Foxglove: The Plant and the Science of Digitalis
Every story about William Withering begins with a plant: the tall purple foxglove, Digitalis purpurea, which grows wild on the banks, heaths and woodland edges of Britain. In 1785 Withering published his Account of the Foxglove, describing how its dried leaf could clear the swelling of “dropsy” and noting, in the last of his nine Inferences, that it had “a power over the motion of the heart, to a degree yet unobserved in any other medicine.” He did not know why. The explanation took another century and a half to arrive, and it came from chemistry and cell biology rather than from the bedside.
This page follows that natural source from the hedgerow to the heart cell: what kind of plant foxglove is and why its leaves varied in strength, how country people used it long before Withering, the related “heart plants” such as squill and strophanthus, the isolation of the glycosides digitoxin and digoxin, the discovery of the sodium-potassium pump that they block, how blocking that pump strengthens the heartbeat and slows the pulse, why the gap between a useful and a poisonous amount is so small, and why foxglove is still listed among dangerous garden plants. The story of Withering’s own cases is told on the An Account of the Foxglove page, and the modern trials on the Legacy and Later Research page. Nothing here is dosing information.
Table of Contents
- Digitalis purpurea in the Hedgerow
- A Biennial and Its Leaves
- Folk Medicine Before Withering
- Squill, Strophanthus and Other Heart Plants
- Isolating the Glycosides: Digitoxin and Digoxin
- The Sodium-Potassium Pump
- How a Pump Inhibitor Strengthens the Heart
- Slowing the Pulse
- A Narrow Margin
- Foxglove as a Poisonous Plant
- Key Research Papers
- Connections
- Featured Videos
1. Digitalis purpurea in the Hedgerow
The purple foxglove is one of the most recognisable wild flowers of western Europe. In its flowering year it sends up a single stem, often taller than a child, hung along one side with drooping, tube-shaped flowers in shades of purple and pink, each one spotted darker inside the throat. Bees crawl right into the tubes to reach the nectar. The leaves are soft, downy and wrinkled, broadest near the ground and smaller up the stem.
The scientific name comes from the shape of the flower: digitalis comes from the Latin digitus, a finger, because a flower fits over a fingertip like the finger of a glove. Purpurea simply means purple. Older botany books placed foxglove in the figwort family (Scrophulariaceae); modern classifications based on DNA move it to the plantain family (Plantaginaceae). Withering, who was a serious botanist and had published the first British flora arranged on the Linnaean system in English in 1776, knew the plant well before he ever thought of it as a medicine.
Other members of the genus matter to this story too. The woolly foxglove, Digitalis lanata, a plant of south-eastern Europe with smaller, pale, brown-veined flowers, became the industrial source of digoxin in the twentieth century. But for Withering, and for the folk healers before him, “foxglove” meant the purple plant of the English countryside.
2. A Biennial and Its Leaves
Foxglove is a biennial: it normally lives for two years. In the first year it forms only a low rosette of leaves pressed close to the ground. In the second year it puts up its flowering spike, sets thousands of tiny seeds, and usually dies. A plant gathered from the wild may therefore be a young rosette, a plant coming into flower, or one already going to seed, and its leaves do not all carry the same amount of active substance.
Withering understood that this mattered, even without knowing the chemistry. His colleague Dr Ash had told him that Dr Cawley, principal of Brasenose College in Oxford, had been cured of “Hydrops Pectoris” (fluid in the chest) by foxglove root. Withering rejected the root: in a biennial plant, he reasoned, its strength would be too uncertain. He chose instead the leaves, gathered “when it was in its flowering state, and carefully dried”. Drying turned a variable fresh herb into a powder that could be weighed, and in the summer of 1776 he had leaves dried specifically so that doses could be ascertained. He also moved away from long boiling (a decoction) toward an infusion, reasoning by analogy with tobacco that prolonged heat might weaken the leaf.
Even so, the leaf remained a natural product. Its strength still varied with the plant’s age, the season and place of gathering, and the way it was dried and stored. Writing on Withering’s legacy, the pharmacologist Alasdair Breckenridge listed this as the first of four lessons from the foxglove story: preparations that were not standardised produced unpredictable effects. That problem was only solved when the active compounds themselves could be purified and measured (section 5).
3. Folk Medicine Before Withering
Foxglove did not enter medicine through the learned tradition. Groves and Bisset, reviewing the older sources in 1991, found that digitalis was apparently unknown to the physicians of ancient Greece and Rome. By the Middle Ages, however, it was widely used in European folk medicine. One of the ways it was used surprised later readers: it was applied to the skin by inunction (rubbing a preparation in), and some of the effects recorded from that practice are recognisable today as signs of glycoside overdose. People were absorbing enough through the skin to feel it.
A Danish history of cardiac glycosides by Norn and Kruse points to the Physicians of Myddfai, a Welsh family of healers whose remedies are associated with the thirteenth century (around 1250), as an early written source mentioning foxglove. Folk healers in Britain used it for many purposes, from wounds and swellings to other ailments, often in mixtures where its role was hard to separate from everything else in the pot.
Withering’s own starting point came out of this world. In 1775 his opinion was asked about “a family receipt for the cure of the dropsy” that had long been kept a secret by “an old woman in Shropshire”; the mixture contained twenty or more herbs, and he judged that “the active herb could be no other than the Foxglove”. (The popular name “Mother Hutton” for that woman is a later legend, traced on the Account page.) Kinne-Saffran and Kinne, in their 2002 review “Herbal diuretics revisited: from ‘wise women’ to William Withering”, set his work at the end of more than 2,000 years of plant diuretics described by Pliny, Dioscorides, Hildegard of Bingen, Matthioli and Fuchs, and of the knowledge held by peasant “wise women” whose remedies for swelling were passed down outside the universities.
4. Squill, Strophanthus and Other Heart Plants
Foxglove is not the only plant that makes heart-active glycosides. Several unrelated plants around the world evolved similar chemicals, probably as a defence against grazing animals, and people found them independently.
- Squill (sea onion, Drimia or Urginea maritima), a large Mediterranean bulb, was known to the Romans, according to Norn and Kruse. It was the standard diuretic of Withering’s day, and it shaped his early mistakes: physicians believed squill worked best when it made the patient sick, and Withering admitted in the Account that by this analogy he first gave foxglove “in doses very much too large” and “urged its continuance too long”. He later concluded that the diuretic effect did not depend on causing nausea at all.
- Strophanthus, an African climbing plant, supplied arrow poisons used for hunting; the same Danish history describes its role as a source of glycosides. Its purified glycoside, ouabain, later became a standard laboratory tool for studying the sodium pump.
- Other plants with related compounds include lily of the valley and oleander, both familiar as garden plants and both on lists of poisonous plants for the same reason as foxglove.
The repeated discovery of these plants, on different continents and in different traditions, reflects a single shared target in the animal body. That target was not identified until the 1950s (section 6). The site also covers gentler traditional heart herbs, such as hawthorn and motherwort, which are not cardiac glycoside plants.
5. Isolating the Glycosides: Digitoxin and Digoxin
The active substances in foxglove belong to a chemical family called cardiac glycosides. Each molecule has two parts. One part is a steroid-shaped core carrying a small ring (a lactone) — this is the business end that fits the target in the heart cell. The other part is a short chain of sugars (“glycoside” means a compound with sugar attached), which changes how the molecule dissolves, how it is absorbed and how long it stays in the body.
Through the nineteenth century, chemists in France and Germany worked to pull crystalline active fractions out of the leaf, and the main glycoside of the purple foxglove became known as digitoxin. It is fat-soluble, well absorbed, and very long-lasting in the body, leaving the body only slowly.
In 1930 Sydney Smith, a chemist working in Britain, published a short paper in the Journal of the Chemical Society titled “Digoxin, a new digitalis glucoside”. Digoxin, the compound he described, differs from digitoxin by a single extra oxygen-bearing (hydroxyl) group on the steroid core. That small change makes it more water-soluble, shorter-acting and cleared mainly by the kidneys. Digoxin became the most widely used digitalis drug of the twentieth century, extracted industrially from the woolly foxglove, Digitalis lanata.
Purification answered Withering’s old problem. Instead of a dried leaf of uncertain strength, physicians now had a single, weighable chemical. Digitoxin itself never disappeared: in 2025 the DIGIT-HF trial tested it in modern heart failure care, a result described on the Legacy and Later Research page.
6. The Sodium-Potassium Pump
Every living cell keeps sodium mostly outside and potassium mostly inside. The difference is maintained by a protein machine in the cell membrane that burns the cell’s energy currency, ATP, to push three sodium ions out and pull two potassium ions in on each cycle. This gradient powers nerve signals, muscle contraction, and the transport of many nutrients into cells. The site’s animated explainer of the sodium-potassium pump shows the cycle step by step.
The pump was discovered by the Danish physiologist Jens Christian Skou. Working on nerves from shore crabs, he described in 1957 an enzyme in the nerve membrane that broke down ATP only when both sodium and potassium were present — a sodium- and potassium-activated ATPase, now called the Na+/K+-ATPase. Skou shared the 1997 Nobel Prize in Chemistry for the discovery.
It soon became clear that this enzyme was the long-sought target of the foxglove glycosides. As Vardaman Buckalew summarises in his history of the field, the Na,K-ATPase carries a highly conserved receptor site for plant-derived digitalis glycosides, and these compounds inhibit the pump when they bind it. The same site has been preserved across the animal kingdom, which explains why squill, strophanthus and foxglove all act in the same way, and why the plants are toxic to so many species. It also prompted a long search for the body’s own “digitalis-like” molecules, described on the Legacy and Later Research page.
7. How a Pump Inhibitor Strengthens the Heart
It seems strange that a poison which partly jams a vital pump can make a weak heart beat more strongly. The explanation, built up over decades of muscle research reviewed by Wray, Eisner and Allen on the foxglove’s bicentenary in 1985, runs through calcium.
- Sodium builds up a little. When some of the heart cell’s sodium pumps are blocked, slightly more sodium stays inside the cell.
- Calcium leaves more slowly. Heart cells remove much of their calcium through a second membrane protein, the sodium-calcium exchanger, which lets sodium run in to push calcium out. With more sodium already inside, the inward push is weaker, so less calcium is removed after each beat.
- The calcium store fills. The extra calcium is taken up into the cell’s internal store (the sarcoplasmic reticulum).
- Each beat releases more calcium. On the next beat the fuller store releases more calcium onto the contractile proteins, and the muscle contracts more forcefully.
Pharmacologists call this a positive inotropic effect: a stronger squeeze. In a failing, enlarged heart, a stronger squeeze moves more blood forward, which helps explain the effect Withering saw. When the heart pumps better, blood flow to the kidneys improves, more urine is made, and the fluid that had collected in the legs, abdomen and lungs (edema, the old “dropsy”) can be cleared. Withering believed the leaf acted chiefly on the kidneys; as historians such as Bessen have noted, he was largely unaware of the cardiac mechanism behind the diuresis he recorded.
Minerals take part in this chemistry. Calcium is the final messenger of the stronger beat, and potassium competes with glycosides at the pump, a point that returns in section 9.
8. Slowing the Pulse
The second great effect of digitalis is on the heart’s rhythm. Withering recorded in his list of effects that, with large and quickly repeated amounts, the pulse could become very slow — “even as slow as 35 in a minute”. That observation pointed to something the glycosides do to the heart’s electrical system.
The heart’s beat starts in a natural pacemaker in the upper chambers and passes to the lower chambers through a relay station called the atrioventricular (AV) node. Cardiac glycosides slow conduction through this node in two ways: directly, through their effects on the heart cells, and indirectly, by increasing the activity of the vagus nerve, the body’s natural brake on the heart rate. Pharmacologists call this a negative chronotropic and dromotropic effect.
This is why, long after its use for dropsy, digitalis became a mainstay for atrial fibrillation. In atrial fibrillation the upper chambers quiver chaotically and bombard the AV node with signals; by slowing the node, the glycosides let fewer of those signals through, so the lower chambers beat more slowly and fill better. A modern trial in Withering’s own city, Birmingham, compared low-dose digoxin with another rate-slowing drug in permanent atrial fibrillation; its results are on the Legacy and Later Research page.
9. A Narrow Margin
The same mechanism that helps the heart can poison it. If too many pumps are blocked, calcium overloads the heart cells, and they can begin to fire on their own between normal beats. Together with the slowed conduction through the AV node, this produces the dangerous arrhythmias of digitalis poisoning: extra beats, heart block, and in the worst cases fatal rhythms. Pharmacologists describe digitalis as having a narrow therapeutic index: the amount that helps is not far below the amount that harms.
Withering described the warning signs from experience. With very large, quickly repeated doses he listed “sickness, vomiting, purging, giddiness, confused vision, objects appearing green or yellow; increased secretion of urine … slow pulse, even as slow as 35 in a minute, cold sweats, convulsions, syncope, death.” Emergency physician H. A. Bessen, reviewing the Account two centuries later, noted that it already recorded striking examples of digitalis toxicity. The green-yellow tint to vision is still taught today as a classic sign.
Several things narrow the margin further, and they are described repeatedly in the pharmacology literature:
- Low potassium. Potassium and the glycosides compete for the pump. When blood potassium is low (for example after strong diuretics), more glycoside binds, and toxicity appears at lower amounts. High potassium is also a feature of severe poisoning, because blocked pumps let potassium leak out of cells.
- Low magnesium and high calcium, which change heart cell excitability (see magnesium).
- Reduced kidney function, which slows the removal of digoxin.
- Individual differences in sensitivity and in how the drug is handled, the second of Breckenridge’s four lessons from Withering.
In 1969 Thomas Smith, Vincent Butler and Edgar Haber published a radioimmunoassay that could measure therapeutic and toxic serum digoxin concentrations in the blood, and in 1976 the same group reported reversing advanced digoxin poisoning with antibody fragments. Those developments, which finally put numbers on Withering’s narrow margin, are told on the Legacy and Later Research page.
10. Foxglove as a Poisonous Plant
Outside the pharmacy, foxglove is a poisonous plant. Every part of it contains cardiac glycosides, and the narrow margin described above applies to the raw plant with none of the control that a purified, measured medicine allows. The leaf strength varies from plant to plant, which is exactly why Withering struggled with it and why the medicine was eventually purified.
Poisoning from the wild plant is reported mostly in three situations: children or animals eating parts of a garden plant; herbal teas or home preparations; and mistaken identity. Before it flowers, a foxglove rosette of soft, hairy leaves looks similar to the leaves of comfrey, a plant long used as a herbal remedy. In 2017 Wu and colleagues reported a case of fatal cardiac glycoside poisoning in which foxglove had been mistaken for comfrey. Reports of this kind describe the same signs Withering listed in 1785 — nausea and vomiting, disturbed vision, a slow or irregular pulse — along with high blood potassium.
The plant’s place in medicine and its place on lists of poisonous plants are the same fact seen from two sides. Withering’s achievement, as his own book makes plain, was to take a powerful country poison and show, case by case and with every failure recorded, that it could be “converted to salutary ends”. The careful measurement that his method demanded is the reason the foxglove’s chemistry is still in use more than two centuries later.
Key Research Papers
- Groves MJ, Bisset NG. A note on the use of topical digitalis prior to William Withering. J Ethnopharmacol. 1991;35(2):99-103. PubMed PMID: 1809828
- Kinne-Saffran E, Kinne RK. Herbal diuretics revisited: from “wise women” to William Withering. Am J Nephrol. 2002;22(2-3):112-8. PubMed PMID: 12097727
- Norn S, Kruse PR. [Cardiac glycosides: From ancient history through Withering’s foxglove to endogeneous cardiac glycosides]. Dan Medicinhist Arbog. 2004:119-32. PubMed PMID: 15685783
- Krikler DM. The foxglove, “The old woman from Shropshire” and William Withering. J Am Coll Cardiol. 1985;5(5 Suppl A):3A-9A. PubMed PMID: 3886750
- Smith S. Digoxin, a new digitalis glucoside. J Chem Soc. 1930:508-510. DOI: 10.1039/JR9300000508
- Skou JC. The influence of some cations on an adenosine triphosphatase from peripheral nerves. Biochim Biophys Acta. 1957;23(2):394-401. PubMed PMID: 13412736
- Buckalew VM. Endogenous digitalis-like factors: an overview of the history. Front Endocrinol (Lausanne). 2015;6:49. PubMed PMID: 25918512
- Wray S, Eisner DA, Allen DG. Two hundred years of the foxglove. Med Hist Suppl. 1985;(5):132-50. PubMed PMID: 3915521
- Bessen HA. Therapeutic and toxic effects of digitalis: William Withering, 1785. J Emerg Med. 1986;4(3):243-8. PubMed PMID: 3543113
- Breckenridge A. William Withering’s legacy--for the good of the patient. Clin Med (Lond). 2006;6(4):393-7. PubMed PMID: 16956147
- Smith TW, Butler VP Jr, Haber E. Determination of therapeutic and toxic serum digoxin concentrations by radioimmunoassay. N Engl J Med. 1969;281(22):1212-6. PubMed PMID: 5388455
- Wu IL, Yu JH, Lin CC, Seak CJ, Olson KR, Chen HY. Fatal cardiac glycoside poisoning due to mistaking foxglove for comfrey. Clin Toxicol (Phila). 2017;55(7):670-673. PubMed PMID: 28463019
PubMed Topic Searches
- https://pubmed.ncbi.nlm.nih.gov/?term=Digitalis+purpurea+cardiac+glycosides
- https://pubmed.ncbi.nlm.nih.gov/?term=cardiac+glycosides+Na%2CK-ATPase+mechanism
- https://pubmed.ncbi.nlm.nih.gov/?term=foxglove+poisoning
- https://pubmed.ncbi.nlm.nih.gov/?term=history+of+digitalis
Further Reading
- Withering W. An Account of the Foxglove, and some of its Medical Uses: with Practical Remarks on Dropsy, and Other Diseases. Birmingham; 1785. Project Gutenberg eBook #24886. https://www.gutenberg.org/ebooks/24886
- Lee MR. William Withering (1741–1799): a biographical sketch of a Birmingham Lunatic. James Lind Library Bulletin; 2005. https://www.jameslindlibrary.org/articles/william-withering-1741-1799-a-biographical-sketch-of-a-birmingham-lunatic/