Retracted: Safety and efficacy of tramadol in the treatment of idiopathic detrusor overactivity: a double-blind, placebo-controlled, randomized study


M. R. Safarinejad MD, PO Box 19395–1849, Tehran, Iran. Tel: 98 21 2245 4499 Fax: 98 21 2245 6845



To evaluate the efficacy and safety of tramadol in patients with idiopathic detrusor overactivity (IDO).


A total of 76 patients 18 years or older with IDO were randomly assigned to receive 100 mg tramadol sustained release (group 1, n = 38) or placebo (group 2, n = 38) every 12 h for 12 weeks. Clinical evaluation was performed at baseline and every 2 weeks during treatment. All patients underwent urodynamics and ice water test at baseline and 12-week treatment. Main outcome measures were number of voids per 24 h, urine volume per void and episodes of urge incontinence per 24 h on a frequency volume chart and detailed recording of adverse effect.


After 12 weeks of treatment mean number of voids per 24 h ± SD decreased from 9.3 ± 3.2 to 5.1 ± 2.1 (< 0.001 vs. placebo) [95% confidence interval (CI) −5.1-−0.4]. At that time mean urine volume per void increased from 158 ± 32 to 198 ± 76 ml (< 0.001 vs. placebo) (95% CI 8-22), while mean number of incontinence episodes per 24 h decreased from 3.2 ± 3.3 to 1.6 ± 2.8 (< 0.001 vs. placebo) (95% CI −2-0.3). Tramadol induced significant improvements in urodynamic parameters. More adverse effects were associated with tramadol treatment than with placebo (< 0.05). The main adverse event with tramadol was nausea.


In patients with non-neurogenic IDO tramadol provided beneficial clinical and urodynamic results. Further studies are required to draw final conclusions on the efficacy of this drug in IDO.


The International Continence Society has updated the terminology of the symptoms, signs, urodynamic observations and conditions associated with lower urinary tract dysfunction [1]. Overactive bladder (OAB) is one with inappropriate and involuntary contractions, causing urgency, frequency and/or urinary incontinence. It has an overall prevalence of 16% in Western Europe [2] and 17% in the USA [3]. Idiopathic detrusor overactivity (IDO) is defined as detrusor overactivity with no known precipitating cause [4].

Detrusor contractility is mediated by cholinergic muscarinic receptors; therefore, antimuscarinic agents are often used for treatment of IDO. Their therapeutic efficacy is limited by their troublesome side-effects resulting in the discontinuation of treatment in a significant number of patients [5, 6]. The development of new drugs should therefore target alternative pathways affecting detrusor overactivity.

Tramadol is a centrally acting analgesic with two distinct mechanisms of action: one enantiomer exerts a predominantly weak µ opioid effect, whereas the other inhibits noradrenaline and serotonin reuptake, activating descending monoaminergic inhibitory pathways [7]. The effect of the non-opioid component of tramadol is mediated through α2-agonistic and serotoninergic activities, by inhibiting the re-uptake of noradrenaline and 5-hydroxytryptamine (5-HT) [8].

There is evidence that unmyelinated C fibres become predominant in mediating detrusor reflex in cases of chronic spinal cord lesions [9] and possibly in idiopathic detrusor hyperactivity [10]. C fibre hyperactivity of the bladder has been identified in patients with detrusor overactivity due to various nonspinal lesions [11]. Through desensitization of sensory fibres by capsaicin or resiniferatoxin the micturition reflex mediated by C fibres can be suppressed and detrusor overactivity can be successfully treated [12, 13]. In a study by Carlsson and Jurna [14], the effects of tramadol on motor and sensory responses of the spinal nociceptive system were tested in the rat. Tramadol at 100 and 200 µg injected intrathecally depressed both the spontaneous activity in ascending axons and their activities because of stimulation of C fibres and activation from Aδ fibres in the sural nerve.

Tramadol also inhibits the function of cholinergic receptors, such as muscarinic M1 and M3 receptors [15, 16]. Tramadol was found to inhibit micturition in rats at doses below that resulting in analgesia [17]. Recently, Pehrson et al.[18] demonstrated that tramadol effectively suppresses apomorphine-induced IDO in doses shown to have analgesic activity in rats. This supports a dopaminergic mechanism. Tramadol may be acting on IDO through several mechanisms, namely inhibition of cholinergic muscarinic receptors, inhibition of serotonin and noradrenaline reuptake with resultant bladder relaxation through stimulation of β-adrenoceptors, changes in dopamine receptor activation and/or stimulation of µ- and δ-opioid.

Opening of ATP-sensitive K+-channels (KATP) in smooth muscle cells will lead to potassium efflux and hyperpolarization, which in turn leads to reduced opening probability of Ca2+ channels and thus decreased influx of calcium. The use of KATP channel openers has shown such inhibitory effects in the detrusor of several species [19, 20]. Thus KATP channel openers may be an attractive way of treating bladder overactivity, by eliminating unwanted contraction during the filling phase. It has also been known that the modulation of both K+ channels and nitrergic pathways has a role in the perception of pain and the effects of analgesics. More recently, Yalcin et al. [21] demonstrated that nonspecific voltage-dependent K+ channels and nitrergic systems may have a role in the antinociceptive effects of tramadol in hot-plate test in mice.

These rationales and pharmacological properties prompted us to investigate the urodynamic and clinical effects of oral tramadol in the treatment of IDO.

Materials and methods

Seventy-six patients aged ≥ 18 years with cystometrically proved detrusor overactivity (idiopathic instability or detrusor hyperreflexia, or uninhibited phasic detrusor contractions with an amplitude of ≥ 10 cm H2O) and average urinary frequency of eight or more voids per 24 h on the baseline frequency–volume chart entered the study.

None of the patients had received other treatment for IDO for at least 4 weeks before the start of the study. The methods of urodynamic study were in accordance with the recommendations of the International Continence Society [22]. All subjects gave informed consent and the study protocol and questionnaire were reviewed and approved by local ethics committees. All patients underwent preliminary assessment, including a medical history, physical examination, serum chemistries, urinalysis, urine culture and assessment of urinary tract by intravenous urography (IVU) and voiding cystourethrography (VCUG).

In urodynamic studies, first uninhibited detrusor contraction, maximum bladder capacity and maximum pressure of uninhibited detrusor contractions were recorded. The bladder was emptied with a 12-Fr Nélation’s catheter. Urodynamics were done with a six-channel urodynamic apparatus using a three-way 8-Fr cystometry catheter and a 12-Fr all silicone double lumen rectal balloon catheter. The urethral pressure was sited at the highest basic pressure and calculated constantly during filling. The urodynamic catheters were fixed on the body surface with adhesive tape and their location repeatedly checked on radioscopy. Continuous bladder filling was done at 30 ml min−1 with sterile contrast media at body temperature with patients sitting. Patients were asked to report and explain all sensations they experienced in the bladder during filling. During the course of cystometry, the following four bladder characteristics were recorded: (i) capacity, (ii) sensation, (iii) compliance and (iv) the occurrence of involuntary contractions.

Exclusion criteria included stress urinary incontinence (urine escaping from urethra during coughing when the bladder was stable), maximum flow rate < 10 ml s−1, postvoid residual urine > 20% of functional bladder capacity, urinary tract infection, any proved lower urinary tract disease, intermittent catheterization, average total voided volume > 3000 ml per 24 h and, for women, being perimenopausal or menopausal.

Each eligible patient was given a randomization number using an interactive voice response system, which followed a randomization table generated by the method of random permuted blocks [23]. Persons who were geographically and operationally independent of the study investigator did the randomization of the study. During a 4-week wash-out and run-in period patients were given a diary and requested to record accurately the frequency–volume chart.

Patients were randomly assigned to receive either 100 mg tramadol sustained release (Tramadol LP) twice daily (group 1) or placebo (group 2) for 12 weeks. Sustained-release tablets release the active ingredient over a period of 12 h, and have a bioavailability of 87–95% compared with capsules [17]. Peak plasma concentrations (Cmax) in urine occurred at nearly the same time as in plasma. The urine concentrations are 43- to 46-fold higher than the corresponding plasma concentrations [24]. Tramadol is mainly excreted via kidneys (approximately 90%) [25]. The mean elimination half-life is about 5–6 h [24].

To accomplish blinding of the medication, the pharmacy prepared identical-appearing capsules of 100 mg sustained release tramadol and placebo before randomization. According to the manufacturer, tramadol and placebo were indistinguishable as to appearance, colour and flavour.

Patients were seen at entry, at baseline after run-in, and after 2, 4, 6, 8 and 12 weeks of treatment. Dose reduction was not permitted. Efficacy was assessed using 2-week voiding diaries. The primary end-point was the change in the number of episodes of urge incontinence per 24 h. Secondary end-points included changes in number of voids per 24 h and volume of urine per void.

The efficacy of the placebo and tramadol was also assessed every 2 weeks during treatment and at the end of the 12-week treatment, by determining the proportion of patients who achieved normal voiding frequency (less than eight voids per 24 h), complete response of incontinence (complete dryness on voiding diary chart) or complete response (less than eight voids per 24 h and complete dryness on the chart). For subjective judgement of improvement of bladder symptoms we used a 6-point Likert scale at baseline and week 12 [26]. Patients rated the bladder condition with choices between ‘my bladder condition does not cause me any problems at all’ and ‘my bladder condition causes me very severe problems’ on a scale of: 0, no problems; 1, very minor problems; 2, minor problems; 3, moderate problems; 4, severe problems; and 5, many severe problems. Post-treatment improvement in symptoms was defined as a score decrease of ≥ 1 point.

Statistical analysis

The study sample size of 76 patients (38 per group) was powered for a difference of approximately 1 SD between the tramadol and placebo group with a 95% confidence interval (α = 0.05) a statistical power of 85% (β = 0.15) and assuming an overall 10% dropout rate.

All efficacy measures were evaluated by comparing baseline values with those obtained after 12 weeks of treatment. Intention-to-treat analyses were performed on all efficacy variables. Data were expressed as mean ± SD. As data were not always normally distributed, nonparametric tests (Kruskal–Wallis test, Mann–Whitney Wilcoxon test or Wilcoxon signed rank test for paired test comparing a change to 0) were used to test the relations.The independent sample two-tailed t-test was used to compare the efficacy variables by tramadol and placebo. Comparison of the incidence of side-effects, percentages of patients with normalized voiding frequency, complete response of incontinence and complete response was made using the χ2 test. The level of significance was defined as α = 0.05 (two-sided). For voiding variables we also calculated 95% confidence intervals (CIs) for the differences among treatment groups in the mean change from baseline. Statistical analysis was performed using the computer statistical package SPSS release 10.1.1 for Windows (SPSS, Chicago, IL, USA; 1999) and SAS/6.4 (SAS Institute Cary, NC, USA).


Seventy-six patients were recruited, but only 69 (91%) completed the whole randomized trial study. Of these, seven withdrew from the study before week 12, including four (10.5%) on placebo and three (8%) on tramadol. Reasons for noncompletion included two adverse effects (tramadol group), three lack of effect on IDO (placebo group) and two lost to follow-up (one from each group). The difference in dropout rates was not significant between the groups. At baseline, the study and placebo groups were similar in demographic and clinical characteristics (Table 1).

Table 1. Patient characteristics
VariablesTramadol n = 35Placebo n = 34 P-value
  1. NS, Not significant.

 Mean age, years (range)39 (21–59)37 (19–57)NS
 Men:women (%)31 : 6933 : 67NS
 Median weight, kg (range) 56 (48–81)54 (49–79)NS
Clinical characteristics, %
 Mean ± SD duration of symptoms (years)  5 ± 3.2   5 ± 2.8NS
 Mean ± SD urge incontinence episodes per 24 h 3.2 ± 3.3 3.3 ± 3.1NS
 Mean ± SD voids per 24 h 9.3 ± 3.2 9.6 ± 3.1NS
 Mean ± SD urine volume per void (ml)158 ± 32151 ± 28NS

The subjects were evaluated at baseline, every 2 weeks during treatment and at the end of the 12-week treatment. The beneficial effects of tramadol were apparent at the end of the 2-week treatment and continued throughout the 12-week treatment period.

Table 2 shows voiding variables at baseline and at 12 weeks’ treatment. Patients receiving tramadol demonstrated a significant decrease in the number of voids per 24 h compared with those receiving placebo. Mean ± SD pretreatment numbers of voids per 24 h were 9.3 ± 3.2 and 9.6 ± 3.1 in groups 1 and 2, respectively. They decreased to 5.1 ± 2.1 and 8.7 ± 3.2, respectively (< 0.001) (95% CI − 5.1 - − 0.4). This benefit continued throughout the treatment period (Figure 1). The mean pretreatment urine volume was 158 ± 32 and 151 ± 28 ml per void in groups 1 and 2, respectively. The mean of urine volume per void was 198 ± 76 and 159 ± 31 ml in the tramadol and placebo groups, respectively (< 0.001) (95% CI 8-22). This benefit also continued throughout the treatment period (Figure 2). For the number of episodes of urge incontinence per 24 h tramadol was more effective than placebo in decreasing incontinence. The baseline mean number of incontinence episodes per 24 h decreased from 3.2 ± 3.3 and 3.3 ± 3.1 to 1.6 ± 2.8 and 3.1 ± 3 at 12 weeks’ treatment in groups 1 and 2, respectively (< 0.001) (95% CI − 2-− 0.3) (Table 2, Figure 3).

Table 2. Voiding and urodynamic variables at baseline and after 12 weeks of treatment
VariablesTramadolPlacebo P-value95% CI*
  1. NS, Not significant. *95% confidence Interval.

At baseline
 Mean ± SD volume per void (ml)158 ± 32 151 ± 28NS
 Mean ± SD no of incontinence episodesper 24 h  3.2 ± 3.3  3.3 ± 3.1 NS
 Mean ± SD no of voidsper 24 h  9.3 ± 3.2  9.6 ± 3.1 NS
 Patients with incontinence, n (%)  15 (42.9)  13 (38.2)NS
 Mean ± SD volume at initial contraction (ml)152 ± 43157 ± 36NS
 Mean ± SD max. bladder capacity (ml)178 ± 61173 ± 68NS
 Mean ± SD max. pressure of uninhibited detrusor contractions (cm H2O) 46 ± 14 42 ± 16NS
 Mean ± SD postvoid residual vol. (ml) 33 ± 18 35 ± 17NS
At 12-week treatment schedule
 Mean ± SD volume per void (ml)198 ± 76159 ± 31<0.001(8-22)
 Mean ± SD no of incontinence episodes per 24 h 1.6 ± 2.8 3.1 ± 3.0<0.001(− 2- 0.3)
 Mean ± SD no of voids per 24 h 5.1 ± 2.1 8.7 ± 3.2<0.001(− 5 .1- − 0.4)
 Patients with incontinence, n (%)  5 (14.3) 13 (38.2)<0.05(7 -15)
 Mean ± SD vol. at initial contraction (ml)228 ± 38 164 ± 41<0.01(14 -28)
 Mean ± SD max. bladder capacity (ml)252 ± 98181 ± 56<0.05(12 -24)
 Mean ± SD max. pressure of uninhibited detrusor contractions (cm H2O) 31 ± 18  37 ± 16<0.01(5 -19)
 Mean ± SD postvoid residual vol. (ml) 31 ± 21 36 ± 20NS
Figure 1.

Mean change from baseline in number of voids per 24 h. Tramadol (▪), placebo (◆)

Figure 2.

Mean change from baseline in mean urine volume per void. Tramadol (▪), placebo (◆)

Figure 3.

Mean change from baseline in mean number of incontinence episodes per 24 h. Tramadol (▪), placebo (◆)

Normalizing voiding frequency was achieved in significantly more patients on tramadol than on those on placebo (63%vs. 14%, χ2 test P < 0.001). Before treatment, 28 (41%) of the 69 patients reported having incontinence. Complete urinary continence was achieved by 10 of the 15 patients (mean age 29.6 years) at the end of treatment with tramadol. Urinary continence was not achieved in 13 out of 13 patients (mean age 28.8 years) with placebo (< 0.001) (95% CI 7-15). Complete response was achieved in 29% of women who received tramadol. There was no complete response at the end of the placebo study (< 0.001).

Table 2 shows the urodynamic parameters at baseline and at 12 weeks’ treatment. The mean pretreatment threshold in the first uninhibited detrusor contraction was markedly increased from baseline (152 ± 43 ml) in the tramadol group to 228 ± 38 ml compared with only a gradual and mild increase in the placebo group (from 157 ± 36 to 164 ± 41 ml) (< 0.01) (95% CI 14 -28). The mean pretreatment maximum bladder capacity was 178 ± 61 ml for tramadol compared with 173 ± 68 ml for placebo. Tramadol was superior in increasing mean maximum bladder capacity. The mean maximum bladder capacity at 12 weeks’ treatment was 252 ± 98 and 181 ± 56 ml for tramadol and placebo, respectively (< 0.05) (95% CI 12-24). The baseline and 12-week mean maximum pressures of uninhibited detrusor contractions were 46 ± 14, 42 ± 16 cm H2O and 31 ± 18, 37 ± 16 cm H2O in groups 1 and 2, respectively. The tramadol group showed a significantly decreased mean maximum pressure of uninhibited detrusor contractions than that in the placebo group (< 0.01) (95% CI 5-19). For the subjective evaluation of problems relating to the bladder conditions, 68% and 18% of patients who received tramadol and placebo, respectively, reported an improvement (< 0.05).

More adverse effects were associated with tramadol treatment (< 0.05) (Table 3). Thirty-four percent and 16% in the tramadol and placebo groups, respectively, reported adverse events. Nausea was the most frequently reported adverse event in the tramadol group, as noticed by seven (18.4%) patients. Both tramadol- and placebo-associated side-effects were mild in affected patients but two patients in the tramadol group dropped out of the study because of nausea.

Table 3. Adverse events
  n, Tramadol (patients %) n, Placebo 2 (patients %) P-value
  1. NS, Not significant.

Adverse events13 (34)6 (15.8)<0.05
 Nausea  7 (18.4)2 (5.3)<0.05
 Vomiting 3 (7.9)2 (5.3)NS
 Dizziness 2 (5.3)1 (2.6)NS
 Constipation 1 (2.6)0NS


Tramadol is widely used as an analgesic. For moderate pain it is as effective as morphine, without the respiratory depression associated with opioids [27]. Morphine is effective in blocking micturition reflexes within the central nervous system [28] and reuptake inhibitors of noradrenaline and/or 5-HT can influence micturition [29].

Detrusor overactivity in, for example, Parkinson’s disease and cerebral infarction may be due to changes in dopamine receptor activation. In a rat model of detrusor overactivity due to such changes, tramadol abolished the response to dopamine receptor stimulation [21]. In a recent study it was also found that tramadol normalized detrusor overactivity in middle cerebral artery-occluded rats [30].

During the last decade there have been many efforts to provide therapeutic alternatives to standard anticholinergics in the treatment of IDO. Our study represents the first observation of the effects of tramadol on bladder function in humans.

IDO is a common and troublesome condition. To our knowledge, an effective and well-tolerated medication for treating IDO is not yet available. Our placebo-controlled trial demonstrates that sustained release tramadol at doses of 100 mg twice daily is efficacious with respect to frequency–volume chart parameters. Clinical onset of the effect of tramadol was rapid. Patients achieved approximately 70% of the maximum effect within 2 weeks of beginning treatment (Figures 1–3). This finding is useful, since after patients achieve symptomatic relief they may tolerate better fluid intake, as shown by the improved volume per void.

In this study tramadol induced a significant decrease in first uninhibited detrusor contraction (< 0.01) and in maximum bladder capacity (< 0.05), and a significant decrease in maximum pressure of uninhibited detrusor contractions (< 0.01) compared with placebo at 12 weeks’ treatment. As the increase in micturition volume after (±)-tramadol was abolished by pretreatment with naloxone, µ-opioid receptor stimulation was suggested to be the main inhibitory mechanism of (±)-tramadol on normal rat micturition [31]. However, within the brain, opioid, serotonin and noradrenaline pathways are intimately coupled [32], suggesting a potential interaction between opioid and monoaminergic mechanisms in the action of (±)-tramadol.

Tramadol was well tolerated and no serious safety concerns were identified. There were few serious adverse events or withdrawals owing to adverse events, which were typically associated with tramadol, namely nausea, which occurred infrequently. Compliance was high in most patients. About 18% (seven) of patients reported nausea. We did not observe any significant difference in the incidence of adverse effects, except for nausea, which occurred significantly more frequently with tramadol. Two patients receiving tramadol withdrew from treatment due to nausea. Given the favourable tolerability profile of tramadol, future trials should evaluate its efficacy and safety with different regimens and doses to improve further the symptoms of detrusor overactivity.

In conclusion, tramadol was effective and well tolerated in treating IDO. These promising results should motivate the development of further double-blind, placebo-controlled trials including more patients and a long-term study to confirm our findings and determine the acceptability of this treatment over a longer follow-up.