Does the placebo effect modulate drug bioavailability? Randomized cross-over studies of three drugs.

Background: Medication effect is the sum of its drug, placebo, and drug*placebo interaction effects. It is conceivable that the interaction effect involves modulating drug bioavailability; it was previously observed that being aware of caffeine ingestion may prolong caffeine plasma half-life. This study was set to evaluate such concept using different drugs.

Methods: Balanced single-dose, two-period, two-group, cross-over design was used to compare the pharmacokinetics of oral cephalexin, ibuprofen, and paracetamol, each described by its name (overt) or as placebo (covert). Volunteers and study coordinators were deceived as to study aim. Drug concentrations were determined blindly by in-house, high performance liquid chromatography assays. Terminal-elimination half-life (t_½) (primary outcome), maximum concentration (C_max), C_max first time (T_max), terminal-elimination-rate constant (λ), area-under-the-concentration-time-curve, to last measured concentration (AUC_T), extrapolated to infinity (AUC_I), or to T_max of overt drug (AUC_Overttmax), and C_max/AUC_I were calculated blindly using standard non-compartmental method. Covert-vs-overt effect on drug pharmacokinetics was evaluated by analysis-of-variance (ANOVA, primary analysis), 90% confidence interval (CI) using the 80.00-125.00% bioequivalence range, and percentage of individual pharmacokinetic covert/overt ratios that are outside the +25% range.

Results: Fifty, 30, and 50 healthy volunteers (18%, 10%, and 6% females, mean (SD) age 30.8 (6.2), 31.4 (6.6), and 31.2 (5.4) years) participated in 3 studies on cephalexin, ibuprofen, and paracetamol, respectively. Withdrawal rate was 4%, 0%, and 4%, respectively. Eighteen blood samples were obtained over 6, 10, and 14 h in each study period of the three drugs, respectively. ANOVA showed no significant difference in any pharmacokinetic parameter for any of the drugs. The 90% CIs for AUC_T, AUC_I, C_max, AUC_Overttmax, and C_max/AUC_I were within the bioequivalence range, except for ibuprofen C_max (76.66-98.99), ibuprofen C_max/AUC_I (77.19-98.39), and ibuprofen (45.32-91.62) and paracetamol (51.45-98.96) AUC_Overttmax. Out of the 126 individual covert/overt ratios, 2.0-16.7% were outside the +25% range for AUC_T, 2.0-4.2% for AUC_I, 25.0-44.9% for C_max, 67.3-76.7% for AUC_Overttmax, and 45.8-71.4% for T_max.

Conclusions: This study couldn't confirm that awareness of drug ingestion modulates its bioavailability. However, it demonstrates the trivial effect of blinding in bioequivalence studies and the extent of bio-variability that would be expected when comparing a drug product to itself.

Trial registration: ClinicalTrials.gov identifier: NCT01501747 (registered Dec 26, 2011).