Cat’s Whiskers: Metabolic, Blood Pressure and Other Research

Away from the kidney, cat’s whiskers has acquired a second reputation — for blood pressure, blood sugar, uric acid and inflammation. That literature is much larger than the urinary one, much more modern, and much less useful than its size suggests, for a reason that is easy to state and easy to miss: almost none of it studied the tea. It studied purified flavones in an organ bath, ethanolic extracts in rodents, and, in the two human blood-pressure trials, a multi-ingredient supplement in which Orthosiphon was one component among four.

This page sorts that file. It also lands on a finding that says something uncomfortable about how the herb is sold: the single best piece of human evidence this plant has is for an indication nobody buys it for, at a dose no cup of tea approaches, using a preparation standardised to a different compound from the one on the pharmacopoeial label. That is worth ranking at the top rather than burying, because it is true and because the popular narrative gets it backwards.

The Shape of the Research File
Blood Pressure: Two Human Trials, Both Combinations
Who Gets the Credit for Three Millimetres?
The Vasodilation Mechanism and Its Solvent Problem
The Best Single-Herb Trial Is About Something Else
Standardised to What?
Blood Sugar and Metabolic Claims
Uric Acid and Gout
Anti-Inflammatory and Antioxidant Work
Reading Preclinical Work Without Being Fooled
What Would Actually Change the Picture
Evidence Ledger for This Page
Key Research Papers
Connections

The Shape of the Research File

Three structural features explain most of what follows.

It is concentrated by geography. An unusually large share of the modern pharmacology comes from Malaysian universities and institutes, which have invested heavily in standardising misai kucing extracts for an export market. That is not a criticism — it is careful work, and without it this plant would have almost no modern file at all. But a literature produced largely by one country’s research programme around one commercial crop is a literature with a direction of travel, and independent replication elsewhere is thin.

It is split across names. Papers appear under Orthosiphon stamineus (the older synonym, and still the commonest in pharmacology), Orthosiphon aristatus (the currently accepted name, used by the European monograph), Orthosiphon grandiflorus (the Thai stone trial), Orthosiphon spicatus and Clerodendranthus spicatus (Chinese and Japanese sources). Searching one name silently misses most of the field. Every study cited on this page has its binomial named for that reason.

It studied extracts, not the drink. The traditional and licensed preparation is a hot-water infusion. The pharmacology overwhelmingly used ethanolic or hydroalcoholic extracts, or purified single compounds. Those are different interventions, and for this plant the difference is chemically consequential: the marker flavones are polymethoxylated, which raises lipophilicity and lowers water solubility, so they extract poorly into hot water and well into ethanol. This point is developed on the diuretic page; it applies to every mechanism below.

Blood Pressure: Two Human Trials, Both Combinations

Evidence tier: randomised human trials — of multi-ingredient products.

Two randomised trials from Bologna are the best human cardiovascular data this plant appears in.

The first is reported as Effect of a combined nutraceutical containing Orthosiphon stamineus on blood pressure and metabolic syndrome components in hypertensive dyslipidaemic patients: a randomized clinical trial, in Complementary Therapies in Clinical Practice, 2012 — PubMed search. Participants were hypertensive and dyslipidaemic and already taking calcium-channel blockers or ACE inhibitors. The intervention was a combined nutraceutical of which orthosiphon leaf extract was one ingredient. The word “combined” is in the title.

The second is the CONDOR study: Effect of dietary supplementation with Diuripres on blood pressure, vascular health, and metabolic parameters in individuals with high-normal blood pressure or stage I hypertension, in Phytotherapy Research, 2023 — PubMed search. Sixty people, a four-week run-in on a sodium-restricted Mediterranean-style diet, then eight weeks of a supplement containing magnesium, standardised orthosiphon, hawthorn and hibiscus, or placebo. Reported against placebo at eight weeks: systolic pressure down 3.4 ± 0.9 mmHg, diastolic down 3.8 ± 0.7 mmHg, central aortic pressure down 5.2 ± 1.0 mmHg, with a small fall in high-sensitivity C-reactive protein.

These are real trials, properly randomised and placebo-controlled, and the results are internally consistent. They are also, as evidence about cat’s whiskers, close to uninterpretable — which the next section explains rather than merely asserts.

Who Gets the Credit for Three Millimetres?

Two separate problems sit on top of the CONDOR result, and both must be cleared before anyone can attribute it to this herb.

Problem one: three of the four ingredients have their own blood-pressure literature.

Hibiscus has been tested repeatedly as a single agent for blood pressure, with meta-analyses reporting reductions of a similar order to CONDOR’s entire result. See hibiscus and blood pressure and PubMed: hibiscus and blood pressure, meta-analysis.
Magnesium supplementation has its own meta-analytic literature on blood pressure — PubMed search. See also Magnesium.
Hawthorn has a long cardiovascular literature of its own — see Hawthorn.

So the one ingredient with no single-agent blood-pressure trial is orthosiphon, and it is the one being credited on supplement labels. This is textbook formula substitution, and it is worse than the usual case: normally a formula’s components are unstudied and the attribution is merely unfounded. Here the other components are studied and plausibly sufficient to explain the whole effect. Citing CONDOR as evidence that java tea lowers blood pressure is not a stretch — it is a misattribution with a named alternative culprit.

Problem two: the effect size deserves calibration, not excitement. A 3–4 mmHg fall in systolic pressure is genuine and, spread across a population, not trivial. For an individual it is roughly what a few kilograms of weight loss achieves, or a gram a day less sodium, and it is a fraction of what one antihypertensive drug does. It is also measured after a run-in diet that had already changed the participants’ sodium intake. Nobody should stop or reduce a prescribed antihypertensive on the strength of it. See Hypertension.

A last note in the herb’s favour, for balance: nothing here shows orthosiphon contributed nothing. It may have contributed some of the effect, possibly through the mild diuresis that gives the product its name. The honest verdict is that the trial cannot tell us, and that its design was never intended to.

The Vasodilation Mechanism and Its Solvent Problem

Evidence tier: ex vivo tissue and animal.

The mechanistic case is set out in Mechanistic insights into the synergistic vasodilatory actions of eupatorin, sinensetin, and 3′-hydroxy-5,6,7,4′-tetramethoxyflavone in ex vivo aortic ring model and their antihypertensive efficacy in in vivo rat model, in Hypertension Research, 2025 — PubMed search. Three of the plant’s polymethoxyflavones were applied to isolated rat aortic rings and then given to hypertensive rats, with synergistic vasodilation and a blood-pressure-lowering effect reported.

This is good, clean pharmacology, and it is worth understanding exactly what it establishes and what it does not.

An aortic ring bath is a clean experiment about a molecule. The tissue is suspended in a buffer and the compound is added at a known concentration. Nothing is absorbed, metabolised, protein-bound or excreted. It answers “can this molecule relax this vessel?” — a real and useful question.
It says nothing about whether the molecule ever gets there. Between a cup of tea and a human arteriole lie extraction, gut absorption, microbial transformation, first-pass metabolism and plasma protein binding, and for these compounds none of that is established at drinkable doses.
The compounds tested are exactly the poorly water-soluble ones. This is the sharp edge of the argument. The study did not test an infusion; it tested purified polymethoxyflavones. Methoxylation is what makes them lipophilic, and lipophilic is what makes them extract badly into hot water. So the better this mechanism looks, the worse the case for the tea — because the mechanism belongs to the fraction the traditional preparation is least equipped to deliver.

That inversion is worth stating plainly: a strong flavone mechanism is an argument for a standardised ethanolic extract and, simultaneously, an argument against the drink that carries all the tradition. Very few herbs have a cleaner example of this tension.

The Best Single-Herb Trial Is About Something Else

Evidence tier: phase II randomised, double-blind, placebo-controlled — single herb.

One genuinely single-herb human trial exists, and it is not about the kidney, blood pressure or blood sugar. It is A phase II randomized, double-blind, placebo-controlled study of Nuvastatic (C50SEW505OESA), a standardized rosmarinic acid-rich polymolecular botanical extract formulation to reduce cancer-related fatigue in patients with solid tumors, in Supportive Care in Cancer, 2024 — PubMed search. A hundred and ten patients with stage II–IV solid tumours undergoing chemotherapy received a standardised rosmarinic-acid-rich O. stamineus extract, 1000 mg three times daily, or placebo, for nine weeks. Fatigue scores on the Brief Fatigue Inventory improved significantly in the extract group.

By design quality this is comfortably the best evidence cat’s whiskers has: randomised, double-blind, placebo-controlled, multicentre, single-herb. It deserves to be reported at the top of the file rather than at the bottom, which is where a page ordered by popularity would put it.

Four qualifications belong with it, and none of them is a dismissal.

Phase II, industry-linked, unreplicated. That is the ordinary status of a promising early result, and most such results do not survive phase III.
A subjective primary outcome. Fatigue questionnaires are the correct instrument here and are validated, but they are also the outcome type most responsive to expectation — which blinding mitigates rather than eliminates.
The dose is 3 g of concentrated extract per day. Depending on the drug-to-extract ratio that corresponds to a very large quantity of leaf — far beyond the European daily tea dose of 6–12 g of herb. We are not going to state the leaf equivalent as a number, because the ratio for this specific product is not something we can verify. The direction is unambiguous: no cup of tea reaches this exposure.
It is a specific claim in a specific population. Reduced chemotherapy-related fatigue in people with cancer is not evidence for energy, wellbeing or “vitality” in anybody else.

And now the awkward part, which is a chemistry problem rather than a trial problem. The extract was standardised to be rosmarinic-acid-rich. Rosmarinic acid is the mint family’s signature polyphenol and is not remotely specific to this plant: rosemary, lemon balm, perilla, sage and spearmint all carry it, several in greater quantity. If rosmarinic acid is doing the work, the active is a common Lamiaceae compound and the plant is a delivery vehicle rather than the reason. Attributing the result to Orthosiphon as such is compound substitution. See Lemon Balm, Rosemary and Perilla, which carry the same compound and their own claims.

Standardised to What?

The previous section exposes an incoherence at the centre of how this herb is sold, and it is worth isolating.

The European Pharmacopoeia identifies Orthosiphonis folium partly by a minimum content of sinensetin, and supplement labels advertise sinensetin standardisation accordingly. But the one good human trial used a rosmarinic-acid-rich extract. Buying the standardised product does not buy you the trialled product. Two different compounds, two different extraction targets, one plant name on both bottles.

Three further points about marker compounds, which apply well beyond this herb.

A marker is chosen for measurability, not for activity. Sinensetin earns its role by being reasonably abundant, reasonably characteristic and easy to quantify by chromatography. Nothing about being a good analytical handle makes a compound the active principle, and the pharmacopoeia does not claim it does.
A marker assay is a potency check, not an identity check. Sinensetin is not unique to this genus — it is a polymethoxyflavone also found in citrus, and its name records that origin. A certificate of analysis reporting sinensetin therefore tells you a sample contains sinensetin; it does not tell you the sample is Orthosiphon leaf. Only DNA-based methods settle identity, which matters because “cat’s whiskers” is also an English name for the unrelated vegetable Cleome gynandra. This is developed on the safety page.
Standardisation says nothing about the rest of the extract. Two products matched for sinensetin can differ completely in solvent, drug-to-extract ratio, diterpene content and potassium load.

Blood Sugar and Metabolic Claims

Evidence tier: preclinical, plus combination-product human data.

Antidiabetic activity is one of the commonest modern claims for this plant, and the underlying work is rodent and in vitro: glucose-lowering in diabetic rat models, inhibition of carbohydrate-digesting enzymes in assay systems, and effects on markers of insulin resistance. The topic search PubMed: Orthosiphon stamineus antidiabetic and hypoglycaemic returns the field.

On the human side, the two Bologna trials reported effects on metabolic syndrome components — and carry the same combination problem as their blood-pressure results, with magnesium in particular having its own metabolic literature.

So the verdict is absent, not refuted: no adequate randomised trial has tested single-herb cat’s whiskers on fasting glucose, HbA1c or insulin sensitivity in people with or without diabetes. That is a gap, not a failure, and it should not be reported as either a proven benefit or a disproven one. Anyone taking glucose-lowering medication should treat an unquantified herbal hypoglycaemic claim as a reason for caution rather than enthusiasm — not because harm has been demonstrated, but because an unmeasured effect on top of a measured drug is exactly the situation in which surprises happen.

Uric Acid and Gout

Evidence tier: animal; no human trial.

Gout is a traditional indication almost everywhere the plant grows, and there is a coherent thread linking it to the diuretic story: the rat work reviewed by the European regulator reported improved uric acid elimination alongside increased urine output.

The modern contribution is Elucidating the therapeutic mechanism of Orthosiphon aristatus in hyperuricemic nephropathy: an integrated microbiome–metabolomics approach, in the Journal of Ethnopharmacology — PubMed search. It reports changes in gut bacterial populations and purine-related metabolites in an animal model of hyperuricaemic kidney disease. This is sophisticated, hypothesis-generating and entirely preclinical.

There is no human trial of cat’s whiskers for gout or hyperuricaemia. That absence matters more here than for most indications, because untreated hyperuricaemia damages joints and kidneys permanently, and allopurinol and febuxostat are cheap, effective and well studied. Using java tea instead of urate-lowering therapy would be a bad trade made on no evidence. Using it alongside, as a fluid-forward habit, is a different and far more defensible proposition. See Gout, Uric Acid testing and, for the same claim on another traditional plant, chanca piedra and uric acid.

Anti-Inflammatory and Antioxidant Work

Evidence tier: in vitro and animal.

The orthosiphols — staminane- and isopimarane-type diterpenes structurally distinctive to the genus, present at roughly 0.2% of the dried leaf — are the compounds most often invoked in anti-inflammatory work, alongside the flavones. The constituent map in Analysis of the chemical constituents and their metabolites in Orthosiphon stamineus Benth. (PLoS One, 2024) is a useful modern inventory — PubMed search.

Two cautions apply to this whole category, and they are not specific to this plant.

Chemical antioxidant assays measure chemistry in a tube. A high radical-scavenging score describes how a compound behaves in a cuvette. It does not establish that anything is scavenged in a person, at a reachable concentration, at a site where it matters. Antioxidant capacity has repeatedly failed to predict clinical benefit across the whole supplement field.
Cell-culture concentrations routinely exceed achievable plasma levels by orders of magnitude, and cells in a dish are not exposed to the conjugated, microbially transformed metabolites that actually circulate.

None of this means the compounds are inert. It means “anti-inflammatory” on a label, backed by cell work, is a hypothesis about a mechanism, not a claim about a person.

Reading Preclinical Work Without Being Fooled

Since most of this file is preclinical, here are the three design questions worth asking of any animal study you meet for this plant. They catch more errors than a blanket “it was only a rat study” caveat.

Did the positive control behave? A comparator drug exists to prove the model responds, not to rank the treatments. When an extract appears to beat a reference drug in rodents, the usual explanation is that the drug performs badly in that model. The diuresis experiments reviewed by the European regulator pass this test honestly — furosemide produced about 2.5-fold urine volume against the extract’s 1.3-fold, so the assay worked and the herb ranked below the drug.
Was the extract given before or after the insult? Rodent protection studies frequently dose the extract before the toxin. That tests prevention of an injury that has not happened yet, which is not the question a reader with an existing condition is asking. If you follow the topic search PubMed: Orthosiphon stamineus hepatoprotective, read each design before its conclusion.
Does the model model the disease? A chemically induced rodent condition often shares a name with a human disease and little else. Hyperuricaemic nephropathy induced in a rat over weeks is not gouty kidney disease accumulated in a person over decades.

A fourth question is specific to this plant and worth adding to the list: which solvent? The animal diuresis data alone contain a 50% ethanolic extract that worked and a 70% ethanolic extract that did nothing. If a study does not state its solvent and drug-to-extract ratio, its result cannot be connected to anything you can buy.

What Would Actually Change the Picture

Naming the missing experiments is more useful than repeating that evidence is limited. Four trials would settle most of this page, and every one of them is ordinary.

Single-herb, placebo-controlled ambulatory blood pressure. Standardised orthosiphon extract alone against matched placebo, with 24-hour ambulatory monitoring rather than clinic readings, in untreated stage 1 hypertension. This is the trial that would separate the herb from the hibiscus.
A dose-ranging pharmacokinetic study. Plasma and urinary concentrations of sinensetin, eupatorin and rosmarinic acid and their conjugates after a defined infusion and after a defined extract, in humans. Without this, no mechanism on this page can be connected to a dose, and every “it reaches active levels” claim in either direction is a guess.
Tea versus extract, head to head. The single most decision-relevant experiment for a reader, and nobody has run it: does the traditional infusion deliver anything the standardised capsule does, or vice versa?
Replication of the fatigue trial in phase III, ideally by an independent group, with an active comparator where one exists.

Evidence Ledger for This Page

Randomised, double-blind, placebo-controlled, single herb — positive, phase II, unreplicated. A standardised rosmarinic-acid-rich extract at 3 g/day reduced cancer-related fatigue in 110 patients on chemotherapy. The strongest evidence this plant has, for an indication it is not sold for, at a dose no tea reaches.
Randomised human trials — uninterpretable for this herb. Two Italian trials of multi-ingredient products reported small blood-pressure and metabolic effects; the products also contained magnesium, hawthorn and hibiscus, all with their own evidence.
Effect size, calibrated. The 3–4 mmHg systolic reduction in CONDOR is comparable to a few kilograms of weight loss and a fraction of one antihypertensive drug.
Ex vivo and animal — positive. Three polymethoxyflavones produced synergistic vasodilation in aortic rings and lowered blood pressure in hypertensive rats.
Chemistry — adverse to the tea. Those same flavones are the least water-soluble constituents, so a strong flavone mechanism argues for the ethanolic extract and against the traditional infusion.
Animal — positive but preliminary. Improved uric acid excretion in rats; microbiome and metabolomic changes in a hyperuricaemic nephropathy model.
Preclinical only. Antidiabetic, anti-inflammatory and antioxidant activity, in rodents and cell systems, at concentrations not shown to be achievable in people.
Absent, not refuted. Single-herb human data for blood pressure, blood sugar, uric acid, gout, inflammation or fatigue outside cancer; any human pharmacokinetics; any tea-versus-extract comparison.
Internal incoherence. The herb is standardised and sold on sinensetin; the one good trial used a rosmarinic-acid-rich extract. The label and the evidence point at different compounds.
Refused. A leaf-equivalent figure for the 3 g/day trial dose, because the drug-to-extract ratio of that product is not something we can verify; and any claim that this herb lowers blood pressure as a single agent, which no trial has tested.

Key Research Papers

Citations are PubMed topic searches with title, journal and year in prose, so a link can never resolve to the wrong paper.

A phase II randomized, double-blind, placebo-controlled study of Nuvastatic (C50SEW505OESA), a standardized rosmarinic acid-rich polymolecular botanical extract formulation to reduce cancer-related fatigue in patients with solid tumors. Supportive Care in Cancer, 2024. The best single-herb trial. PubMed search
Effect of dietary supplementation with Diuripres on blood pressure, vascular health, and metabolic parameters in individuals with high-normal blood pressure or stage I hypertension: the CONDOR randomized clinical study. Phytotherapy Research, 2023. Four ingredients, one of them orthosiphon. PubMed search
Effect of a combined nutraceutical containing Orthosiphon stamineus on blood pressure and metabolic syndrome components in hypertensive dyslipidaemic patients: a randomized clinical trial. Complementary Therapies in Clinical Practice, 2012. PubMed search
Mechanistic insights into the synergistic vasodilatory actions of eupatorin, sinensetin, and 3′-hydroxy-5,6,7,4′-tetramethoxyflavone in ex vivo aortic ring model and their antihypertensive efficacy in in vivo rat model. Hypertension Research, 2025. PubMed search
Elucidating the therapeutic mechanism of Orthosiphon aristatus in hyperuricemic nephropathy: an integrated microbiome–metabolomics approach. Journal of Ethnopharmacology. Animal model. PubMed search
Analysis of the chemical constituents and their metabolites in Orthosiphon stamineus Benth. via UHPLC-Q Exactive Orbitrap-HRMS and AFADESI-MSI techniques. PLoS One, 2024. The modern constituent inventory. PubMed search
Gut microbiota of pigs metabolizes extracts of Filipendula ulmaria and Orthosiphon aristatus — herbal remedies used in urinary tract disorders. Planta Medica, 2022. Why circulating compounds differ from label compounds. PubMed search
Meta-analytic literature on Hibiscus sabdariffa and blood pressure — the named alternative explanation for the CONDOR result. PubMed search
Meta-analytic literature on magnesium supplementation and blood pressure — the second named alternative. PubMed search
Topic search — Orthosiphon stamineus antidiabetic and hypoglycaemic. The rodent and in vitro glucose literature.
Topic search — orthosiphol diterpenes and inflammation. The genus-specific diterpenes.
Topic search — rosmarinic acid pharmacokinetics in humans. Read this before crediting any rosmarinic-acid mechanism to any one plant.
Topic search — Orthosiphon aristatus clinical trial and Orthosiphon stamineus clinical trial. Run both names; the human literature is small enough to read in an afternoon.

Connections

All Herbs
Cat’s Whiskers / Java Tea — the main page, with constituents, doses and the monograph.
Diuretic and Kidney Stone Claims — where the solvent and solubility argument is set out in full.
Urinary Tract Complaints — the antibacterial file, and what “flushing” licenses.
Kidney Cautions and Interactions — potassium, prescribed diuretics and product identity.
Hibiscus and Blood Pressure — the ingredient with its own trials, sharing the same capsule.
Hawthorn — the third cardiovascular ingredient in the CONDOR product.
Magnesium — the fourth, with its own blood-pressure meta-analyses.
Hypertension — what a 3–4 mmHg change is worth, and what actually treats high blood pressure.
Gout — the traditional indication with no human evidence and proven drugs available.
Chanca Piedra for Uric Acid and Gout — the same claim on another traditional plant.
Uric Acid Testing — the measurement any urate claim should be judged against.
Lemon Balm — a cheaper source of the same rosmarinic acid.
Rosemary — likewise, and the plant the compound is named after.
Perilla — where this site examined the rosmarinic-acid attribution problem before.

Back to Table of Contents