↵^a k = number of subject groups;

↵^b Q_M = heterogeneity statistic (Cochran's Q) – tests whether any coefficient (not including the intercept) is significantly different than 0;

↵^c I² = % of total variability due to heterogeneity;

↵^d R² = % of total heterogeneity explained by the covariate(s);

↵* p-Value<0.0001; CI, confidence interval; —, value does not need to be determined.

↵^a coefficient of variation;

↵^b percent fluctuation between peak and trough plasma concentration values over a given dosing interval (i.e. [peak – trough]/trough);

↵^c Data reported by Pedersen and Steffensen on Day 6 following oral maintenance (q24h) therapy with a sustained release theophylline formulation (Noctelin – Riker Labs Inc, Loughborough, UK) [n = 14];

↵^d Data reported for 10 of 14 children investigated by Pedersen and Steffensen on Day 7 following oral maintenance (q24h) therapy with a sustained release theophylline formulation (Noctelin – Riker Labs) [n = 10 – same study group as depicted above; data for 4 children was unavailable];

↵^e PBPK models were not parameterized to include intradose variability. (i.e. once steady-state was achieved, concentration-time values were congruent between dosing intervals). As such, simulated data is only provided as a single value obtained on day 7 of theophylline maintenance dosing [n = 50];

↵^f 25% reduction in small intestinal transit time (SITT) from adult values. SITT for normal adults was parameterized as 2.1h - the default PK-Sim® v5.2 value. This represent the time span between gastric emptying of 63% of a nonabsorbable marker and localization of 90% of the marker within the caecum.

↵^g 25% reduction in large intestinal transit time (LITT) from adult values. LITT for normal adults was parameterized as 44.2h - the default PK-Sim® v5.2 value. This represent the time span between 90% of a nonabsorbable marker reaching the caecum and localization of 70% of the marker within the feces;