Analytic Turnover Responses to Step Changes

From a Step Change to a Concentration Trajectory

A change in biomarker production or degradation does not usually make the measured concentration jump immediately to its new equilibrium. The production and loss rates change, but the concentration already present in the system remains and is carried forward. A first-order turnover model makes this distinction explicit: after an abrupt change, concentration moves exponentially toward a new limiting value.

This analytic solution is useful for understanding the direction and shape of a response, checking calculations, and recognizing what a simple turnover model predicts. It applies when the post-change production rate and first-order loss rate are constant. It is not a general solution for time-varying inputs or nonlinear loss.

Model and Assumptions for a Step

Suppose the system is at steady state before time \(t=0\), with production \(P_0\), loss-rate constant \(k_0\), and concentration \(R_0=P_0/k_0\). At \(t=0\), production and/or loss changes abruptly to new constant values \(P_1\) and \(k_1\). For times \(t\geq 0\), the model is

The step means the model parameters switch instantaneously. It does not mean that concentration itself jumps: in the absence of an instantaneous addition or removal of biomarker, \(R(0^+)=R(0^-)=R_0\). The model also assumes a well-mixed system, constant volume, constant post-step rates, and loss proportional to concentration. It excludes an explicit delay, an additional pool, and saturation of the loss process.

The post-step limiting concentration is obtained by setting the derivative to zero:

This is the concentration the model approaches if the new rates remain in place. It need not equal the concentration at the time of the step.

Deriving the Analytic Response

Subtract the new steady-state concentration from both sides of the differential equation. Because \(P_1=k_1R_{\mathrm{ss},1}\), the deviation \(D(t)=R(t)-R_{\mathrm{ss},1}\) satisfies a simple first-order equation:

Separating variables and integrating from 0 to \(t\) gives the solution. For positive \(D\), integration yields \(\ln(D(t)/D(0))=-k_1t\); the resulting exponential expression also holds when \(D\) is zero or negative.

The response is therefore determined by three quantities: the concentration at the step, the new limiting concentration, and the post-step loss-rate constant. At time \(t\), the remaining difference from the new limit is the initial difference multiplied by \(e^{-k_1t}\). Equivalently, the fraction of the initial concentration gap that has closed is \(1-e^{-k_1t}\).

Differentiating the solution shows why the trajectory is monotone under these assumptions:

Since \(k_1>0\) and the exponential is positive, the derivative has the same sign at every finite time as \(R_{\mathrm{ss},1}-R_0\). The concentration rises toward the limit if that difference is positive, falls if it is negative, and remains constant if it is zero. The initial slope is also \(P_1-k_1R_0\), which provides a direct check on the direction of change.

Applying the Solution and Checking Its Implications

For any proposed step response, calculate the new limit before calculating a concentration at a particular time. Then check whether the predicted value lies between the initial concentration and that limit. For \(k_1>0\), \(e^{-k_1t}\) lies between zero and one for \(t>0\), so the solution is a weighted average of \(R_0\) and \(R_{\mathrm{ss},1}\). It cannot cross the limit or overshoot it under this model.

A useful computational check is to evaluate both the starting value and the long-time limit: the formula gives \(R(0)=R_0\), and as \(t\) grows, \(e^{-k_1t}\) approaches zero, giving \(R(t)\to R_{\mathrm{ss},1}\). The units also provide a check: \(k_1t\) is dimensionless, and the result \(R(t)\) has concentration units.

When a model has several abrupt changes, the solution can be applied interval by interval. At each new step, use the concentration carried forward from the preceding interval as the initial concentration, and use the new rates to calculate that interval’s limiting value and trajectory. Concentration remains continuous across a rate change unless the model explicitly includes an instantaneous addition or removal.

Diagnostics, Alternatives, and Limitations

The single-step model predicts a smooth, monotone approach to its new limit. Repeated observations that show a clear delay, an overshoot, or a sustained reversal of direction are inconsistent with this prediction if the rates truly remain constant after one step. Such patterns may indicate that the step assumption is inadequate, that loss is not first order, or that additional biological processes must be represented. Measurement error can also create apparent departures, so a single noisy observation is not by itself evidence against the model.

An analytic solution is preferable when the model is linear with constant rates over the interval: it is exact, easy to evaluate, and makes the limiting concentration and response shape explicit. Numerical integration can instead approximate trajectories when rates vary with time or the differential equations are more complex, but it does not remove the need to specify an appropriate model. A multi-pool, delayed, or nonlinear model may be needed when the data show features the single first-order equation cannot produce; those additional structures require their own assumptions and parameters.

A key limitation is that a step is an idealization. If production or loss changes gradually, treating it as an abrupt switch may misrepresent the early trajectory. Also, a good fit to an exponential curve does not by itself establish that the biological mechanism is first-order turnover: different mechanisms can produce similar observed responses. The analytic result describes the consequences of the stated model assumptions; it does not prove those assumptions from the curve alone.