What does half-life mean?

Half of what remains

Imagine a concentration of 100 units with a six-hour half-life. In a simple model, 50 units remain after six hours, 25 after 12 hours, 12.5 after 18 hours and 6.25 after 24 hours. Each interval halves what remains. The amount removed gets smaller, even though the proportion stays the same.

This is an invented mathematical example, not a measurement for a medicine or a dosing guide. In a simple first-order model, four or five half-lives leave only a small fraction of the starting concentration, but not necessarily zero. The example also assumes the drug is already in the place being measured. After a dose, absorption can make the concentration rise before elimination dominates.6

The assumptions matter

The simple curve assumes a constant fraction is eliminated over each interval in the part of the curve being described. Half-life is related to clearance—how the body removes a drug—and the apparent volume over which it is distributed. In a simplified one-compartment model, half-life is approximately 0.693 times the apparent distribution volume divided by clearance. This describes a model; it is not a personal calculator.

Real concentration curves can have an early distribution phase and a slower later phase. Absorption can be delayed, breakdown products can remain active, and elimination may not follow the simple model. FDA labelling guidance says a long terminal half-life should be explained when relevant to safety or effectiveness. It says the reported half-life should usually be the effective half-life based on time to steady state. Ask which phase, sample and analysis produced the number.1246

Concentration and effect are different measurements

A blood-plasma half-life describes how concentration changes over time. It does not by itself tell you when an effect starts, peaks or ends. An effect may lag while a drug reaches its target or changes a later biological process. Target binding, a lasting biological change or an active breakdown product may extend an effect after the blood concentration falls. Conversely, measurable drug can remain after a clinically important effect ends.

FDA labelling guidance asks for the onset, peak, reversibility and duration of effects when relevant. A long half-life can reduce fluctuations, but can also prolong exposure after an unwanted effect or interaction. “Half has gone” does not necessarily mean “half the benefit has gone.”34

A number can hide more than one half-life

A reported half-life depends on what was measured and which part of the concentration curve was analysed. After an intravenous dose, a drug may leave the bloodstream quickly as it spreads into tissues, then decline more slowly as it is eliminated. The early distribution half-life and later terminal half-life answer different questions.

After an oral dose, slow absorption can also shape the apparent half-life; this is sometimes called flip-flop kinetics. A result may describe plasma, whole blood, a breakdown product or an effect compartment. Two studies can report different numbers without either being arithmetically wrong. Compare the route, sample, measurement period and analysis.146

What the concept is useful for

With repeated doses, some of the previous dose may remain when the next arrives. Concentrations rise and fall toward a repeated pattern called steady state, where average input and elimination balance. Under an ideal first-order model, about half of the eventual change has happened after one half-life, and four to five half-lives often gets close. The dose, interval, absorption and formulation still affect peaks and troughs.

After stopping, four or five half-lives may reduce the original concentration to roughly 6% or less in that simple model. This does not prove there is no drug or effect left. Kidney or liver function, age, interacting medicines and disease can change clearance.

FDA guidance on multiple-dose studies in people with kidney impairment discusses considering lower or less frequent doses to limit accumulation and a loading strategy when a greatly prolonged half-life slows reaching steady state. These are study-design considerations, not a rule for every patient or medicine. Do not use half-life alone to calculate your schedule, decide when another substance is safe or change prescribed treatment.51