Zero-Order Elimination

Why the Elimination Structure Matters

A concentration can fall by a fixed amount per hour or by a fixed fraction per hour. Those descriptions sound similar, but they imply different concentration-time curves. In zero-order elimination, the amount removed in equal time intervals is constant while the model applies. In first-order elimination, the amount removed depends on how much drug is present. Choosing between these structures therefore changes predictions about the decline, the time required to reach a concentration of interest, and how profiles from different starting concentrations compare.

Zero-order elimination is a structural assumption for a defined range of conditions, not a claim that a drug must be eliminated at one fixed rate at every possible concentration. This tutorial derives the simple zero-order profile and compares its predictions with first-order elimination. It does not address the more detailed mechanisms that can produce concentration-dependent elimination.

Definition and Assumptions

Definition: In a zero-order elimination model, the amount of drug in the modeled body compartment decreases at a constant rate \(R_0\) per unit time while drug remains in the compartment. With no input, the amount balance is \(\frac{dA}{dt}=-R_0\), where \(A\) is the amount of drug and \(R_0\) has units of amount per time.

To express that amount decline as a concentration profile, assume a single, well-mixed compartment with constant volume \(V\), no drug input during the interval, and a constant elimination rate \(R_0\) over the modeled range. The concentration is \(C=A/V\). The model also assumes that distribution and changes in volume do not create additional changes in concentration during the interval. These assumptions make the derivation direct; if they are not plausible, a straight-line concentration decline need not describe elimination alone.

The rate \(R_0\) is an amount per time, such as mg/h. It is not itself a concentration slope. Under constant \(V\), the concentration slope is \(-R_0/V\), with units such as mg/L/h. Keeping these quantities distinct prevents a common units error.

Deriving the Concentration Profile

Let \(A_0\) be the amount at time \(t=0\), and let \(C_0=A_0/V\). With no input and constant \(R_0\), integrate the amount balance from time 0 to time \(t\):

\int_{A_0}^{A(t)} dA = \int_0^t -R_0\,dt

Since \(R_0\) is constant, the integrals give \(A(t)-A_0=-R_0t\), or \(A(t)=A_0-R_0t\). Dividing by the constant volume \(V\) gives the concentration profile:

C(t)=C_0-\frac{R_0}{V}t

Thus, while the model applies and \(C(t)>0\), concentration declines linearly with time. Its slope is \(-R_0/V\). The model predicts depletion at \(t_{\mathrm{dep}}=A_0/R_0=VC_0/R_0\). This is a boundary of the simple model: extending the line beyond that time would give a negative amount and concentration, which are not physically meaningful. The zero-order description must stop at depletion or be replaced by a model that represents the behavior outside its stated range.

The units verify the concentration equation: \((R_0/V)t\) has units \((\text{mg/h})/(\text{L})\times\text{h}=\text{mg/L}\), the same as \(C_0\). If volume is not constant, then \(C=A/V\) does not imply a constant concentration slope even when the amount is eliminated at a constant rate.

Worked Example: Predicting a Zero-Order Decline

Suppose, hypothetically, a one-compartment model has \(C_0=12.0\) mg/L, \(V=20.0\) L, and a constant elimination rate \(R_0=30.0\) mg/h. Assume no input, constant volume, and that zero-order elimination applies over the time considered. Find the concentration after 4.00 h and the predicted depletion time.

First calculate the concentration slope:

-\frac{R_0}{V}=-\frac{30.0\ \text{mg/h}}{20.0\ \text{L}}=-1.50\ \text{mg/L/h}

The initial amount is \(A_0=VC_0=(20.0\ \text{L})(12.0\ \text{mg/L})=240\) mg. Applying the concentration equation at 4.00 h:

C(4.00\ \text{h})=12.0\ \text{mg/L}-(1.50\ \text{mg/L/h})(4.00\ \text{h})=6.00\ \text{mg/L}

The predicted depletion time is \(A_0/R_0=240\ \text{mg}/(30.0\ \text{mg/h})=8.00\) h. The 4.00-hour prediction is before depletion, so it is within the mathematical range of the model; the model should not be extended to predict a negative concentration after 8.00 h.

Comparison With First-Order Elimination

First-order elimination removes a constant fraction of the amount per unit time. In a one-compartment model with constant volume and no input, its amount balance is \(\frac{dA}{dt}=-kA\), where \(k\) has units of inverse time. The resulting concentration is \(C(t)=C_0e^{-kt}\). In contrast, zero-order elimination removes a constant amount per unit time and gives a straight-line concentration decline under the assumptions stated above.

Feature	Zero-order elimination	First-order elimination
Amount eliminated per unit time	Constant \(R_0\) while the model applies	Proportional to the amount present
Concentration profile, constant \(V\), no input	Linear decline: \(C(t)=C_0-(R_0/V)t\)	Exponential decline: \(C(t)=C_0e^{-kt}\)
Loss across equal time intervals	Equal absolute amounts, and therefore equal concentration decrements if \(V\) is constant	Equal fractions of the remaining amount, not equal absolute amounts
Time to reach zero in the simple model	Finite: \(VC_0/R_0\)	Concentration approaches zero asymptotically in the mathematical model

The comparison concerns the model structures under their stated assumptions. It does not imply that observed data must follow either ideal curve exactly. Measurement variability, distribution, input, changing volume, or an elimination process that changes across the observed range can all alter the observed profile.

Worked Example: Comparing Equal-Interval Changes

Consider a hypothetical comparison with the same initial concentration, \(C_0=10.0\) mg/L. For the zero-order model, let the concentration decline be 2.00 mg/L/h. For the first-order model, let \(k=0.200\) h\(^{-1}\). Assume constant volume and no input for both models. Compare the concentration after 1.00 h and 2.00 h.

For zero-order elimination:

C(t)=10.0\ \text{mg/L}-(2.00\ \text{mg/L/h})t

Therefore, \(C(1.00\ \text{h})=8.00\) mg/L and \(C(2.00\ \text{h})=6.00\) mg/L. Each hour produces the same decrement, 2.00 mg/L.

For first-order elimination:

C(t)=(10.0\ \text{mg/L})e^{-(0.200\ \text{h}^{-1})t}

At 1.00 h, \(C=10.0e^{-0.200}=8.187\) mg/L. At 2.00 h, \(C=10.0e^{-0.400}=6.703\) mg/L. The first-hour decrement is \(10.0-8.187=1.813\) mg/L, whereas the second-hour decrement is \(8.187-6.703=1.484\) mg/L. The absolute decrement gets smaller because the first-order model removes a constant fraction of the amount remaining, not a constant amount.

These calculations compare model predictions, not evidence from a real study. Their contrast is useful: equal absolute decrements are the ideal zero-order pattern, while equal proportional decrements are the ideal first-order pattern.

Using Concentration Data and Checking the Assumptions

For a constant-volume, no-input interval, a zero-order model predicts that plotting concentration against time gives a straight line with a negative slope. If that slope is \(m\), then \(m=-R_0/V\). Concentration observations are not expected to lie perfectly on the line when measurements vary, so the relevant question is whether a straight-line decline is a reasonable description over the interval, given the data and the structural assumptions.

A visual check should consider the timing and spread of observations, not only whether the first and last concentrations define a line. Curvature in a concentration-versus-time plot may indicate that the constant-rate assumption is inadequate, but it can also arise from distribution or other changes that the simple one-compartment derivation excludes. On a logarithmic concentration scale, a first-order decline is linear under its assumptions; this contrast can help describe the shape, but a plot alone does not establish a mechanism.

In population pharmacokinetics, subjects may have different values of parameters such as \(R_0\) and \(V\). The derivation applies to a subject’s profile conditional on that subject’s parameters. A pooled set of concentrations can therefore show variation around a shared structural pattern, or appear curved because subjects differ. The simple equation should not be interpreted as asserting that all individuals have identical elimination rates or volumes.

Worked Example: Interpreting a Measured Slope

Suppose a hypothetical participant has concentrations of 9.0 mg/L at 1.0 h and 5.0 mg/L at 5.0 h. Assume the observations are adequately described by a zero-order decline over this interval, with no input and constant volume \(V=15.0\) L. Calculate the implied elimination rate and the model-predicted concentration at 3.0 h.

The slope between the two observations is:

m=\frac{5.0-9.0\ \text{mg/L}}{5.0-1.0\ \text{h}}=-1.00\ \text{mg/L/h}

Because \(m=-R_0/V\), the implied amount elimination rate is \(R_0=-mV\):

R_0=(1.00\ \text{mg/L/h})(15.0\ \text{L})=15.0\ \text{mg/h}

The linear model predicts \(C(3.0\ \text{h})=9.0\ \text{mg/L}+(-1.00\ \text{mg/L/h})(3.0-1.0\ \text{h})=7.0\) mg/L. This is an interpolation between the two observations. Two points determine a line, so they cannot by themselves show that the zero-order assumption is appropriate; additional observations and a check for departures from a straight-line pattern are needed.

Common Pitfalls and Limitations

Confusing amount rate with concentration slope. \(R_0\) has units of amount per time. The concentration slope is \(-R_0/V\), and depends on volume.
Extending the line below zero. The equation gives a negative value after its predicted depletion time. That is a mathematical extrapolation, not a physically valid concentration prediction.
Calling a straight segment proof of zero-order elimination. A short interval or sparse observations can look linear by chance. Distribution, changing volume, measurement error, or other structural features can produce a similar appearance.
Assuming the rate is constant at all concentrations. The model only claims a constant rate over the range and conditions to which it is applied. Its assumptions should be assessed for the interval being modeled.
Ignoring input or distribution. The derived profile assumes no input and a single well-mixed compartment. If drug is entering the compartment or concentrations are changing because of distribution, the observed slope is not necessarily the elimination slope derived here.

The zero-order model is useful when a constant absolute amount loss provides an adequate description over the relevant range. Its simplicity is also a limitation: it does not explain why the elimination rate is constant, and it can make unrealistic predictions if applied beyond its valid range. First-order elimination is the main contrasting structure: it predicts a concentration-dependent absolute rate and an exponential decline rather than a linear one. Neither shape should be chosen solely because it is familiar; the model must be compatible with the assumptions and the observed profile.

Key takeaway: With no input and constant volume, zero-order elimination gives \(C(t)=C_0-(R_0/V)t\): a constant amount is removed per unit time, producing a constant concentration decrement. The line is valid only while its structural assumptions apply and the predicted amount remains nonnegative.

Check Your Understanding

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

A hypothetical model has \(R_0=18.0\) mg/h and \(V=12.0\) L. What is the concentration slope under the zero-order assumptions?
For \(C_0=8.0\) mg/L and a zero-order concentration decline of 1.0 mg/L/h, what concentration does the model predict at 6.0 h? Is the prediction before depletion?
In a first-order model, does the amount eliminated in each equal time interval remain constant? Explain the contrast with zero-order elimination.
Why can two concentration observations not establish that a zero-order model is appropriate?
What problem arises if a zero-order concentration equation predicts a value below zero?

Solutions. (1) The slope is \(-R_0/V=-(18.0\ \text{mg/h})/(12.0\ \text{L})=-1.50\) mg/L/h. (2) \(C(6.0\ \text{h})=8.0-(1.0)(6.0)=2.0\) mg/L. Depletion is predicted at \(8.0/(1.0)=8.0\) h, so 6.0 h is before depletion. (3) No. First-order elimination removes an amount proportional to the amount present, so the absolute amount removed over equal intervals generally decreases as the amount falls; zero-order elimination posits a constant amount removed per unit time while it applies. (4) Any two distinct observations define a straight line, so they cannot reveal whether additional observations depart systematically from linear decline. (5) A negative concentration is physically impossible and signals that the equation has been extended beyond its depletion boundary or outside its valid range.

What You'll Learn