Half-Life and Concentration-Time Interpretation

What Half-Life Means in a One-Compartment Model

A drug concentration does not fall by the same absolute amount in each equal time interval under first-order elimination. Instead, in a one-compartment model after an intravenous bolus, it falls by the same proportion over each equal interval. The half-life is the time required for that concentration to decrease to one-half its value, provided the model's assumptions apply.

This definition connects a useful clinical interpretation to a specific mathematical model. Half-life is not automatically a fixed property of every observed concentration curve: the exponential relationship, and therefore the simple half-life interpretation, depends on the structural model and elimination assumptions.

Definition: For a one-compartment IV bolus model with first-order elimination, concentration at time \(t\) after the bolus is \(C(t)=C_0 e^{-kt}\), where \(C_0\) is concentration at time zero and \(k>0\) is the elimination rate constant, in inverse time units. The elimination half-life, \(t_{1/2}\), is the time interval that satisfies \(C(t+t_{1/2})=C(t)/2\). It is measured in the same time units used for \(t\).

The definition applies at any time on this model curve: after one half-life, the concentration is half the concentration at the beginning of that interval. It does not mean that half of the drug is necessarily removed from the entire body at that moment under every possible model. Here, concentration tracks the amount in a single, well-mixed compartment with a fixed volume.

Deriving the Half-Life Formula

Starting with the one-compartment exponential equation, apply the definition at an arbitrary time \(t\). The concentration \(t_{1/2}\) later is \(C_0 e^{-k(t+t_{1/2})}\), and it must equal one-half of \(C_0 e^{-kt}\).

C_0e^{-k(t+t_{1/2})}=\frac{1}{2}C_0e^{-kt}

Cancel the positive common factor \(C_0e^{-kt}\). Taking natural logarithms then gives the half-life formula:

e^{-kt_{1/2}}=\frac{1}{2} \quad\Longrightarrow\quad -kt_{1/2}=\ln\left(\frac{1}{2}\right)=-\ln 2 \quad\Longrightarrow\quad t_{1/2}=\frac{\ln 2}{k}

In the one-compartment model, the elimination rate constant is \(k=CL/V\), where \(CL\) is clearance and \(V\) is the compartment volume. Substituting gives \(t_{1/2}=(\ln 2)V/CL\). This relationship shows that, with clearance held fixed, a larger volume corresponds to a longer half-life; with volume held fixed, a larger clearance corresponds to a shorter half-life. The units confirm the interpretation: if \(CL\) is in L/h and \(V\) is in L, \(CL/V\) is in h\(^{-1}\), so \(t_{1/2}\) is in hours.

Key result: Under one-compartment disposition and first-order elimination, \(t_{1/2}=\ln(2)/k\). Since \(\ln(2)\approx0.6931\), the half-life is approximately \(0.6931/k\). If \(k=CL/V\), the equivalent expression is \(t_{1/2}=0.6931V/CL\). These expressions require a positive, constant \(k\).

Worked Example: Calculating Half-Life From Clearance and Volume

Suppose a hypothetical one-compartment IV bolus model has \(CL=5.00\) L/h and \(V=50.0\) L. Calculate the elimination rate constant and half-life, then predict the concentration after one half-life if the dose is 500 mg.

Check the model conditions. For this calculation, assume the drug is distributed instantaneously through the single compartment, clearance and volume are constant, and elimination is first-order. The dose is given as an IV bolus, with no input continuing after time zero.

First calculate the elimination rate constant:

k=\frac{CL}{V} =\frac{5.00\ \mathrm{L/h}}{50.0\ \mathrm{L}} =0.100\ \mathrm{h}^{-1}

Then calculate half-life:

t_{1/2}=\frac{0.6931}{0.100\ \mathrm{h}^{-1}} =6.931\ \mathrm{h}

Under the stated model, initial concentration is dose divided by volume:

C_0=\frac{500\ \mathrm{mg}}{50.0\ \mathrm{L}} =10.0\ \mathrm{mg/L}

After 6.931 h the predicted concentration is \(10.0/2=5.00\) mg/L. The result is a model prediction, not a guarantee that a measured concentration will be exactly 5.00 mg/L.

Estimating Half-Life From Concentrations

If \(k\) is not given, the exponential equation can be rearranged using two concentrations observed at times \(t_1\) and \(t_2\). The estimate requires positive concentrations and \(t_2>t_1\). Under a mono-exponential decline, the slope of log concentration against time is \(-k\); a two-point calculation estimates that slope from the two observations.

\ln C(t)=\ln C_0-kt \qquad\Longrightarrow\qquad \hat{k}=\frac{\ln C(t_1)-\ln C(t_2)}{t_2-t_1} \qquad\Longrightarrow\qquad \hat{t}_{1/2}=\frac{\ln 2}{\hat{k}}

The hat indicates an estimate from observations rather than a known model parameter. The expression for \(\hat{k}\) is positive when concentration decreases between the observations. With real data, measurement error can make any pair of points unrepresentative. When several observations lie on the same log-linear segment, fitting a line to log concentration versus time uses more information than relying on one pair. The fitted slope is still interpretable as \(-k\) only if that segment is reasonably mono-exponential and the observations are appropriate for the model.

Worked Example: Estimating Half-Life From Two Observations

In a hypothetical IV bolus profile, concentrations are 8.00 mg/L at 2.00 h and 4.00 mg/L at 8.00 h. Estimate \(k\) and the half-life from these two observations.

Check the model conditions. Assume both observations are on the same one-compartment, first-order elimination phase, with no ongoing drug input and positive concentrations. The observed concentrations decline, as required for a positive elimination rate constant.

The time difference is \(8.00-2.00=6.00\) h. Substituting the concentrations:

\hat{k} =\frac{\ln(8.00\ \mathrm{mg/L})-\ln(4.00\ \mathrm{mg/L})}{6.00\ \mathrm{h}} =\frac{\ln(2)}{6.00\ \mathrm{h}} =\frac{0.6931}{6.00\ \mathrm{h}} =0.1155\ \mathrm{h}^{-1}

The concentration units cancel inside the logarithmic ratio. The estimated half-life is:

\hat{t}_{1/2} =\frac{0.6931}{0.1155\ \mathrm{h}^{-1}} =6.00\ \mathrm{h}

Here the estimate is exactly the observation interval because the second concentration is exactly half the first. With noisy data, the ratio may not be exactly one-half, and the estimate from just two points may be sensitive to measurement error.

Predicting Concentration Changes Across Half-Lives

The exponential equation can be written in terms of half-life by replacing \(k\) with \(\ln(2)/t_{1/2}\). If \(m\) half-lives have elapsed, concentration is multiplied by \(2^{-m}\). The number \(m\) need not be an integer, so the same relationship predicts concentrations between whole half-life intervals.

\frac{C(t)}{C_0} =e^{-kt} =e^{-(\ln 2)t/t_{1/2}} =2^{-t/t_{1/2}}

Elapsed time	Fraction of starting concentration remaining
1 half-life	1/2 = 50%
2 half-lives	1/4 = 25%
3 half-lives	1/8 = 12.5%
4 half-lives	1/16 = 6.25%

These percentages describe concentration relative to its starting value under this model. They do not indicate that concentration reaches zero after a finite number of half-lives. For any finite \(t\), the exponential \(C_0e^{-kt}\) is positive if \(C_0>0\), although it may become very small.

Worked Example: Predicting Concentration After Fractional Half-Lives

Suppose a hypothetical one-compartment concentration is 12.0 mg/L at the start of a period and the applicable half-life is 4.00 h. Predict the concentration 10.0 h later and calculate the percentage decrease.

Check the model conditions. The calculation assumes the same first-order, one-compartment elimination phase applies throughout the 10.0 h, with no additional drug input and no change in the model parameters.

The elapsed time represents \(10.0/4.00=2.50\) half-lives. Thus:

C(10.0\ \mathrm{h}) =12.0\ \mathrm{mg/L}\times 2^{-2.50} =\frac{12.0\ \mathrm{mg/L}}{5.657} =2.121\ \mathrm{mg/L}

The fraction remaining is \(2^{-2.50}=0.1768\), or 17.68%. Therefore, the percentage decrease is \(100\%-17.68\%=82.32\%\). The concentration has not been reduced to zero; it is predicted to be 2.121 mg/L under the assumptions.

When the Interpretation Works—and When It Does Not

A single constant half-life is justified when concentration follows a mono-exponential decline with a constant positive rate constant over the time interval of interest. For the model taught here, that means a one-compartment system, first-order elimination, a constant volume and clearance, and no input after the IV bolus. The relationship is a description of this model's concentration curve, not a universal rule for all dosing situations.

A practical diagnostic is to plot positive concentrations against time on a logarithmic concentration scale. A straight declining segment is consistent with an exponential concentration decline. Curvature or distinct slopes suggest that a single \(k\), and therefore one half-life over that whole interval, may not describe the observations adequately. A straight-looking plot is not proof of the model, but it is a useful check. Samples near or below a quantification limit can also make log-concentration calculations unreliable.

For two positive observations, a decline gives \(\hat{k}>0\). If concentrations are equal, the two-point estimate is zero and the corresponding half-life is undefined (formally, it tends to infinity); if the later concentration is higher, the calculated rate constant is negative, which contradicts the assumed elimination-only decline. Those results are warnings to check the observations, timing, and model conditions rather than report an ordinary elimination half-life.

The main alternatives answer different questions. A two-point calculation is simple and transparent but uses only two observations. A line fitted to multiple log concentrations can summarize a mono-exponential phase using more points, but it still assumes that the selected observations share one log-linear slope. A more general model may be needed when the observed curve is not mono-exponential; the simple one-compartment half-life should not be forced onto such data. Infusion or repeated input also changes the concentration-time pattern, so the post-bolus decline interpretation does not apply unchanged while input is occurring.

Common pitfall: “After five half-lives, the drug is gone.” This is incorrect. Under the one-compartment exponential model, after five half-lives the fraction remaining is \(2^{-5}=1/32=3.125\%\), not zero. Whether that remaining concentration is clinically important depends on the drug, the concentration, and the context; half-life alone does not settle that question.

Half-life is useful for translating a rate constant into an intuitive time scale and for predicting proportional concentration changes. Its usefulness depends on stating which concentration-time segment and which model support the estimate. In particular, a half-life estimated from two observations is a model-based summary of those data, not proof that every part of the profile follows the same exponential decline.

Check Your Understanding

Answer each question, then compare your reasoning with the solutions below.

A one-compartment model has \(k=0.200\) h\(^{-1}\). What is its half-life?
Under a first-order one-compartment decline, what fraction of the starting concentration remains after three half-lives?
A hypothetical concentration falls from 6.00 mg/L to 3.00 mg/L over 5.00 h. Estimate \(k\) and the half-life.
Why is a single half-life not a suitable summary if the log-concentration versus time plot has two clearly different slopes?
Under the exponential model, does concentration reach exactly zero after a finite number of half-lives? Explain.

Solutions. (1) \(t_{1/2}=0.6931/(0.200\ \mathrm{h}^{-1})=3.466\) h. (2) The fraction is \(2^{-3}=1/8=12.5\%\). (3) The concentrations are positive and decline, so the two-point estimate is \(k=\ln(6.00/3.00)/(5.00\ \mathrm{h})=0.6931/5.00=0.1386\ \mathrm{h}^{-1}\), giving \(t_{1/2}=0.6931/0.1386=5.00\) h. (4) Different slopes imply different apparent elimination rates across the segments; a single constant \(k\) and half-life do not describe both slopes. (5) No. For finite \(t\), \(C(t)=C_0e^{-kt}>0\) when \(C_0>0\) and \(k>0\); the concentration approaches zero but does not equal zero at a finite time.