Life With What
There's a whole family waiting for you →

Malvaceae

Hibiscus sabdariffa

Roselle · Hibiscus · Karkade · Sorrel

Trial support Caution
Part used
flowerfruit
Native range
West Africa; cultivated across the tropics
Morphology
An annual shrub to two and a half metres with lobed leaves and pale yellow flowers, whose fleshy deep-red calyces swell after flowering and are the part harvested.

What is in it

anthocyanins delphinidin-3-sambubioside
organic acids hibiscus acid, citric acid
polyphenols

Evidence

Among the better-supported entries. Multiple randomised trials and meta-analyses show a modest but consistent reduction in systolic and diastolic blood pressure, with effect sizes smaller than antihypertensive drugs and comparable to some dietary interventions.

Safety

A food and generally well tolerated. Because the blood pressure effect is real, it is additive with antihypertensive treatment. Reduces levels of some drugs. Traditionally avoided in pregnancy.

Interactions: Additive with antihypertensives; Reduces plasma levels of hydrochlorothiazide and of chloroquine

Regulatory: Permitted as a food and as a listed medicine ingredient.

Historical use

Drunk as a cold or hot infusion across west and north Africa, Egypt, the Caribbean and Mexico — karkade, bissap, agua de Jamaica. Used for blood pressure and as a cooling drink.

Recorded as history and attributed where possible. Long use establishes that a plant was widely used, not that it worked, and not that it was safe.

What it is taken for, and who says so

Sold as a tea and extract for blood pressure and cholesterol, and drunk very widely as a beverage across West Africa, the Caribbean and Latin America.

Recorded because it is what people believe, not because it is what the plant does. Whether any of it holds up is the Evidence section above, and the two frequently disagree. A sincere account of feeling better is compatible with the plant doing nothing — see the placebo effect.

Worth knowing

Roselle is one of the few plants in this collection where a large body of ordinary daily use and a respectable trial literature agree, and the finding is modest and useful rather than dramatic.

It also demonstrates the point that a real effect creates a real interaction. Because hibiscus genuinely lowers blood pressure, someone already on antihypertensive medication who starts drinking it daily is adding to their treatment without anyone recording it — and it separately reduces blood levels of hydrochlorothiazide, a common antihypertensive, which pushes in the other direction.

An inert tea cannot do either of those things. This is what it looks like when a herb works.

Nomenclature verified against GBIF backbone taxonomy.