Effects of bariatric surgery on untreated lower urinary tract symptoms: a prospective multicentre cohort study Serge Luke, Ben Addison*, Katherine Broughton†, Jonathan Masters*, Richard Stubbs‡ and Andrew Kennedy-Smith§ Department of Urology, Dunedin Hospital, Dunedin, *Department of Urology, Auckland City Hospital, †Department of General Surgery, Middlemore Hospital, Auckland, ‡Wakefield Hospital, and §Department of Urology, Wellington Hospital, Wellington, New Zealand
Objective
Results
To evaluate the effects of bariatric surgery on lower urinary tract symptoms (LUTS) in a prospective cohort study.
In all, 86 patients were recruited and 82% completed at least one follow-up after surgery. There was significant weight loss and reduction of BMI after surgery (P < 0.001). At 6 weeks, there was a significant reduction in overall symptom score (P < 0.001) and this improvement was sustained at 1 year. Linear regression analysis showed that total symptoms score at baseline, HOMA-IR, preoperative insulin level and change in insulin level postoperatively were predictive of the change in total symptoms score while the amount of weight loss was not.
Patients and Methods Patients undergoing bariatric surgery were recruited into the study. LUTS were assessed using the International Prostate Symptoms Score (IPSS) in men and Bristol Female Lower Urinary Tract Symptoms Score Questionnaire (BFLUTS) in women. Serum glucose, insulin and prostate-specific antigen (PSA) levels were recorded; insulin resistance was quantified using the Homeostasis Model Assessment (HOMA-IR) method. Patients were assessed before surgery, and at 6–8 weeks and 1 year after surgery. Weight loss, change in body mass index (BMI), total symptoms score as well as individual symptoms were tested for statistical significance with correction for multiple testing using Bonferroni method. Linear regression analysis was performed with total symptoms score change at 1 year as the outcome variable and BMI, age, total symptoms score before surgery, HOMA-IR, glucose level before surgery, insulin level before surgery, change in insulin level after surgery, weight loss and BMI loss as predictor variables.
Introduction Obesity complicates the management of many conditions and is a major cause of morbidity, with effects on multiple organ systems. It is increasingly prevalent across all age groups, particularly in the developed world and therefore presents a significant public health challenge. In New Zealand, the 2008–2009 adult nutrition survey indicated 69% of men and 61% of women were overweight (BMI 30–35 kg/m2) or obese (BMI >35 kg/m2) [1]. In the USA, obesity affects >78 million people [2]. LUTS of frequency, nocturia, poor flow and urinary incontinence are reported to be associated with obesity in both men and women [3–6]. Prevalence of these © 2014 The Authors BJU International © 2014 BJU International | doi:10.1111/bju.12943 Published by John Wiley & Sons Ltd. www.bjui.org
Conclusions The study confirms the improvement in LUTS after weight loss but there is no correlation between the improvement and the time course or degree of weight loss. Rather there is a suggestion that the improvement in symptoms is linked to improvement in insulin resistance seen as a result of both bariatric surgery and weight loss.
Keywords LUTS, bariatric surgery, metabolic syndrome
symptoms has been associated with metabolic syndrome, which is a condition linked to insulin resistance and obesity. However, details of the association between obesity and insulin resistance are poorly understood [7]. Weight loss whether after bariatric surgery or conservative measures has been shown to improve urinary incontinence in previously obese women [8]. However, there is little data on the mechanisms linking obesity and LUTS and how LUTS improve after weight loss. Bariatric surgery is the most reliable way of achieving major weight loss in severely obese individuals and has been shown to have a profound and weight loss-independent effect on
BJU Int 2014 wileyonlinelibrary.com
Luke et al.
insulin resistance [9,10]. It can therefore provide us with a means of exploring one possible mechanism of improvement in LUTS after weight loss. We hypothesised that morbidly obese patients presenting for bariatric surgery have LUTS that change after bariatric surgery. We aimed to explore the relationship between any change in LUTS and the weight loss that follows bariatric surgery, recognising weight loss continues for ≈12 months after such surgery. Further we hypothesised that any improvement in LUTS seen after bariatric surgery may well relate to improvements in insulin resistance, rather than specifically to weight loss.
Patients and Methods This prospective multicentre cohort study was approved by the New Zealand Central Regional Ethics Committee. Men and women undergoing bariatric surgery between November 2009 and May 2011 in three New Zealand hospitals were recruited at their last preoperative consultation by the bariatric surgeon or clinical nurse specialist. The invitation to participate was generally non-selective but influenced by factors such as: involvement in other studies at that centre, likely reliable attendance at follow-up clinics, and perceived willingness to participate. All patients received verbal and written information about the study before written informed consent was obtained. The bariatric procedures performed included open gastric bypass, laparoscopic gastric bypass and laparoscopic sleeve gastrectomy, which were the standard procedures being undertaken at the participating hospitals. Patients were excluded from the study if they (a) had previous urological surgery, (b) had previously diagnosed urological disease other than BPH and overactive bladder, (c) were taking α-blockers or 5α-reductase inhibitors, (d) had undergone any urological intervention (e.g. cystoscopy, prostate biopsy) in the month before recruitment, (e) had clinically significant renal impairment and (f) had UTI detected by dipstick urine test. Symptoms were assessed using validated questionnaires at three time points: (a) preoperatively, (b) 6–8 weeks postoperatively before major weight loss had occurred and (c) 1 year after surgery by which time most weight loss had been achieved. Men completed the IPSS and women the BFLUTS at each time point. These questionnaires have previously been shown to have strong validity and reliability and are considered excellent instruments to measure changes after interventions [11,12]. LUTS are considered to be mild in patients if the IPSS is in the range of 1 to 7, moderate in the range of 8 to 19 and severe in the range of 20 to 35. The BFLUTS questionnaire assesses the domain of incontinence, as well as voiding and filling but there is no universally accepted classification of severity of the symptoms. Patients who failed to attend follow-up appointments were sent questionnaires by mail and were encouraged to respond using prepaid return envelopes. © 2014 The Authors 2 BJU International © 2014 BJU International
Non-responders were contacted by telephone and questionnaires were completed over the phone. Wherever possible, fasting blood tests were also obtained. Laboratory investigations undertaken at each time point included fasting serum insulin and fasting plasma glucose, and PSA levels were measured in men before surgery. Insulin resistance was quantified using Homeostasis Model Assessment method (HOMA-IR) described by Matthews et al. [13]. This indirect method uses fasting insulin and glucose levels to predict fasting steady-state glucose and insulin concentrations for a wide range of possible combinations of insulin resistance and β-cell function and is widely used in research [14]. Statistical analysis was performed to assess the differences in the mean score for each variable in the questionnaire, as well as the total symptoms score at each time point, using a paired two-tail Student’s t-test with Bonferroni correction for multiple testing. We also compared the change in total symptoms score at 6 weeks and 1 year between patients who had gastric bypass and those who had laparoscopic sleeve gastrectomy. Levels of insulin, glucose and HOMA-IR were not normally distributed and therefore changes were assessed using a Wilcoxon signed-rank test. Simple linear regression was performed with change in total symptoms score as the dependent variable. The following explanatory variables were evaluated: preoperative BMI, age, fasting glucose, fasting insulin, HOMA-IR, change in fasting insulin at 6 weeks, preoperative total symptoms score, weight loss at 1 year and BMI loss at 1 year. We also performed multiple linear regression analysis with change in total symptoms score as the dependent variable. For this model preoperative BMI, age, HOMA-IR, total preoperative symptoms score and weight loss at 1 year were used as predictor variables. A P < 0.05 was considered to indicate statistical significance.
Results In all, 86 patients were recruited into the study including 29 men and 57 women. Laparoscopic sleeve gastrectomy was undertaken in 62 patients, laparoscopic gastric bypass in four, and open gastric bypass in 20. Outcome data were pooled from all procedure types for analysis. Follow-up assessment was accomplished at 6 weeks in 67 patients (78%) and in 60 (71%) patients at 1 year after surgery. In all, 70 patients (82%) underwent surgery and completed at least one follow-up questionnaire. The other 16 patients (six men, 10 women) were withdrawn from the study. Their preoperative data was not included in the analysis. In all, 11 of these were not contactable, had moved out of the study area or declined to continue participation. Three patients did not proceed with surgery and two died within 1 year of surgery. Baseline patient characteristics of the 70 patients included for analysis are shown in Table 1. Patients excluded from the study had no
Effects of bariatric surgery on untreated LUTS
Table 1 The patient's baseline characteristics.
Fig. 2 Symptoms score change (BFLUTS) in women. P-values for comparison of baseline and at 6 weeks and comparison of baseline and
Characteristic
Men
Number of patients Ethnicity, % European Maori Pacific Islanders Other Mean (SD): Age, years Weight, kg BMI, kg/m2 Glucose level, mmol/L Insulin level, mU/L Total symptoms score PSA level, ng/dL
Women
23
47
65 17 13 5
72 15 1 12
49.06 (9.68) 146.66 (27.35) 46.04 (6.63) 7.35 (4.50) 32.52 (50.07) 6.65 (5.21) 0.71 (0.51)
44.29 (9.29) 117.86 (16.60) 43.98 (6.26) 6.52 (2.65) 20.80 (22.16) 9.13 (5.19)
at 1 year are shown in brackets. Significant P-values have been adjusted for multiple comparisons using Bonferroni correction.
Total Score (
Objective
Results
To evaluate the effects of bariatric surgery on lower urinary tract symptoms (LUTS) in a prospective cohort study.
In all, 86 patients were recruited and 82% completed at least one follow-up after surgery. There was significant weight loss and reduction of BMI after surgery (P < 0.001). At 6 weeks, there was a significant reduction in overall symptom score (P < 0.001) and this improvement was sustained at 1 year. Linear regression analysis showed that total symptoms score at baseline, HOMA-IR, preoperative insulin level and change in insulin level postoperatively were predictive of the change in total symptoms score while the amount of weight loss was not.
Patients and Methods Patients undergoing bariatric surgery were recruited into the study. LUTS were assessed using the International Prostate Symptoms Score (IPSS) in men and Bristol Female Lower Urinary Tract Symptoms Score Questionnaire (BFLUTS) in women. Serum glucose, insulin and prostate-specific antigen (PSA) levels were recorded; insulin resistance was quantified using the Homeostasis Model Assessment (HOMA-IR) method. Patients were assessed before surgery, and at 6–8 weeks and 1 year after surgery. Weight loss, change in body mass index (BMI), total symptoms score as well as individual symptoms were tested for statistical significance with correction for multiple testing using Bonferroni method. Linear regression analysis was performed with total symptoms score change at 1 year as the outcome variable and BMI, age, total symptoms score before surgery, HOMA-IR, glucose level before surgery, insulin level before surgery, change in insulin level after surgery, weight loss and BMI loss as predictor variables.
Introduction Obesity complicates the management of many conditions and is a major cause of morbidity, with effects on multiple organ systems. It is increasingly prevalent across all age groups, particularly in the developed world and therefore presents a significant public health challenge. In New Zealand, the 2008–2009 adult nutrition survey indicated 69% of men and 61% of women were overweight (BMI 30–35 kg/m2) or obese (BMI >35 kg/m2) [1]. In the USA, obesity affects >78 million people [2]. LUTS of frequency, nocturia, poor flow and urinary incontinence are reported to be associated with obesity in both men and women [3–6]. Prevalence of these © 2014 The Authors BJU International © 2014 BJU International | doi:10.1111/bju.12943 Published by John Wiley & Sons Ltd. www.bjui.org
Conclusions The study confirms the improvement in LUTS after weight loss but there is no correlation between the improvement and the time course or degree of weight loss. Rather there is a suggestion that the improvement in symptoms is linked to improvement in insulin resistance seen as a result of both bariatric surgery and weight loss.
Keywords LUTS, bariatric surgery, metabolic syndrome
symptoms has been associated with metabolic syndrome, which is a condition linked to insulin resistance and obesity. However, details of the association between obesity and insulin resistance are poorly understood [7]. Weight loss whether after bariatric surgery or conservative measures has been shown to improve urinary incontinence in previously obese women [8]. However, there is little data on the mechanisms linking obesity and LUTS and how LUTS improve after weight loss. Bariatric surgery is the most reliable way of achieving major weight loss in severely obese individuals and has been shown to have a profound and weight loss-independent effect on
BJU Int 2014 wileyonlinelibrary.com
Luke et al.
insulin resistance [9,10]. It can therefore provide us with a means of exploring one possible mechanism of improvement in LUTS after weight loss. We hypothesised that morbidly obese patients presenting for bariatric surgery have LUTS that change after bariatric surgery. We aimed to explore the relationship between any change in LUTS and the weight loss that follows bariatric surgery, recognising weight loss continues for ≈12 months after such surgery. Further we hypothesised that any improvement in LUTS seen after bariatric surgery may well relate to improvements in insulin resistance, rather than specifically to weight loss.
Patients and Methods This prospective multicentre cohort study was approved by the New Zealand Central Regional Ethics Committee. Men and women undergoing bariatric surgery between November 2009 and May 2011 in three New Zealand hospitals were recruited at their last preoperative consultation by the bariatric surgeon or clinical nurse specialist. The invitation to participate was generally non-selective but influenced by factors such as: involvement in other studies at that centre, likely reliable attendance at follow-up clinics, and perceived willingness to participate. All patients received verbal and written information about the study before written informed consent was obtained. The bariatric procedures performed included open gastric bypass, laparoscopic gastric bypass and laparoscopic sleeve gastrectomy, which were the standard procedures being undertaken at the participating hospitals. Patients were excluded from the study if they (a) had previous urological surgery, (b) had previously diagnosed urological disease other than BPH and overactive bladder, (c) were taking α-blockers or 5α-reductase inhibitors, (d) had undergone any urological intervention (e.g. cystoscopy, prostate biopsy) in the month before recruitment, (e) had clinically significant renal impairment and (f) had UTI detected by dipstick urine test. Symptoms were assessed using validated questionnaires at three time points: (a) preoperatively, (b) 6–8 weeks postoperatively before major weight loss had occurred and (c) 1 year after surgery by which time most weight loss had been achieved. Men completed the IPSS and women the BFLUTS at each time point. These questionnaires have previously been shown to have strong validity and reliability and are considered excellent instruments to measure changes after interventions [11,12]. LUTS are considered to be mild in patients if the IPSS is in the range of 1 to 7, moderate in the range of 8 to 19 and severe in the range of 20 to 35. The BFLUTS questionnaire assesses the domain of incontinence, as well as voiding and filling but there is no universally accepted classification of severity of the symptoms. Patients who failed to attend follow-up appointments were sent questionnaires by mail and were encouraged to respond using prepaid return envelopes. © 2014 The Authors 2 BJU International © 2014 BJU International
Non-responders were contacted by telephone and questionnaires were completed over the phone. Wherever possible, fasting blood tests were also obtained. Laboratory investigations undertaken at each time point included fasting serum insulin and fasting plasma glucose, and PSA levels were measured in men before surgery. Insulin resistance was quantified using Homeostasis Model Assessment method (HOMA-IR) described by Matthews et al. [13]. This indirect method uses fasting insulin and glucose levels to predict fasting steady-state glucose and insulin concentrations for a wide range of possible combinations of insulin resistance and β-cell function and is widely used in research [14]. Statistical analysis was performed to assess the differences in the mean score for each variable in the questionnaire, as well as the total symptoms score at each time point, using a paired two-tail Student’s t-test with Bonferroni correction for multiple testing. We also compared the change in total symptoms score at 6 weeks and 1 year between patients who had gastric bypass and those who had laparoscopic sleeve gastrectomy. Levels of insulin, glucose and HOMA-IR were not normally distributed and therefore changes were assessed using a Wilcoxon signed-rank test. Simple linear regression was performed with change in total symptoms score as the dependent variable. The following explanatory variables were evaluated: preoperative BMI, age, fasting glucose, fasting insulin, HOMA-IR, change in fasting insulin at 6 weeks, preoperative total symptoms score, weight loss at 1 year and BMI loss at 1 year. We also performed multiple linear regression analysis with change in total symptoms score as the dependent variable. For this model preoperative BMI, age, HOMA-IR, total preoperative symptoms score and weight loss at 1 year were used as predictor variables. A P < 0.05 was considered to indicate statistical significance.
Results In all, 86 patients were recruited into the study including 29 men and 57 women. Laparoscopic sleeve gastrectomy was undertaken in 62 patients, laparoscopic gastric bypass in four, and open gastric bypass in 20. Outcome data were pooled from all procedure types for analysis. Follow-up assessment was accomplished at 6 weeks in 67 patients (78%) and in 60 (71%) patients at 1 year after surgery. In all, 70 patients (82%) underwent surgery and completed at least one follow-up questionnaire. The other 16 patients (six men, 10 women) were withdrawn from the study. Their preoperative data was not included in the analysis. In all, 11 of these were not contactable, had moved out of the study area or declined to continue participation. Three patients did not proceed with surgery and two died within 1 year of surgery. Baseline patient characteristics of the 70 patients included for analysis are shown in Table 1. Patients excluded from the study had no
Effects of bariatric surgery on untreated LUTS
Table 1 The patient's baseline characteristics.
Fig. 2 Symptoms score change (BFLUTS) in women. P-values for comparison of baseline and at 6 weeks and comparison of baseline and
Characteristic
Men
Number of patients Ethnicity, % European Maori Pacific Islanders Other Mean (SD): Age, years Weight, kg BMI, kg/m2 Glucose level, mmol/L Insulin level, mU/L Total symptoms score PSA level, ng/dL
Women
23
47
65 17 13 5
72 15 1 12
49.06 (9.68) 146.66 (27.35) 46.04 (6.63) 7.35 (4.50) 32.52 (50.07) 6.65 (5.21) 0.71 (0.51)
44.29 (9.29) 117.86 (16.60) 43.98 (6.26) 6.52 (2.65) 20.80 (22.16) 9.13 (5.19)
at 1 year are shown in brackets. Significant P-values have been adjusted for multiple comparisons using Bonferroni correction.
Total Score (
