Abstract
Life-threatening hyponatremia in psychotic patients is common and typically is attributable to either antipsychotic medication or to acute psychosis in those with the polydipsia-hyponatremia syndrome. The preferred treatment for one situation may worsen the hyponatremia if caused by the other situation. Hence it is critical to distinguish between these two possibilities. Case reports and series were identified through electronic databases. Fifty-four cases of hyponatremia without recognized causes in psychotic patients were divided into those with dilute (plasma osmolality) urine. The distribution of urine concentration and measures likely to be associated with psychotic illness and its treatment were compared in both groups. Naranjo׳s scale was utilized to determine the probability hyponatremia was drug-induced. Urine osmolality fit a bimodal distribution (intersection 219mOsm/kg) better than a unimodal distribution. 'Probable' drug-induced cases occurred 6.8 (95%CI=1.6-28.9) times more often in those with concentrated urine. Acute psychotic exacerbations occurred 4.5 (95%CI=0.4-54.1) times more often in those with dilute urine. These findings, as well as several other trends in the data, indicate that measures of urine concentration can help distinguish between antipsychotic-induced and psychosis-induced hyponatremia.