What Is a Half-Life?

A half-life is the time it takes for half of something to disappear. If a compound has a half-life of four hours, then four hours after it arrives, half of it is gone. That is the whole idea, and it is probably the single most useful number in this subject.

It is also misunderstood in two specific places. The first is arithmetic: people expect the second half to leave the way the first half did, and it does not. The second is a word problem, and it is the reason this article exists. Half-life describes how long a molecule survives inside a living body. It keeps getting confused with how long a product stays good in storage. Those are different clocks, driven by completely different things, and the last section here is about nothing else.

Count it out with real numbers

Abstract explanations of half-life are much harder than counting, so let us count. Imagine a thousand units of something arriving in the bloodstream, and imagine its half-life is four hours.

Read down that last column and the shape of the thing is obvious. The first four hours remove five hundred units. The next four hours remove two hundred and fifty. By twenty hours, five half-lives have passed and about three percent of the original is left.

That last figure is where a familiar rule of thumb comes from. People say a compound is effectively gone after four or five half-lives, and the table is the reason 1. Strictly speaking it never reaches zero, since halving something forever never quite empties it. Practically, it falls below the point where anyone can measure it.

The same arithmetic runs backwards, and this surprises people. If something is given repeatedly rather than once, it takes about the same four or five half-lives for the amount in the body to climb to a steady level, where what arrives each interval matches what leaves 1. The number that tells you how long something takes to clear also tells you how long it takes to build up.

One clarification that saves confusion later. A half-life describes an enormous population of molecules, not the fate of any one of them. Some are removed in the first minute and some last far longer than average. The number describes the shape of the crowd.

A constant proportion, not a constant amount

Here is the mistake almost everyone makes on first meeting, and it is worth making deliberately so you can watch it fail.

Five hundred units left in the first four hours. So another five hundred should leave in the next four hours, and the whole thing should be gone by hour eight. Neat, symmetrical, and wrong. At hour eight there are two hundred and fifty units left, and they are in no hurry.

The reason is that removal is not a fixed quota. The machinery doing the removing — enzymes cutting the molecule apart, kidneys filtering it out — works on whatever is in front of it. When a lot is present, a lot passes through per hour. When little is present, little does. What stays constant is not the amount removed but the fraction of what remains.

A bucket with a hole in the bottom behaves the same way. Fill it and it drains quickly, because the weight of water above the hole pushes hard. As the level drops the flow slows, and it slows in proportion. The time it takes for the depth to halve is the same whether you start with a full bucket or a half-full one. That is the whole mechanism, and there is nothing more mysterious behind it.

This produces a genuinely counterintuitive consequence. Doubling the starting amount does not double the time to clear it. It adds exactly one half-life. Two thousand units instead of one thousand does not mean forty-eight hours instead of twenty-four; it means twenty-eight, because the extra thousand disappears in the first four hours and then you are back on the original schedule.

Behaviour like this — proportional, always the same fraction — is called first-order, and most compounds at most levels follow it. One honest exception is worth knowing. If more molecules are present than the removal machinery can handle at once, that machinery saturates: it runs flat out at a fixed maximum rate, removing a constant amount per hour rather than a constant proportion, and the tidy halving breaks down until levels fall far enough for it to resume 1. Alcohol is the familiar example of something that leaves at a roughly fixed rate, which is why quoting a single half-life for it is not straightforward.

What a short half-life means in practice

Peptides tend towards the short end, and some of them are startlingly short. The hormone GLP-1, which your own gut releases while you eat, is trimmed and switched off by an enzyme in blood within a couple of minutes of being released 3.

Think about what that does to anyone trying to study it. A molecule with a two-minute half-life cannot be given once and observed over an afternoon, because within the hour there is essentially nothing left to observe. It has to be delivered continuously, or measured in a very tight window, and every blood sample has to be handled fast enough that the enzymes in the tube do not keep working on it.

So half-life shapes research before any question about effect is even reached. It determines how often something can be given, how often samples are taken, how long a session runs, and sometimes whether a study is feasible at all. When you read that a compound was infused rather than injected, the half-life is usually the reason.

It also shapes what levels look like over time. A short half-life gives a sharp peak and a rapid fall: a brief period of high concentration and then very little. A long half-life gives something flatter, but with a cost at both ends, since it takes several half-lives to build up to a stable level and several more to leave once it stops. Short-lived is inconvenient and, in exactly one respect, forgiving: whatever is happening will stop happening soon. A long-lived molecule cannot be recalled.

One caution matters more than almost anything else here. A half-life belongs to a molecule in a particular species, not to the molecule alone. Small animals clear things faster, so a compound's half-life in a mouse is usually far shorter than in a person, and for many research peptides no human figure has ever been published at all.

Why some peptides are built to last longer

If a molecule does something useful but vanishes in two minutes, the obvious move is to make a version that does the same thing for longer. There are three main ways to do it, and none of them require changing what the molecule does when it arrives 2.

• Protect the cut site. Enzymes that chop peptides usually recognise a short stretch of the chain and cut at a specific point in it. Swapping a single amino acid at that point can make the chain unrecognisable to that enzyme, so the cut never happens.
• Hitch a ride on something long-lived. Albumin is the most abundant protein in blood and survives for weeks. Attaching a fatty tail to a peptide makes it cling to albumin, which shelters it from enzymes, keeps it too bulky for the kidneys to filter out, and releases it gradually.
• Make it bigger. Kidneys filter by size, so anything below a certain size is removed quickly. Attaching a large inert chemical chain, or fusing the peptide to part of an antibody, pushes it above that threshold.

The best-known worked example combines the first two. Natural GLP-1 lasts minutes; the modified analogue semaglutide was designed with an alteration at the enzyme's cut site and a fatty acid attachment that binds albumin, and the result is a molecule intended for weekly rather than continuous delivery 3.

Notice what did not change. The receptor it fits is the same receptor. What it does on arrival is the same thing. All the engineering went into durability. This is worth internalising, because half-life is largely an engineering property rather than a statement about how strong or how useful a molecule is, and the two are often quietly conflated in write-ups.

Half-life in the body is not shelf life in the vial

This is the confusion that sends people wrong most often, and it is entirely understandable, because both numbers are stated in units of time, both get filed under the word stability, and both sound like an answer to how long does it last?

They are answers to different questions, and neither one predicts the other.

Half-life measures how quickly a living body removes a molecule. It requires a body: enzymes to cut, kidneys to filter, a liver to alter, blood to carry things around. Take away the living system and the number has nothing to describe. A sealed container on a shelf has no half-life, because nothing in there is removing anything.

Shelf life measures something else entirely: how long the material in a container remains chemically what the label says it is. No biology is involved at all. The threats are ordinary chemistry and physics — water, oxygen, light, heat, and molecules bumping into one another over months 4.

Those threats have names worth recognising. Hydrolysis is water slowly cutting the chain — the same bond a protease attacks, but taking months rather than seconds because no enzyme is speeding it along. Oxidation is oxygen attacking particular amino acids. Deamidation is a slow spontaneous change at certain positions that alters the molecule without breaking it in two. Aggregation is molecules clumping into larger assemblies that no longer behave as individual molecules do 45. All of these run faster when it is warm, and most run faster in the presence of water.

Put the two side by side and the independence becomes obvious. A peptide with a two-minute half-life can sit perfectly stable as a dry powder in a freezer for years, because there is no body present to remove it. A peptide painstakingly engineered for a week-long half-life will still degrade in a badly stored container, because protecting a molecule from enzymes does nothing whatsoever about oxygen.

One last distinction inside the distinction. A stated shelf life is a claim about one particular preparation kept in one particular way. Change the container, the temperature, or whether the material is dry or dissolved, and the number changes with it, which is why storage conditions are always quoted alongside it. A half-life is closer to a property of the molecule itself, but even that shifts with the species it is measured in and with how well a given individual's kidneys and liver are working.

So when you meet a duration attached to a compound, the useful habit is to finish the question rather than accept the number. How long does it last where — inside a living body, or sealed in a container on a shelf? Almost every muddle about these two numbers dissolves the moment somebody asks that out loud.