Exercise, Protein, and Inflammaging: What the Evidence Shows

Exercise plus protein may lower several inflammatory biomarkers in older adults, but the evidence does not support every proposed explanation. A 2025 meta-analysis of randomized trials found that the combination performed better than exercise alone for C-reactive protein (CRP) and tumor necrosis factor-alpha (TNF-α), while exercise alone also outperformed protein alone for interleukin-6 (IL-6). The gut-barrier theory, however, failed its direct test in active older adults. The more credible working model involves repeated muscle-derived anti-inflammatory signals and changes in aging T-cell populations.

That distinction matters. Lower CRP, IL-6, and TNF-α are biomarker findings, not proof that a program prevents disease or extends life. The available evidence also cannot tell us how large the average changes were, because the primary meta-analysis abstract reports p-values rather than pooled effect sizes and confidence intervals. Still, the pattern gives older adults a practical reason to pair regular training with adequate protein instead of treating either as a standalone inflammation supplement.

The combination produced the most consistent cytokine signal

Nejati Bervanlou et al.'s 2025 meta-analysis searched randomized controlled trials indexed in PubMed, Scopus, Cochrane, and SPORTDiscus through January 2024. It compared exercise, protein supplementation, and combined exercise-plus-protein interventions in the context of inflammaging.

For CRP, combined exercise and protein outperformed exercise alone (P<0.00001), and pooled experimental subgroups outperformed controls (P=0.0002). For IL-6, exercise alone beat protein alone (P=0.001), as did the combined intervention (P=0.05); pooled experimental arms also outperformed protein-only arms (P=0.0002). The TNF-α result drew the sharpest line: only the exercise-plus-protein group showed a significant reduction versus exercise alone (P<0.00001). Protein alone was ineffective, and exercise alone did not significantly suppress TNF-α in that comparison.

This looks like a combination effect, especially for TNF-α. It does not prove biological synergy in the strict statistical sense, because the meta-analysis did not report a formal interaction test. Nor can readers judge clinical importance from p-values alone. A statistically significant shift in a blood marker may be small, and the abstract does not provide standardized mean differences, 95% confidence intervals, or a way to compare changes with a clinically meaningful threshold.

Interventions also differed across the pooled trials: exercise modality, intensity, duration, protein dose, timing, and protein source were not uniform. The result supports the general pairing of exercise and protein; it does not establish one ideal shake, dose, or schedule. A broader 2026 meta-analytic compendium on exercise and aging is consistent with anti-inflammatory adaptation, although the available summary does not provide effect-size detail for this question.

Muscle sends an anti-inflammatory signal during exercise

IL-6 can confuse this story because chronically elevated resting IL-6 is part of the inflammaging pattern. During exercise, contracting muscle also releases IL-6 as a myokine. The two situations differ in timing and biological context. Acute exercise-induced IL-6 can suppress TNF-α and IL-1β while stimulating anti-inflammatory mediators including IL-1 receptor antagonist and IL-10.

Kistner et al. (2024) found a dose-response relationship between the duration of moderate-intensity walking and acute muscle-derived IL-6 secretion. The study gives the mechanism a real-world anchor: more time spent walking at moderate intensity produced more acute IL-6 release. It did not specifically study people aged 50 or older, so applying its myokine kinetics to inflammaging remains an inference.

A 2025 review by Garg et al. describes other exercise-induced myokines, including irisin, IL-15, and CXCL1, as part of broader systemic immune signaling. Reviews can organize plausible biology, but they do not replace intervention trials in older adults. The reasonable conclusion is narrower: repeated bouts of muscle contraction offer a plausible route to longer-term anti-inflammatory adaptation. They do not establish that any one myokine caused the cytokine changes in the meta-analysis.

Training may also reshape an aging immune system

Inflammaging involves more than circulating cytokines. With age, naïve T-cell compartments contract while dysfunctional, pro-inflammatory T-cell subsets can accumulate. These include terminally differentiated effector memory RA-positive cells, usually called T-EMRA cells, and CD28-null T cells.

In a 2021 randomized trial by Despeghel et al., 40 previously inactive older adults with a mean age of 70.3 years completed six weeks of low-dose combined resistance and endurance training twice weekly. The intervention produced measurable shifts in T-cell subpopulations, including the CD4+/CD8+ ratio and T-EMRA populations. This is encouraging because the program was modest and short. It was also small: 30 participants received training and 10 served as controls after a pandemic-driven 3:1 allocation. That design limits confidence in subgroup and secondary-endpoint estimates, and the study did not show whether the changes persisted after training stopped.

Intensity appears to matter. Niemiro et al. (2022) randomized older women at high risk of breast cancer to moderate- or high-intensity aerobic training. Moderate intensity reduced T-cell phenotypes associated with immunosenescence; high intensity did not. This does not mean hard exercise harms immunity, but it rejects a simple more-intensity-is-better rule for this endpoint. The participants' breast-cancer risk also limits generalization to healthy older adults and men.

The pattern extends beyond aerobic exercise. A 2022 systematic review by Salimans et al. concluded that resistance exercise can modulate immune-cell subsets in humans. A separate systematic review and meta-analysis by Sardeli et al. found that acute exercise mobilizes and reduces circulating senescent lymphocyte counts. Acute reductions could reflect redistribution rather than permanent clearance, so they should not be treated as proof that exercise deletes senescent cells. Across repeated sessions, though, these findings fit the longer-term T-cell changes seen in training trials.

The gut-barrier explanation did not hold up in its direct test

The gut-barrier hypothesis is appealing: exercise and protein might improve intestinal integrity, reduce bacterial endotoxin entering circulation, and thereby reduce inflammation. But the one direct randomized test in this evidence base did not support that sequence in active older adults.

Huschtscha et al. (2022) enrolled 32 active adults with a mean age of 62 years who already exercised 211±91 minutes per week. Participants were assigned to dairy milk, exercise plus dairy milk, exercise alone, or control. Over 12 weeks, the exercise groups performed whole-body progressive resistance training three times weekly, and the milk intervention provided 30 g/day of high-protein dairy milk. Neither training nor the protein-plus-training intervention significantly changed lipopolysaccharide-binding protein (LBP) or soluble CD14 (sCD14), indirect markers of endotoxin translocation.

This was a prospectively designed mechanistic test, so the null result deserves weight. It does not rule out every gut-related effect. LBP and sCD14 are downstream, indirect biomarkers; the trial did not measure intestinal fatty acid-binding protein, zonulin, fecal calprotectin, or other direct indicators of epithelial injury or permeability. The marker analysis also had roughly six to seven participants per group after four incomplete samples were excluded. A small study can miss a real effect.

Even so, the result does not justify calling the intestinal epithelium the upstream gatekeeper for the exercise-protein cytokine findings. Participants were already active, so their baseline permeability or endotoxin exposure may have been low, but the study did not measure either. Gut-barrier benefits in sedentary people starting exercise remain a plausible question, not an established mechanism.

What to do with this evidence

The best-supported strategy is pleasantly unglamorous: perform regular combined resistance and endurance exercise at a sustainable moderate intensity, and meet protein needs consistently. The studies do not establish a universal protein prescription. The most specific protocol here used 30 g/day of high-protein dairy milk alongside resistance training, but no head-to-head trial in this evidence base shows whether dairy, whey, casein, or plant protein produces superior inflammatory outcomes in adults over 50.

• Use moderate effort as the default for immune-focused training; the available randomized comparison found benefits with moderate, not high, intensity for immunosenescence-associated T cells.
• Include resistance work. Human evidence links it to immune-cell modulation, and combined resistance-endurance training shifted T-cell subsets within six weeks in previously inactive older adults.
• Pair training with adequate protein, while recognizing that the evidence supports the combination more clearly than a precise product or dose.
• Do not use gut-barrier claims as the main sales pitch. In active older adults, 12 weeks of resistance training with or without dairy protein did not change the measured endotoxin-translocation biomarkers.

The evidence supports exercise plus protein as a sensible anti-inflammaging strategy, with cytokine and immune-cell biomarkers as the measured outcomes. It has not yet established the size of the benefit, the best protein source, or a gut-barrier mechanism in active older adults.