Costanzo, Erin S. MA; Antes, Lisa M. MD; Christensen, Alan J. PhD
Posm = serum osmolality;, Uosm = urine osmolality.
Psychogenic or primary polydipsia, characterized by excessive thirst and compulsive water drinking, is a common problem among psychiatric populations, affecting 6% to 20% of patients (1, 2). Additionally, polydipsia is a symptom of diabetes insipidus, a disorder in which there is either diminished or absent production of vasopressin or lack of kidney response to vasopressin. It is unknown whether diabetes insipidus and psychogenic polydipsia are related. Authors of a case study of a patient with both disorders suggested that the chronic intake of excess fluid associated with psychogenic polydipsia could suppress the hypothalamic-pituitary axis and subsequently result in dysregulated release of vasopressin (3). The patient described in the current case report was diagnosed with both paranoid-type schizophrenia and central diabetes insipidus. It is likely that her polydipsia and polyuria had both psychogenic and physiological components.
Treatment of a person with both etiologies is necessarily more complex; interventions must target physiological and behavioral components. Central diabetes insipidus is generally treated medically with desmopressin. Pharmacological treatments for primary or psychogenic polydipsia in psychiatric patients include antihypertensive medications like propranolol and angiotensin-converting enzyme inhibitors, opioid antagonists, clozapine, and risperidone (1, 4). Behavioral treatment has also been used to decrease excess fluid consumption among inpatients with polydipsia. Most behavioral interventions target psychiatric or mentally retarded inpatients and involve components of frequent weighing to monitor diurnal weight gain associated with fluid intake (4–7), involuntary restriction of access to fluid (4, 6, 8), reinforcement such as tokens that can be exchanged for pleasurable items or activities (4, 6–9), and “time out” in a room or removal of tokens for violations of program restrictions (5, 6). Most published cases of nonpsychiatric patients with primary polydipsia conduct treatment on inpatient units with fluid restriction strictly managed by staff, although additional components such as instruction in relaxation techniques have been introduced (10, 11). Such inpatient interventions consume a great deal of time, may be unnecessarily restrictive for nonpsychiatric patients and high-functioning psychiatric patients, and may not be feasible for use with outpatients. Few behavioral interventions that have potential for outpatient use have been previously reported in the literature (12–14).
The patient was a college-educated 60-year-old woman with a DSM-IV diagnosis of schizophrenia, paranoid type (15). Her symptoms included auditory hallucinations and ideas of reference. At the time of our intervention, delusions and hallucinations were not prominent, and her symptoms could be classified as residual. She had been treated for schizophrenia since her early 20s and took fluphenazine from 1966 to 1994. She was monitored by her psychiatrist every 2 to 3 months for medication management and was prescribed 16 mg Risperdal (Janssen Pharmaceutical), 3 mg lorazepam, and 1 mg benztropine daily. At the time of our intervention, the patient was generally functioning well. She was living independently in an apartment and receiving some assistance with shopping and other household tasks. She spent part of each weekday at the local mental health center attending a prevocational program and volunteered part-time at a local cooperative grocery store. She regularly saw a therapist for assistance developing daily living and social skills and strategies to cope with her psychiatric symptoms.
She was referred to a renal physician for evaluation of polyuria and polydipsia, which had been long-standing since the 1980s, but which the patient had found more troublesome recently. She also had a history of urinary incontinence with detrusor instability, previously treated with Ditropan (Ortho-McNeil Pharmaceutical). By her report, she was considerably incapacitated by her frequent urination, which was worsened by her urinary incontinence, and she wore a diaper both during the day and at night to avoid wetting herself. By her own admission, she drank fluids “constantly” both day and night.
On presentation to the renal clinic, her serum sodium was 145 mEq/l, serum osmolality (Posm) was 297 mosm/kg, and urine osmolality (Uosm) was 92 mosm/kg. Although both compulsive water drinking and diabetes insipidus are characterized by dilute urine, baseline hyperosmolarity is diagnostic of the latter. To confirm the diagnosis, the patient underwent a water deprivation test, during which her urine outputs averaged 250 to 300 ml per hour (thus approximating 7 l/d). Three hours into the test, her arginine vasopressin level was less than 1 pg/ml (ie, undetectable), Posm was 296 mosm/kg, serum sodium was 147 mEq/l, and Uosm was 228 mosm/kg (in healthy people, arginine vasopressin is stimulated at Posm above 285 mosm/kg.) She was then given aqueous vasopressin with marked decline in her urine output to 60 to 90 ml/h and further increase in her Uosm to 496 mosm/kg. Her baseline elevated Posm and serum sodium, and her response to aqueous vasopressin after dehydration, confirmed the diagnosis of central diabetes insipidus.
Her renal physician prescribed a trial of low-dose desmopressin, first intranasally (which she did not tolerate because of local irritation), then oral desmopressin. She was instructed on the side effects of desmopressin, including water intoxication, should she not be able to limit her fluids while on this therapy. Her clinical and laboratory status were monitored closely as an outpatient. She responded to low-dose desmopressin with some increase in her Uosm from a baseline average of 70 mosm/kg to approximately 300 mosm/kg with associated decreased urine output. However, she continued to ingest excess fluid, resulting in a significant decline in her serum sodium level to 125 mEq/l. The desmopressin therefore had to be discontinued. Because it was thought that compulsive water drinking was probably contributing to her excess fluid intake, the patient was referred to our clinical health psychology service for treatment of chronic polydipsia with the goal of substantially reducing her fluid intake to permit resumption of desmopressin.
On psychological interview, the patient reported constant thirst, polyuria, and frequent urinary incontinence. An initial behavioral self-monitoring exercise indicated that she consumed more than 6 l of fluid each day, including tea, coffee, juice, soda, and water. She also reported that her refrigerator was filled with dozens of bottles of water so that she would always have a large supply of cold water readily available. The patient’s excessive fluid intake appeared to be related to several eccentric and even delusional beliefs. For example, the patient erroneously believed that not consuming large amounts of fluid could actually be dangerous to her health. She believed that she would be unable to empty her bladder if she did not continuously consume fluid to “push the urine out.” She also expressed a belief that she would need a large quantity of water readily available in case of a health crisis wherein her life would be endangered if she did not drink substantial amounts of fluid.
Behavioral Treatment Program
We attempted a behavioral self-control intervention with the patient. The intervention was adapted from a protocol initially designed for and successfully used in a study by our group involving the reduction of fluid intake in hemodialysis patients (16). The goal for the present patient was to reduce and maintain her fluid intake at 3 l or less per day to reinstate pharmacotherapy for diabetes insipidus without endangering her sodium level. She met with the therapist weekly for 30 to 45 minutes for 12 weeks. Specific components of the intervention are described below.
The patient was provided with a booklet in which she could record time, type of fluid, amount, and situation or activity for each beverage consumed. 1 Records were reviewed weekly by the patient and her therapist and generally appeared to be thorough and complete. The patient also used a 500-ml water bottle as a device for measuring fluid intake. She was instructed to fill the water bottle 6 times per day at prearranged times approximately 2.5 hours apart. If she chose to drink something other than the water in her water bottle, she was instructed to pour out the amount consumed from the water bottle. She practiced using the water bottle as a measuring tool during sessions, and she was provided with written instructions to which she could refer if she became confused. The self-monitoring techniques were modified somewhat after the patient became more adept at using her water bottle to measure and limit fluid intake. Because she found that she could avoid drinking for long periods during some parts of the day but desired more fluid at mealtimes and after exercise, the specified fill times were eliminated, allowing the patient more control over diurnal variation in her fluid intake pattern.
The initial use of the water bottle as a measuring device also served a stimulus control function, limiting to some extent the large variety of beverage containers the patient used. She was also encouraged to remove most beverages from her refrigerator, but was very reluctant to do so because she worried that she would not have water available in the event of a crisis. She ultimately agreed to remove the majority of her water bottles and place them in a bag inside of a closet where they were neither visible nor readily available, but could be accessed should a crisis occur.
The patient was instructed on coping strategies for managing her continuous sensation of thirst. First, the patient and therapist generated a list of enjoyable activities in which the patient could engage to distract herself from her thirst. Second, substitutes for the sensation of drinking fluid were used. Although sucking on hard candy did not alleviate her thirst, she found that ice cubes were helpful. Because of the increased behavioral effort and stimulus control function associated with sucking on ice (in contrast to beverage consumption) and the relatively diminished fluid content in ice, she was encouraged to substitute ice for fluid as often as she was able. Finally, she was encouraged to drink fluid as slowly as possible, taking small sips.
The patient was asked to reward herself with a pleasurable activity at the end of the day when she was able to successfully keep her fluid intake under 3 l. However, she had difficulty identifying appropriate rewards and effectively implementing this component. Therefore, positive feedback from the therapist and renal physician and feedback on laboratory test results served as reinforcers. By her report, the patient was also reinforced by a reduction in urinary frequency and substantial improvement in her urinary incontinence.
The patient appeared to be motivated and committed to the outpatient behavioral intervention. She engaged readily with the therapist during each session and expressed pride in her improvement over the course of the intervention. Initial self-monitoring records indicated that the patient was consuming 6 to 7 l of fluid each day. According to her subsequent self-monitoring records and as illustrated in Figure 1, she was able to decrease her fluid intake to less than 3 l per day within 2 weeks of implementing the behavioral treatment program and was able largely to maintain this decrease through the end of the intervention period (week 12) and at a 3-week postintervention follow-up assessment (week 15). Her fluid consumption did fluctuate to some extent with a brief relapse during week 9, but she recovered from this relapse and consistently consumed less than 3 l of fluid per day throughout the remainder of the intervention and follow-up period. At the end of the intervention and at follow-up, the patient reported decreases in thirst, polyuria, and urinary incontinence. She also reported less subjective distress and improved ability to function in her daily activities as a result of improvement in her urinary incontinence.
Three weeks after implementation of this behavioral feedback intervention, repeat laboratory work was drawn to assess objectively how well the patient was complying with the fluid restriction. At that time, her Uosm was 146 mosm/kg and her serum sodium was 146 mEq/l (Figure 2). This rise in serum sodium corroborated her self-monitoring records, demonstrating that she was in fact drinking substantially less fluid, and her low Uosm was consistent with deficiency in arginine vasopressin. Given these results and her committed motivation to continue with the behavioral intervention, she was given a second trial of desmopressin at week 3. The patient was initially prescribed only doses at bedtime, and she subsequently reported fewer wet diapers overnight. Because her serum sodium remained on average within the normal range of 134 to 140 mEq/l and her urine remained quite dilute, the dose of desmopressin was increased, and a morning dose was added at week 11. At week 12, her Uosm increased to 435 mosm/kg, resulting in less frequent voids. However, her serum sodium fell to 127 mEq/l, which may have been caused in part by indiscretions in fluid intake and failure to self-monitor her fluid consumption carefully. Therefore, her self-reported fluid intake during that week was likely underestimated (Figure 1). Within 3 weeks of reinstituting a more stringent fluid intake log, her serum sodium had improved to 129 mEq/l (Figure 2). This represented an approximate decrease in free fluid intake of 1.5 l per day when compared with the value of 125 mEq/l obtained on her initial trial of desmopressin before the behavioral intervention. Because the patient was able to control her excess fluid consumption moderately, her desmopressin was continued with close monitoring of her clinical status and serum sodium levels.
The current case report illustrates the successful use of an outpatient behavioral intervention in a patient with psychotic illness and polydipsia associated with both diabetes insipidus and psychogenic components. The behavioral treatment was implemented successfully in an outpatient setting with as little as 30 minutes per week of contact with a therapist and a much less restrictive intervention than previous inpatient behavioral interventions for polydipsia (6, 8). The current intervention used self-monitoring, stimulus control, coping skills training, and reinforcement components. It is likely that similar interventions could be applied with high-functioning psychiatric patients or nonpsychiatric patients with polydipsia, and with people with diabetes insipidus who have excess fluid intake while taking desmopressin.
The current case report also highlights the efficacy of combining behavioral and medical interventions. After behavioral treatment, the patient was able to benefit from medication to treat her diabetes insipidus that would have been unsuccessful and potentially dangerous had the patient been unable to control her excess fluid intake. The successful interface of psychology and medicine was a result of frequent contact between the psychologist and physician and exchange of information regarding the patient’s behavioral and medical progress. The case demonstrates the utility and efficacy of combining medical and behavioral components to treat long-standing polydipsia.
The authors thank John Stokes, MD, for consultation about the patient’s medical diagnosis.
1 Previous behavioral interventions for polydipsia have used daily diurnal weight gain as a tool for tracking fluid intake among inpatients (4–7). Because the patient in the current report was seen once each week on an outpatient basis, this strategy was not feasible. However, patient self-monitoring of weight, in addition to fluid intake, may be a useful strategy for some future outpatient behavioral programs. Cited Here...
1. Verghese C, de Leon J, Josiassen RC. Problems and progress in the diagnosis and treatment of polydipsia and hyponatremia. Schizophr Bull 1996; 22: 455–64.
2. Evenson RC, Jos CJ, Mallya AR. Prevalence of polydipsia among public psychiatric patients. Psychol Rep 1987; 60: 803–7.
3. Fricchione G, Kelleher S, Ayyala M. Coexisting central diabetes insipidus and psychogenic polydipsia. J Clin Psychiatry 1987; 48: 75–6.
4. Kern RS, Marshall BD, Kuehnel TG, Mintz J, Hayden JL, Robertson MJ. Letter to the editor: effects of risperidone on polydipsia in chronic schizophrenia patients. J Clin Psychopharmacol 1997; 17: 432–5.
5. Baldwin LJ, Beck NC, Menditto AA, Arms TS, Cormier JF. Decreasing excessive water drinking by chronic mentally ill forensic patients. Hosp Commun Psychiatry 1992; 43: 507–11.
6. Bowen L, Glynn SM, Marshall BD, Kurth CL, Hayden JL. Successful behavioral treatment of polydipsia in a schizophrenic patient. J Behav Ther Exp Psychiatry 1990; 21: 53–61.
7. Pavalonis D, Shutty M, Hundley P, Leadbetter R, Vieweg V, Downs M. Behavioral intervention to reduce water intake in the syndrome of psychosis, intermittent hyponatremia, and polydipsia. J Behav Ther Exp Psychiatry 1992; 23: 51–7.
8. McNally RJ, Calamari JE, Hansen PM, Kaliher C. Behavioral treatment of psychogenic polydipsia. J Behav Ther Exp Psychiatry 1988; 19: 57–61.
9. Bentz JE, Bergeron JA. Letter to the editor: polydipsia and hyponatremia. Am J Psychiatry 1989; 146: 405–6.
10. Klonoff EA, Moore DJ. Compulsive polydipsia presenting as diabetes insipidus: a behavioral approach. J Behav Ther Exp Psychiatry 1984; 15: 353–8.
11. Oldenburg B. Behavioural treatment of compulsive polydipsia: two case studies. Behav Change 1987; 4: 23–9.
12. Waller G, Hyde CE, Thomas CS. A “biofeedback” approach to the treatment of chronic polydipsia. J Behav Ther Exp Psychiatry 1993; 24: 255–9.
13. Gangdev PS. Cognitive behavioral treatment of compulsive water drinking: a case report. J Cogn Psychother Int Q 1993; 7: 63–8.
14. Thomas JL, Howe J, Gaudet A, Brantley PJ. Behavioral treatment of chronic psychogenic polydipsia with hyponatremia: a unique case of polydipsia in a primary care patient with intractable hiccups. J Behav Ther Exp Psychiatry 2001; 32: 241–50.
15. American Psychiatric Association. Diagnostic and statistical manual of mental disorders, 4th edition, text revision. Washington, DC: American Psychiatric Association; 2000.
16. Christensen AJ, Moran PJ, Wiebe JS, Ehlers SL, Lawton WJ. Effect of a behavioral self-regulation intervention on patient adherence in hemodialysis. Health Psychol 2002; 21: 393–7.