Codeine and Testosterone Concentration Patterns During Co-Administration of Codeine and Testosterone Undecanoate in Male Wistar Rats

Background: Three widely separated sampling days cannot establish conventional pharmacokinetic parameters. We therefore examined exploratory whole-blood codeine and plasma total-testosterone concentration patterns during co-administration of codeine and testosterone undecanoate in male Wistar rats. Methods: Thirty-two rats were allocated to saline (S), codeine (C), testosterone undecanoate (T), or combined codeine plus testosterone undecanoate (CT) groups (eight animals per group). Codeine (10 mg/kg by oral gavage) was administered once daily from Day 0 through Day 5 (six total doses) and testosterone undecanoate (100 mg/kg intramuscularly) on Day 3. On Day 0, whole blood was collected two hours after codeine administration; therefore, Day 0 was a post-dose sampling point rather than a pre-dose baseline. No codeine was administered on Day 6; the Day 6 sample was collected after completion of the six-dose treatment schedule. Codeine dose volumes were adjusted according to each rat’s baseline body weight. Measurements unavailable because of insufficient sample volume were treated as missing without imputation. Codeine was quantified only in C and CT; the corresponding model therefore estimated the association of adding testosterone undecanoate among codeine-treated animals rather than a full factorial drug–drug interaction. Results: All 14 available Day 0 samples had instrument-reported whole-blood codeine concentrations (range, 0.339–4.705 ng/mL), consistent with their post-dose status. The C-versus-CT group-by-day test using all instrument-reported values yielded Wald chi-square = 8.487 (df = 2, p = 0.014). With the verified LOQ of 0.25 ng/mL, only CT7 Day 14 (0.08 ng/mL) was below the LOQ while remaining above the LOD of 0.07 ng/mL; excluding this observation yielded Wald chi-square = 7.856 (df = 2, p = 0.020). Plasma testosterone availability varied by group and day (n = 3–8); the raw-scale group-by-day test was p = 0.0076, whereas the log-scale sensitivity test was p = 0.052. Conclusions: The observed numerical patterns are exploratory. Sparse sampling, one below-LOQ codeine value, an influential CT7 observation, unequal plasma-testosterone availability, and sensitivity to analytical and modeling choices prevent causal, pharmacokinetic, metabolic, endocrine-mechanistic, or clinical conclusions.