Risk factors for polyuria in a cross-section of community psychiatric lithium-treated patients.

OBJECTIVES: Polyuria increases the risk of dehydration and lithium toxicity in lithium-treated patients. Risk factors have been inconsistently described and the variance of this adverse effect remains poorly understood. This study aimed to establish independent risk factors for polyuria in a community, secondary-level lithium-treated sample of patients.

METHODS: This was a cross-sectional study of the lithium-treated patients attending a general adult and an old age psychiatry service. Participants completed a 24-hour urine collection. Urine volume and the presence of polyuria were the outcomes of interest. The relationship between outcome and the participant's demographic and clinical characteristics was explored with univariable and multivariable analysis.

RESULTS: A total of 122 participants were included in the analysis, with 38% being diagnosed with polyuria. Female gender and increased body weight independently predicted the presence of polyuria (standardized regression coefficient 1.01 and 0.94, respectively; p = 0.002 and p = 0.003, respectively). Female gender and increased body weight, lithium dose, and duration of lithium treatment independently predicted higher 24-hour urine volumes (standardized regression coefficients 0.693, p < 0.0005; 0.791, p < 0.0005; 0.276, p = 0.043; 0.181, p = 0.034, respectively). Of three different weight metrics, lean body weight was the most predictive.

CONCLUSIONS: Female gender and increased body weight explain part of the variance of this adverse effect. Both risk factors offer fresh insights into the pathophysiology of this potentially reversible and dangerous adverse effect of lithium treatment. Future research should focus on understanding the differences between the genders and between different body compositions in terms of lithium pharmacokinetics and pharmacodynamics.