Drug Half-Life

What half-life means

The half-life of a drug, t_½, is the time required for plasma concentration to fall by 50%. For drugs that follow first-order kinetics — and most do, at therapeutic doses — this number is a property of the drug-and-patient pair, not of the dose. It tells you two essential things at once: how long the drug persists after stopping, and how quickly it accumulates to steady state when dosed repeatedly.

The formula that derives half-life from the two more fundamental pharmacokinetic parameters is:

t_½ = 0.693 × V_d / CL

Where:

• V_d (volume of distribution) is an apparent volume — the volume that would be required to hold the entire administered dose at the measured plasma concentration. Small V_d (~4 L) means the drug stays in plasma; large V_d (1,000 L) means it hides in tissues like fat or muscle.
• CL (clearance) is the volume of plasma cleared of drug per unit time, usually by liver and kidney combined.
• 0.693 = ln(2) appears because half-life is defined relative to 50%, not to 1/e (the natural decay constant).

A long half-life can arise either because V_d is enormous (amiodarone is sequestered into fat — V_d ~5,000 L) or because clearance is small (warfarin is slowly metabolized — CL ~5 mL/min). Two different mechanisms, same downstream consequence: a drug that hangs around.

The math of exponential decay

First-order kinetics means the rate of elimination is proportional to the amount present. The plasma concentration over time is:

C(t) = C₀ × e^−kt = C₀ × (½)^t/t½

Where k = ln(2) / t_½ is the elimination rate constant. After integer numbers of half-lives:

• 1 t_½ → 50.0% remaining
• 2 t_½ → 25.0%
• 3 t_½ → 12.5%
• 4 t_½ → 6.25%
• 5 t_½ → 3.125%
• 6 t_½ → 1.56%
• 7 t_½ → 0.78%

Clinicians use "5 half-lives" as the rule of thumb for complete clearance — the residual ~3% rarely matters. The same number applies in reverse for accumulation to steady state. After 5 half-lives of repeat dosing, plasma concentration sits at about 97% of its eventual steady-state value.

Half-lives of common drugs

• Adenosine — ~10 seconds. Used for IV cardioversion of supraventricular tachycardia; the dose is given as a fast IV push because by the time it would reach the heart through a slow infusion, it's gone.
• Nitroglycerin — ~2 minutes. Sublingual tablet bursts in fast, washes out fast — perfect for an angina attack, useless for chronic prevention without a sustained-release formulation.
• Aspirin — ~15–20 minutes (parent compound). Its active metabolite salicylate has a longer half-life that varies with dose (4 hours at 600 mg, up to 20 hours at toxic doses due to saturable elimination).
• Ibuprofen — ~2 hours. Dosed 3–4 times daily for chronic pain.
• Acetaminophen — ~2–3 hours. Dosed every 4–6 hours.
• Metoprolol — ~3–7 hours (immediate-release tartrate). Extended-release succinate is dosed once daily through formulation engineering.
• Warfarin — ~36–42 hours. Steady state takes 5–7 days. INR doesn't catch up to dose changes for 3–5 days, which is why warfarin dosing is so slow and unforgiving.
• Fluoxetine — ~2–3 days (parent), 7–15 days (norfluoxetine metabolite). Because of the long t_½, no taper is needed when switching to another SSRI — the washout is automatic.
• Diazepam — ~30–60 hours, with active metabolites (nordiazepam, oxazepam) that extend the effective duration. In elderly patients, accumulation can take days to manifest.
• Methadone — ~15–60 hours, highly variable. The slow elimination is responsible both for methadone's long-lasting analgesia and for the risk of late respiratory depression — peak respiratory effects can occur after pain has worn off.
• Amiodarone — ~25–100 days. Effects continue for weeks after stopping. Drug interactions and side effects (pulmonary fibrosis, thyroid dysfunction) outlast discontinuation.

Steady state and dosing interval

On repeat dosing, each dose adds to whatever drug remains from the previous one. If you dose every half-life and the daily dose equals the daily clearance, the system converges to a steady state where peak and trough are constant from dose to dose. Average steady-state concentration is:

C_ss,avg = (F × Dose) / (CL × τ)

Where F is bioavailability, Dose is the dose per interval, CL is clearance, and τ is the dosing interval.

The interval τ is usually set close to t_½. Choose τ << t_½ and the drug accumulates dangerously. Choose τ >> t_½ and plasma concentration crashes between doses, dipping below the therapeutic window. The exception: extended-release formulations let short-half-life drugs be dosed once daily by slowing absorption to match elimination, smoothing out the peaks and troughs.

Loading dose

Waiting 5 half-lives to reach steady state is fine for chronic prevention of angina or hypertension. It's unacceptable in sepsis, status epilepticus, or new-onset atrial fibrillation where you need the drug working now.

The fix is a loading dose — a one-time front-loaded dose that fills V_d to the target concentration, after which maintenance dosing replaces what is lost each interval. The basic formula:

Loading dose ≈ V_d × C_target

For amiodarone (V_d ~66 L/kg), loading is over 6,000 mg in the first day. For digoxin (V_d ~5 L/kg in adults), loading is 8–12 µg/kg IV. For vancomycin (V_d ~0.7 L/kg), loading is 25–30 mg/kg in severe sepsis. The loading dose is not just "the same drug, more of it" — it is specifically chosen to saturate V_d.

Why half-lives vary so much across drugs

The 10⁹-fold range from adenosine (10 s) to amiodarone (100 d) reflects the underlying biology:

• Adenosine is destroyed by adenosine deaminase, ubiquitous in red blood cells and vascular endothelium. CL is enormous; V_d is tiny because it's metabolized everywhere.
• Warfarin binds tightly to albumin (V_d small) but is slowly metabolized by CYP2C9 (CL small). Net effect: long t_½.
• Amiodarone is so lipophilic that V_d is in the thousands of liters. Even with adequate clearance, the math gives a half-life in months.
• Acetaminophen is rapidly conjugated. CL is high; V_d is moderate. Half-life is short.
• Lithium is renally cleared with no metabolism. Its half-life depends almost entirely on glomerular filtration; renal impairment dramatically extends it.

First-order vs zero-order kinetics

Most drugs follow first-order kinetics at therapeutic doses — a constant fraction is eliminated per unit time, and half-life is constant. But some drugs saturate their elimination enzymes within or near the therapeutic range, switching to zero-order kinetics (constant amount eliminated per time). Half-life is no longer constant.

Important zero-order or capacity-limited examples:

• Ethanol — alcohol dehydrogenase saturates around 5 mg/dL. Above that, elimination proceeds at a fixed rate of ~10–20 mg/dL per hour regardless of starting concentration.
• Phenytoin — CYP2C9 saturates at therapeutic concentrations. Small dose increases can cause large jumps in plasma level; clinical decisions use the Michaelis-Menten Vmax/Km model.
• Aspirin at high doses — salicylate glucuronidation saturates, extending half-life from 4 to 20+ hours in overdose. This drives the need for alkaline diuresis in salicylate poisoning.
• Theophylline at high concentrations — partially capacity-limited.

For these drugs, "half-life" is a moving target. Clinicians use direct concentration measurement and population PK models instead.

Where the concept came from

The fundamentals of drug elimination kinetics were worked out by Torsten Teorell in Sweden in the 1930s, treating the body as a series of compartments with first-order transfer rates. The modern formal expression of clearance and the linkage to half-life came from Malcolm Rowland, Leslie Benet, and others in the 1960s and 1970s, who built the framework now taught in every pharmacy and medical school. Rowland and Tozer's Clinical Pharmacokinetics textbook (first edition 1980) made it standard curriculum.

The half-life concept itself is borrowed from radioactive decay — the same mathematics. Radium-226 has a half-life of 1,600 years; aspirin has a half-life of 15 minutes; they're governed by the same first-order differential equation, just with different rate constants.

Common misconceptions

• "The drug is completely gone after one half-life." Half-life means 50% gone, not 100%. Five half-lives is the practical "complete elimination" benchmark (~97%).
• "Half-life and duration of action are the same." Not for irreversible drugs. Aspirin's t_½ is ~15 minutes but its antiplatelet effect lasts the platelet's lifetime (~7–10 days) because it permanently acetylates COX-1. Omeprazole has a short t_½ but persistent acid suppression because it irreversibly inhibits the proton pump. For these, the relevant time constant is enzyme/protein turnover, not drug t_½.
• "All drugs reach steady state in the same number of half-lives." True for first-order drugs, but only between repeat-dose intervals matched to t_½. Drugs given as long IV infusions can reach steady state more linearly. Drugs with zero-order kinetics never reach a clean steady state.
• "Half-life is constant for a drug across all patients." No — Vd and CL vary with age, weight, organ function, genetics, drug interactions, pregnancy, and disease. Population PK estimates have wide ranges. "Warfarin t_½ 36–42 hours" is an average; individuals span 20–80 hours.
• "The longer the half-life, the safer the drug." Not at all. Long-t_½ drugs are harder to fix when they cause toxicity — you cannot just stop and have the effect wear off. Amiodarone toxicity (pulmonary, thyroid, cornea) persists for weeks after stopping.
• "Crushing extended-release tablets just speeds onset." It can also produce immediate dumping of a multi-day dose, blowing past the relationship between t_½ and dosing interval. Tragic crushed-OxyContin and crushed-Lithium-CR overdoses are well documented.

Half-lives and dosing of representative drugs

Drug	Half-life	Typical interval	Time to steady state	Notes
Adenosine	~10 s	IV push, single	n/a	Must be IV bolus; degraded in blood
Nitroglycerin	~2 min	SL PRN; IV drip	~10 min (IV)	Tachyphylaxis with continuous use
Aspirin (parent)	~15–20 min	q4–6h	~1 hr	Antiplatelet effect lasts platelet life (~7–10 d)
Ibuprofen	~2 h	q6–8h	~10 h	—
Metoprolol IR	~3–7 h	q12h	~1–2 d	ER succinate dosed q24h
Warfarin	~36–42 h	q24h	~5–7 d	INR lags dose changes by 3–5 d
Fluoxetine + metabolite	2–3 d + 7–15 d	q24h	~4–6 wk	Self-tapering on discontinuation
Diazepam (+ metabolites)	30–60 h+	q12–24h	~7–10 d	Accumulates in elderly
Methadone	15–60 h	q24h	~5–10 d	Late respiratory depression risk
Amiodarone	25–100 d	q24h after load	Weeks – months	Loading dose essential