1228 BIOL PSYCHIATRY 1992;31:1228-1231 Estimating Daily Urine Volume in Psychiatric Patients: Empiric Confirmation Morris B. Goldman, Robert C. Marks, Lesley Blake, Miljana Petkovic, Donald Hedeker, and Daniel J. Luchins Introduction Polyuria and polydipsia, common symptoms in psychiatric patients, cause significant morbid- ity. No convenient and reliable method exists to diagnose, or determine the severity, of these disorders in water balance (Vieweg and Lead- better 1990). Accurate measures of intake or output are difficult to obtain (Vieweg and David 1985); estimates based on patient or staff reports are unreliable (lllowsky and Kirch 1988); and changes in body weight or serum sodium are insensitive. Many rely on spot measures of urine dilution (Vieweg et al 1986), though these meth- ods have only been validated in a single abstract (Koczapski et al 1989). These author's found that early morning and late afternoon urine cre- atinine concentrations in male polydipsic inpa- tient chronic schizophrenics, adjusted for body weight, are highly correlated to measured 24- hr urine volumes. The goals of this study were to test Koczapski and colleagues' method in polydipsic and non- polydipsic subjects, and to examine the influ- ence of other putative factors on the relationship between urine creatinine concentration and mea- sured 24-hr urine output. We also calculated From the University of Chicago Pritzker School of Medicine (MBG, RCM, LB, MP, DJL), Illinois State Psychiatric Institute (MBG, RCM, LB, DH, DJL), the Elgin Mental Health Center (MBG), and the University of Illinois (DH), Chicago, Illinois. Address reprint requests to Morris B. Goldman, MD, University of Chicago Medical Center, 5841 South Maryland, Box 411, Chi- cago, IL 60637. Received February I I, 1991; revised February 18, 1992. preliminary indicators of polyuria/polydipsia that can be tested in future studies. Method Subjects were I I (10 men, l woman)chronic schizophrenics (Research Diagnostic Criteria [RDC] Diagnosis); and 7 patients with (3 men, 4 women) major affective disorder, manic types between the ages of 18 and 59. Eight were thought to be polydipsic. Schizophrenic patients were clinically stable and receiving fluphenazine HC! (mean _.+ SD)(28.7 _+ 26.7 rag/day). The af- fective patients were acutely manic. Six were drug-free, one was receiving 30 mg haloperidol daily. Collections took place on an inpatient psy- chiatric research unit with intake of cooked meat varying from 80 to 160 gm/day. Urine collection began after voiding at 7:30 AM, and ended with the 7:30 AM sample the next day. Body weight was determined in the morning. A multiple linear regression model to predict measured 24-hr urine volume was constructed with the parameter that Koczapski empirically found provided the best fit: Daily urine volume (liters) = 0.875 x (8 AM body weight (Kg)/urine creatinine concentra- tion (mg/dl)s AM + body weight/urine cre- atinine concentration4 I'M). We used 7:3t3 AM instead of 8 AM samples for the morning urine. Gender and age were se- quentially added to the model because of evi- © 1992 Society of Biological Psychiatry 0006-3223/921505.00