AI Summary & Key Takeaways
AI InsightThese findings suggest that the unadjusted eGFR difference is not robust to adjustment for baseline demographic and clinical characteristics, while a confounder-adjusted difference in normalized renal volume between the lithium group and controls represents a novel finding that warrants confirmation in prospective studies.
📌 3 Key Takeaways
Abstract
Aims: To evaluate the association between lithium and renal structure using computed tomography (CT)-based volumetric methods in smoking patients with bipolar disorder, and its relationship with renal function indicators. Methods: This retrospective, cross-sectional, single-center study included 165 participants with contrast-enhanced abdominal CT: smoking patients with bipolar disorder receiving lithium (n=55), lithium-naïve smoking patients with bipolar disorder (n=55), and smoking controls without known psychiatric or renal disease (n=55). Renal parenchymal volume, dimensions, cortical thickness at four sites, and cystic changes were assessed by two radiologists using 3D Slicer software; volumes were normalized to body surface area (BSA). Serum creatinine, estimated glomerular filtration rate (eGFR), and other biochemical parameters were recorded. Results: Interobserver agreement was high (ICC=0.81). In unadjusted analysis, eGFR differed among groups (p=0.037) and was lower in the lithium group than in controls (difference between group medians, 8.8 ml/min/1.73 m²; p=0.011), but this difference was no longer significant after adjustment for age, gender, body-mass index (BMI), and pack-years (p=0.503); creatinine did not differ (p=0.586). In unadjusted comparisons, renal parenchymal volume, dimensions, and BSA-normalized volume did not differ among groups (all p>0.09). Right lower-pole cortical thickness showed an isolated unadjusted difference between the lithium and non-lithium groups (p=0.016) that did not survive correction for the four regional comparisons. Renal cyst frequency was similar among groups (p=0.572). Total renal volume correlated positively with height, weight, and BMI. In these adjusted models, BSA-normalized renal volume was significantly lower in the lithium group than in controls (B=-22.8 cm³/m², 95% CI -41.0 to -4.7; p=0.014), a difference not apparent in the unadjusted comparison. Conclusion: Although unadjusted eGFR was lower in lithium-treated patients with bipolar disorder than in controls, this difference did not persist after covariate adjustment for age, gender, BMI, and smoking exposure. Renal parenchymal volume, dimensions, and cyst frequency did not differ significantly on unadjusted comparison, although BSA-normalized renal volume was significantly lower in the lithium group after covariate adjustment. These findings suggest that the unadjusted eGFR difference is not robust to adjustment for baseline demographic and clinical characteristics, while a confounder-adjusted difference in normalized renal volume between the lithium group and controls represents a novel finding that warrants confirmation in prospective studies.