Control of bladder sensations: An fMRI study of brain activity and effective connectivity

Abstract

When the bladder is fairly full, the desire to void can be suppressed, but it can also be called forth deliberately. We studied brain activity during such intentional modulations of bladder sensation in 33 healthy volunteers (17 women, 16 men). The supplementary motor area, midcingulate cortex, insula, frontal operculum, and right prefrontal cortex were consistently more active when the desire to void was enhanced without allowing urine to pass (“attempted micturition”) than during a baseline task when bladder sensations were suppressed. The right anterior insula and midbrain periaquaeductal grey (PAG) were more active at higher than at lower bladder volumes. Responses of the right thalamus and several other right-hemispherical regions were stronger in women than in men. Using the psychophysiological interaction (PPI) method, we found that the midcingulate cortex had stronger connectivity (indicated by parallel co-variations of the activation time series) with the PAG and medial motor areas during “attempted micturition” than during the baseline task, possibly reflecting monitoring of urethral sphincter contractions. Conversely, the left and right insula showed decreased connectivity with many other brain regions during “attempted micturition”, possibly due to predominant processing of bladder-afferent input. Intentional modulations of the desire to void change the effective connectivity of supraspinal regions involved in bladder control.

Introduction

Control of urinary continence involves several levels of the central nervous system (Blok, 2002). Information on the intravesical pressure is encoded in the activity of thin myelinated afferent fibres (Häbler et al., 1993) and conveyed via the spinothalamic tract to the mesencephalic periaquaeductal grey (PAG), which in turn projects to the pontine micturition centre (PMC). In case of the voiding reflex, the PAG excites the PMC once the afferent input becomes strong enough (Griffiths, 2002). Efferent signals of the PMC then cause the bladder to contract and the urethral sphincter to relax until voiding takes place. The voiding reflex is normally controlled by cortical centres (Griffiths and Tadic, 2008). If necessary, healthy adults can therefore postpone micturition despite a full bladder and can deliberately initiate voiding even if the bladder is nearly empty. Previous neuroimaging studies have identified several cortical regions that are involved with the control of continence, namely the insula, the cingulate gyrus, and the prefrontal cortex (for reviews, see Griffiths and Tadic, 2008, Kavia et al., 2005).

Intact bladder control requires conscious perception of bladder filling (Griffiths, 2007). In general, the intensity of bladder sensation increases with bladder volume, progressing from a first sensation of filling, via a first desire to void, to a strong or even painful desire to void. However, the relationship between bladder volume and the intensity of the sensation is not fixed (Kavia et al., 2005). When the bladder is fairly full, the desire to void can be suppressed and ignored, but it can also be called forth deliberately. A previous fMRI study identified brain regions that were active during such intentional modulations of the desire to void in healthy female volunteers (Kuhtz-Buschbeck et al., 2005). During an urge (U) task, the women directed their attention to the sensations arising from the bladder and urethra and increased the desire to void as one does when initiating micturition. Since they did not allow urine to pass, this U-task represents “attempted micturition”. The task was performed ten times because fMRI requires averaging of frequently repeated events. The supplementary motor area (SMA), the cingulate cortex, the prefrontal and posterior parietal cortex, the insula and frontal operculum were more active during “attempted micturition” than during suppression of bladder sensations (Kuhtz-Buschbeck et al., 2005). Possible effects of the bladder volume on the task-related brain activity during “attempted micturition” were not previously analysed and corresponding data for men are lacking up to now.

The present fMRI study was therefore undertaken with the following goals: firstly, to identify brain regions that are consistently active during “attempted micturition”, using a conservative statistical threshold in a large group of normal volunteers (16 men, 17 women). Secondly, to test for gender-related differences in brain activity during this task. Thirdly, to study the effect of different bladder volumes by comparing measurements made at larger vs. smaller bladder volumes. Fourthly, to explore the effective connectivity of cortical regions involved in bladder control during “attempted micturition”. The importance of network interactions between brain regions in mediating sensorimotor and cognitive tasks is increasingly acknowledged (Horwitz et al., 2005). Up to now, however, only one very recent study has examined the connectivity of supraspinal bladder control regions by physiophysiological interaction analysis (Tadic et al., 2008).

Attention to visual stimuli is known to modulate the cortico-cortical connectivity of visual areas and posterior parietal regions (Friston et al., 2003). Could attention to bladder sensations have similar effects on the relevant brain regions? “Attempted micturition” requires interoceptive awareness and selective attention, which yield perceptual prominence to the desire to void. To investigate task-related changes of effective connectivity, we performed psychophysiological interaction (PPI) analyses (Friston et al., 1997). A PPI means that the contribution of one brain area (source region) to another area (target region) changes significantly with the psychological or cognitive context (here: attention to bladder sensations). Four source regions known to be involved in bladder control were chosen a priori, namely the cingulate cortex, the PAG, and the left and right insula. Among other functions, the cingulate cortex is implicated in interoceptive awareness and pain processing (Craig, 2002, Critchley et al., 2004), modulation of bodily arousal and autonomic responses (Critchley et al., 2003), and it is active during micturition (Blok et al., 1998, Nour et al., 2000). The PAG links afferent and efferent pathways of bladder control (Kavia et al., 2005). Afferent signals from the bladder and urethra are relayed via the PAG and thalamus to the insula, possibly with a predominance of the right side (Griffiths and Tadic, 2008). We expected the effective connectivity between these source regions and other brain areas (target regions) to change during “attempted micturition”, in comparison to a baseline (B) task, in which the desire to void was suppressed. Hence we looked for task-dependent influences from the source regions to other (“target”) brain regions.