Obesity, defined as a body mass index (BMI) equal to or greater than 30 kg/m2, has been increasing worldwide. Over 42% of adult Americans have obesity, and the worldwide obesity prevalence is reaching epidemic proportions [1]. There are significant comorbidities associated with obesity including coronary artery disease, hypertension, certain types of cancer, non-insulin-dependent diabetes mellitus, dyslipidemia, osteoarthritis, and pulmonary diseases [2].

Obesity has also been linked to the pathophysiology of multiple pelvic floor disorders, including urinary incontinence (UI) [3]. Several studies have investigated the association between UI and obesity, suggesting that increased intra-abdominal pressure may weaken pelvic muscles and increase urinary bladder pressure with urethral hypermobility. Additionally, prolonged effects of strain of the pelvic musculature have a chronic impact on pelvic organ function [4]. Women with more severe obesity are at increased odds of having at least one pelvic floor disorder, indicating that the severity of obesity is also associated with these conditions [5].

The prevalence of UI in women with obesity seeking weight loss surgery is as high as 67% [6]. Modest weight loss of 5–10% total body weight has shown to improve UI and quality of life [7]. Weight loss is strongly considered a first line treatment for UI, however, patient adherence and long-term efficacy of non-surgical weight loss programs are poor. Bariatric surgery has become widely accepted as an effective treatment option for obesity, with significant, sustainable, long-term weight loss and improvement or remission of many obesity-related comorbidities [8]. We hypothesized that bariatric surgery would significantly improve obesity-related UI symptoms as well as improve quality of life.

Methods

Study design

This is an interim analysis of an ongoing, prospective, single-institution observational study of adult women (age ≥ 18 years) undergoing bariatric surgery from January 2019 to present at the University of Oklahoma Metabolic and Bariatric Surgery program. Institutional IRB approval for this study was obtained. Inclusion criteria are: female sex, adult age (≥ 18 years), and active enrollment in our multidisciplinary bariatric surgery program (Table 1). At baseline, all patients were evaluated for bariatric surgery per our standard protocols. Patients who underwent treatment for urinary symptoms within 3 months of enrollment, known urinary tract anomaly or malignancy, previous urinary diversion, or history pelvic radiation were excluded.

Table 1 Baseline demographic & clinical data
Full size table

Participants completed the Pelvic Floor Distress Inventory (PFDI-20), International Consultation on Incontinence Questionnaire—Urinary Incontinence Short Form (ICIQ-UI-SF), King’s Health Questionnaire (KHQ), and Patient Global Impression of Improvement (PGI-I). Questionnaires were administered upon enrollment, pre-operatively, and at 3, 6, and 12 months post-operatively. The preoperative time point was included to assess change in UI symptoms and quality of life-related dietary weight loss alone, which is a mandated requirement of our program. Patients who had an enrollment BMI between 40 and 49.9 were required to lose 5% total body weight (TBWL) and patients with a BMI ≥ 50 were required to lose 10% TBWL. Patient with an enrollment BMI < 40 did not have a defined pre-surgical weight loss requirement but were encouraged to lose weight prior to surgery.

Demographic and clinical data were collected at all study time points. Study data were stored in a HIPAA-compliant RedCap database and de-identified after collection and prior to data analysis. The primary outcomes of the study were changes in UI survey scores and perceived quality of life related to UI symptoms. Secondary outcomes were change in BMI and %TBWL.

Scoring

PFDI-20

The PFDI-20 is a validated questionnaire consisting of 20 questions regarding pelvic floor-related symptoms. This survey was selected primarily due to the Urinary Distress Index (UDI-6) component, which contains questions related to urinary symptoms. The two additional components include questions related to pelvic organ prolapse and defecatory dysfunction (Pelvic Organ Prolapse Distress Index 6, Colorectal-Anal Distress Inventory 8). Scoring is focused on the mean score of the three subsections, and then combined to produce an overall score from 0 to 300, with higher scores indicated more severe symptoms.

ICIQ-UI-SF

The ICIQ-UI-SF is a validated survey used to evaluate the frequency, severity, and impact on quality of life from UI symptoms. It is a 4-question survey utilized to obtain a brief yet comprehensive summary of the level and impact of UI. Scoring ranges from 0 to 21, with higher scores indicating more severe symptoms.

KHQ

The KHQ is a validated survey designed to evaluate the subjective impact of UI on quality of life. It addresses nine different domains with 21-questions using Likert scales. The domains are general health perception, incontinence impact, role limitations, physical and social limitations, personal relationships, emotions, sleep/energy, and severity coping measures. Scores for each domain are calculated through a complex system that manages missing values. Results range from 0 to 100, with higher scores indicated a worse quality of life.

PGI-I

The PGI-I is a validated survey for women with stress UI who have undergoing UI treatment and is intended to assess perception of treatment response. It is a single question survey comparing symptoms before and after the therapeutic intervention and is scored from 1, “very much better”, to 7 “very much worse.” Because this is a measure of improvement after intervention, it was not assessed at the initial time of enrollment in the bariatric surgery program.

Operative procedure

Patients underwent either one of two primary bariatric procedures—laparoscopic Roux-en-Y gastric bypass (RYGB) or laparoscopic vertical sleeve gastrectomy (VSG)—or a revision of a previous bariatric procedure to a RYGB or VSG for the purpose of obesity treatment. UI symptom severity had no influence on procedure selection. All patients were operated on by one of two bariatric surgeons at a single-institution. There was no unintentional bladder or urinary tract injury at time of surgery for any of the procedures.

Statistics

Chi square test was used to examine categorical data and Student’s t-test for continuous data. All baseline data at time of enrollment was used as reference for t-tests, except for PGI-I which used pre-operative data (symptom improvement related to dietary weight loss) as reference for evaluation of symptom improvement related to surgical weight loss. Univariate and multivariable analyses were used to identify risk factors for UI symptoms. P-values < 0.05 were considered significant. Mean and standard deviation were reported for continuous variables, and the frequency and percentage for categorical variables.

Results

At interim analysis, 108 patients had enrolled in the study and 60% had progressed to surgery. We analyzed the following surveys: enrollment (n = 108), pre-operative (n = 43), 3-month (n = 29), 6-month (n = 26), and 1-year (n = 27). Mean BMI decreased from 49.8 ± 10.0 to 31.1 ± 6.5 at 1-year (35.1% ± 9.5% TBWL). All surveys showed significant improvement in UI symptoms over time (Fig. 1). The only notable risk factor for UI symptoms was parity (p = 0.044). Starting BMI, age, race, surgery type, and complications did not affect overall UI symptom severity (Table 2).

Fig. 1
figure 1

Urinary incontinence (UI) symptoms and quality of life percent improvement following bariatric surgery when compared to baseline evaluation with body mass index (BMI). Improvement is seen in all categories following bariatric surgery. Average BMI and questionnaire scores measuring UI symptoms and UI-related quality of life were compared over time at pre-determined points including baseline enrollment in bariatric program, pre-operative, and post-operative evaluations. As PGI-I is a measure of improvement, scores were compared to the pre-operative timepoint instead of baseline. PFDI-20 Pelvic Floor Distress Inventory, ICIQ-UI-SF International Consultation on Incontinence Questionnaire—Urinary Incontinence Short Form, KHQ King’s Health Questionnaire, PGI-I Patient Global Impression of Improvement

Full size image
Table 2 Multivariable predictors of UI by PFDI-20 score
Full size table

Overall, UI symptoms (PFDI-20 overall score) were correlated with BMI at time of survey and %TBWL (p = 0.03, p = 0.019) with the urinary leakage component (UDI-6) having the strongest relationship (p < 0.0001). PFDI-20 scores improved from 43.8 ± 36.0 at baseline to 35.0 ± 30.0 at 1-year, with the most improved score noted at 6-months (27.5 ± 26.8). Statistical significance in score is noted only at the 3-month and 6-month time points, however (Table 3). From time of surgery to the 6-month time point is when the most rapid weight loss occurred (22.9% TBWL), as compared to the following 6-month time period to the 1-year assessment (5.6% TBWL) (Table 3).

Table 3 Body mass index (BMI) and questionnaire scores at all time points (mean ± SD, p-value)
Full size table

When all components of the PFDI-20 questionnaire were analyzed separately, the urinary leakage component (UDI-6) has the strongest relationship with BMI, %TBWL, and percent excess body weight loss (%EBWL) (p < 0.0001, p < 0.0001, p < 0.0001, respectively). Urinary leakage scores improved over time, with a baseline score of 24.2 ± 21.3 to an improved 14.3 ± 14.6 at 1-year (p = 0.0241). Additionally, there is greatest improvement in symptoms noted at 6-months postoperatively (9.0 ± 15.1) (p = 0.0008). Interestingly, scores significantly improved at all postoperative time points (3-month p = 0.0008, 6-month p = 0.0008, 1-year p = 0.0241), but not at the preoperative evaluation (p = 0.6280).

Similarly, when measuring the impact of urinary incontinence (ICIQ-UI-SF), there is no improvement in symptom severity at the preoperative time point. Improvement is seen, however, at all postoperative time points (Table 3). Significant improvement, once again, is only noted at 3 and 6-month postoperatively (3-month p = 0.0339, 6-month p = 0.0198) with a non-significant increase in score severity at 1-year compared to other post-operative time points (p = 0.1318).

Quality of life related to UI symptoms (KHQ) significantly improved after surgery (3-month p = 0.0047, 6-month p = 0.0042, 1-year p = 0.0165), with general health perception having the strongest correlation with %TBWL (p = 0.025). Quality of life was not significantly changed after weight loss associated with diet and lifestyle modification alone (preoperative time point) (p = 0.9260) (Table 3). When all nine categories are individually analyzed, social and physical limitations affecting quality of life worsened throughout the bariatric pre-operative program (p = 0.0009, p = 0.0012, respectively). While no significant improvement is seen in these limitations by 1-year following surgery, there is a general improvement in the raw score of these subcategories over time (p = 0.0599, p = 0.3261 ,respectively) that may be underpowered to show significance. Additionally, perception of symptom improvement with surgery (PGI-I) improved over time when compared to pre-operative diet and lifestyle modification alone (3-month p = 0.0289, 6-month p = 0.0024, 1-year p = 0.0035). Overall, there is a significant difference in quality of life related to UI symptoms and perception of symptom following surgery, as compared with diet and lifestyle modification alone.

Discussion

UI is a common problem among women worldwide and can be challenging to treat due to the complexity of the condition [9]. Obesity is a known modifiable risk factor for UI [10]. Our study demonstrated that there is a significant positive impact on UI symptoms and perceived quality of life following bariatric surgery in patients with obesity.

Previous studies report a baseline prevalence of UI from 18 to 66%, with prevalence decreasing to 9.8–18.3% at 1-year post-bariatric surgery [11, 12]. Pelvic floor symptoms at 1-year postoperatively, when measured via PFDI-20 questionnaire, improved 31% when compared to baseline score of 43.8 ± 36.0 (12-month score 35.0 ± 30.0), with the UI specific UDI-6 improving 40.9% at 1-year. Additionally, score improvement is directly correlated to both BMI and %TBWL at all postoperative time points (p = 0.030, p = 0.019, respectively). When comparing PFDI-20 to preoperative BMI however, there is no significant improvement in UI symptoms (p = 0.112). This shows that diet and lifestyle modification alone prior to surgical intervention does not improve UI despite an average 6.6% TBWL in our patient population after completion of the preoperative bariatric program. Our findings stand in contrast to previous urogynecology data suggesting that 5–10% TBWL is sufficient to improve UI [7]. We also saw an absence of improvement with dietary weight loss alone for quality of life related to UI. While there is no significant improvement seen in the KHQ at the preoperative assessment, there is a strong improvement following bariatric surgery (Table 3). Our data suggests that the weight loss associated with bariatric surgery is an intervention that positively impacts UI and UI-associated quality of life.

Furthermore, previous urogynecology studies have shown significant improvement in stress UI following surgical intervention with a mid-urethral sling [13]. This was measured using UDI-6, and ultimately determined that a minimal clinically important difference (MCID) of 5.8 was needed to perceive improvement in symptoms [13]. MCID for UDI-6 was converted from established MCID for the 22 question long form Urinary Distress Index [14, 15]. At 1-year, UDI-6 scores following bariatric surgery improved by 9.9. No surgical intervention was performed on pelvic anatomy, yet results are clinically meaningful and superior to outcomes following specific pelvic procedures (mid-urethral sling). This may drive future clinical decisions for operative preference in women with symptoms of UI and concomitant obesity.

To further support our findings that bariatric surgery-associated weight loss improves UI and quality of life when compared to diet and lifestyle modification alone, perceived UI symptom improvement is significant (p < 0.05) at all postoperative time points when compared to the preoperative assessment (Table 3). While this supports our hypothesis that bariatric surgery can improve UI symptoms, additional lifestyle modifications required as part of the bariatric surgery program may also contribute to self-reported symptom perception. While limited, there is some evidence that suggests fluid restriction and elimination of carbonated beverages can improve UI [16]. Our bariatric surgery center strongly encourages elimination of carbonated beverages following bariatric surgery. Additionally, one of the most common complications following surgery is dehydration due to inability to consume adequate amounts of fluid due to surgical alteration of gastrointestinal anatomy. In our patient population, 4 out of 65 patients (6.15%) who progressed to surgery returned to the emergency department due to oral intake intolerance and dehydration, with no patients requiring readmission. This may contribute to UI symptom improvement in the early postoperative period; however, additional studies would need to be performed before attributing fluid restriction as the driver of post-bariatric UI symptom improvement.

Our study additionally shows that perceived quality of life (KHQ) is most significantly improved at 3 and 6 months when compared to baseline (Table 3). This may be attributable to the most rapid and highest degree of weight loss occurring in the first 6 months after bariatric surgery [17]. While the KHQ is a verified questionnaire for perceived quality of life based on urinary symptoms alone, there may be other attributable factors that improve perceived quality of life after rapid weight loss. While the KHQ questionnaire focuses on bladder function affecting social life, sex life, ability to complete household tasks, and physical activities, weight loss after bariatric surgery has also shown to improve quality of life in all of these categories. Previous data reports over 70% of patients report overall quality of life after surgery as good or very good [18].

While overall UI symptoms and quality of life improved following bariatric surgery, there is a small decline in perception of symptoms at 1-year following bariatric surgery despite additional weight loss at this time point. Previous data have shown a trend in UI relapse following bariatric surgery, however, this has been directly correlated with weight regain. With each 5 kg increase in weight, the risk of incontinence relapse was shown to be 29.4% [19]. Perhaps the decline in perception of UI symptoms is related to the taper in weight loss that occurs at 12–18 months with a focus on transition to weight-maintenance. Dissatisfaction with weight loss maintenance, compared to rapid weight loss in the early post-operative period, may contribute to overall perception of quality of life. Complex changes in psychological and social functioning occur after significant weight loss, and the psychosocial aspect associated with bariatric surgery, including loss of identity and change in self-perception, may additionally contribute to our small decline in perceived symptom improvement [20, 21]. As weight begins to stabilize around 1-year, patients may shift their focus from a tangible improvement in BMI to their psychosocial relationships that also contribute to quality of life. Previous literature has established that bariatric surgery is associated with increased incidence of divorce and separation [22]. As interpersonal relationships change, this may contribute to the change in symptom perception seen at the 1-year post-operative evaluation. Additionally, previous data have shown that bariatric surgery, specifically Roux-en-Y gastric bypass, is associated with improvement in UI among women independent of weight loss [23]. Further long-term data are needed to analyze our suggestion that interpersonal relationships and psychosocial changes may also affects symptoms and QOL related to UI.

Despite our limited sample size, our study incorporated four different validated questionnaires for the evaluation of UI. By elucidating specific perception and QOL-related data, which are subjective in nature, we are better able to understand the psychosocial relationship between weight loss and UI in our population beyond objective symptoms evaluation. This differs from larger multicenter trials that have evaluated UI following bariatric surgery [19].

Limitations

Our study is limited by number of participants. As this is an interim analysis, our longitudinal association cannot be completely analyzed as there are 13% of participants who have not yet undergone bariatric surgery, as well as participants who continue to enroll. Additionally, 26.9% of participants did not complete our required pre-operative bariatric program for various reasons, and thus did not progress to surgery. This limits out ability to effectively correlate data at each time point, limiting our longitudinal analysis. Any research study relying on patient return at designated time points suffers from loss to follow-up, and our study is no different. 60% of participants who have undergone surgery have either been lost to follow-up, had telemedicine follow-up, or chose not to answer the 1-year surveys. Combined with the patients who did not complete our program and have surgery, our longitudinal data were therefore limited to 28 of 108 (26%) of initial study enrollees.

We did not utilize urodynamic evaluation or pelvic exam findings of UI for measurement of UI due to its level of invasiveness. However, our administered questionnaires are validated and accurate tools for the assessment of UI when compared to objective testing [24]. Given the nature of subjective assessment, there is always inherent possibility of error due to each participant’s education level and understanding of the questionnaire. Further, there was no control group used in our study.

Additionally, we did not study the mechanism of relationship between weight loss and UI. We did not differentiate between the multiple types of UI and how the pathophysiology of each subtype may affect the relationship between UI and weight loss. To prove that weight loss alone can be an effective treatment for UI, a stronger understanding of the physiology surrounding this phenomenon is required. While it is well studied that decreased intraabdominal pressure from weight loss contribute to UI symptom improvement, hormonal changes frequently occur weight loss, and can also be a contributory factor in symptom improvement [25, 26]. Lifestyle modifications such as smoking cessation and incorporation of pelvic floor strengthening exercise post-bariatric surgery may also contribute to improvement in UI symptoms. These modifications were not followed in our study.

Conclusion

Although the relationship is complex and likely depends on many factors, weight loss after bariatric surgery is associated with improvement in UI symptoms. Additionally, there is a significant improvement in perception of UI symptoms and in quality of life related to UI. Bariatric surgery can play a role in the long-term treatment of UI in women with obesity to improve not only physical symptoms but also quality of life. Bariatric surgery may negate the need for further invasive incontinence therapy and further studies should investigate the specific factors related to post-bariatric improvement in UI.