Distress due to lithium-induced polyuria: Exploratory study


Subho Chakrabarti, MD, FRCPsych, Department of Psychiatry, Postgraduate Institute of Medical Education and Research (PGIMER), Chandigarh 160012, India. Email: subhochd@yahoo.com


Lithium-induced polyuria, although common, often goes unrecognized. The purpose of the present study was therefore to investigate the complaints of polyuria, and distress and functional impairment associated with polyuria, in 56 patients with bipolar disorder on long-term lithium treatment. All participants underwent 24-h urine collection, and renal function tests. Polyuria (24-h urine volume > 3 L) was found in 70% of subjects. Unless directly enquired about, polyuria was underreported. Impairment in work and daily routine due to increased urine output/frequency was associated with 24-h urine volumes. Polyuria is a highly prevalent, distressing and impairing side-effect of long-term lithium treatment, requiring due attention.

POLYURIA (24-H URINE VOLUME > 3 L) occurs in 20–33% of patients receiving lithium.1 Although mild degrees of polyuria may be tolerated, more severe polyuria increases the risk of dehydration and lithium intoxication during illness/surgery.2 It can also cause marked distress/discomfort and may interfere seriously with patients' socio-occupational functioning.3 These aspects of lithium-induced polyuria, however, have seldom been investigated, which prompted the current study.



Consecutive sampling for 1 year yielded 56 suitable subjects with bipolar affective disorder (eight exclusions), from the lithium clinic. All patients were on lithium for more than 1 year, adherent, in remission and on a stable medication regimen, with no pre-existing renal pathology.

The sample of 56 patients consisted predominantly of men (n = 41), with an average age of approximately 40 years (39.5 ± 11.2 years), with longstanding bipolar disorder (for 16.5 ± 9.5 years) and multiple episodes (12.5 ± 9.3). They had been on lithium treatment for approximately 11 years (10.7 ± 6.4 years), on average. Mean doses of lithium were 986 ± 168 mg/day; average serum levels were 0.66 mmol/L and the majority (n = 38) were on divided doses. Thirty-eight patients were receiving only lithium, the rest were on additional mood stabilizers, antipsychotics, antidepressants or other psychotropics.


The protocol was approved by the local ethics committee. Written informed consent was obtained from all participants.

Medical records were examined to obtain clinical and demographic details, and to specifically determine if the subjects had ever complained of polyuria. Patients were subsequently interviewed to determine if they had ever experienced increased urine volume or increased frequency of micturition. If polyuria was detected its duration, severity, and the attendant distress and impairment in work/daily routine were recorded. A checklist was used to note the presence of any additional side-effects. A complete physical examination was then followed by collection of 24-h urine, while patients maintained their usual fluid intake. Twenty-four-hour urine volumes, creatinine and protein levels were determined. Fasting blood sugars, serum urea, creatinine and lithium levels were also monitored.

Statistical analysis

Data were analyzed using the SPSS, version 12 (SPSS, Chicago, IL, USA). Unpaired t- and χ2 tests were used to compare groups with or without polyuria. Pearson's and Spearman's correlation coefficients were used for correlation analysis. Significance was set at P < 0.05, and two-tailed tests were used.


Table 1 shows that polyuria (24-h urine volume > 3 L) was found in 39 (70%) of the subjects. According to their records, however, only 53% (n = 30) had ever spontaneously reported having problems due to increased frequency/volume of urine. In contrast, 86% of the patients (n = 48) reported these problems, when specifically asked about them; 35 of these had 24-h urine volumes > 3 L. Thus, 90% of the group with polyuria reported problems/distress due to increased volume/frequency. A majority of the patients (n = 35; 63%) also reported significant impairment in their daily routine/work due to polyuria. Twenty-seven of these patients had 24-h urine volumes > 3 L, which meant that 69% of the group with polyuria reported problems in daily routine /work due to polyuria.

Table 1.  Lithium-induced polyuria: Associated distress, impairment and correlates
ParametersSubjects (n = 56)
24-h urine volume ≤ 3 L24-h urine volume > 3 L
  • *

    P < 0.01(two-tailed).

  • Severity of impairment: present, 1; absent, 0.

No. subjects1739
Duration of illness (years), mean ± SD11.4 ± 8.118.7 ± 9.4*
Total no. episodes (mean ± SD)9.7 ± 9.813.8 ± 8.9
No. manic episodes (mean ± SD)4.6 ± 5.77.1 ± 4.9
No. depressive episodes (mean ± SD)4.7 ± 4.36.5 ± 5.3
Duration of lithium treatment (years), mean ± SD8.5 ± 5.511.7 ± 6.6
Lithium dose (mg/day), mean ± SD935 ± 2001008 ± 150
Serum lithium level (mmol/L), mean ± SD0.63 ± 0.10.67 ± 0.1
Once daily schedule, n414
Complaints about increased volume of urine passed, n (%)7 (41)23 (59)
Complaints about increased frequency of micturition, n (%)7 (41)20 (51)
Duration of polyuria (months), mean ± SD42.1 ± 2350.8 ± 37
Frequency of day-time micturition (mean ± SD)5.9 ± 2.37.2 ± 3.2
Frequency of night-time micturition (mean ± SD)1.8 ± 0.82.4 ± 1.4
No. additional side-effects (mean ± SD)1.3 ± 0.81.4 ± 1.2
Serum creatinine level (mg/dL), mean ± SD0.94 ± 0.21.2 ± 0.7
24-h urinary creatinine excretion (mg), mean ± SD1097.4 ± 385.41235.1 ± 288.8
Distress due to increased frequency/volume of urine passed, n (%)13 (76)35 (90)
Severity of impairment in work/daily routine (mean ± SD)0.1 ± 0.30.5 ± 0.5*

As shown in Table 1, polyuria was significantly associated with a longer duration of illness (t = 2.78; P < 0.05) and greater impairment in daily routine/work (t = 3.23; P < 0.01). Such impairment was also significantly associated with greater 24-h urine volumes (r = 0.34; P < 0.05).

Additionally, the duration of illness (r = 0.40; P < 0.01) and lithium treatment (r = 0.36; P < 0.05), the total number of episodes (r = 0.31; P < 0.05), and the number of depressive episodes (r = 0.58; P < 0.001), were all significantly associated with greater 24-h urine volumes. Distress was significantly associated with the frequency of micturition over 24 h (r = 0.29; P < 0.05). None of the other parameters was significantly associated with either polyuria or 24-h urine volumes.


Rates of lithium-induced polyuria have varied widely from 20% to 70% across previous studies.4 The high prevalence of polyuria in the current sample was probably due to the presence of chronic recurrent illness, and prolonged treatment with lithium, all of which were significantly associated with polyuria and 24-h urine volumes in the present study, as well as in earlier reports.5

Despite having clinically significant polyuria for several years, only approximately half the patients had spontaneously voiced this complaint. This indicated that patients had either learnt to compensate for polyuria by increasing their water intake, or were not aware of polyuria being a side-effect of lithium. Nevertheless, such underreporting is of some concern, because it increases the chances of the condition remaining undetected until patients are hospitalized for another illness or surgery, and their access to water is restricted. This puts them at risk of developing dehydration, hypernatremia and lithium intoxication during such crises.2

In contrast, on detailed enquiry a much larger proportion of patients reported problems of increased volume/frequency of urine, which they found to be distressing. Approximately 90% of the patients with 24-h urine volumes > 3 L reported considerable distress, and more than two-thirds of them reported impairment in their daily routine/work due to urinary problems. Severity of impairment was also significantly greater in this group and correlated with 24-h urine volumes, while distress correlated with the frequency of micturition. Self-rated distress due to increased urinary volume/frequency, however, did not have a one-to-one correlation with 24-h volumes. This was not surprising, because the relationship between subjective reporting of side-effects and objective abnormalities is usually a complicated one, and subjective distress is often dependent on additional parameters such as affective morbidity, mood state and personality traits.6

The high prevalence of polyuria and the attendant distress and impairment all indicate that polyuria should not be dismissed as a minor side-effect of little clinical significance. Rather, it should be regarded as an early indicator of possible renal damage, which requires proper evaluation and consideration for therapeutic intervention.3

These findings, however, are based on a small number of patients with chronic and severe illness. Thus, they should be regarded as preliminary, until replicated among larger and more diverse groups of patients on lithium.