Within-Patient Association Between Creatinine Clearance and Serum Lithium Concentrations During Routine Monitoring

Although renal elimination determines lithium exposure, the magnitude of the within-patient association between renal function and serum lithium concentration during routine monitoring is uncertain.

This report describes a single-center retrospective repeated-measures study of 530 lithium measurements from 57 adults treated between April 2020 and March 2025. Creatinine clearance (CCr) was decomposed into within-patient deviation and patient-mean components. A linear mixed-effects model with a correlated patient-specific random slope included prescribed dose and clinical covariates; a sensitivity model omitted dose.

In the dose-adjusted model, a 10 mL/min higher CCr relative to a patient's mean was associated with a 0.041 mEq/L lower lithium concentration (β=-0.041; 95% CI, -0.061 to -0.020; P<0.001). The dose-omitted estimate was smaller but remained negative (β=-0.022; 95% CI, -0.043 to -0.001; P=0.042). Estimates remained negative across random-effects structures and all 57 leave-one-patient-out analyses.

Repeated measurements separated within-patient fluctuations from between-patient differences. The primary estimate describes the concurrent association at the same recorded dose and covariate values. These estimates contextualize paired CCr and lithium changes but do not replace direct therapeutic drug monitoring.