First-Order Oral Absorption

Why Model Oral Input?

After an oral dose, drug does not enter the systemic circulation all at once. A structural model must represent both the delay associated with absorption and the subsequent change in drug amount in the body. In the first-order absorption model, the rate of transfer out of a dosing depot is proportional to the amount remaining there. Paired with a one-compartment disposition model, this assumption gives a concentration-time profile that rises after the dose, reaches a peak, and then declines.

The model is useful when this simple shape is a reasonable description of the data. It also makes the roles of the absorption rate and the one-compartment disposition rate explicit. The equations below describe a single dose; they do not describe accumulation from repeated dosing.

Definition: Let \(D\) be a single oral dose, \(F\) the fraction of that dose assumed to reach the systemic circulation, \(V\) the apparent volume of the one-compartment model, \(k_a\) the first-order absorption rate constant, and \(k\) the one-compartment disposition rate constant. Both \(k_a\) and \(k\) have units of inverse time. The model assumes \(D>0\), \(0<F\leq 1\), \(V>0\), \(k_a>0\), and \(k>0\). Here, \(FD\) is the dose available to the systemic circulation; \(k=CL/V\) when clearance \(CL\) and volume \(V\) are used to parameterize the one-compartment disposition model.

Model Formulation and Concentration-Time Profile

Represent the oral dose as entering a depot outside the central compartment. For convenience, initialize the depot with the systemically available amount \(FD\). Let \(A_g(t)\) be the amount remaining in that depot, \(A(t)\) the amount in the central compartment, and \(C(t)=A(t)/V\) the central-compartment concentration. Time \(t\) is measured from the dose, so \(A_g(0)=FD\) and \(A(0)=0\). The model equations are:

\frac{dA_g(t)}{dt}=-k_a A_g(t), \qquad \frac{dA(t)}{dt}=k_a A_g(t)-kA(t).

The first equation says that absorption removes depot drug at a rate proportional to the amount remaining. Its solution is \(A_g(t)=FD e^{-k_a t}\), so the rate of systemic input at time \(t\) is \(k_a FD e^{-k_a t}\). The second equation balances that input against loss from the central compartment at rate \(kA(t)\).

Substitute the depot solution into the central-compartment equation. Multiplying the resulting differential equation by \(e^{kt}\) gives an integrating factor that allows the equation to be integrated directly:

\frac{dA}{dt}+kA=k_a FD e^{-k_a t}, \qquad \frac{d}{dt}\left(Ae^{kt}\right) =k_a FD e^{(k-k_a)t}.

When \(k_a\ne k\), integrate from \(0\) to \(t\), using \(A(0)=0\). Dividing the resulting amount by \(V\) gives the concentration:

A(t)=\frac{FD k_a}{k_a-k} \left(e^{-kt}-e^{-k_a t}\right), \qquad C(t)=\frac{FD k_a}{V(k_a-k)} \left(e^{-kt}-e^{-k_a t}\right),\quad t\geq 0.

The coefficient has units of concentration: \(FD/V\) is dose per volume, and \(k_a/(k_a-k)\) is dimensionless. Thus the equation predicts \(C(t)\) in, for example, milligrams per litre when \(D\) is in milligrams and \(V\) is in litres. Although the sign of \(k_a-k\) can be positive or negative, the complete expression predicts a nonnegative concentration: the difference of exponentials changes sign with the denominator.

Worked Example: Calculating Concentration After One Oral Dose

Suppose, hypothetically, that \(D=500\ \text{mg}\), \(F=0.80\), \(V=40\ \text{L}\), \(k_a=1.20\ \text{h}^{-1}\), and \(k=0.20\ \text{h}^{-1}\). Assume the single-dose, first-order absorption and one-compartment model applies. Calculate the predicted concentration at \(t=1\ \text{h}\).

Since \(k_a\ne k\), use the derived equation. The coefficient is:

\frac{FD k_a}{V(k_a-k)} = \frac{(0.80)(500\ \text{mg})(1.20\ \text{h}^{-1})} {(40\ \text{L})(1.20-0.20)\ \text{h}^{-1}} =12.0\ \text{mg/L}.

At one hour, \(e^{-0.20}=0.8187308\) and \(e^{-1.20}=0.3011942\). Therefore:

C(1) =12.0(0.8187308-0.3011942) =6.210\ \text{mg/L}.

The model predicts a concentration of \(6.210\ \text{mg/L}\) one hour after the dose, under the stated assumptions. This is a model prediction, not a measured concentration.

Time and Concentration of the Peak

For distinct positive rates, the peak time can be found by differentiating \(C(t)\) and setting the derivative equal to zero. The coefficient \(FD k_a/[V(k_a-k)]\) is positive or negative together with the difference of exponentials, so the derivative can be set to zero using the bracketed terms:

\frac{dC}{dt} = \frac{FD k_a}{V(k_a-k)} \left(-k e^{-kt}+k_a e^{-k_a t}\right)=0.

Rearrange the zero-derivative equation to obtain \(k_a e^{-k_a t}=k e^{-kt}\). Taking logarithms and solving for time yields:

\ln\left(\frac{k_a}{k}\right)=(k_a-k)t, \qquad t_{\max}=\frac{\ln(k_a/k)}{k_a-k}.

This time is positive for either ordering of the rates. If \(k_a>k\), both the logarithm and denominator are positive; if \(k_a<k\), both are negative. The concentration at the peak is obtained by substituting \(t_{\max}\) into \(C(t)\). The derivative changes from positive to negative at this time, so it is a maximum: the concentration rises before \(t_{\max}\) and falls afterward.

Worked Example: Finding the Predicted Peak

Consider a hypothetical single dose with \(D=250\ \text{mg}\), \(F=0.80\), \(V=40\ \text{L}\), \(k_a=0.80\ \text{h}^{-1}\), and \(k=0.25\ \text{h}^{-1}\). The two rates are positive and distinct, so the peak-time formula applies.

First calculate the peak time:

t_{\max} =\frac{\ln(0.80/0.25)}{0.80-0.25} =\frac{\ln(3.2)}{0.55} =2.11482\ \text{h}.

The concentration coefficient is:

\frac{FD k_a}{V(k_a-k)} = \frac{(0.80)(250\ \text{mg})(0.80\ \text{h}^{-1})} {(40\ \text{L})(0.55\ \text{h}^{-1})} =7.272727\ \text{mg/L}.

At the calculated peak time, \(kt_{\max}=0.528705\) and \(k_at_{\max}=1.691856\). Thus \(e^{-kt_{\max}}\approx0.5894\) and \(e^{-k_at_{\max}}\approx0.1842\). Substituting these values gives:

C(t_{\max}) =7.272727(0.5894-0.1842) \approx2.947\ \text{mg/L}.

Under this model and these hypothetical parameter values, the predicted peak occurs \(2.11482\ \text{h}\) after dosing and is approximately \(2.947\ \text{mg/L}\). The reported concentration is rounded; the arithmetic uses the unrounded rates and peak time.

When the Two Rate Constants Are Equal

The formula with \(k_a-k\) in its denominator cannot be evaluated directly when \(k_a=k\). This is a removable mathematical singularity, not a prediction of infinite concentration. Return to the integrated expression for \(A(t)\). As \(k_a\) approaches \(k\), the exponential difference approaches zero at the same rate as the denominator. Equivalently, integrate the integrating-factor equation with \(k_a=k\), for which the exponential on the right becomes one. The equal-rate profile is:

A(t)=FD k t e^{-kt}, \qquad C(t)=\frac{FD}{V}kt e^{-kt} \quad\text{when }k_a=k.

Differentiating this expression shows that its peak occurs at \(t=1/k\), with peak concentration \(FD/(Ve)\), where \(e\) is the base of the natural logarithm. This special case is also useful numerically: when the two rates are very close, subtracting nearly equal exponentials and dividing by a small difference can lose numerical precision. A stable implementation should use the equal-rate limit or a numerically stable equivalent calculation.

Worked Example: Applying the Equal-Rate Limit

Suppose, hypothetically, \(D=250\ \text{mg}\), \(F=0.80\), \(V=40\ \text{L}\), and \(k_a=k=0.50\ \text{h}^{-1}\). The distinct-rate formula does not apply, so use the equal-rate expression.

At \(t=2\ \text{h}\):

C(2) =\frac{(0.80)(250\ \text{mg})}{40\ \text{L}} (0.50\ \text{h}^{-1})(2\ \text{h})e^{-(0.50)(2)} =2.5e^{-1} =1.839\ \text{mg/L}.

The peak time is \(1/k=1/0.50=2\ \text{h}\), so the calculated value is also the peak concentration. The units are consistent: \(kt\) is dimensionless, leaving dose per volume as the concentration unit.

Assumptions, Diagnostics, and Alternatives

The profile follows from specific structural assumptions, not from the oral route alone. It assumes a single dose at time zero; first-order transfer from one depot; a single, well-mixed central compartment with constant \(V\); and loss from that compartment at a constant first-order rate \(k\). It also assumes that \(F\) is constant for the dose and that the amount entering the central compartment is represented by \(FD\). These are modeling assumptions. Observed data may not be sufficient to verify each one, especially if samples are sparse.

Useful checks begin with the predicted shape and the data collection design. The model predicts zero central concentration at the instant of the dose, a rise to one peak, and a subsequent decline toward zero. Systematic residual patterns around the rising phase, peak, or late decline can signal that the structural model does not capture the observations. A peak systematically earlier or later than observed, for example, may indicate that the assumed input profile is inadequate. Residuals should be considered in relation to sampling times and assay limits, rather than interpreted as proof of a particular alternative model.

The absorption rate is informed especially by observations during the rising phase and around the peak. If those times are poorly sampled, different parameter values may produce similar predictions over the observed times, leaving \(k_a\) weakly determined. Sparse late observations can also make the decline difficult to characterize. Such limitations affect what can be learned from a fitted profile; a numerically successful fit does not by itself establish that the assumptions are correct.

Model structure	Input representation	What distinguishes it from this model
One-compartment IV bolus	Drug is placed in the central compartment at time zero.	It does not include a separate oral depot or a first-order absorption phase.
One-compartment oral model with first-order absorption	Drug enters from a depot at a rate proportional to the amount remaining.	It predicts a gradual rise and a peak after the dose, as derived here.
Other oral-input structures	Represent the timing or shape of input differently.	They may be considered when the first-order depot profile does not adequately describe the observed data.

The one-compartment IV bolus model is a useful contrast because its dose begins in the central compartment, whereas this oral model has an additional depot state. Other oral-input structures are alternatives when the timing or shape predicted here is inadequate; their details are separate model choices. Model comparison should be guided by the dosing process, the observation pattern, and diagnostic evidence, not by adding parameters solely to improve fit.

Common Pitfalls

Using the wrong sign. Writing \(e^{-k_at}-e^{-kt}\) while retaining the denominator \(k_a-k\) makes the predicted concentration negative. The order of the exponentials and denominator must match the derived equation.
Treating the dose as an immediate central-compartment amount. That would remove the depot and describe a different input structure. In this model, the central amount starts at zero and increases as drug is absorbed.
Applying the distinct-rate equation when the rates are equal. A zero denominator does not mean an infinite concentration; use \(C(t)=(FD/V)kt e^{-kt}\) when \(k_a=k\).
Interpreting a fitted curve as proof of the assumptions. A plausible fit is not enough to establish that absorption is first order or that one compartment is adequate. Examine the data coverage and residual patterns.
Reporting a peak time without units or context. The rate constants must use reciprocal time units, and \(t_{\max}\) is then reported in the corresponding time unit.

Key takeaway: A single oral dose with first-order depot absorption and one-compartment disposition gives a difference-of-exponentials concentration profile. Its peak time follows from setting the derivative to zero, and when the two rates coincide, the correct profile is the finite equal-rate limit.

Check Your Understanding

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

For a single dose with \(D=100\ \text{mg}\), \(F=1\), and \(V=20\ \text{L}\), what is the predicted central concentration at time zero? Why?
If \(k_a=1.0\ \text{h}^{-1}\) and \(k=0.20\ \text{h}^{-1}\), calculate \(t_{\max}\).
For \(k_a=k=0.40\ \text{h}^{-1}\), calculate the time of the peak.
Name one data-design limitation that could make the absorption rate difficult to estimate, and explain why.

Solutions. (1) \(C(0)=0\ \text{mg/L}\), because the initial condition places no drug in the central compartment; drug is initially in the depot. (2) \(t_{\max}=\ln(1.0/0.20)/(1.0-0.20)=\ln(5)/0.80=2.012\ \text{h}\). (3) For equal rates, \(t_{\max}=1/k=1/0.40=2.50\ \text{h}\). (4) If samples are sparse during the rising phase and near the peak, the observations provide little information about the timing of absorption, so multiple absorption-rate values may yield similar predictions at the recorded times.

What You'll Learn