Digoxin

Digoxin at a Glance

Digoxin is a prescription cardiac drug, and among the most-prescribed and most-argued-about medicines of the twentieth century. It is a cardenolide glycoside that inhibits the sodium–potassium pump, strengthening cardiac contraction and slowing conduction through the AV node.

Its evidence has a distinctive shape. The DIG trial randomised 6,800 patients with heart failure and found mortality completely unchanged — 34.8% versus 35.1% — while hospitalisation for worsening heart failure fell from 34.7% to 26.8% ^{1Reference 1Digitalis Investigation Group · 1997RCTThe effect of digoxin on mortality and morbidity in patients with heart failure — randomised controlled trialView study →}. That is a real result and a narrow one, and the drug’s reputation has run ahead of it in both directions ever since.

Its main source plant is not the one people assume. Digoxin is chiefly a molecule of Digitalis lanata, the woolly foxglove, and is only a minor component of the common foxglove D. purpurea — though it is detectable in purpurea poisonings ^{14,15Reference 14Ravi et al. · 2020High-resolution tandem mass spectrometry dataset reveals fragmentation patterns of cardiac glycosides in leaves of the foxglove plantsView study →Reference 15Kwon et al. · 2011HPLC method validation for Digitalis and its analogue by pulsed amperometric detectionView study →}. The famous digitalis literature is largely digoxin’s, and it gets silently attributed to the wrong plant.

It also carries a signal that runs against its own laboratory reputation. Cardiac glycosides look antiproliferative in cell culture; digoxin in people is a phyto-estrogen, and large Danish cohorts link current use to increased incidence of oestrogen-sensitive cancers, with risk returning to baseline when the drug is stopped ^{7,8,10Reference 7Biggar et al. · 2011ObservationalDigoxin use and the risk of breast cancer in women — cohort studyView study →Reference 8Biggar et al. · 2012ObservationalDigoxin use and the risk of cancers of the corpus uteri, ovary and cervix — cohort studyView study →Reference 10Biggar · 2012ReviewMolecular pathways: digoxin use and estrogen-sensitive cancers—risks and possible therapeutic implications — reviewView study →}.

None of this is a use for foxglove. The leaf cannot be assayed to the resolution a drug with this therapeutic index requires — content swings several-fold with season, part and provenance ^{14Reference 14Ravi et al. · 2020High-resolution tandem mass spectrometry dataset reveals fragmentation patterns of cardiac glycosides in leaves of the foxglove plantsView study →} — and crude Digitalis preparations were abandoned once the glycosides could be purified.

The Confounding Problem

Digoxin is the clearest example in cardiology of a drug that looks dangerous in observational data and neutral in randomised data, and the reason is not subtle: the sickest patients are the ones who get it.

The 2015 BMJ meta-analysis pooled 52 studies and 621,845 patients — over four million patient-years — and reported the pooled risk ratio for death by study design ^{3Reference 3Ziff et al. · 2015Meta-analysisSafety and efficacy of digoxin: systematic review and meta-analysis of observational and controlled trial data — meta-analysisView study →}:

Analysis type	Pooled risk ratio for death (95% CI)
Unadjusted observational	1.76 (1.57–1.97)
Adjusted observational	1.61 (1.31–1.97)
Propensity-matched	1.18 (1.09–1.26)
Randomised controlled trials	0.99 (0.93–1.05)

The gradient runs one way, and it does not stop until the design removes the confounding entirely. The same analysis found that digoxin users were 2.4 years older than controls, with a lower ejection fraction (33% versus 42%), more diabetes, and greater use of diuretics and antiarrhythmics — and that studies with better methods and lower risk of bias were significantly more likely to report a neutral association ^{3Reference 3Ziff et al. · 2015Meta-analysisSafety and efficacy of digoxin: systematic review and meta-analysis of observational and controlled trial data — meta-analysisView study →}. Across all study types, digoxin reduced all-cause hospital admission (risk ratio 0.92, 0.89–0.95) ^{3Reference 3Ziff et al. · 2015Meta-analysisSafety and efficacy of digoxin: systematic review and meta-analysis of observational and controlled trial data — meta-analysisView study →}.

The observational meta-analyses that report harm are not wrong about their own data. Pooling 19 adjusted studies in 326,426 patients gives a hazard ratio for all-cause mortality of 1.21, rising to 1.29 in the atrial-fibrillation subgroup ^{4Reference 4Vamos et al. · 2015Meta-analysisDigoxin-associated mortality: a systematic review and meta-analysis of the literature — meta-analysisView study →}; a 2019 update across 825,061 patients found 1.17 overall, 1.23 in atrial fibrillation and 1.11 in heart failure ^{5Reference 5Vamos et al. · 2019Meta-analysisMeta-Analysis of Effects of Digoxin on Survival in Patients with Atrial Fibrillation or Heart Failure: An Update — meta-analysisView study →}. Adjustment is what these analyses lack the information to do, not what they failed to attempt.

What the randomised data does support is a dose-response for harm at higher exposure. A post hoc analysis of the DIG trial in 3,782 men found mortality of 29.9% at serum concentrations of 0.5–0.8 ng/mL, 38.8% at 0.9–1.1 ng/mL and 48.0% at 1.2 ng/mL or above. Compared with placebo, the lowest band had a 6.3% lower absolute mortality; the highest band had an 11.8% higher one ^{2Reference 2Rathore et al. · 2003RCTAssociation of serum digoxin concentration and outcomes in patients with heart failure — randomised controlled trialView study →}. The modern practice of targeting a low serum concentration comes from that analysis, not from the trial’s headline.

Digoxin vs. Digitoxin

The two drugs share a mechanism, a toxidrome and a therapeutic index, and differ almost entirely in how the body handles them ^{12Reference 12Fender et al. · 2024ReviewFender, J., Klöcker, J., Boivin-Jahns, V., et al. (2024). “Cardiac glycosides”-quo vaditis?-past, present, and future? — review. Naunyn Schmiedebergs Arch Pharmacol. https://pubmed.ncbi.nlm.nih.gov/39007928/View study →}:

	Digoxin	Digitoxin
Principal plant source	Digitalis lanata	Digitalis purpurea
Oral bioavailability	55–75%, variable	>90%, consistent
Elimination	Predominantly renal	Hepatic + enterohepatic; largely renal-independent
Half-life	~30–50 hours	~4–7 days
Consequence	Levels track kidney function — an advantage when clearing an overdose, a hazard when renal function falls	Stable in renal impairment; an overdose takes a week to clear

The practical rule that follows is that renal function governs digoxin dosing in a way it does not govern digitoxin’s, and a falling eGFR in a patient on a stable digoxin dose is a rising digoxin level ^{12,13Reference 12Fender et al. · 2024ReviewFender, J., Klöcker, J., Boivin-Jahns, V., et al. (2024). “Cardiac glycosides”-quo vaditis?-past, present, and future? — review. Naunyn Schmiedebergs Arch Pharmacol. https://pubmed.ncbi.nlm.nih.gov/39007928/View study →Reference 13Whayne TF Jr · 2018Systematic reviewClinical Use of Digitalis: A State of the Art Review — systematic reviewView study →}.

A “Digoxin Level” Does Not Mean Digoxin

Digoxin immunoassays cross-react with other cardenolides, which turns them into a general screening test for cardiac-glycoside poisoning — and a source of misleading results.

The assay’s usefulness for detecting foxglove and oleander ingestion is established, and a study across five commercial digoxin immunoassays showed they also detect convallatoxin, the principal glycoside of lily of the valley ^{21Reference 21Fink et al. · 2014In vitroRapid detection of convallatoxin using five digoxin immunoassays — in vitroView study →}. A published case series records three patients with detectable “digoxin” concentrations after eating lily of the valley mistaken for ramps — none of whom had taken digoxin ^{22Reference 22Trebach et al. · 2022Case reportDetectable Digoxin Concentrations in 3 Patients with Ramps Misadventure — case reportView study →}.

Two things follow, in opposite directions. A positive digoxin level in someone with a cardiac-glycoside toxidrome and no digoxin prescription is useful evidence of a plant ingestion, and should be read that way rather than dismissed as an error. And a numerical “digoxin concentration” in that setting is a qualitative signal, not a quantitative one — the assay is calibrated for digoxin and the cross-reacting molecule has its own potency, so the number cannot be used to gauge severity or to dose the antidote.

What It's Used For

The evidence splits into three tiers that are rarely kept apart: one large randomised trial with a modest positive result, a much larger observational literature that is confounded, and a laboratory literature whose human epidemiology points the other way.

Evidence by IndicationStrength of support · Beta

01Heart Failure With Reduced Ejection FractionPromising66%DIG, n=6,800 — hospitalisation reduced, mortality unchanged ^{1,2Reference 1Digitalis Investigation Group · 1997RCTThe effect of digoxin on mortality and morbidity in patients with heart failure — randomised controlled trialView study →Reference 2Rathore et al. · 2003RCTAssociation of serum digoxin concentration and outcomes in patients with heart failure — randomised controlled trialView study →}

02Rate Control in Atrial FibrillationPromising55%RATE-AF vs bisoprolol; guideline-recognised niche where blood pressure is marginal ^{6,13Reference 6Kotecha et al. · 2020RCTEffect of Digoxin vs Bisoprolol for Heart Rate Control in Atrial Fibrillation on Patient-Reported Quality of Life: The RATE-AF Randomized Clinical Trial — randomised controlled trialView study →Reference 13Whayne TF Jr · 2018Systematic reviewClinical Use of Digitalis: A State of the Art Review — systematic reviewView study →}

03Antiproliferative ActivityShaky8%In-vitro class-level work, contradicted by the human epidemiology ^{10,17Reference 10Biggar · 2012ReviewMolecular pathways: digoxin use and estrogen-sensitive cancers—risks and possible therapeutic implications — reviewView study →Reference 17Bejček et al. · 2021ReviewNa(+)/K(+)-ATPase Revisited: On Its Mechanism of Action, Role in Cancer, and Activity Modulation — reviewView study →}

04Antiviral ActivityShaky10%Cell culture, class-level repurposing reviews; no human data ^{18Reference 18Škubník et al. · 2021ReviewRepurposing Cardiac Glycosides: Drugs for Heart Failure Surmounting Viruses — reviewView study →}

Mechanisms

Pharmacokinetics

Oral bioavailability is 55–75% and variable between preparations and patients. Plasma protein binding is low, and elimination is predominantly renal — largely unchanged drug through glomerular filtration and tubular secretion, with P-glycoprotein as the transporter involved. Elimination half-life is about 30–50 hours in normal renal function and lengthens substantially as it falls ^{12Reference 12Fender et al. · 2024ReviewFender, J., Klöcker, J., Boivin-Jahns, V., et al. (2024). “Cardiac glycosides”-quo vaditis?-past, present, and future? — review. Naunyn Schmiedebergs Arch Pharmacol. https://pubmed.ncbi.nlm.nih.gov/39007928/View study →}.

Two clinical consequences dominate. First, renal function is the dose: a falling eGFR raises the level on an unchanged prescription, which is the commonest route into chronic digoxin toxicity. Second, because the therapeutic window is defined by concentration rather than tablet strength, digoxin is one of the few oral drugs routinely monitored by serum level — with the DIG post hoc analysis pointing to 0.5–0.8 ng/mL rather than the older, higher targets ^{2,13Reference 2Rathore et al. · 2003RCTAssociation of serum digoxin concentration and outcomes in patients with heart failure — randomised controlled trialView study →Reference 13Whayne TF Jr · 2018Systematic reviewClinical Use of Digitalis: A State of the Art Review — systematic reviewView study →}.

Clinical Trials

Last checked: September 2026.

Isolate vs. Plant Studies

Every result above is the isolated drug at a measured dose, monitored by serum concentration. No controlled trial has ever tested a Digitalis leaf preparation, and the reason is the same one that ended the crude drug’s career: leaf glycoside content is low and swings several-fold with season, plant part, developmental stage and provenance ^{14Reference 14Ravi et al. · 2020High-resolution tandem mass spectrometry dataset reveals fragmentation patterns of cardiac glycosides in leaves of the foxglove plantsView study →}, so a weighed quantity is not a dose. Before chemical assay was possible, pharmacies standardised digitalis leaf by biological assay against animal hearts ^{19Reference 19Norn et al. · 2004Norn, S., Kruse, P. R. (2004). [Cardiac glycosides: From ancient history through Withering’s foxglove to endogeneous cardiac glycosides]. Dan Medicinhist Arbog. https://pubmed.ncbi.nlm.nih.gov/15685783/View study →}. A drug that requires a live animal to establish its strength cannot be dosed to a 0.5–0.8 ng/mL window, and once the glycosides could be purified and measured the leaf preparations were displaced entirely.

The species question is a second layer of the same problem. Digoxin is chiefly a Digitalis lanata glycoside; the common foxglove D. purpurea is dominated by digitoxin, gitoxin and gitaloxin, and carries digoxin only as a minor component — though enough to be detected in purpurea poisonings ^{14,15Reference 14Ravi et al. · 2020High-resolution tandem mass spectrometry dataset reveals fragmentation patterns of cardiac glycosides in leaves of the foxglove plantsView study →Reference 15Kwon et al. · 2011HPLC method validation for Digitalis and its analogue by pulsed amperometric detectionView study →}. The most famous digitalis evidence is therefore not evidence about the most famous digitalis plant. Any statement that “foxglove is the source of digoxin” needs the species qualifier, or the wrong plant inherits the literature.

Prevalence in Nature

Digoxin is the principal cardenolide of Digitalis lanata (woolly foxglove, a Balkan and central European species) and the reason that plant, rather than the ornamental D. purpurea, became the commercial source. In D. purpurea it is a minor constituent alongside digitoxin, gitoxin and gitaloxin — the last being the best chemical marker separating the two species ^{14Reference 14Ravi et al. · 2020High-resolution tandem mass spectrometry dataset reveals fragmentation patterns of cardiac glycosides in leaves of the foxglove plantsView study →}.

Cardenolides of the same class occur widely outside Digitalis: convallatoxin in Convallaria majalis, oleandrin in Nerium oleander, ouabain in Strophanthus, scillaren in Urginea maritima, and bufadienolides in toad skin. Mammals are also thought to produce endogenous digitalis-like compounds acting as hormonal regulators of Na⁺/K⁺-ATPase, though the measurement difficulties at those concentrations keep the field contested ^{12Reference 12Fender et al. · 2024ReviewFender, J., Klöcker, J., Boivin-Jahns, V., et al. (2024). “Cardiac glycosides”-quo vaditis?-past, present, and future? — review. Naunyn Schmiedebergs Arch Pharmacol. https://pubmed.ncbi.nlm.nih.gov/39007928/View study →}.

No content figure is quotable for either species. Leaf glycoside levels vary several-fold with season, part, developmental stage and provenance ^{14Reference 14Ravi et al. · 2020High-resolution tandem mass spectrometry dataset reveals fragmentation patterns of cardiac glycosides in leaves of the foxglove plantsView study →}; the analytical methods to measure them are well established ^{14,15Reference 14Ravi et al. · 2020High-resolution tandem mass spectrometry dataset reveals fragmentation patterns of cardiac glycosides in leaves of the foxglove plantsView study →Reference 15Kwon et al. · 2011HPLC method validation for Digitalis and its analogue by pulsed amperometric detectionView study →}, but the plant does not hold still.

Discovery & Synthesis

Foxglove entered orthodox medicine through William Withering’s Account of the Foxglove (1785), and the glycoside chemistry followed a century later. Digitoxin was purified from D. purpurea first; digoxin was isolated from Digitalis lanata in 1930, some sixty years later, and rapidly displaced both digitoxin and crude leaf across most of the world on the strength of its shorter half-life and easier titration ^{19Reference 19Norn et al. · 2004Norn, S., Kruse, P. R. (2004). [Cardiac glycosides: From ancient history through Withering’s foxglove to endogeneous cardiac glycosides]. Dan Medicinhist Arbog. https://pubmed.ncbi.nlm.nih.gov/15685783/View study →}.

Purification is what ended the plant preparations. Once the glycoside could be weighed and assayed, standardisation against a frog’s or pigeon’s heart became indefensible ^{19Reference 19Norn et al. · 2004Norn, S., Kruse, P. R. (2004). [Cardiac glycosides: From ancient history through Withering’s foxglove to endogeneous cardiac glycosides]. Dan Medicinhist Arbog. https://pubmed.ncbi.nlm.nih.gov/15685783/View study →}.

Manufacture is still extraction. Digoxin is obtained from Digitalis lanata leaf; the cardenolide skeleton with its three digitoxose sugars has never been an economical synthetic target, and the plant makes it well. Prescribing has fallen sharply across Europe over the past three decades as heart-failure therapy has expanded ^{20Reference 20Ecker et al. · 2026Critical analysis of digitalis glycosides: declining use but increasing poison-center exposure cases for digitoxinView study →}.

Patents: not yet researched.

Toxicity & Safety

Digoxin is flagged poisonous — a description of the purified drug as much as of the plant. Therapy and toxicity are the same pump inhibition at slightly different exposures.

Safety at a Glance

☠️ PoisonousCardiac cautionOverdose riskDrug interactionsPhytoestrogenic

• Narrow therapeutic index, defined by serum concentration rather than dose. Mortality rose stepwise with concentration across the DIG cohort ^{2Reference 2Rathore et al. · 2003RCTAssociation of serum digoxin concentration and outcomes in patients with heart failure — randomised controlled trialView study →}.
• Cardiac toxidrome — bradycardia, AV block, ventricular ectopy, bidirectional ventricular tachycardia. Hyperkalaemia in acute overdose ^{12Reference 12Fender et al. · 2024ReviewFender, J., Klöcker, J., Boivin-Jahns, V., et al. (2024). “Cardiac glycosides”-quo vaditis?-past, present, and future? — review. Naunyn Schmiedebergs Arch Pharmacol. https://pubmed.ncbi.nlm.nih.gov/39007928/View study →}.
• Renal function is the dose. A falling eGFR on an unchanged prescription is a rising digoxin level — the commonest route into chronic toxicity ^{12Reference 12Fender et al. · 2024ReviewFender, J., Klöcker, J., Boivin-Jahns, V., et al. (2024). “Cardiac glycosides”-quo vaditis?-past, present, and future? — review. Naunyn Schmiedebergs Arch Pharmacol. https://pubmed.ncbi.nlm.nih.gov/39007928/View study →}.
• Hypokalaemia potentiates toxicity without changing the level. A normal serum digoxin in a hypokalaemic patient is not reassurance ^{12Reference 12Fender et al. · 2024ReviewFender, J., Klöcker, J., Boivin-Jahns, V., et al. (2024). “Cardiac glycosides”-quo vaditis?-past, present, and future? — review. Naunyn Schmiedebergs Arch Pharmacol. https://pubmed.ncbi.nlm.nih.gov/39007928/View study →}.
• A real antidote exists — digoxin-specific Fab fragments reverse life-threatening digitalis toxicity ^{16Reference 16Antman et al. · 1990Clinical trialTreatment of 150 cases of life-threatening digitalis intoxication with digoxin-specific Fab antibody fragmentsView study →}.
• Xanthopsia — a yellow-green cast to vision with halos — is characteristic and easily dismissed ^{12Reference 12Fender et al. · 2024ReviewFender, J., Klöcker, J., Boivin-Jahns, V., et al. (2024). “Cardiac glycosides”-quo vaditis?-past, present, and future? — review. Naunyn Schmiedebergs Arch Pharmacol. https://pubmed.ncbi.nlm.nih.gov/39007928/View study →}.
• Interactions are transporter-mediated (P-glycoprotein), not CYP450. Verapamil, amiodarone, quinidine, macrolides and ketoconazole raise levels ^{12Reference 12Fender et al. · 2024ReviewFender, J., Klöcker, J., Boivin-Jahns, V., et al. (2024). “Cardiac glycosides”-quo vaditis?-past, present, and future? — review. Naunyn Schmiedebergs Arch Pharmacol. https://pubmed.ncbi.nlm.nih.gov/39007928/View study →}.
• An oestrogenic off-target effect. Current use is associated with increased incidence of oestrogen-sensitive cancers and with gynaecomastia in men ^{7,8,10Reference 7Biggar et al. · 2011ObservationalDigoxin use and the risk of breast cancer in women — cohort studyView study →Reference 8Biggar et al. · 2012ObservationalDigoxin use and the risk of cancers of the corpus uteri, ovary and cervix — cohort studyView study →Reference 10Biggar · 2012ReviewMolecular pathways: digoxin use and estrogen-sensitive cancers—risks and possible therapeutic implications — reviewView study →}.

Full Safety & Interactions Detail

The toxidrome. Poisoning opens with nausea, vomiting and abdominal pain, moves to visual disturbance — classically xanthopsia, a yellow-green cast with halos around lights — and confusion, and becomes dangerous in the cardiac and electrolyte findings: hyperkalaemia in acute overdose (because the inhibited pump is what moves potassium into cells), profound bradycardia and heart block, and ventricular arrhythmias including the characteristic bidirectional ventricular tachycardia ^{12Reference 12Fender et al. · 2024ReviewFender, J., Klöcker, J., Boivin-Jahns, V., et al. (2024). “Cardiac glycosides”-quo vaditis?-past, present, and future? — review. Naunyn Schmiedebergs Arch Pharmacol. https://pubmed.ncbi.nlm.nih.gov/39007928/View study →}. Serum potassium was the classic predictor of death in acute poisoning before specific antibody fragments existed.

The antidote. Digoxin-specific antibody fragments (Fab) bind circulating glycoside and reverse toxicity; the definitive multicentre series covered 150 cases of life-threatening digitalis intoxication ^{16Reference 16Antman et al. · 1990Clinical trialTreatment of 150 cases of life-threatening digitalis intoxication with digoxin-specific Fab antibody fragmentsView study →}. Their availability is why cardiac-glycoside poisoning is among the more survivable serious poisonings. (The long-taught warning that intravenous calcium causes a fatal “stone heart” in digitalis toxicity is widely repeated and contested in the modern literature — disputed rather than settled.)

Interactions run through transporters, not cytochromes — the standing mislabelling in this class. Digoxin is a P-glycoprotein substrate: verapamil, amiodarone, quinidine, macrolide antibiotics and ketoconazole raise serum levels. Separately, potassium-depleting diuretics, laxatives and corticosteroids potentiate toxicity without raising the level at all, and hypomagnesaemia and hypercalcaemia act in the same direction. Because both routes exist, neither a normal drug level nor an unchanged prescription rules out toxicity.

Renal function. Predominantly renal elimination means the dose must track the kidney. Acute kidney injury, dehydration, and the gradual decline of eGFR with age all raise exposure on an unchanged prescription, and elderly patients are correspondingly over-represented in cardiac-glycoside exposure reports ^{12,20Reference 12Fender et al. · 2024ReviewFender, J., Klöcker, J., Boivin-Jahns, V., et al. (2024). “Cardiac glycosides”-quo vaditis?-past, present, and future? — review. Naunyn Schmiedebergs Arch Pharmacol. https://pubmed.ncbi.nlm.nih.gov/39007928/View study →Reference 20Ecker et al. · 2026Critical analysis of digitalis glycosides: declining use but increasing poison-center exposure cases for digitoxinView study →}. Where renal function is poor and unstable, digitoxin — cleared largely independently of the kidney — is the alternative that has been proposed for exactly this reason ^{12,13Reference 12Fender et al. · 2024ReviewFender, J., Klöcker, J., Boivin-Jahns, V., et al. (2024). “Cardiac glycosides”-quo vaditis?-past, present, and future? — review. Naunyn Schmiedebergs Arch Pharmacol. https://pubmed.ncbi.nlm.nih.gov/39007928/View study →Reference 13Whayne TF Jr · 2018Systematic reviewClinical Use of Digitalis: A State of the Art Review — systematic reviewView study →}.

The oestrogenic effect is a real pharmacological property, not an epidemiological artefact. Digoxin binds oestrogen receptors and causes gynaecomastia. In a Danish nationwide cohort, current users had a breast-cancer risk ratio of 1.39 (95% CI 1.32–1.46), with no increase in former users — risk normalised on stopping — and no increase at all in a control group of women taking angina drugs, which rules out “being under cardiac care” as the explanation ^{7Reference 7Biggar et al. · 2011ObservationalDigoxin use and the risk of breast cancer in women — cohort studyView study →}. Uterine cancer, also oestrogen-sensitive, showed a risk ratio of 1.48 (1.32–1.65) in current users, while ovarian (1.06) and cervical (1.00) cancers — relatively oestrogen-insensitive — showed none ^{8Reference 8Biggar et al. · 2012ObservationalDigoxin use and the risk of cancers of the corpus uteri, ovary and cervix — cohort studyView study →}. Men on digoxin have a lower prostate-cancer risk (0.76), the direction an oestrogenic mechanism predicts ^{10Reference 10Biggar · 2012ReviewMolecular pathways: digoxin use and estrogen-sensitive cancers—risks and possible therapeutic implications — reviewView study →}. A separate UK nested case-control study (4,113 incident pancreatic cancers against 16,072 matched controls) found an adjusted odds ratio of 1.41 (1.16–1.72) for pancreatic cancer in active digoxin users, with no effect on survival once diagnosed (HR 0.97, 0.81–1.18) ^{11Reference 11Boursi et al. · 2021ObservationalDigoxin use is associated with pancreatic cancer risk but does not affect survival — case-control studyView study →}. In women already diagnosed with breast cancer, overall relapse risk was not significantly raised (HR 1.13, 0.88–1.46), though first-year relapse in oestrogen-receptor-positive tumours was (2.51, 1.39–4.55) ^{9Reference 9Biggar et al. · 2013ObservationalBreast cancer in women using digoxin: tumor characteristics and relapse risk — cohort studyView study →}.

Assay interference. Digoxin immunoassays cross-react with other cardenolides, so a measurable “digoxin level” can arise from foxglove, oleander or lily-of-the-valley ingestion in someone who has never taken the drug ^{21,22Reference 21Fink et al. · 2014In vitroRapid detection of convallatoxin using five digoxin immunoassays — in vitroView study →Reference 22Trebach et al. · 2022Case reportDetectable Digoxin Concentrations in 3 Patients with Ramps Misadventure — case reportView study →}. The result is diagnostically useful and numerically meaningless: it identifies a cardiac glycoside, but the number cannot gauge severity or guide the antidote dose.

Dosage

Prescription-only, and monitored by serum concentration rather than by dose. The figures below are the doses used in the trials.

The dose is a starting point and the serum concentration is the target — the DIG post hoc analysis found mortality rising stepwise across concentration bands, with harm above 1.2 ng/mL ^{2Reference 2Rathore et al. · 2003RCTAssociation of serum digoxin concentration and outcomes in patients with heart failure — randomised controlled trialView study →}. Renal function, age, body composition and interacting drugs all move the concentration on a fixed dose.

No plant-material equivalent can be stated. Digitalis leaf glycoside content varies several-fold ^{14Reference 14Ravi et al. · 2020High-resolution tandem mass spectrometry dataset reveals fragmentation patterns of cardiac glycosides in leaves of the foxglove plantsView study →}, the species differ in which glycoside dominates ^{14,15Reference 14Ravi et al. · 2020High-resolution tandem mass spectrometry dataset reveals fragmentation patterns of cardiac glycosides in leaves of the foxglove plantsView study →Reference 15Kwon et al. · 2011HPLC method validation for Digitalis and its analogue by pulsed amperometric detectionView study →}, and a drug titrated to a 0.5–0.8 ng/mL window cannot be approached with plant material at all.

References

1. Digitalis Investigation Group (1997). The effect of digoxin on mortality and morbidity in patients with heart failure — randomised controlled trial. N Engl J Med. https://pubmed.ncbi.nlm.nih.gov/9036306/
2. Rathore, S. S., Curtis, J. P., Wang, Y., et al. (2003). Association of serum digoxin concentration and outcomes in patients with heart failure — randomised controlled trial. JAMA. https://pubmed.ncbi.nlm.nih.gov/12588271/
3. Ziff, O. J., Lane, D. A., Samra, M., et al. (2015). Safety and efficacy of digoxin: systematic review and meta-analysis of observational and controlled trial data — meta-analysis. BMJ. https://pubmed.ncbi.nlm.nih.gov/26321114/
4. Vamos, M., Erath, J. W., Hohnloser, S. H. (2015). Digoxin-associated mortality: a systematic review and meta-analysis of the literature — meta-analysis. Eur Heart J. https://pubmed.ncbi.nlm.nih.gov/25939649/
5. Vamos, M., Erath, J. W., Benz, A. P., et al. (2019). Meta-Analysis of Effects of Digoxin on Survival in Patients with Atrial Fibrillation or Heart Failure: An Update — meta-analysis. Am J Cardiol. https://pubmed.ncbi.nlm.nih.gov/30539748/
6. Kotecha, D., Bunting, K. V., Gill, S. K., et al. (2020). Effect of Digoxin vs Bisoprolol for Heart Rate Control in Atrial Fibrillation on Patient-Reported Quality of Life: The RATE-AF Randomized Clinical Trial — randomised controlled trial. JAMA. https://pubmed.ncbi.nlm.nih.gov/33351042/
7. Biggar, R. J., Wohlfahrt, J., Oudin, A., et al. (2011). Digoxin use and the risk of breast cancer in women — cohort study. J Clin Oncol. https://pubmed.ncbi.nlm.nih.gov/21422417/
8. Biggar, R. J., Wohlfahrt, J., Melbye, M. (2012). Digoxin use and the risk of cancers of the corpus uteri, ovary and cervix — cohort study. Int J Cancer. https://pubmed.ncbi.nlm.nih.gov/21913187/
9. Biggar, R. J., Andersen, E. W., Kroman, N., et al. (2013). Breast cancer in women using digoxin: tumor characteristics and relapse risk — cohort study. Breast Cancer Res. https://pubmed.ncbi.nlm.nih.gov/23421975/
10. Biggar, R. J. (2012). Molecular pathways: digoxin use and estrogen-sensitive cancers—risks and possible therapeutic implications — review. Clin Cancer Res. https://pubmed.ncbi.nlm.nih.gov/22368159/
11. Boursi, B., Huber, J. S., Haynes, K., et al. (2021). Digoxin use is associated with pancreatic cancer risk but does not affect survival — case-control study. Cancer Causes Control. https://pubmed.ncbi.nlm.nih.gov/33064241/
12. Fender, J., Klöcker, J., Boivin-Jahns, V., et al. (2024). “Cardiac glycosides”-quo vaditis?-past, present, and future? — review. Naunyn Schmiedebergs Arch Pharmacol. https://pubmed.ncbi.nlm.nih.gov/39007928/
13. Whayne TF Jr (2018). Clinical Use of Digitalis: A State of the Art Review — systematic review. Am J Cardiovasc Drugs. https://pubmed.ncbi.nlm.nih.gov/30066080/
14. Ravi, B. G., Guardian, M. G. E., Dickman, R., et al. (2020). High-resolution tandem mass spectrometry dataset reveals fragmentation patterns of cardiac glycosides in leaves of the foxglove plants. Data Brief. https://pubmed.ncbi.nlm.nih.gov/32300626/
15. Kwon, H. J., Sim, H. J., Lee, S. I., et al. (2011). HPLC method validation for Digitalis and its analogue by pulsed amperometric detection. J Pharm Biomed Anal. https://pubmed.ncbi.nlm.nih.gov/20800985/
16. Antman, E. M., Wenger, T. L., Butler VP Jr, et al. (1990). Treatment of 150 cases of life-threatening digitalis intoxication with digoxin-specific Fab antibody fragments. Final report of a multicenter study — clinical trial. Circulation. https://pubmed.ncbi.nlm.nih.gov/2188752/
17. Bejček, J., Spiwok, V., Kmoníčková, E., et al. (2021). Na(+)/K(+)-ATPase Revisited: On Its Mechanism of Action, Role in Cancer, and Activity Modulation — review. Molecules. https://pubmed.ncbi.nlm.nih.gov/33800655/
18. Škubník, J., Bejček, J., Pavlíčková, V. S., et al. (2021). Repurposing Cardiac Glycosides: Drugs for Heart Failure Surmounting Viruses — review. Molecules. https://pubmed.ncbi.nlm.nih.gov/34577097/
19. Norn, S., Kruse, P. R. (2004). [Cardiac glycosides: From ancient history through Withering’s foxglove to endogeneous cardiac glycosides]. Dan Medicinhist Arbog. https://pubmed.ncbi.nlm.nih.gov/15685783/
20. Ecker, P. A. M., Schaper, A., Bräunig, H. J., et al. (2026). Critical analysis of digitalis glycosides: declining use but increasing poison-center exposure cases for digitoxin. Eur J Clin Pharmacol. https://pubmed.ncbi.nlm.nih.gov/42334625/
21. Fink, S. L., Robey, T. E., Tarabar, A. F., et al. (2014). Rapid detection of convallatoxin using five digoxin immunoassays — in vitro. Clin Toxicol (Phila). https://pubmed.ncbi.nlm.nih.gov/24980812/
22. Trebach, J., Calleo, V., Akbar, S., et al. (2022). Detectable Digoxin Concentrations in 3 Patients with Ramps Misadventure — case report. Wilderness Environ Med. https://pubmed.ncbi.nlm.nih.gov/35691768/

View all 22 references▾