The association of emotional illness with disturbances in water balance has been recognized for almost 50 years. In a large and well-controlled study published in 1936, Sleeper and Jellinek1 first observed that many schizophrenics excrete abnormally large volumes of urine. They also found that the polyuria ceased when drinking was prevented and postulated that it was due to excessive water intake. This conclusion was fully substantiated by subsequent investigators, especially Barlow and de Wardener,2 who described in detail the psychologic and antidiuretic function of nine patients with varying kinds of emotional illness. These investigators showed convincingly that the polyuria was not due to neurogenic or nephrogenic diabetes insipidus and labeled the condition "compulsive water drinking," noting that these patients usually gave reasons other than thirst to explain their polydipsia.
Barlow and de Wardener2 also found that, in most compulsive water drinkers, plasma osmolality remained in the normal