Meal Timing and Gastric State Transitions

Why meal timing must be represented over time

A dose taken with food and a dose taken before a meal do not necessarily experience the same gastric conditions throughout their residence in the stomach. A meal eaten after dosing can change the emptying rate of drug still present in the stomach. Likewise, a dose taken after a meal may begin emptying in a fed state, then continue emptying after the stomach returns to a fasted state. The relevant question is therefore not only whether food was present at dosing, but which gastric state the dose experiences over time.

The earlier material on fed-state physiology and emptying models provides possible state-specific emptying parameters. Here, meal timing determines when those parameters apply. The examples use a first-order emptying hazard to make the transitions explicit; the same timeline logic can be used with other emptying models, provided their state dependence is defined appropriately.

Set dosing time to \(t=0\), and let \(G(t)\) be the amount of the dose remaining in the stomach, in milligrams. In the first-order model, the state-specific emptying hazard is \(k_F\) in the fasted state and \(k_{\mathrm{fed}}\) in the fed state, both in inverse hours. The hazard is the instantaneous emptying rate divided by the amount still in the stomach. The amount leaving the stomach per unit time is \(k_{q(t)}G(t)\).

Calculating a piecewise emptying profile

Within any interval where the state is constant, the gastric amount satisfies a first-order differential equation. If the state changes at time \(\tau\), calculate the amount at \(\tau\) using the state before the transition, then continue from that same amount using the new state-specific hazard. Thus, the remaining amount is continuous at the transition even though the emptying rate may change abruptly.

When the state is constant on successive intervals, the integral is a sum of hazard-times-duration terms. If interval \(j\) has hazard \(k_j\) and duration \(\Delta t_j\), then \(G(t)=G(0)\exp(-\sum_j k_j\Delta t_j)\). The cumulative amount emptied by time \(t\) is \(G(0)-G(t)\), and the instantaneous emptying rate within an interval is \(k_jG(t)\).

A short derivation

For a constant hazard \(k\), separation of variables gives \(dG/G=-k\,dt\). Integrating from the start \(t_a\) of an interval to time \(t\) gives \(\ln G(t)-\ln G(t_a)=-k(t-t_a)\), so \(G(t)=G(t_a)e^{-k(t-t_a)}\). At a transition, the amount at the end of the first interval becomes the initial amount for the next. Repeating this calculation across intervals multiplies the exponential terms, which is equivalent to adding their hazard-times-duration contributions in the exponent. This establishes the piecewise formula without resetting the dose at a meal.

Meals eaten before dosing

For a meal before dosing, the fed interval may already be underway at \(t=0\). In that case, the dose starts with the fed-state hazard, and a later transition to fasted state changes the hazard for whatever dose remains. The meal time itself is not the dose time: both must be placed on the same timeline.

Multiple meals and separated fed intervals

If the modeled state returns to fasted between meals, each fed interval must be represented separately. Extending one fed interval from the first meal to the last would incorrectly apply the fed-state hazard during the intervening fasted period. The cumulative-hazard form makes the calculation straightforward: identify each interval, multiply its hazard by its duration, and add those contributions.

Choosing and checking the state representation

The piecewise model is useful when state-specific hazards and transition times are available or can be specified. Its simplicity also makes its assumptions visible. A static fed-or-fasted model assigns one state to the entire dose trajectory, often based only on the condition at dosing. A time-varying piecewise model instead updates the hazard when the defined state changes. A model with gradual transitions can represent a progressive change in gastric conditions rather than an abrupt switch, but it requires a time-varying hazard or additional state description. These are alternative representations, not interchangeable calculations: the appropriate choice depends on the physiological information and the intended level of detail.

A practical check is to write a timeline before calculating anything. Mark dosing time, meal times, the assumed onset of each fed interval, and the assumed return to fasted state. Then verify that every time after dosing belongs to exactly one state. For a first-order model with nonnegative hazards, \(G(t)\) must be nonincreasing and must never exceed \(G(0)\). The amount emptied plus the amount remaining must equal the initial gastric amount:

At each transition, check that the calculated gastric amount is unchanged while the rate updates to the new hazard times that amount. A jump in the amount itself usually indicates that the interval calculation was restarted incorrectly or that a transition was treated as a mass transfer. If a numerical calculation produces increasing gastric mass under positive emptying hazards, check the sign in the differential equation and the time units. For example, hazards in \(\mathrm{h}^{-1}\) must be multiplied by durations in hours.

A further limitation is the memoryless assumption. If emptying depends on how long the dose has been in the stomach, a state-specific constant hazard may not capture the process. Lagged or transit-time descriptions may require tracking residence time or another model-specific quantity through the state change. A meal-state timeline alone does not specify those details. Nor does a gastric emptying profile by itself establish how much drug reaches the systemic circulation; that requires the relevant downstream absorption and disposition model.