Exercise for Inflammaging: Which Training Mode Fits Which Marker?

If inflammaging is your target, the exercise mode matters. In older adults, resistance training, aerobic training, and combined programs all lower C-reactive protein (CRP). But the largest meta-analysis in this area found a more specific pattern for two other markers: aerobic training had the clearest signal for tumor necrosis factor-alpha (TNF-α), while combined aerobic-plus-resistance training had the clearest signal for interleukin-6 (IL-6).

This is a useful prescription map, not a promise of disease prevention. CRP, TNF-α, and IL-6 are resting blood biomarkers, not cardiovascular events, disability, or years of life. Exercise changes them by small-to-moderate average amounts over trials that generally lasted 12 to 24 weeks. That is worthwhile, but it does not make exercise a substitute for medical treatment in someone with high cardiovascular risk or active inflammatory disease.

The best current map: CRP for all modes, TNF-α for aerobic, IL-6 for combined

Khalafi et al.'s 2023 systematic review and meta-analysis pooled 49 randomized controlled trials involving 1,898 adults aged 65 and older, including participants with and without chronic disease. Compared with nonexercise control conditions, structured exercise significantly reduced CRP by a standardized mean difference (SMD) of −0.45 (95% CI, −0.61 to −0.29; p<0.001), TNF-α by −0.30 (95% CI, −0.46 to −0.13; p=0.001), and IL-6 by −0.17 (95% CI, −0.32 to −0.02; p=0.02).

Those figures describe relative standardized differences, not an absolute change in milligrams per liter or a personal forecast. By conventional benchmarks, CRP showed the largest pooled effect and the narrowest confidence interval of the three. The analysis supports a real average reduction in these markers; it does not establish how much an individual person's lab result will fall.

The modality subgroups give the result its practical value. CRP fell significantly with resistance training, aerobic training, and combined training. TNF-α reduction emerged preferentially in the aerobic-training subgroup. IL-6 reduction emerged specifically in the combined-training subgroup. Separate CRP effect sizes for each mode were not available, so the evidence does not tell us whether lifting, cardio, or their combination lowers CRP the most.

For the broadest biomarker coverage, combine aerobic and resistance work. If one mode must carry the program, aerobic training has the best evidence for TNF-α, while resistance training has solid support for CRP.

Resistance training has a narrower but useful case

Resistance training deserves more credit than the usual "cardio for inflammation" shorthand gives it. Kim et al.'s 2022 meta-analysis focused on 18 randomized trials of resistance training alone, totaling 539 older adults with a mean age of 70.2 years. It found statistically significant reductions in CRP and IL-6. The median trial enrolled only 30 participants, however, so its subgroup analyses cannot settle fine-grained questions about the best frequency, intensity, or duration.

That result complicates the larger analysis rather than overturning it. Khalafi et al. found that resistance training did not emerge as the significant IL-6-driving subgroup when exercise modes were compared across 49 trials; Kim et al. found an IL-6 reduction after pooling resistance-only trials. Both analyses point toward reduction, but they differ on how confidently clinicians can prescribe resistance training specifically for IL-6.

The sensible reading is cautious: resistance training may lower IL-6, but combined training has the stronger large-sample warrant if IL-6 is the explicit target. For someone already lifting, adding aerobic work is the most evidence-supported adjustment.

Programs using at least 60% of one-repetition maximum and lasting at least 12 weeks were consistently represented in the resistance-training trials. The evidence cannot reliably distinguish two versus three sessions per week. Small trials are poor referees for that kind of contest.

Do not build a program around the IL-10 result

Kim et al. also reported an IL-10 estimate of SMD −3.34 with resistance training. That result is too anomalous for prescription-level use. IL-10 was among the least commonly measured markers, the trial sizes were small, and a pooled effect of that size is roughly eight times the CRP effect reported in the much larger Khalafi analysis.

IL-10 also changes acutely after exercise, while a resting measurement after a training intervention answers a different question. The available evidence does not establish whether timing of blood collection distorted this estimate, but it is a credible source of measurement noise. A dramatic pooled shift in an anti-inflammatory cytokine should not outweigh the more reproducible CRP, TNF-α, and IL-6 findings.

Chronic disease raises the stakes and muddies the comparison

Inflammation often starts higher in people with chronic disease, but that does not produce a simple rule that exercise works better or worse. In Khalafi et al., CRP fell in both healthy participants and those with chronic disease, while IL-6 and TNF-α reductions were more reliably detected in chronic-disease subgroups. By contrast, Bautmans et al.'s 2021 systematic review of 23 recent studies reported that anti-inflammatory effects appeared attenuated in populations with greater disease burden.

These findings assess different things. Higher starting inflammation can make a reduction easier to detect statistically, while disease biology can still limit the proportional response. Neither review resolves whether a person with advanced chronic disease gains more or less than a healthy peer.

Type 2 diabetes studies support combined training in a high-inflammation clinical setting. Yang et al.'s 2023 meta-analysis found that combined exercise reduced CRP and IL-6 in middle-aged and older adults with type 2 diabetes. In a separate randomized trial of middle-aged and older women with type 2 diabetes, Su et al. likewise found lower CRP and IL-6 after aerobic exercise plus resistance training. These studies do not provide a matched healthy older-adult comparison, so they cannot tell us whether the effect is larger or smaller than in healthier populations.

Why combined training is the practical default

Combined training is the best broad-spectrum choice because it covers the three markers most consistently linked to exercise in this literature: CRP, TNF-α, and IL-6. The conclusion comes mainly from subgroup patterns, not a direct head-to-head trial proving that a mixed program beats every single-mode program for every marker.

Other evidence points in the same direction without answering the older-adult question as precisely. Silva et al.'s 2024 systematic review and meta-analysis found that combined training reduced inflammatory markers in sedentary adults, although that population was not restricted to older adults. Nejati Bervanlou et al.'s 2025 review examined exercise plus protein intake and found effects on inflammaging markers, but it could not fully separate the contribution of exercise from protein. That matters because many resistance-training programs in later life include protein supplementation.

Aerobic work may contribute metabolic and cardiovascular adaptations, while resistance work may add muscle-related metabolic effects. That explanation fits the observed subgroup pattern, but the cited analyses did not directly test it as a mechanism. Biomarker movement is evidence of an effect, not proof of the biological route.

What the evidence cannot yet tell you

The strongest modality analysis included only adults aged 65 and older. It cannot confirm that the same cytokine pattern applies to people aged 60 to 64. The reviews also pooled studies with differing CRP assays and blood-draw timing. Acute post-exercise IL-6 and IL-10 behavior is not the same outcome as a resting value after weeks of training.

Most importantly, no study here compares these biomarker effects with statins, anti-inflammatory drugs, or cardiovascular-event outcomes. An SMD of −0.45 for CRP may be biologically relevant, but these trials cannot show that it prevents heart attacks, slows disease progression, or offsets a medication indication.

The most defensible plan is therefore ordinary, not exotic: use a sustainable mix of aerobic and resistance exercise if you want the widest anti-inflammatory signal; choose aerobic training when TNF-α is the particular research target; and use resistance training confidently for CRP, especially when preserving strength and muscle function also matter. Recheck the rest of the clinical picture with a clinician when inflammation, diabetes, cardiovascular disease, kidney disease, cancer treatment, or medication decisions are involved.