species abundance in opposite directions, and an 8.6-month randomized trial in people with NAFLD and prediabetes linked exercise and diet to more diversified keystone gut bacteria.

References

1. Shaoxian Liang, Yaling Shao, Hairui Qin, et al. (2026). A cohort study of physical activity, gut microbiota, serum metabolites, and the risk of steatotic liver disease. Annals of Hepatology. https://doi.org/10.1016/j.aohep.2026.102443
2. Kim Sui Wan, Wah-Kheong Chan, Muhammad Fadhli Mohd Yusoff, et al. (2024). Correspondence: “A multisociety Delphi consensus statement on new fatty liver disease nomenclature”. Annals of Hepatology. https://doi.org/10.1016/j.aohep.2024.101485
3. Symptoms & Causes of NAFLD & NASH – NIDDK. NIH. https://www.niddk.nih.gov/health-information/liver-disease/nafld-nash/symptoms-causes
4. Gabriel Baldanzi, Sergi Sayols-Baixeras, Elin Ekblom-Bak, et al. (2024). Accelerometer-based physical activity is associated with the gut microbiota in 8416 individuals in SCAPIS. eBioMedicine. https://doi.org/10.1016/j.ebiom.2024.104989
5. Runtan Cheng, Lu Wang, Shenglong Le, et al. (2022). A randomized controlled trial for response of microbiome network to exercise and diet intervention in patients with nonalcoholic fatty liver disease. Nature Communications. https://doi.org/10.1038/s41467-022-29968-0