Mind-body exercise may help calm the inflammatory biology tied to depression, anxiety, schizophrenia, and Parkinson's disease. The catch is that the answer changes with the biomarker. A 2026 dose-response network meta-analysis by Zheng et al. compared Tai Chi, qigong, mindfulness-based stress reduction (MBSR), and yoga across inflammatory and neurotrophic markers in neuropsychiatric disorders. That is new and useful. But a 2024 randomized controlled trial by Lindsay et al. found that MBSR reduced proinflammatory gene regulation without lowering circulating CRP or IL-6 protein levels. If one trial says the genes move and the blood proteins do not, then "anti-inflammatory" is not one simple bucket.
A lot of this literature relies on protein assays such as ELISA for CRP, IL-6, TNF-alpha, and related cytokines. Those are real measures, but they may miss earlier or upstream effects. So the live question is not whether mind-body practices touch inflammation at all. The better question is what level of biology they change, how fast, and whether different practices work through different routes.
What the 2026 meta-analysis adds, and what it cannot yet settle
The Zheng et al. paper is the first systematic review and model-based Bayesian dose-response network meta-analysis to compare Tai Chi, qigong, MBSR, and yoga on TNF-alpha, IL-6, IL-1beta, CRP, IL-10, and BDNF in people with neuropsychiatric disorders. The biological rationale is familiar. Depression and schizophrenia often track with higher IL-6, IL-1beta, and TNF-alpha, while IL-10 and BDNF can act as counterweights. The statistical approach also fits the problem. There are no head-to-head randomized trials that directly compare all these modalities in the same diagnostic group, so a network meta-analysis is one of the few ways to estimate relative rankings.
Still, this paper has a major limit for anyone looking for a neat prescription. The available publication details confirm the framework and methods, but the retrievable text supplied here does not provide the quantitative results: no verified dose thresholds, no trustworthy modality ranking, and no reproducible effect sizes for each biomarker. That rules out strong claims about how many weekly hours of yoga or Tai Chi produce a given drop in IL-6. Any article that claims otherwise is doing fan fiction with cytokines.
The paper also pools a broad set of conditions, including depression, schizophrenia, Alzheimer's disease, and Parkinson's disease. That increases statistical reach, but it also mixes populations with different inflammatory drivers. Stress-linked immune dysregulation in major depression is not the same process as synuclein-linked microglial activation in Parkinson's disease. A ranking that looks good across the whole network may not travel cleanly from one diagnosis to another.
The strongest mechanistic clue comes from MBSR, and it complicates the story
The most precise mechanistic evidence in this set comes from Lindsay et al. in 2024. In that randomized controlled trial, older adults completed a standard 8-week MBSR program or a control condition. The investigators measured both proinflammatory gene expression and systemic inflammatory proteins. MBSR significantly reduced proinflammatory gene regulation, including NF-kappaB-related signaling, but did not reduce circulating CRP or IL-6 at the protein level.
This is not a trivial null result. It suggests that MBSR can shift inflammatory regulation upstream, at the level of transcription, before those changes show up in standard blood protein panels, or without producing large enough downstream protein changes during a typical 8-week trial. That is a different claim from "MBSR lowers IL-6 in the blood." The study does not support that stronger statement. It supports a narrower one: MBSR altered proinflammatory gene regulation in a randomized trial.
Once you see that split, some of the broader literature gets harder to read casually. If MBSR mainly acts through transcriptional regulation and stress-system normalization, while more movement-heavy practices also recruit exertion-related anti-inflammatory pathways, then a pooled cytokine network may be mixing different biological routes into one outcome class. That can blur both dose-response curves and modality rankings.
Stress biology is part of the mechanism
A second line of evidence points to the hypothalamic-pituitary-adrenal axis rather than direct cytokine suppression. In a 2022 study by Saban et al., an 8-week MBSR program in women veterans measured psychological well-being, cortisol, and inflammatory markers. The study found cortisol normalization alongside changes in inflammatory pathways. That pattern fits a plausible model: chronic stress pushes the HPA axis out of tune, immune cells become less responsive to glucocorticoid signaling, and inflammatory tone stays elevated. MBSR may partly work by reducing that stress load and restoring regulation.
This does not prove that every downstream inflammatory protein will fall on schedule in every population. It does show that mind-body practice can act on systems upstream of the cytokine panel. That distinction matters if you are using a single CRP result as judge, jury, and executioner. Biology rarely works that way.
Movement-based practices may add something different
The case for physical mind-body practices does not rest on MBSR alone. A 2022 systematic review and meta-analysis by Ma et al. examined randomized controlled trials of physical activity in people with mild cognitive impairment. Mild cognitive impairment is not a primary neuropsychiatric disorder, so this is supporting context rather than direct proof for depression or schizophrenia. But it still matters. The review found that physical activity interventions modulated blood inflammatory cytokines and neuroprotective factors in a cognitively vulnerable population.
That strengthens the biological case that movement itself can shift neuroimmune signaling. It also raises a reasonable mechanistic distinction. Tai Chi, qigong, and yoga are not just seated attention training with nicer pants. They add physical work, balance, respiratory pacing, and probably some autonomic effects that do not map perfectly onto MBSR. If those movement-based practices change circulating proteins more readily than MBSR does, the difference may reflect biology rather than trial noise.
Parkinson's disease is relevant, but it is not a shortcut to all mental disorders
For Parkinson's disease, the best direct evidence in this set is a 2025 randomized clinical trial by Kwok et al. on meditation and yoga. The trial measured anxiety, depression, and chronic inflammation in patients with Parkinson's disease. That makes it clinically interesting because Parkinson's combines neurodegeneration with common psychiatric symptoms, and inflammation is part of the disease biology.
The cautious reading is the right one. This trial supports the idea that meditation and yoga may help at the intersection of mood symptoms and chronic inflammation in Parkinson's disease. It does not mean the same intervention will have the same inflammatory effects in major depression, generalized anxiety disorder, or schizophrenia. Parkinson's disease has a distinct neuroinflammatory substrate, including microglial activation and synuclein pathology. Useful evidence, yes. No universal template.
Where the evidence agrees, and where it still falls apart
Across these studies, three points hold up reasonably well. First, chronic neuroinflammation is part of the biology of several neuropsychiatric disorders, with repeated attention to IL-6, IL-1beta, TNF-alpha, BDNF, and related markers in the literature summarized by Zheng et al.. Second, mind-body practices can modulate stress biology and inflammatory signaling. The Lindsay trial supports a gene-regulatory effect, and the Saban study supports HPA-axis involvement. Third, movement-based interventions can alter inflammatory and neuroprotective markers in related clinical populations, as summarized by Ma et al..
The disagreements are just as important. We do not yet know whether reductions in proinflammatory gene expression are mechanistically equivalent to reductions in circulating cytokine proteins. We do not know whether one modality reliably beats the others across diagnoses, because the direct head-to-head randomized trials do not exist in the evidence provided here. We also do not know the optimal dose. The 2026 network meta-analysis was designed to answer that question, but the accessible text does not provide the numbers needed to make a credible dosing recommendation.
Trial design adds more uncertainty. Participants usually know what intervention they got, follow-up is often short, and biomarker timing varies. An 8-week program may be enough to shift transcription and cortisol while still being too short to produce a clean change in CRP or IL-6. If so, a "null" protein result may mean "not yet measurable here," not "nothing happened."
The practical verdict
The cleanest conclusion is narrower than the marketing copy you will see elsewhere. Mind-body practices appear biologically active in neuropsychiatric and neurologically adjacent populations, and MBSR has randomized evidence for reducing proinflammatory gene regulation even when CRP and IL-6 protein levels do not budge over the same window. Yoga and meditation also have randomized trial support in Parkinson's disease for anxiety, depression, and chronic inflammation outcomes. What the evidence does not yet justify is a precise ranking of Tai Chi versus qigong versus yoga versus MBSR, or a confident dose target for each biomarker across disorders.
For readers deciding what to do now, the safest evidence-based advice is boring in the best way. Pick a structured program you can keep doing for at least 8 weeks. Treat consistency as more reliable than modality tribalism. If you are expecting a dramatic drop in CRP after a short mindfulness program, the current evidence says that benchmark may be too crude or too early. The biology may move first at the level of gene regulation and stress signaling, long before a routine blood panel reflects it.