What Loop Direction Represents
A concentration-effect plot can show that the same plasma concentration is associated with different responses at different times. The difference between the rising and falling branches is called hysteresis. To interpret its direction, specify both the axes and the order in which observations occur. In this tutorial, plasma concentration \(C_p\) is on the horizontal axis and an effect measure that increases with drug effect is on the vertical axis. Observations are traced forward in time.
With this convention, a counterclockwise loop means that, at a plasma concentration encountered once while \(C_p\) is rising and again while it is falling, the effect is greater on the falling branch. A clockwise loop means it is lower on the falling branch. This describes the orientation of a trajectory, not a cause: loop direction alone cannot establish which biological mechanism produced it.
The effect-compartment model provides a standard explanation for counterclockwise hysteresis. It represents delayed equilibration between plasma and a hypothetical effect site. The effect-site concentration \(C_e\) changes according to the first-order equation
Here \(k_{e0}\) is the effect-site equilibration rate constant, with units of inverse time. The response is assumed to follow \(C_e\) through a fixed relationship \(R=f(C_e)\). For the counterclockwise interpretation, that relationship must be increasing over the concentrations being considered: a larger \(C_e\) gives a larger value of the plotted effect measure. This is the setting developed in the earlier material on the effect site as a latent equilibration state and the first-order effect-compartment equation.
Why Effect-Site Lag Creates a Counterclockwise Loop
When \(C_p\) rises, the effect site does not jump immediately to the new plasma concentration. If plasma concentration is increasing from a low initial value, \(C_e\) generally trails below \(C_p\). After plasma concentration turns downward, \(C_e\) may continue to rise briefly; later it falls, still reflecting earlier plasma exposure. Thus the response can continue increasing even while plasma concentration has begun to decline.
The branch comparison can be justified directly. Suppose \(C_p\) has one rising phase followed by one falling phase, and the same concentration \(c\) is reached at times \(t_r\) and \(t_f\), respectively. Assume \(C_p(t)\ge c\) between those crossings, and that \(C_e(t_r)\le c\). The solution of the effect-compartment equation from \(t_r\) to \(t_f\) is
Because \(C_p(s)\ge c\) throughout that interval, the integral term is at least \(c[1-e^{-k_{e0}(t_f-t_r)}]\). Therefore,
The last inequality follows from \(C_e(t_r)\le c\). If the concentration is strictly above \(c\) for part of the interval, the effect-site concentration on the falling crossing is strictly greater. An increasing response function \(R=f(C_e)\) then gives \(R(t_f)>R(t_r)\): the falling branch lies above the rising branch at that plasma concentration. Tracing the trajectory forward in time yields a counterclockwise loop.
This argument depends on the stated conditions. It is not a universal claim about every pair of observations called “rising” and “falling.” A complicated concentration profile, an inconsistent initial effect-site state, a nonmonotonic response relationship, or additional time-dependent mechanisms can defeat the simple branch comparison.
Worked Example: Calculating the Two Branches for a Rising and Falling Profile
Consider a hypothetical plasma profile that rises linearly from 0 to 10 mg/L over 20 minutes, then falls linearly to 0 mg/L over the next 20 minutes. Let \(k_{e0}=0.10\ \text{min}^{-1}\), \(C_e(0)=0\), and let the effect measure be \(R=C_e/(2+C_e)\), with concentrations in mg/L. Compare the response at \(C_p=5\) mg/L on the rising and falling branches.
The rising-phase slope is \(a=10/20=0.50\ \text{mg/L/min}\), so \(C_p(t)=at\). Solving the effect-compartment equation for this ramp gives \(C_e(t)=a[t-(1-e^{-k_{e0}t})/k_{e0}]\). The rising crossing of 5 mg/L occurs at \(t=10\) min:
The corresponding response is \(R=1.839/(2+1.839)=0.479\). The falling crossing occurs at \(t=30\) min, or 10 minutes after the peak. At the peak, the rising-ramp solution gives
During the falling ramp, \(C_p(20+s)=10-0.50s\). Its effect-site solution is \(C_e(20+s)=C_p(20+s)+0.50/0.10+[C_e(20)-10-0.50/0.10]e^{-0.10s}\). At \(s=10\) min:
At the same plasma concentration of 5 mg/L, the effect measure is 0.479 while concentration is rising and 0.766 while it is falling. The higher falling-branch response is consistent with effect-site lag and a counterclockwise loop.
Reading Data in Time Order
A plot without time ordering can be ambiguous. Establish which observations occur first, locate the increasing and decreasing concentration phases, and compare responses at concentrations represented on both branches. If the effect measure increases with \(C_e\), a higher response on the falling branch is the pattern predicted by a delayed effect-site concentration. The separation is not itself an estimate of \(k_{e0}\); quantifying the loop and estimating the equilibration rate require additional methods.
Worked Example: Interpreting Matched Concentrations in Hypothetical Observations
In a hypothetical experiment, a response fraction is recorded as plasma concentration first rises and then falls. At \(C_p=4.0\) mg/L, the rising-branch response is 0.32 and the falling-branch response is 0.61. Assume the same response definition and measurement method apply at both times, and larger response fractions represent greater effect.
The branch difference is \(0.61-0.32=0.29\) response-fraction units. Since it is positive, the response is greater on the falling branch at the matched plasma concentration. Following the time course, the trajectory moves toward higher concentration before returning at a higher response, the counterclockwise pattern expected when effect lags plasma concentration.
This interpretation is conditional: it describes the observed ordering, and it is compatible with effect-site delay. It does not prove that delay is the mechanism. For example, a changing response over time or a mismatch in the timing of concentration and response measurements could also separate the branches.
Worked Example: Choosing the Direction for an Inhibitory Endpoint
Suppose a hypothetical drug lowers heart rate. At \(C_p=5\) mg/L, heart rate is 82 beats/min on the rising branch and 70 beats/min on the falling branch. If raw heart rate is plotted vertically, the falling-branch point is lower. It would be misleading to call that a clockwise drug-effect loop without considering what the endpoint means.
Define an effect-increasing measure as the reduction from a baseline heart rate of 90 beats/min: \(E=90-\text{heart rate}\). On the rising branch, \(E=90-82=8\) beats/min; on the falling branch, \(E=90-70=20\) beats/min. At matched \(C_p\), the effect measure is greater on the falling branch, by \(20-8=12\) beats/min. With this drug-effect orientation, the pattern is counterclockwise and is consistent with effect lag. The choice of vertical axis must be explicit because reversing the endpoint direction reverses the visual orientation.
Assumptions, Diagnostics, and Alternatives
The simple interpretation is most defensible when plasma concentration has a clear rise and fall, sampling represents both phases, the effect measure is consistently oriented, and the response-concentration relationship is stable and monotonic over the observed range. The effect-compartment explanation also assumes that \(C_e\) follows the stated first-order equilibration equation and that response follows \(C_e\) without an additional delay or history dependence.
Before interpreting a loop, check whether the same plasma concentrations are observed on both branches and whether the sampling times establish their order. Sparse measurements can miss the peak or make a branch appear to turn in the wrong place. A response that saturates may produce little visible vertical separation even when \(C_e\) differs. If the response relationship changes with time, or if the endpoint is plotted in a direction that decreases with effect, visual direction alone can mislead.
The main alternatives differ in what they imply. An instantaneous equilibrium response model predicts a single concentration-effect relationship: at a given \(C_p\), response is the same on both branches, apart from observation error. An effect-compartment model attributes separation to delayed equilibration of \(C_e\). Other delayed biological processes, such as response turnover, can also produce temporal separation but do not have the same interpretation as plasma-to-effect-site equilibration. Time-varying sensitivity or measurement timing problems can produce apparent loops without the proposed effect-site mechanism.
Accordingly, loop direction is a useful description of temporal ordering, not a stand-alone causal test or model-selection result. Counterclockwise orientation is consistent with response lagging plasma concentration under the assumptions above; it does not uniquely identify an effect compartment. Conversely, failure to see a clear counterclockwise loop does not by itself rule out delay, especially when the data are sparse, the response is near saturation, or the concentration profile does not cover both branches well.
Common Mistake
Check Your Understanding
Answer each question using the stated axis convention: plasma concentration on the horizontal axis and an effect-increasing measure on the vertical axis.
- At a matched concentration, the falling-branch response is greater than the rising-branch response. What loop direction does this indicate?
- In one sentence, explain why a first-order effect compartment can have a greater \(C_e\) on the falling crossing than on the rising crossing.
- A response measure decreases as drug effect increases. What should be done before using its plotted loop direction to infer temporal ordering?
- Does a counterclockwise loop prove that effect-site equilibration lag caused the observed separation? Why or why not?
- What does an instantaneous equilibrium response model predict at a plasma concentration encountered on both branches?
Solutions. 1. Counterclockwise, under the stated axis convention. 2. The effect site averages recent plasma concentrations, so during the interval between the two crossings it receives input at or above the matched concentration and retains the earlier exposure. 3. Reorient or transform the endpoint so that larger plotted values represent greater drug effect, and state the axis convention. 4. No; the direction is compatible with effect-site lag but can also arise from other time-dependent mechanisms or measurement-timing problems. 5. It predicts the same response on both branches at the same plasma concentration, because response depends on the current concentration through a single equilibrium relationship.