Can Chinese Herbs Rewire the Cell's Power Grid? A New Review Says the Evidence Isn't There Yet
Mitochondria are often called the cell's powerhouses, and the metaphor has done a lot of quiet damage. A powerhouse is a single building; mitochondria are better imagined as a network — hundreds of small generators wired together, trading calcium, lipids, and stress signals with the rest of the cell and even with neighbouring cells. When a heart attack or a stroke cuts off oxygen, that network is what fails first, and it is also what reperfusion — the restoration of blood flow that saves the tissue — damages next, in a burst of oxidative stress.
It is therefore no surprise that traditional Chinese medicine (TCM) researchers have spent years reporting that herbal formulas and plant compounds protect mitochondria after heart attack and stroke. The surprise, in a critical review published in Frontiers in Pharmacology in November 2026, is how little of that literature tests what it claims to test.
The review, by researchers at Henan University of Chinese Medicine, screened 1,663 records and kept 29 high-confidence reports. Its question was narrower and harder than the usual one. It did not ask whether herbal interventions change mitochondrial markers. It asked whether they restore communication — the exchange of signals between mitochondria themselves, between mitochondria and other organelles, and between cells. Changing a marker is easy. Restoring a network is a different claim, and it needs different evidence.
A Network, Not a Powerhouse
The authors organise the evidence around four levels of what they call the mitochondrial communication network. Inside a single mitochondrion, enzymes and energy production must stay coordinated. Between mitochondria, the organelles fuse, split, and pass material to one another. Between mitochondria and other organelles — the endoplasmic reticulum, lysosomes, lipid droplets, the nucleus — membrane contact sites shuttle calcium and lipids. And between cells, mitochondria and their fragments can be transferred or cleared away by immune cells.
That framework matters because most published studies operate at the easiest level: they measure a static protein or a downstream injury marker and conclude that "mitochondrial function" improved. A fall in a stress marker tells you the tissue is less injured, not that communication across the network was repaired.
Within TCM theory, heart attack and stroke are discussed in terms of patterns such as qi deficiency, blood stasis, phlegm, and heat, and the classical strategies are to supplement qi, activate blood, and open the collaterals. The review is careful on this point: pattern concepts should not be translated into one-to-one molecular equivalents. Qi deficiency is not synonymous with ATP depletion, and blood stasis cannot be reduced to a single clotting marker. The honest translational question is whether chemically defined herbal interventions reproducibly change measurable biological processes relevant to the traditional therapeutic context.
What the Evidence Actually Shows
Across the 29 reports, findings clustered around calcium and redox balance, energy metabolism, mitochondrial dynamics, mitophagy (the clearance of damaged mitochondria), and immune-metabolic responses. The strongest support was for individual compounds:
- Salvianolic acid B, from Danshen (Salvia miltiorrhiza), reduced myocardial damage and suppressed activation of the NLRP3 inflammasome in animal models, apparently through SIRT1-AMPK-PGC-1α signalling. But no study showed it binding directly to a mitochondrial calcium transporter or remodelling a contact site.
- Astragaloside IV, from Huangqi (Astragalus), improved mitochondrial shape and metabolism and reduced calcium-related cell death — yet a fall in cytosolic calcium cannot by itself prove that mitochondrial calcium transport was preserved.
- Ginsenosides Rb1, Rg1, and compound K, from ginseng and notoginseng, modulated mitochondrial fusion proteins and energy metabolism in stroke models. The Rb1 evidence for astrocytes delivering mitochondria to neurons rested heavily on a fluorescent dye, MitoTracker — and dyes can cross cell boundaries on their own, so a glowing signal is not proof that an intact organelle was transferred.
- Berberine, puerarin, baicalin, and curcumin influenced general autophagy markers without establishing whether the effect was restored communication or simply a downstream response to less tissue death.
The multi-herb formulas fared no better under the stricter lens. Buyang Huanwu Decoction — the classical post-stroke formula built around Huangqi — showed experimental links to mitochondrial quality control and axonal mitochondrial recruitment. Compound Danshen Dripping Pills and Qishen Yiqi Dripping Pills showed favourable changes in lipid metabolism, ferroptosis, and mitochondrial biogenesis in rodent models. But the review is blunt about synergy: the traditional idea that herbs work together under the Jun-Chen-Zuo-Shi (sovereign-minister-assistant-courier) hierarchy cannot be equated with verified pharmacological synergy. That requires authenticated ingredients, decomposition and add-back experiments, and quantitative interaction modelling — none of which these studies provided.
The Clinical Evidence Is Disconnected
The gap widens in human studies. Systematic reviews of Buyang Huanwu Decoction during stroke rehabilitation reported possible improvements in neurological function and daily living, but the underlying trials carried high risks of bias, variable herbal compositions, small samples, and inconsistent reporting of TCM patterns. Trials of Compound Danshen Dripping Pills after emergency coronary intervention and of Qishen Yiqi Dripping Pills in chronic ischemic heart failure (the CACT-IHF trial) reported functional benefits. Yet none of these trials measured a single mitochondrial biomarker — no circulating mitochondrial DNA, no acylcarnitine profile, no immune-cell respiration. Without such measures, no one can say whether any benefit came through mitochondria.
The preclinical literature has its own structural problems, and the review names them plainly. A systematic appraisal of 45 Buyang Huanwu studies found widespread failures on standard reporting and bias checklists: allocation concealment was rare, outcome assessment was usually unblinded, and sample-size calculations were mostly absent. Most animal experiments used young, healthy, male rodents given the herb before the induced stroke or heart attack — the opposite of clinical reality, where an elderly patient arrives hours after the vessel closes. And the doses used in cell experiments often exceed anything achievable in human blood, while compounds like berberine can inhibit cytochrome P450 enzymes, risking unmonitored interactions with standard antiplatelet or anticoagulant drugs.
What Would Count as Proof?
The review does not conclude that the idea is wrong — only that mitochondrial communication is, for now, a useful framework for generating hypotheses, not an established clinical mechanism of TCM. It lays out what would have to change:
- Authenticated, characterised preparations. The botanical identity and chemical fingerprint of what was tested, so that one laboratory's "Danshen" is the same as another's.
- Clinically achievable doses, given after the injury rather than before it, in animals that resemble real patients.
- Dynamic, communication-specific assays — sensors that report contact-site activity in living cells, or flux measurements — instead of static protein levels.
- Causal tests. Genetic rescue experiments that switch a proposed mediator off and on, and show the effect following it both ways.
- Human trials with mitochondrial biomarkers built in from the start, so that a functional improvement can actually be attributed to a mitochondrial mechanism.
Readers of this site will recognise step four. It is exactly the design that made the fatty-liver formula study we covered earlier worth reading: the researchers switched off the gene the formula was supposed to work through, watched the benefit vanish, then switched the downstream gene back on and watched it return. That is the standard this review is asking the mitochondria literature to meet — and, for now, the literature does not meet it.
The Takeaway
When a field produces hundreds of positive papers, it is tempting to assume the weight of numbers settles the question. It does not. A hundred studies measuring the same downstream marker are still one kind of evidence, and the kind that matters — direct, causal, communication-specific, measured at a dose a patient could actually receive — is mostly missing.
That is not a reason to dismiss the field. The four-level framework is genuinely useful: it tells the next researcher exactly which experiment to run, and the reader exactly which question to ask. When you next read that a herb "protects mitochondria" after a heart attack or stroke, ask which level of the network was actually measured, whether the dose was one a human could reach, and whether anyone switched the mechanism off and put it back. If the answers are a static protein, a preventive dose in a young healthy rat, and no, you are looking at a hypothesis, not a finding — and a clearly labelled hypothesis is exactly where good science begins.
Source: Wan Z, Wang X, Xu Z, Ji R, Wan H. "Traditional Chinese medicine interventions and mitochondrial communication in myocardial and cerebral ischemia: a critical review." Frontiers in Pharmacology. 2026;17:1992115. DOI: 10.3389/fphar.2026.1992115.
If you are interested in how the classical Chinese medical tradition thinks about health more broadly — its ideas of balance, prevention, and the body's patterns — our book Huangdi Neijing: A Modern Introduction offers a beginner-friendly guide to the foundational text of TCM.
This article is for educational purposes only and is not medical advice. If you are concerned about heart disease or stroke, please consult a qualified healthcare professional.