Search

Objectives: Early kidney damage during lithium treatment in bipolar disorder is still hypothetical. We aimed at identifying the determinants of a decreased measured glomerular filtration rate (mGFR) and the accuracy of kidney MRI imaging in its detection.Methods: In this cross-sectional cohort study, 217 consecutive lithium-treated patients underwent mGFR and kidney MRI with half-Fourier turbo spin-echo and Single-shot with long echo time sequences.Results: Median age was 51 [27–62] years, and median lithium treatment duration was 5 [2–14] years. 52% of patients had a stage 2 CKD. In multivariable analysis, the determinants of a lower mGFR were a longer lithium treatment duration (β −0.8 [−1; −0.6] ml/min/1.73 m2 GFR decrease for each year of treatment), a higher age (β −0.4 [−0.6; −0.3] ml/min/1.73 m2 for each year of age, p < 0.001), albuminuria (β −3.97 [−6.6; −1.3], p = 0.003), hypertension (β −6.85 [−12.6; −1.1], p = 0.02) and hypothyroidism (β −7.1 [−11.7; −2.5], p = 0.003). Serum lithium concentration was not associated with mGFR. Renal MRI displayed renal microcyst(s) in 51% of patients, detected as early as 1 year after lithium treatment initiation. mGFR and lithium treatment duration were strongly correlated in patients with microcyst(s) (r = −0.64, p < 0.001), but not in patients with no microcysts (r = −0.24, p = 0.09). The presence of microcysts was associated with the detection of an mGFR