The pharmacokinetics of 400 µg of oral desmopressin in elderly patients with nocturia, and the correlation between the absorption of desmopressin and clinical effect


Gitte M. Hvistendahl, Institute of Experimental Clinical Research, Aarhus University Hospital-Skejby, Dk-8200 Aarhus N, Denmark.



To investigate the pharmacokinetic profile of oral desmopressin in elderly patients with nocturia, and to analyse any possible correlation between the absorption and clinical effect.


In all, 32 patients were screened to determine the baseline number of nocturnal voids and the nocturia index; of these, 24 fulfilled the inclusion criteria and were enrolled for a pharmacokinetic evaluation of oral desmopressin 400 µg. A double-blind, randomized, placebo-controlled, crossover-effect evaluation period was then used to test the association between the absorption of desmopressin and pharmacodynamic effect. Serial plasma samples were collected for 8 h for a pharmacokinetic analysis of desmopressin. The pharmacodynamics after an equivalent oral dose before bedtime were assessed by measuring changes in the number of nocturnal voids, time to first nocturnal void and nocturnal diuresis, from placebo to active treatment.


There was a linear relationship between plasma desmopressin at 2 h after dosing and the area under the plasma concentration curve from 0 to infinity (Pearson's ρ 0.923, P < 0.001). Women had a significantly higher plasma desmopressin concentration than men (P = 0.0012) and more adverse events. There was no correlation between plasma desmopressin at 2 h after dosing and the within-patient response in any of the effect variables. Generally, the number of nocturnal voids and nocturnal diuresis were half that with placebo. The time to the first nocturnal void was almost doubled compared with placebo.


There seems to be a relationship between gender, plasma level of desmopressin and the incidence of adverse events. Plasma desmopressin at 2 h after dosing cannot be used to predict the pharmacodynamic response, although desmopressin lowers the nocturnal diuresis and the number of nocturnal voids.


area under the curve


Vasopressin is produced in the anterior hypothalamus and is involved in regulating body water homeostasis (antidiuresis). Various physiological stimuli of the osmotic receptors in the lamina terminalis induce the release of vasopressin. The most important stimuli are increased plasma osmolality and hypotension. Vasopressin receptors have been identified in, e.g. the kidney, liver, brain, pituitary gland, aortic smooth muscle and on platelets. These receptors are divided into three subtypes, V1, V2 and V3. The V2 receptors in the basolateral membrane of the collecting duct cells are the basis for regulating urine output and hence body water balance.

Desmopressin (1-desamino-8-d-argenine vasopressin) is a synthetic analogue (V2-receptor agonist) of vasopressin. Contrary to vasopressin, desmopressin has very poor affinity for V1 receptors and therefore no pressor effect [1]. This specific antidiuretic effect makes desmopressin useful for managing several disorders involving the regulation of the urine production, e.g. nocturia, nocturnal enuresis and diabetes insipidus.

Nocturia, defined as voiding that disturbs sleep, has many causes [2]; there is a gradual increase in the prevalence of nocturia with age in both men and women [3–5]. Before starting treatment for nocturia with an antidiuretic drug it is important to distinguish between nocturnal polyuria (>35% of daily urine volume in people with no 24-h polyuria) and nocturnal frequency of other causes. There are several studies investigating the effect of antidiuresis on nocturnal polyuria [6–10] and the pharmacokinetics of desmopressin have been studied in healthy adult volunteers [7,7,11–16], in adult patients with central diabetes insipidus [17] and in children with nocturnal enuresis [18]. All these studies were in subjects aged <60 years, so the information on the bioavailability and pharmacokinetics of desmopressin in the highly relevant group of elderly people is sparse.

A few years ago a pharmacokinetic study was conducted using of 200 µg of oral desmopressin in a group of healthy volunteers aged 55–70 years (unpublished). Unfortunately the oral dose used resulted in significantly lower plasma levels than reported in other studies, albeit that the clinical effect was evident, and consequently the pharmacokinetic results for oral desmopressin were limited and inconclusive. The efficacy studies of desmopressin in nocturia have shown that doses of 400 µg are sometimes needed to obtain the full effect [10], although other studies claim that the pharmacodynamic response does not increase when the dose is increased from 200 to 400 µg [19].

Thus the aim of the present study was to investigate the pharmacokinetic profile of 400 µg of oral desmopressin in elderly men and women complaining of nocturia, and to analyse any possible correlation between the absorption of desmopressin and the pharmacodynamic effect.


Thirty-two patients reporting a large nocturnal diuresis were screened; of these, 24 (15 men and nine women) fulfilled the inclusion criteria and were included in the study. The inclusion criteria were age ≥ 65 years, nocturia more than twice per night and a nocturia index of >1 (defined as the mean nocturnal urine volume during the screening period divided by the largest voided volume). The exclusion criteria were any clinical significant renal, hepatic, gastrointestinal, pulmonary, cardiovascular, endocrinological or neurological disorder, any significant symptoms from the urinary tract apart from nocturia, medical treatment with drugs known or suspected to interact with desmopressin, and diuresis during the screening period of >40 mL/kg body weight. The study was conducted according to the Declaration of Helsinki and ICH Good Clinical Practice. Each patient signed a statement of informed consent approved by the local Ethics Committee before entering the trial.

After giving informed consent the patients’ health was confirmed by a complete physical examination, including dipstick testing a urine sample, uroflowmetry and ultrasonography after voiding to exclude residual urine. The patients then entered a 1-week screening period, during which they were asked to register the time and volume of each day- and night-time void, and the time and volume of fluid intake for at least three 24-h cycles. For the remaining nights of the week, the time and volume of each nocturnal void, including the first morning void, were recorded. From the registered data the number of nocturnal voids and the nocturia index were calculated. After the screening period, patients who fulfilled the inclusion criteria were included in the trial.

The trial was in two parts, i.e. part A, a pharmacokinetic evaluation of one oral dose of desmopressin 400 µg, and part B, a randomized, placebo-controlled, crossover-effect evaluation period (Fig. 1). The patients received information to drink no more than enough to satisfy their thirst from 1 h before to 8 h after taking the drug. The intake of coffee, tea or caffeinated beverages and other diuretic liquids were standardized as much as possible during the study days.

Figure 1.

The study design.

On the day of the pharmacokinetic evaluation the patients arrived at the laboratory in the morning. The first blood sample (0 h) was taken just before drug administration and then at 0.5, 0.75, 1, 1.5, 2, 3, 4, 6 and 8 h afterward. A standardized lunch was served at 3 h and an afternoon snack at 7 h. To evaluate the effects the patients arrived at 18.00 hours and were hospitalized for 14 h on 3 consecutive days, followed by a 7–14-day wash-out period before entering the second period of 3 days; the daytime was spent as outpatients. Desmopressin or placebo was given during 3 nights each in a randomized crossover design. On the test nights drug intake was at a standardized bedtime at 23.00 hours. Only one blood sample to determine plasma desmopressin was taken at 2 h after drug administration. The time and volume of each nocturnal void, and the time and volume of the morning void, was registered. The time of rising in the morning was standardized to 07.00 hours.

Safety was assessed during the study, as measurements of serum sodium, weight, blood pressure and pulse, taken before drug administration and 8 h afterward. The patient was withdrawn from the trial if the serum sodium declined to <125 mmol/L or there was symptomatic hyponatraemia with water retention (e.g. weight increase, oedema), cerebral symptoms or convulsions. Other intolerable adverse events, as judged by the patient or by the doctor (e.g. persisting nausea, headache or tiredness, feeling sick), or significant protocol violations (e.g. high diuresis >40 mL/kg) led to withdrawal.

All blood samples for desmopressin analysis were stored immediately on ice and then centrifuged at 1550 g within 45 min of collection. The plasma fraction was obtained and stored in labelled tubes at − 70°C pending desmopressin analysis at the Department of Bioanalytical Chemistry, Ferring AB, Denmark, using a validated radioimmunoassay method. The lower limit of quantification for human plasma was 2.50 pg desmopressin/mL plasma. The mean intra- and interassay coefficients of variation of spiked human EDTA-plasma at 5, 10 and 100 pg/mL was 10.1%, 7.0% and 5.33%, and 18.1%, 10.6% and 8.65%, respectively.

The common pharmacokinetic characteristics of desmopressin were calculated using noncompartmental analysis. The area under the plasma desmopressin concentration vs time curve (AUCt) was calculated using the linear trapezoidal method and AUCt extrapolated to infinity (AUCinf) according to the following equation:

AUCinf = AUCt + Clastz

where Clast denotes the last measurement for the patient in question and λz the estimated slope from a log-linear regression on the last measurements from the patient in question. The terminal half-life was calculated as t1/2 = ln(2)/λz. The concentrations used for estimating λz, and consequently t½, were chosen after inspecting the ln-transformed plasma drug concentration plotted against time; the values in the tail of the curve where the transformed concentrations seem to follow a linear elimination rate were used. A descriptive statistical analysis, e.g. geometric mean and corresponding percentage coefficient of variation, median and range, and harmonic mean with corresponding interquartile range, was generated for the pharmacokinetic variables.

The correlation between one plasma desmopressin value 2–3 h after intake and AUCinf was assessed graphically and using Pearson's correlation coefficients with corresponding 95% CI. A logistic regression analysis was used to describe correlations between the response (reduction in number of nocturnal voids) and absorption (2–3 h after dosing), baseline mean voided nocturnal volume or baseline nocturnal diuresis. A response was defined as a ≥ 45% reduction in the number of nocturnal voids from placebo treatment. Pearson's and Spearman's correlation coefficients were calculated to detect any pairwise correlation between absorption and the desmopressin-placebo difference in time to first void and the desmopressin-placebo reduction in nocturnal diuresis. Descriptive statistics, e.g. median and range, were calculated for the effect variables, with the level of significance set at P < 0.05.


Finally 23 patients were included in the analysis of pharmacokinetics, as one was excluded because of protocol violations; in the pharmacodynamic part (A), 24 were included and analysed. The mean (sd) age of the included patients was 71.7 (6.5) years and the body weight 75.1 (11.3) kg. Eleven drug-related adverse events were reported during desmopressin treatment and one with placebo. Four patients, all women, were withdrawn from part B because of adverse events related to low serum sodium values. There were no serious adverse events during the study. The most common adverse events reported were hyponatraemia (below normal range), headache and weight increase (>2% of the initial body weight). All but one adverse event after desmopressin treatment were in the women. Six of the nine women had a gradual decrease in serum sodium after the second and third dose of desmopressin, whereas none of the men had similar reductions during active treatment (Table 1).

Table 1.  Serum sodium values for the nine women during placebo/desmopressin treatment. The values in italics represent those women with no clinically relevant changes in serum sodium below normal range (136–146 mmol/L). Serum sodium was measured before the first dose and 8 h afterward. If the patient had a weight gain of >2% from baseline the serum sodium was checked before giving the next dose
Patient noPlaceboDesmopressin
Dose I a/bDose II a/bDose III a/bDose I a/bDose II a/bDose III a/b
  1. ex, excluded; a, 18.00 hours; b, 07.00 hours.

1137/140–/138–/137138/136  –/131127/ex
2143/142–/142–/141146/142  –/141  –/142
3142/142–/142–/142142/140  –/135  –/132
4139/141–/141–/141140/140  –/140  –/140
5141/141–/142–/141140/138  –/140  –/139
6139/139–/141–/139138/138  –/132131/ex
7137/139–/140–/139138/137  –/132129/ex
8138/141–/141–/140140/138  –/135  –/132

Figure 2 shows the plasma desmopressin concentration vs time curve for all patients (with the mean values) and Fig. 3 the data for women and men, respectively. There was a mean peak concentration within 1–2 h after administration; the plasma concentration then gradually decreased during the following 6–7 h. The plasma concentrations were then still above the lower level of the assay for 16 of the 24 patients. The women had significantly higher AUCinf values than the men. The results of the descriptive statistical analysis of the pharmacokinetic variables are listed in Table 2 for all patients and for AUCinf for men and women in Table 3. Pearson's correlation coefficient (95% CI) between plasma desmopressin at 2 h after dosing and AUCinf was 0.923 (0.824–0.967) (P < 0.001). The relationship between AUCinf and desmopressin concentration at 2 h is shown in Fig. 4 (AUCinf 6.36 × desmopressin + 1.73).

Figure 2.

The desmopressin plasma concentrations with time for all 24 patients, with the mean (sd) superimposed.

Figure 3.

Individual desmopressin profiles in women (A) and men (B).

Table 2.  Mean plasma desmopressin pharmacokinetic variables after one oral dose of 400 µg for all patients and for men and women
VariableGeometric mean (95% CI)% Coefficient of variationHarmonic mean (95% Hodges-Lehman CI)Median (interquartile range)
AUCt63.1 (49.6–80.2)60.255.4 (47.7–84.8)
AUCinf79.1 (61.9–101.2)61.970.7 (62.3–106.5)
Cmax16.0 (12.7–20.2)58.214.2 (11.6–22.6)
t1/23.1 (2.9–3.6) 3.3 (2.7–3.8)
tmax 1.5 (1.0–2.0)
Table 3.  Mean plasma desmopressin pharmacokinetic variables after one oral dose of 400 µg for men and women
VariableGeometric mean (95% CI)Female/male ratio (90% CI)P
AUCinf, pg/mL*h120.7 (84.5–172.4)63.2 (48.7–82.0)1.91 (1.32–2.75)0.0219
AUCinf/body weight  1.8 (1.2–2.5) 0.8 (0.6–1.0)2.20 (1.53–3.15)0.0012
Figure 4.

The relationship between desmopressin plasma concentrations at +2 h and AUCinf, with 95% CI for the 24 individual predicted values.

Of all 23 patients, 13 (57%) were responders (nine men and three women). The response could not be significantly predicted by the plasma desmopressin level 2–3 h after dosing or baseline mean voided nocturnal volume. There was no correlation between the desmopressin − placebo difference in time to first void and absorption, or as desmopressin − placebo difference in nocturnal diuresis and absorption, with ρSpearman of − 0.01 (P = 0.531) and 0.06 (P = 0.612), respectively.

The mean number of nocturnal voids was less on desmopressin than placebo, the mean time from drug intake to first micturition significantly higher (P = 0.0261), and the mean nocturnal diuresis significantly lower (P < 0.001; Table 4). The baseline median nocturnal voided volume decreased during desmopressin treatment by 60 (−213 to 55) mL.

Table 4.  The median (range) of the effect variables for all patients and by gender
Difference (desmopressin – placebo)AllWomenMen
number of nocturnal voids−1 (−3 to 0)−1 (−3 to 0)−1 (−2 to 0)
time to first micturition, h1.9 (−0.6 to 6.4)1.2 (−0.6 to 6.4)2.1 (−0.2 to 5.5)
nocturnal diuresis, mL/min−0.8 (−2.4 to − 0.1)−0.9 (−2.4 to − 0.3)−0.7 (−1.0 to − 0.1)
nocturia index−1.0 (−2.3 to − 0.1)−1.3 (−2.3 to − 0.3)−0.9 (−1.7 to − 0.1)


One of the aims of the present study was to investigate the pharmacokinetic profile of one dose of oral desmopressin in elderly men and women. The relatively high dose was chosen based on a previous pharmacokinetic study in elderly men (unpublished data), where low plasma levels of desmopressin only gave limited information on the pharmacokinetics in this age group. The elderly were chosen as they represent a large group with a medical problem that may be resolved by treatment with desmopressin. Studies on LUTS show that 72% of elderly people have nocturia and find it very bothersome [5,20]. Furthermore, the elderly may have different patterns of absorption and elimination of drugs. Based on the data from this study, the AUCinf strongly depends on the plasma desmopressin concentration at 2 h after dosing. Theoretically it is possible that the pharmacokinetic characteristics may differ with different doses, which makes it difficult to apply this model to other age groups or other doses of desmopressin. Surprisingly, the desmopressin plasma levels in the women were higher than in the men; when adjusting the AUCinf for body weight the difference persisted. This is clinically important and illustrates that extra care is required when elderly women are treated with desmopressin. The plasma level of desmopressin was still over the lower limit of quantification 8 h after drug intake in 16 of the 23 patients, indicating that in these elderly patients desmopressin may have had a long duration of action at this specific dosage. For treating nocturia, a duration of 6–8 h is sufficient. A longer duration is unwarranted because of the risk of water retention and related adverse events. Even though the mean serum sodium level for the whole group was rather stable during the 3 days of desmopressin treatment, most of the women had a decrease in sodium level below the normal range. This may have been incidental, but it suggests that these women were dosed above the therapeutic range and supports the suspicion of an overly long duration of action. In a recent desmopressin dose-titration study in men, serum sodium levels below the normal range were reported in 22% of the patients and 4% had serum sodium levels of <130 mmol/L [10]. Furthermore, patients at the highest risk of developing hyponatraemia were those aged ≥ 65 years. Similar adverse events were reported in other studies of desmopressin treatment in elderly patients [6].

The pharmacodynamic study was intended to determine whether the response (as the difference in the number of nocturnal voids) was predicted by absorption 2 h after drug intake. There was a clear reduction in the number of nocturnal voids, significantly reduced nocturnal diuresis and significantly increased time to first void during desmopressin treatment, matching results from other studies [7], but there was no clear association between absorption and response in any of the variables analysed. Interestingly, the men had a longer median time to first void than women after desmopressin (Table 4). However, the groups were small and the range in each group very large, which may explain the difference. Differences in nocturnal bladder capacity and sleep quality may also influence the results.

The lack of association between response and absorption might be explained by the high dose the patients were given. The plasma levels of desmopressin achieved in this study group were probably much higher than the threshold for antidiuretic action, which contradicts the previously cited study [10]. However, they conducted a dose-titration based on the dynamic response, whereas the present patients all received the same dose of desmopressin. Apart from a large inter-individual variation in the plasma desmopressin at 2 h the patients had a large intra-individual variation apparently with no effect on the response variables. This study shows that high doses of desmopressin primarily increase the duration of action and not the response.

In conclusion, the present results indicate clearly that the AUCinf strongly depends on the plasma level at 2 h after dosing with 400 µg of oral desmopressin, suggesting that in future studies it might be possible to characterize the absorption/elimination of the drug using only a few blood samples. There seems to be a relationship between gender, plasma desmopressin and the incidence of adverse events. Therefore, care should be taken when treating elderly women with desmopressin. Desmopressin about halved the number of nocturnal voids and nocturnal diuresis. The time to first void was higher on desmopressin and the nocturia index was halved on desmopressin. There was no correlation between response and plasma concentrations during oral treatment with 400 µg desmopressin. Further studies are needed to determine if this will occur with lower doses of desmopressin.


The authors thank Ferring Pharmaceuticals for financial support.


A. Riis is a statistician at Ferring Pharmaceuticals. J.P. Nørgaard is a medical director and scientific officer.