Aim: To determine whether lithium concentrations measured in saliva, plasma, and erythrocytes at a standardized morning time (09:00) after brief exposure are associated with plasma potassium, sodium, and creatinine. Subjects and Methods: Seventy-seven inpatients or day-hospital patients with bipolar disorder, treatment-resistant depression, or depression with high suicide risk received oral lithium carbonate 600 mg/day for five days. On day 6 at 09:00, unstimulated saliva, plasma, and erythrocyte samples were collected. Lithium was quantified using spectrophotometric assay. Potassium, sodium, and creatinine were obtained from blood at the same visit. Statistical analyses were conducted using multiple linear regression. Results: Models for potassium (R2 = 0.086; p = 0.497) and sodium (R2 = 0.058; p = 0.662) were not significant. For creatinine, the model explained a modest proportion of variance (R2 = 0.252) and was non-significant (p = 0.053). Within this model, plasma lithium showed the largest (negative) standardized coefficient (β = −0.437; p = 0.057), whereas saliva (β = 0.323; p = 0.096) and erythrocytes (β = −0.152; p = 0.467) were not significant. Conclusion: In the first treatment week, lithium in saliva and erythrocytes was not associated with plasma potassium or sodium and added little beyond plasma lithium for explaining creatinine. These early, single-time-point findings support plasma as the reference matrix and motivate larger, longitudinal, multi-time-point studies pairing noninvasive sampling with renal covariates.
Vlahovic et al. (Wed,) studied this question.