Background
Prior study of patients with urgency urinary incontinence by functional magnetic resonance imaging showed altered function in areas of the brain associated with interoception and salience and with attention. Our randomized controlled trial of hypnotherapy for urgency urinary incontinence demonstrated marked improvement in urgency urinary incontinence symptoms at 2 months. A subsample of these women with urgency urinary incontinence underwent functional magnetic resonance imaging before and after treatment.
Objective
This study aimed to determine if hypnotherapy treatment of urgency urinary incontinence compared with pharmacotherapy was associated with altered brain activation or resting connectivity on functional magnetic resonance imaging.
Study Design
A subsample of women participating in a randomized controlled trial comparing hypnotherapy vs pharmacotherapy for treatment of urgency urinary incontinence was evaluated with functional magnetic resonance imaging. Scans were obtained pretreatment and 8 to 12 weeks after treatment initiation. Brain activation during bladder filling and resting functional connectivity with an empty and partially filled bladder were assessed. Brain regions of interest were derived from those previously showing differences between healthy controls and participants with untreated urgency urinary incontinence in our prior work and included regions in the interoceptive and salience, ventral attentional, and dorsal attentional networks.
Results
After treatment, participants in both groups demonstrated marked improvement in incontinence episodes ( P <.001). Bladder-filling task functional magnetic resonance imaging data from the combined groups (n=64, 30 hypnotherapy, 34 pharmacotherapy) demonstrated decreased activation of the left temporoparietal junction, a component of the ventral attentional network ( P <.01) compared with baseline. Resting functional connectivity differed only with the bladder partially filled (n=54). Compared with pharmacotherapy, hypnotherapy participants manifested increased functional connectivity between the anterior cingulate cortex and the left dorsolateral prefrontal cortex, a component of the dorsal attentional network ( P <.001).
Conclusion
Successful treatment of urgency urinary incontinence with both pharmacotherapy and hypnotherapy was associated with decreased activation of the ventral (bottom-up) attentional network during bladder filling. This may be attributable to decreased afferent stimuli arising from the bladder in the pharmacotherapy group. In contrast, decreased ventral attentional network activation associated with hypnotherapy may be mediated by the counterbalancing effects of the dorsal (top-down) attentional network.
Introduction
Urgency urinary incontinence (UUI), involuntary urine loss associated with urgency, affects millions of women, significantly impairs quality of life, and is commonly treated with medications that have side effects limiting adherence. , Altered perceptual awareness may play a role in UUI and suggests an alternative treatment approach by addressing abnormalities in brain function. Evaluation of patients with UUI using functional magnetic resonance imaging (fMRI) has identified specific brain sites that may function abnormally, potentially serving as targets for brain-centered therapies.
Women with UUI manifest abnormal activation in brain regions that govern interoception, the perception and interpretation of physiological stimuli. Output from interoceptive regions is modulated by interaction with brain regions responsible for executive control, including those responsible for direction of attention. These brain regions likely modify urgency perception in UUI and may be important in its genesis or persistence.
Why was this study conducted?
The specific aim of this study was to determine if treatment of urgency urinary incontinence with hypnotherapy compared with pharmacotherapy would be associated with changes in brain networks that direct attention.
Key findings
After 3 months of treatment, both therapies markedly improved symptoms and were associated with decreased activation of the ventral bottom-up attentional network during bladder filling. However, only hypnotherapy was associated with increased connectivity of the dorsal top-down attentional network.
What does this add to what is known?
Hypnotherapy, a mind-body therapy, may decrease bottom-up attentional activation associated with bladder filling without pharmacologic alteration of bladder function. This may be mediated by changes in function of top-down attention.
Activity in brain regions associated with interoceptive and attentional control can be evaluated using fMRI by measuring the effects of localized changes in blood flow occurring in response to recurrent prompted tasks using blood oxygen level–dependent (BOLD) imaging. Changes in brain blood flow measured over time with BOLD imaging can also assess which brain regions are integrated or connected. Connectivity between regions is most commonly assessed in the absence of a task and is inferred when areas of the brain show temporally coherent neural activation and deactivation constituting functional networks. Each functional network interconnects spatially distinct areas of the brain that together implement a unique component of cognition. Some anatomically defined areas of the brain can be components, or nodes, of more than 1 network; for example, the dorsolateral prefrontal cortex (DLPFC) is a component of both the frontoparietal and dorsal attentional networks. Anatomy and function of the interoceptive network and the salience network, which is responsible for detecting and filtering stimuli, overlap considerably. Relevant networks and their nodes are noted in Table 1 .
| Terminology | Abbreviation | Description anatomic areas and functional |
|---|---|---|
| Anterior cingulate cortex, dorsal anterior cingulate cortex | ACC, dACC | Anterior portion of the cingulate gyrus located in the medial cerebral hemisphere; important in emotion and coordinating responses to internal and external events ( Figure 6 ) |
| Brain oxygen level–dependent | BOLD | fMRI sequence that detects local changes in blood oxygen levels caused by increased cerebral blood flow (hemodynamic response) at sites of increased neural activity |
| Corona radiata | — | Continuation of white matter tracts above level of the internal capsule, terminating in the cerebral cortex |
| Default mode network | DMN | Brain regions with increased activity when mind is not engaged in cognitive tasks and deactivated during cognitive tasks; nodes include the ventromedial PFC and the posterior cingulate cortex |
| Dorsal attentional network | DAN | Brain regions associated with focused attention on external stimuli |
| Dorsolateral prefrontal cortex | DLPFC | Region in PFC; a component of both the frontoparietal control network and the DAN ( Figure 6 ) |
| Frontoparietal control network | — | Network typically activated during cognitively demanding tasks; nodes include the lateral prefrontal cortex including DLPFC, IPS, SPL |
| Midcingulate cortex | MCC | Middle portion of cingulate cortex including anterior MCC (sometimes referred to as dACC and posterior MCC); activated in emotion and pain processing and intentional motor control |
| Posterior medial cortex | PMC | Brain region including the precuneus and posterior cingulate; functions as the posterior node of the DMN |
| interoceptive-salience network | — | Brain regions that integrate visceral, autonomic, emotional sensory data to assess salience; the ACC is a component of this network |
| Temporoparietal junction | TPJ | Brain region at the junction of the temporal and parietal lobes; a component of the ventral attention network ( Figure 6 ) |
| Ventral attentional network | VAN | Brain regions activated in response to unexpected behavioral stimuli, or a bottom-up network; includes the TPJ and ventral frontal or prefrontal cortex |
| Ventrolateral prefrontal cortex | VLPFC | Brain region located in PFC; a region that is part of the ventral attention network ( Figure 6 ) |
Pioneering fMRI evaluation of UUI identified increased activation of the interoceptive-salience network compared with controls and suggested that this activation could be modulated by components of the frontoparietal control network. , Our prior UUI fMRI study demonstrated increased activation in regions in the interoceptive-salience network among patients with UUI and differences between patients with UUI and controls with respect to activation and connectivity of attentional networks. Abnormal function of these brain networks may contribute to UUI symptoms. Because mind-body therapies have the potential to alter function of intereoceptive-salience frontoparietal control, and attentional networks, they warrant exploration as a means of treating UUI.
Mind-body therapies may alter attentional network function by their effect on the DLPFC. This anatomic site is a component of the dorsal attentional network, the network responsible for top-down direction of attention. , , Hypnosis, a mind-body therapy, is defined as “a state of consciousness involving focused attention and reduced peripheral awareness characterized by an enhanced capacity for response to suggestion.” fMRI of subjects under hypnosis has suggested augmented connectivity of the dorsal attentional network and interoceptive networks. , Hypnotherapy’s efficacy in treating UUI in our recent randomized controlled trial (RCT) provides the opportunity to identify neural correlates of therapeutic success.
The objective of this study was to determine if improvement in UUI symptoms following hypnotherapy, a mind-body therapy, is associated with amelioration of abnormal brain function on fMRI. We hypothesized that hypnotherapy treatment of UUI would modulate neurologic interactions between the brain and bladder and that hypnotherapy participants would manifest different changes in brain activation and functional connectivity than participants using UUI medications. Based on our prior work evaluating preintervention fMRI characteristics in a subset of the participants of this trial, we hypothesized that postintervention differences in activation and connectivity would involve the interoceptive-salience and attentional networks.
Materials and Methods
Study design
This study represents a subsample of participants from our RCT comparing hypnotherapy vs pharmacotherapy for the treatment of UUI ( ClinicalTrials.gov # NCT01829425 ) in women. The subsample consists of those participants who consented to undergo pre- and posttreatment fMRI scanning. The study was approved by the University of New Mexico Institutional Review Board (HRRC #09-314, January 24, 2012 initial approval with yearly reapproval), and all participants gave written consent. Methodology for the Hypnotherapy Or Pharmacotherapy Trial has been described previously. In brief, participants were recruited from an academic urogynecology clinic and the community at large between March 2013 and April 2016. Pilot study results comparing overactive bladder and controls suggested effect sizes >1 for between-group differences in interoceptive and salience network activation and functional connectivity. Power calculations made with the assumption of 40% to 60% improvement in abnormal activation and connectivity following hypnotherapy suggested that 60 to 75 participants could achieve a power of 0.8.
UUI participants were nonpregnant women aged 18 years or older who had ≥3 UUI episodes (UUIEs) per week for ≥3 months and were free of significant neurologic illness and pelvic organ prolapse beyond the hymen. Women who participated in this optional fMRI substudy were identified before treatment randomization. Randomization was stratified by UUI severity (≤3 or ≥4 UUIEs on 3-day diary). Baseline data collection included validated questionnaire results and UUIEs on diaries ( Table 2 ). Participants underwent cystometrics to determine their individual “strong desire to void” volumes and bladder volumes eliciting a “persistent desire to pass urine without fear of leakage.”
| Clinical and questionnaire characteristics | Hypnotherapy (n=30) | Medication (n=34) | P value of difference |
|---|---|---|---|
| Age, mean (SD) | 54 (13) | 57 (10) | .30 |
| UUIEa at baseline, mean (SD) | 8 (8) | 7 (12) | .50 |
| UUIEa at 2-mo follow-up, mean (SD) | 1 (4) | 1 (3) | 1.00 |
| Change in UUIE a , median (Q1, Q3) | −5 (−2, −10) | −4.5 (−3, −7) | .60 |
Randomized controlled trial interventions
Hypnotherapy
Participants randomized to hypnotherapy received 8 weekly, 1-hour, one-on-one bladder-directed hypnotherapy sessions delivered by a board-certified hypnotherapist using a standardized format, also described previously. ,
Pharmacotherapy
Participants randomized to medications received extended-release oxybutynin 10 mg/d initially or extended-release tolterodine 4 mg/d. Participants also received 8 weekly one-on-one medication counseling sessions delivered using a standardized format. ,
Functional magnetic resonance imaging task and resting connectivity
fMRI scans were obtained pretreatment and 8 to 12 weeks after treatment initiation. To activate specific sites within the brain during MRI, tasks are performed. In this case, the fMRI task consisted of low-, medium-, and high-volume bladder infusions as previously described ( Supplement A ). Our prior study comparing UUI participants to controls demonstrated differences between groups largely confined to high bladder fill volumes, and therefore, comparisons for this study were confined to the high-volume task.
Resting state connectivity data collection was performed over 5 minutes, initially only with an empty bladder. After study initiation, newly published literature reported UUI vs control differences in resting functional connectivity acquired with a partially filled bladder. Accordingly, the study protocol was amended to include a second resting data collection performed with the bladder filled to half the previously determined “strong desire to void volume” for that particular patient.
The imaging acquisition for our fMRI data has been described previously. Multiple regions of interest (ROIs) were derived from our prior work, identified as showing significant activation differences between control and UUI participants. These included the dorsal anterior cingulate cortex (dACC), midcingulate cortex (MCC), bilateral ventrolateral prefrontal cortex (VLPFC), bilateral temporoparietal junction (TPJ), and posteromedial cortex ( Table 1 ).
Statistical analysis
During the bladder-filling task, functional activation values within the aforementioned ROIs were acquired for participants before and after treatment, and analysis of variance (ANOVA) was performed to evaluate differences in functional activation across groups (hypnotherapy vs pharmacotherapy) and time (pretreatment vs posttreatment). Treatment group, time, and the treatment group × time interactions were the effects of interest in this statistical framework.
The same ROIs were used as seed points for resting state functional connectivity analysis examining pre- to posttreatment group differences. Correlations were calculated between the average time-course of activation in these regions and voxel-wise time-course of activation throughout the brain. ROIs and brain sites showing increased connectivity were assigned to corresponding brain networks based on previously reported Talairach spatial coordinates or Brodmann areas. Whole-brain resting state functional connectivity results were corrected for false positives at P <.05 based on 10,000 Monte Carlo simulations ( P <.001; minimum cluster size=1024 μL).
Between-group differences in baseline values for ROI activation or ROI functional connectivity were evaluated by t tests. If paired t tests suggested significant baseline differences between groups, the pre-post treatment differences were further analyzed with analysis of covariance (ANCOVA), using baseline values as a covariate.
Clinical differences between groups were also assessed. Between-group differences in questionnaire and baseline UUIEs were compared using t tests. Because our prior work showed that change in UUIEs 2 months after treatment was not normally distributed, pre-post treatment changes in UUIE difference between groups were compared using Mann Whitney U test.
Results
A total of 72 participants, 36 treated with hypnotherapy and 36 with pharmacotherapy, enrolled in this imaging study; 3 did not complete posttreatment scanning and 5 were excluded secondary to technical problems with fMRI acquisition or because of excessive head motion (defined as 3 times the interquartile range relative to their cohort based on frame-wise displacement) during the bladder-filling task. Figure 1 displays fMRI data inclusion and exclusion. Task data from 64 participants, empty bladder resting connectivity data from 60 participants, and partially filled bladder resting connectivity data from 52 participants were analyzed. Groups did not differ in participant baseline questionnaire scores or in UUIEs ( Table 2 ). After treatment, both hypnotherapy and the medication groups demonstrated marked improvement in UUIEs (both P <.001), and the degree of improvement did not differ between groups ( P =.60; Figure 2 ).
