Central obesity as measured by waist circumference is predictive of severity of lower urinary tract symptoms
Steven A. Kaplan, James Buchanan Brady Foundation, Department of Urology, Weill Medical College of Cornell University, 525 East 68th Street, F9 West, New York, NY 10065, USA. e-mail: email@example.com
Study Type – Prognosis (cohort)
Level of Evidence 2a
What's known on the subject? and What does the study add?
The metabolic syndrome, or Syndrome X, has traditionally been associated with an increased risk of cardiovascular disease and sexual dysfunction. Emerging data however now suggest that the metabolic syndrome may also have a heretofore unrecognized negative effect on voiding function as well. Weight loss through either behavioural modification or bariatric surgery has been shown to lead to improvement in stress and urge incontinence as well as LUTS. A potential relationship may be drawn between obesity and BPH.
This study adds the knowledge that WC can represent a simple metric not only for elements of the metabolic syndrome but also for worsened voiding. These obese men may be at high risk of male pelvic dysfunction.
- • To determine if central obesity as measured by waist circumference (WC) is a risk factor in metabolic dysfunction, which includes hypertension, dyslipidaemia and type 2 diabetes (DM2).
- • To test the hypothesis that central obesity and WC are associated with and predictive of the severity of voiding dysfunction.
- • Men aged ≥40 years with moderate or severe lower urinary tract symptoms (LUTS, International Prostate Symptom Score ≥ 8) with no previous treatment were included for study.
- • Subjects were divided into three groups according to WC (<90, 90–99 and ≥100 cm).
- • Baseline parameters including International Prostate Symptom Score, prostate volume, serum prostate-specific antigen, presence of erectile dysfunction and ejaculatory dysfunction, and the prevalence of hypertension, coronary artery disease and DM2 were compared among the three WC categories.
- • The association between WC and all parameters assessed was tested using multivariate logistic regression analysis.
- • In the 409 consecutive men analysed, WC was significantly and positively associated with prostate volume, serum prostate-specific antigen and International Prostate Symptom Score.
- • Higher WCs were also significantly associated with a greater prevalence of hypertension, coronary artery disease, DM2 and obesity as well as the presence of erectile dysfunction and ejaculatory dysfunction.
- • Increased WC is associated with worsened voiding.
- • There was a significantly increased prevalence of components of the metabolic syndrome in patients with higher WC.
- • Obese men, in particular those with other features of the metabolic syndrome, are at increased risk of male pelvic dysfunction and can be easily recognized by measurement of WC.
body mass index
type 2 diabetes
The global epidemic of obesity and diabetes has led to a striking increase in the number of people afflicted with the metabolic syndrome. The metabolic syndrome, or Syndrome X, consists of a constellation of abnormalities, including central obesity, glucose intolerance, dyslipidaemia and hypertension [1–4]. Together, these have been traditionally associated with an increased risk of cardiovascular disease and sexual dysfunction. Emerging data however now suggest that the metabolic syndrome may also have a heretofore unrecognized negative effect on voiding function [5,6]. Previous studies have shown that provocative weight loss in women, through either behavioural modification [7,8] or bariatric surgery , results in durable improvement in both stress and urge incontinence as well as LUTS. In men, corresponding studies have pointed to a potential relationship between obesity, waist circumference (WC) and body mass index (BMI), and BPH [5,6,10–15]. In this study, we sought to further examine the relationship between central obesity and problems with voiding function. Specifically, we sought to determine whether the use of a simple metric, WC, could be used as a proxy to predict for the prevalence of LUTS in men.
MATERIALS AND METHODS
After Institutional Review Board approval was obtained, data on all men aged 40 years or more who presented to the Institute for Bladder and Prostate Health at the Weill Medical College of Cornell University between January 2006 and June 2008 for evaluation of LUTS were retrospectively analysed. During the initial evaluation, the following parameters were assessed: (i) detailed medical history including the presence of three comorbid conditions: hypertension, coronary artery disease and type 2 diabetes (DM2), where hypertension was defined by a self-reported history or clinically detected blood pressure of 140/90 mmHg or more, coronary artery disease by a history of myocardial infarction, angina, congestive heart failure, coronary artery bypass, angioplasty stent, or blockage of vessels on cardiac catheterization, and DM2 by a self-reported history or biochemically detected DM2 as defined by the American Diabetes Association; (ii) physical examination with assessment of blood pressure, height, weight, BMI and WC; (iii) sociodemographic factors including age, marital status and cigarette smoking history; (iv) urological parameters including IPSS, self-reported presence of erectile and/or ejaculatory dysfunction, and TRUS measurement of prostate volume; and (v) blood laboratory studies including serum cholesterol and PSA. Ejaculatory dysfunction was defined as any type of ejaculatory dysfunction, including retrograde ejaculation, anejaculation, or reduced ejaculatory volume. All patients were queried regarding the presence or absence of erectile dysfunction and ejaculatory dysfunction.
Age was categorized by decade. Self-reported ethnicity was defined as white, black or Hispanic. BMI was categorized as <25, 25–29 and ≥30 kg/m2. Smoking status was stratified by never versus former versus current smokers. WC was measured by placing a tape measure snugly at the horizontal level between the subcostal margin and pelvic brim. Measurement subjects were then prospectively divided into tertiles according to WC (<90, 90–99 and ≥100 cm) and the incidence of various parameters was compared among the three categories of WC .
Statistical analyses were conducted separately for each category of WC. The prevalence rates of each of the 14 parameters were determined. The one-way anova and chi-squared tests were used to compare the prevalence of characteristics between different WC groups for continuous and dichotomous parameters, respectively. Within-group analysis was performed. The associations between WC as an independent variable and different urological function, sexual function and metabolic parameters and between the presence of chronic illnesses as dependent variables were then assessed by multiple logistic regression to estimate odds ratios (ORs) and 95% CI. The multiple logistic regression was adjusted using age and ethnicity as covariates. All variables were treated continuously or dichotomously in the regression analysis. Analyses were conducted using SAS version 9.1 (SAS Institute, Cary, NC, USA) and sudaan version 9.0.1 (Research Triangle Institute, Research Triangle Park, NC, USA).
A total of 409 men between the ages of 40 and 91 years were analysed. Overall, 153, 137 and 119 patients had WC in the <90, 90–99 and ≥100-cm tertiles, respectively. The overall age distribution was as follows: 40–49 years (n= 66, 16.1% of total study population); 50–59 years (n= 121, 29.6%); 60–69 years (n= 101, 24.7%); 70–79 years (n= 72, 17.6%); ≥80 years (n= 49, 12.0%) (Table 1). The median age of the overall study group was 57.6 years.
Table 1. Patient characteristics by waist circumference by analysis of variance and chi-squared tests
|Number|| ||153||137||119|| |
|Age (years, by decade)||45–49||11.7||8.7||6.2||0.21|
|Body mass index (kg/m2)||<25||37.6||29.2||25.6||0.01|
Age was similar between all WC groups (P= 0.09 to P= 0.31) except for patients aged 70–79 years, where a higher proportion of patients (30.1%) had larger WC (≥100 cm) (P= 0.03). There was a higher proportion of each ethnicity with WC 90–99 cm compared with the other WC groupings. The group with the highest BMI (≥30) had a greater proportion of patients with WC ≥ 100 cm (40.6%, P= 0.02). Conversely, the group with the lowest BMI (<25) had a greater proportion of patients with low WC (<90 cm) (37.6%, P= 0.01). Smoking history was similar between all WC groups (P= 0.17 to P= 0.31).
The prevalence of various urological and sexual function parameters was then recorded and then compared between the different WC groups (Table 2). Totals of 92 (22%), 78 (19%) and 72 (18%) patients showed evidence of hypertension, coronary artery disease and DM2, respectively. Fifty-three (13%) patients completed the IPSS, whereas all (100%) patients underwent TRUS.
Table 2. Urological and metabolic parameters by waist circumference by analysis of variance and chi-squared tests
|Prostate volume (g)|| || || || |
| Prostate-specific antigen (ng/dL)||1.87||2.94||3.96||0.001|
|International Prostate Symptom Score|| || || || |
| Urinary frequency (≥8 voids per 24 h)||16.4||26.7||38.9||0.02|
| Nocturia (≥2 voids per night)||14.7||28.9||44.3||0.001|
| Erectile dysfunction||32.1||49.8||74.5||0.001|
| Ejaculatory dysfunction||21.4||39.8||64.7||0.001|
| Mean serum cholesterol (mg/dL)||148||176||254||0.001|
| Coronary artery disease||8.4||17.3||29.4||0.004|
| Type 2 diabetes||11.3||15.9||32.6||0.001|
Overall prostate volume did not differ significantly among the three WC groups (P= 0.07), although subgroup analysis showed that the group with WC ≥ 100 cm had a higher proportion of men with prostate volume >50 g (P= 0.004), but this effect was not seen for the two smaller prostate volume groups (P= 0.41 and P= 0.21). Mean PSA concentration also differed between the different WC groups, increasing from 1.87 ng/dL to 3.96 ng/dL as WC increased (P= 0.001). The overall prevalence of lower urinary tract symptoms as measured by the IPSS did not differ significantly among the three WC groups (P= 0.10), although subgroup analysis showed that the group with WC ≥ 100 cm had a higher proportion of men with severe symptoms (IPSS 20–35, P= 0.003); this effect was not seen for the two lower IPSS groups (P= 0.07 and P= 0.36). Sexual dysfunction significantly differed with WC. The prevalence of erectile dysfunction increased from 32.1% to 74.5% as WC increased (P= 0.001). The prevalence of ejaculatory dysfunction also increased from 21.4% to 64.7% as WC increased (P= 0.002).
The prevalence of different elements of the metabolic syndrome also varied between the different WC groups. The prevalence of hypertension increased from 14.5% to 33.5% as WC increased (P= 0.02). Mean serum cholesterol increased from 148 mg/dL to 254 mg/dL (P= 0.001) over the same groups and the prevalence of coronary artery disease correspondingly increased from 8.4% to 29.4% (P= 0.004). The prevalence of DM2 increased from 11.3% to 32.6% as WC increased (P= 0.001).
The relationship of WC to each urological and sexual function parameter as well as to the presence of cardiovascular or metabolic comorbidities was then tested and quantified via multiple logistic regression (Table 3). Specifically, as WC increased from <90 cm to 90–99 cm to ≥100 cm, there was a greater likelihood for increased prostate volume (OR = 1.39, P= 0.01) and PSA (OR = 2.11, P= 0.001). There was also a greater likelihood of having a higher IPSS as WC increased (OR = 1.68, P= 0.002). Patients with larger WC were also more likely to have erectile dysfunction and ejaculatory dysfunction with OR = 2.32 and OR = 3.02, respectively (P= 0.01 and P= 0.03). Increasing WC was also associated with a greater likelihood of having hypertension, DM2 and coronary artery disease (OR = 2.31, OR = 2.88 and OR = 3.50 with P= 0.001, P= 0.002 and P= 0.001, respectively). Increasing WC also led to a greater likelihood for obesity, defined as BMI ≥ 30 (OR = 2.45, P= 0.001). The likelihood of having each of these disease elements progressively increased as WC increased from <90 cm to 90–99 cm to ≥100 cm.
Table 3. Multivariate logistic regression analysis of urological and metabolic parameters with waist circumference
|Prostate volume||1.00||1.29 (1.23–1.87)||1.39 (1.17–2.32)||0.01|
|Prostate-specific antigen||1.00||1.57 (1.35–2.28)||2.11 (1.59–2.89)||0.001|
|International Prostate Symptom Score||1.00||1.20 (1.12–1.59)||1.68 (1.37–2.45)||0.002|
|Erectile dysfunction||1.00||1.55 (1.25–2.54)||2.32 (1.26–3.93)||0.01|
|Ejaculatory dysfunction||1.00||1.86 (1.50–2.65)||3.02 (1.31–5.78)||0.03|
|Hypertension||1.00||1.50 (1.25–2.16)||2.31 (1.53–3.87)||0.001|
|Type 2 diabetes||1.00||1.41 (1.23–2.31)||2.88 (1.68–4.32)||0.002|
|Coronary artery disease||1.00||2.06 (1.48–3.56)||3.50 (1.32–5.12)||0.001|
|Body mass index||1.00||1.72 (1.48–3.21)||2.45 (1.09–4.12)||0.001|
The metabolic syndrome has become one of the major public health challenges worldwide. Interest in the syndrome initially arose because of its association with cardiovascular disease and DM2 in the setting of a growing worldwide epidemic of obesity. However, data are emerging that point to a potential relationship between the metabolic syndrome and urological disease [5,6,10–14]. Although the relationship between the metabolic syndrome and erectile dysfunction is well established, the former may also have heretofore unknown impact on voiding dysfunction. Patients may suffer not only from traditional cardiovascular disease and erectile and ejaculatory dysfunction but also LUTS from a common underlying mechanism.
Previous definitions for the metabolic syndrome have been conflicting in terms of inclusion of WC as a parameter, which makes the study of anthropometric measures to predict for the presence of the metabolic syndrome difficult [1–4]. It has consequently also been difficult to link anthropometric factors to potential pathological consequences of the metabolic syndrome. For example, several studies have shown a link between factors such as WC or BMI and the risk of developing BPH [5,6,10–14] but others have not [17–19]. Similarly, many studies have shown associations with hypertension, DM2 and dyslipidaemia [5,6] but this has not been borne out in others .
This study attempts to further examine the link between WC and LUTS. The WC appears to be linked to characteristics associated with BPH, i.e. larger prostate volume and higher PSA level. Patients with larger WC also tended to have more severe symptoms, with a higher IPSS, as would be expected in patients with larger prostates. This effect on prostate volume and severity of symptoms however was most pronounced and specific for men with larger prostates (>50 g) and severe symptoms (IPSS 20–35). As expected, WC correlated well with the presence of hypertension, DM2, dyslipidaemia and obesity defined as BMI ≥ 30, all elements related to the metabolic syndrome; WC also correlated with the established erectile and ejaculatory dysfunction seen in patients suffering from the metabolic syndrome. Overall, WC may therefore represent a simple and easy metric in diagnosing not only the sequelae normally associated with the metabolic syndrome (i.e. hypertension, DM2, dyslipidaemia and erectile and ejaculatory dysfunction) but also occult voiding dysfunction.
Other studies have also bolstered evidence that WC might represent a reasonable metric for denoting voiding dysfunction. Cinar et al., for example reported the results of a cluster analysis of the Boston Area Community Health Survey where 1899 men were studied and classified into five clusters depending on the severity and pattern of urinary symptoms. More symptomatic men tended to have larger WC as well as higher rates of hypertension, DM and cardiovascular symptoms. Similarly, Kim et al. reported that age, serum PSA and WC were factors that predicted prostate volume in a prospective study of 465 men with moderate to severe LUTS (IPSS ≥ 8).
Different potential mechanisms could underlie the relationship between male pelvic disease and central obesity or the metabolic syndrome. Elevated insulin levels could provide a mechanism for male pelvic disease by stimulating or accelerating the growth of BPH. For example, Hammarsten and Hogstedt  conducted a risk factor analysis in 250 Swedish patients, showing a relationship between a faster rate of BPH progression and the presence of DM2, BMI and WC. Other evidence points to inflammation as a mediator between voiding dysfunction and the metabolic syndrome. Rohrmann et al., for example, reported on an analysis of 2337 patients aged 60 years or older from the Third National Health and Examination Survey (i.e. NHANES III), where the presence of LUTS was positively associated with serum levels of C-reactive protein, a non-specific marker of inflammation, even after adjustment for the presence of the metabolic syndrome. Another potential aetiology could lie in pelvic atherosclerotic disease caused by the presence of metabolic syndrome; the resulting ischaemia could lead to primary bladder dysfunction. Azadzoi et al. for example reported increased levels of TGF-β1 expression leading to fibrosis, smooth muscle atrophy and lowered compliance in the bladders of rabbits subjected to atherosclerotic injury and high cholesterol diets. Ischaemia also has been shown to produce similar histopathological changes in the prostate . These histological changes have been clinically supported by Ponholzer et al. who reported that the presence of vascular risk factors was associated with the development of LUTS in 2536 men and women identified during a Viennese health screening project.
Waist circumference may therefore be both useful and important as a measure of success for possible future treatment options for patients with voiding dysfunction. Active reduction of WC via treatment of the metabolic syndrome or treatment of pelvic inflammation and ischaemia may lead to durable and material improvement in voiding health. Subak et al. for example studied 338 overweight and obese women randomized to an intensive weight loss programme or education, and showed that a 6-month weight loss programme reduced the frequency of stress incontinence. A higher proportion of the treatment group experienced clinically relevant reductions of ≥70% in the frequency of all, stress and urge incontinence episodes. Vella et al. reported on 126 morbidly obese women treated with bariatric surgery, showing significant improvements in stress incontinence, frequency and overall quality of life. Similar results were also reported in men. Parsons and Kashefi  performed a meta-analysis of 11 studies with 43 083 patients showing that light, moderate and heavy physical activity all reduced the risk of developing BPH and LUTS. Khoo et al. showed improvements in LUTS in 68 obese men after at least 10% weight loss from an 8-week weight loss programme. Weight loss in this study also significantly improved insulin sensitivity, erectile function and testosterone levels in patients. A rationale for improved treatment of voiding disorders could therefore be potentially established by addressing aspects of the metabolic syndrome. For example, weight loss through the use of diet modification  and exercise , modification of hypertension, insulin sensitivity and dysplipidaemia through pharmacotherapeutics or bariatric surgery  have been shown to improve diseases associated with the metabolic syndrome and could therefore potentially modify the natural history of voiding dysfunction. Indeed, successful treatment of central obesity holds the potential of being useful in treating urological disease, and WC may represent a simple and easy-to-use metric to quantify the success of treatment.
The study suffers from several limitations. First, it constitutes a single-institution, retrospective experience. It represents a high volume, tertiary-care practice whose patient composition may not correlate with that in the general community. Additionally, it is unclear whether WC or BMI or waist-to-hip ratio constitutes the best anthropometric parameter for correlating voiding dysfunction with central obesity. Finally, it is clear that prospective, randomized trials possessing appropriate durations of follow-up are required to validate the concept that treatment of central obesity can in fact lead to improvements in voiding function.
Increased WC is associated with worsened voiding function. Patients with higher WC are therefore at increased risk for male pelvic dysfunction. Waist circumference may represent an easy diagnostic tool to elucidate the presence of occult voiding dysfunction.
CONFLICT OF INTEREST