SIR—Urinary incontinence (UI) after stroke is a strong predictor of mortality, dependency and need for institutional care [1, 2], and so far there is no effective treatment available [3]. The reasons for this are not fully understood. In a recent prospective study, we found two UI categories: [urinary urge incontinence (UUI), n = 27] and incontinence with impaired awareness (IA) of the need to void [impaired awareness-urinary incontinence (IA-UI), n = 38] [4]. Only the latter was an independent risk factor for poor outcome after 3 months [4] and 1 year [5]. The symptom severity ranged from slight unawareness to anosognosia of leakage; those who recognised their new symptom often were little concerned about it. This suggests that besides the sensory cortex, also attentional systems, cognitive areas, and areas for emotional validation may be involved. Thus, the frontoparietal lobes and adjacent subcortical regions are likely to be affected. In a previous paper, we demonstrated a higher rate of parietal lesions on computerised tomography (CT) in patients with anosognosia of leakage, compared to those who recognised it [6]. However, neither old stroke lesions nor periventricular white matter lesions (WMLs) were addressed, and no comparisons were made with the continent group. Many patients experience recurrent strokes [7], which increases the risk of functional and cognitive decline [8]. WMLs are strongly associated with vascular risk factors [9]. White matter contains numerous pathways and facilitates information processing within networks. As voluntary control of micturition is a complex system, WMLs may contribute to the development of UI when a stroke occurs. Therefore, we wished to investigate the relationship of continence status to the site and extent of both new and old vascular brain lesions, including WMLs, in a more comprehensive way. In all 315 out of 563 patients with acute first-ever or recurrent strokes consecutively admitted to our hospital were assessed between day 7 and 12 after symptom onset and followed up during 1 year. The inclusion/exclusion criteria and assessment methods have been reported earlier [4]. Of the 315 patients, 170 were continent before and after the stroke, whereas 65 became incontinent or developed a leakage type that was different from the pre-stroke type. In order to avoid as much assessment bias as possible in this substudy, only patients without pre-stroke storage symptoms and those with an unequivocal change in leakage symptoms were considered. All patients had undergone a CT at least once as part of the clinical management. The time for the onset of stroke symptoms was obtained from the hospital records or by interview with the patients or their carers. In order to increase the possibility of detecting a fresh ischaemic lesion, we used the latest CT taken during the index stay with the patient in a neurologically stable condition, and set the time window between > 5h (ischaemic strokes) and 4 weeks after symptom onset [10]. Eventually, 185 patients were included (131 continent, 23 UUI, 31 IA-UI). For details, see Appendix 1 in the supplementary data on the journal's website http://www.ageing.oxfordjournals.org. One skilled neuroradiologist who was not aware of the patients' clinical data retrospectively rated all CT scans. In difficult cases, consensus was reached by consulting with another experienced colleague [10]. Stroke lesions were assessed according to a modified version of the scale used by Nys et al. [11]. New lesions were differentiated from old lesions by their appearance on CT. Lesion volume was measured by the three largest diameters [12, 13] (mean value from three measurements). For the assessment of WMLs, the rating scale by Wahlund et al. [14] was applied. Table analyses and logistic regression modelling were chosen as statistical methods; for details see Appendix 2 in the journal's website: http://www.ageing.oxfordjournals.org. There were 103 (56%) men, and 82 (44%) were older than 80 years; no sex differences were detected between the continence groups. Patients with UUI differed from the continent group only by a more frequent history of previous strokes; patients with IA-UI were older and more often mentally impaired before the stroke than the continent group (all P<0.01). There were only seven new and six old visible infratentorial lesions. The IA-UI group had right anterior circulation stroke syndromes more frequently, and showed more often extensive new supratentorial lesions (Table 1). Characteristics of stroke lesions according to continence status P ≤ 0.01 (Chi square test, all compared to continent group); n = 185. UI, urinary incontinence; IA, impaired awareness; CT, computerised tomography; BG, basal ganglia. Characteristics of stroke lesions according to continence status P ≤ 0.01 (Chi square test, all compared to continent group); n = 185. UI, urinary incontinence; IA, impaired awareness; CT, computerised tomography; BG, basal ganglia. Patients with UUI had a slightly higher total WML score than the continent group (P = 0.05), mostly due to higher scores in the left frontal region (P = 0.08). The IA-UI group had significantly higher scores in all regions (P<0.01); it differed from the urge group by a higher right parietal and total right WML score (all P<0.05), see Appendix 3 in the journal's website: http://www.ageing.oxfordjournals.org. In bivariate analysis, there were no statistically significant differences between the continent and the UUI group regarding age, gender, stroke volume (new, old, or combined), or region. Thus, a multivariate regression analysis was performed only with the continent versus the IA-UI group. Extensive WMLs, new basal ganglia (BG) involvement and new lesions involving two or more lobes or regions were all strong and independent risk factors; older age was a weaker factor (Table 2). When CT had been taken later than 24 h after stroke onset, BG involvement was no longer statistically significant. Pre-stroke mental impairment significantly increased the effect of BG involvement and total WML score. No other statistically significant interactions were found regarding time to CT, age, gender, or between the other variables in the model (P > 0.1, data not shown). Clinical and neuroradiological factors related to incontinence with impaired awareness Bivariate and multivariate analysis; n = 162. OR, odds ratio; CI, 95% confidence interval. Dichotomised by the median. Clinical and neuroradiological factors related to incontinence with impaired awareness Bivariate and multivariate analysis; n = 162. OR, odds ratio; CI, 95% confidence interval. Dichotomised by the median. To our knowledge, this is the first prospective study comparing continent patients and patients with different clinical subtypes of urinary incontinence after stroke with respect to cerebrovascular lesions. Most other prospective studies include symptomatic patients only [15, 16], are mainly based on urodynamic observations [16, 17], or relate neuroimaging findings to the presence or absence of any micturition disturbances regardless of type [18, 19]. In the IA-UI group, parietal involvement seemed to be particularly important when combined with frontal lobe or BG injury, whereas isolated BG lesions had impact only when they were visible on CT within 24 h, i.e. were more extensive. The role of BG is to receive afferent signals from the bladder and to pass them on to the insular cortex and the limbic structures which map, process and give emotional value to the signals [20]. This could explain the reduced awareness and also the dull emotional reactions of those who recognised their new symptom. Lesions of the parietal lobe can lead to anosognosia for various sensorimotor and cognitive disturbances [21]. This has been shown with respect to hemiplegia [22] and visual problems [23], particularly when the lesions are right-sided, extensive, bilateral or combined with frontal lesions [22]. This coincides with our observation that the parietal lobe was involved in all but one of the cases with anosognosia for incontinence and new extensive right-sided lesions. Another risk factor was the presence of extensive white matter changes in most supratentorial areas, particularly in the right parietal region. Three small studies have shown that elderly persons with WMLs on CT are more likely to have UI than age-matched controls without WMLs [24–26]. The clinical features of UI were not specified in those studies, and no other prospective studies address WML as a risk factor for UI in a stroke population. However, these papers support our observations. Except for a greater WML load compared to continent individuals, which was mostly due to left frontal involvement, no other neuroradiological factors could be identified for patients with UUI. The influence of other frontal regions—particularly the orbitofrontal cortex—on conscious attention to bladder sensations and voluntary suppression of the voiding reflex has been repeatedly demonstrated in clinical and experimental studies [27–29]. We were not able to show this or any other possible relationship, probably because of the limited number of visible new strokes. Apart from the small number of symptomatic patients that renders the risk estimates uncertain, this study has some other limitations. The authors had no means of influencing the neuroimaging technique or the time points for imaging. CT is less sensitive to ischaemic vascular lesions than MRI and only about 50% are ever visualised [30]. In order to partially compensate for that, the latest CT taken during the hospital stay was chosen. The differing time points, in turn, may have weakened the significance of the topical measurements. However, only seven CTs had been taken between 10 days and 4 weeks after symptom onset, a time frame when fresh ischaemic lesions may become less visible [30]. The adjustment for time was assumed to further ameliorate the problem. CT is also less sensitive for WMLs, yet its ability to detect medium and large lesions is nearly equal to MRI when using Wahlund's scale, and suitable for cross-sectional studies. The under-representation of severe left-sided strokes may have biased the results. In summary, post-stroke incontinence with impaired awareness is related to multiregional strokes, frequently involving the frontoparietal lobes and BG, and to extensive supratentorial WMLs. UUI probably reflects a less serious damage. WMLs seem to be important in both UI types. Because acute lesions generally affect structures for attention to and recognition of bodily sensations, awareness training techniques in those who recognise their new symptom might be a possible treatment option during rehabilitation. Post-stroke urinary incontinence can occur with urge symptoms or with impaired awareness (IA) of the need to void. Patients with both subtypes have WMLs more often than continent patients. Incontinence with IA reflects greater cerebral damage than urge incontinence, involving structures for attention and recognition. Awareness training techniques might be a possible treatment option in selected patients. None. None.
No takes yet. Share an insight, caveat, or question.
Pettersen et al. (2008) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: