- Email: [email protected]
Body Mass Index and its Association with Genitourinary Disorders in Men Undergoing Prostate Cancer Screening
2141
Body Mass Index and its Association with Genitourinary Disorders in Men Undergoing Prostate Cancer Screening Naeem Bhojani, MD,*† Paul Perrotte, MD,* Georg Hutterer, MD,‡ Nazareno Suardi, MD,*§ Claudio Jeldres, MD,*† Shahrokh F. Shariat, MD,¶ Umberto Capitanio, MD,* Philippe Arjane, MD,† Hugues Widmer, MD,† Francois Benard, MD,† Francois Peloquin, MD,† Francesco Montorsi, MD,§ and Pierre Karakiewicz, MD* *Cancer Prognostics and Health Outcomes unit, University of Montreal Health Center, Montreal, Canada; †University of Montreal—Urology, Montreal, Canada; ‡Graz Medical University—Urology, Graz, Austria; §Vita-Salute University San Raffaele—Urology, Milan, Italy; ¶University of Texas Southwestern—Urology, Dallas, TX, USA DOI: 10.1111/j.1743-6109.2008.00811.x
ABSTRACT
Introduction. Elevated body mass index (BMI) may predispose to several pelvic pathologies. Aims. We tested the association between BMI and five end points, namely, (i) erectile dysfunction (ED); (ii) lower urinary tract symptoms (LUTS); (iii) chronic prostatitis-associated pain (CPP); and ejaculatory dysfunction that is subdivided between (iv) pain/discomfort on ejaculation; and (v) subjectively decreased ejaculate volume. Methods. Age, height, and weight were prospectively recorded in a cohort of 590 consecutive healthy men undergoing prostate cancer screening. Continuously coded and categorized BMI (World Health Organization classification) were studied. Main Outcome Measures. Age-adjusted analyses relied on logistic and linear regression models, according to data type. Results. The average age was 54.1 years (range 30–83). Of all, 296 were overweight (50.2%, BMI 25–29.9 kg/m2) and 85 were obese (14.4%, BMI ⱖ 30 kg/m2). After age adjustment, elevated continuously coded BMI (P < 0.001) and elevated categorized BMI (P = 0.01) were associated with worse erectile function. Conversely, after age adjustment, elevated continuously coded BMI (P = 0.02) and elevated categorized BMI (P = 0.05) were associated with a lower rate of subjectively decreased ejaculate volume. Finally, after age adjustment, elevated categorically coded BMI was related to lower rates of CPP (P < 0.001) and to a lower rate of pain/discomfort on ejaculation (P = 0.03). Conclusions. In men undergoing prostate cancer screening, the effect of BMI on the five end points is not invariably detrimental. Elevated BMI may predispose to ED, but may also decrease the rate of pain/discomfort on ejaculation and may lower the reported rate of subjectively decreased ejaculate volume. Finally, it appeared to have no effect on LUTS. Bhojani N, Perrotte P, Hutterer G, Suardi N, Jeldres C, Shariat SF, Capitanio U, Arjane P, Widmer H, Benard F, Peloquin F, Montorsi F, and Karakiewicz P. Body mass index and its association with genitourinary disorders in men undergoing prostate cancer screening. J Sex Med 2008;5:2141–2151. Key Words. Body Mass Index; Erectile Dysfunction; Lower Urinary Tract Symptoms; Ejaculatory Pain; Chronic Prostatitis
Introduction
O
besity is common in Western countries [1]. According to the World Health Organization (WHO) criteria, more than 30% of adults in the united States are obese. Moreover, over 70% of Americans over 40 years of age are overweight (body mass index [BMI] ⱖ 25 kg/m2) [2]. Simi-
© 2008 International Society for Sexual Medicine
larly, obesity is a growing problem in Western European countries [3,4]. Central obesity represents one of several components of the metabolic syndrome [5]. The metabolic syndrome may predispose to several pelvic pathologies, such as erectile dysfunction (ED) and others [6]. Several cross-sectional studies have suggested an association between ED and obesity [7–14]. Similarly, a J Sex Med 2008;5:2141–2151
2142 link was reported between the metabolic syndrome and lower urinary tract symptoms (LUTS) [15–17]. Moreover, several investigators reported on the association of urological symptoms, ED, and the metabolic syndrome [18–21]. The link between elevated BMI and other male disorders might be equally strong. However, the presence and the strength of this link have not yet been quantified. Therefore, the purpose of the current study was to test for the presence and the strength of the association between established and the novel urological manifestations of the metabolic syndrome. In order to avoid the confounding effect of categorizing patients into diseased and nondiseased groups, we performed two types of analyses. One rested on categorized and the other on continuously coded data. Aims
We decided to assess the relationship between BMI and five separate end points, namely, (i) ED; (ii) LUTS; (iii) chronic prostatitis-associated pain (CPP); and ejaculatory dysfunction (EjD) that is subdivided between (iv) pain/discomfort on ejaculation; and (v) subjectively decreased (selfreported) ejaculate volume. Methods
The study cohort consisted of 590 consecutive men without evidence of prostate cancer, who participated in an annual prostate cancer screening event, the Prostate Cancer Awareness Days. Age, height, and weight were prospectively recorded in all participants. BMI was calculated and recorded as either a continuous variable or as defined according to the WHO classification (weight [kg]/height squared [m2]), as either nonobese (BMI < 24.9 kg/m2), overweight (BMI 25–29.9 kg/m2), or obese (BMI ⱖ 30 kg/m2). In all analyses, only three BMI categories were used, since very few (11 [1.9%]) severely obese men participated in the event. The study received Institutional Review Board (IRB) approval from University of Montreal Health Center I.R.B. Main Outcome Measures
The effect of either continuously coded or categorized BMI was tested in all analyses. ED represented the first end point. Its prevalence was assessed with the self-administered International J Sex Med 2008;5:2141–2151
Bhojani et al. Index of Erectile Function (IIEF)-erectile function (EF) scale. The relationship between IIEF was examined in linear regression models, where BMI and the IIEF scale score represented continuously coded variables (Table 1). Since IIEF-EF is also frequently categorized, we reexamined the effect of BMI on the IIEF-EF scale score of <16, which defines moderate to severe ED according to Cappelleri et al. [22]. Logistic regression models were used and the effect of categorized BMI was examined (Table 2). LUTS represented the second end point and its prevalence was assessed with the self-administered international prostate symptom score (IPSS). The relationship between the IPSS score was examined in linear regression models, where BMI and the IPSS scale score represented continuously coded variables (Table 1). Since the IPSS score can also be categorized according to LUTS severity, we reexamined the effect of BMI on the IPSS score ⱖ19, which defines severe LUTS [23]. Logistic regression models were used and the effect of categorized BMI was examined (Table 2). The pain scale score of the Chronic Prostatitis Symptom Index (CPSI) quantified CPP and represented the third end point [24]. The relationship between CPP score was examined in linear regression models, where BMI and the CPP scale score represented continuously coded variables (Table 1). We then categorized the CPP scores according to its median and reexamined the effect of BMI on the CPP score above median. Logistic regression models were used and the effect of categorized BMI was examined (Table 2). Subjectively decreased ejaculate volume and presence of pain/discomfort on ejaculation represented, respectively, the fourth and the fifth end points. The prevalence of these conditions that contribute to EjD was measured with items 8 and 10 of the Danish Prostate Symptom Score sexual function questionnaire [25]. Because of their inherent binary format (yes/no), these two end points could only be analyzed in logistic regression models. Two models were fitted. One relied on continuously coded BMI (Table 1). The second relied on categorized BMI (Table 2). Statistical analyses consisted of t-tests and of univariable and age-adjusted logistic and linear regression models. BMI represented the main predictor in all models. Continuously coded variables can be either kept in their original continuously coded format or can be categorized. Stratified data can only be examined in categorical format. When two continuously coded variables are examined,
Continuously coded IPSS score 0.02, 0.8†
Continuously coded IIEF-EF scale score 0.45, <0.001† 0.91, 0.09‡
0.93, 0.02‡
Presence of decreased ejaculate volume
IIEF-EF scale score Moderate to severe ED (<16) 1.88, 0.01† 1.47, 0.3†
IIEF (N = 574)
Presence of of pain/ discomfort on ejaculation 0.36, 0.03† 0.90, 0.9†
0.42, <0.001† 0.55, 0.07†
0.60, 0.2† 1.25, 0.6†
EjD (N = 590) CPSI pain scale score (>median)
CPSI (N = 500)
IPSS score (ⱖ19)
IPSS (N = 575)
0.62, 0.05† 0.54, 0.1†
Presence of decreased ejaculate volume
The relationship between categorically coded BMI and the IIEF-EF scale score, the IPSS score, the CPSI pain scale score and EjD was tested in age-adjusted logistic regression models. *Age-adjusted body mass index. †Odds ratio, P value. BMI = body mass index; LUTS = lower urinary tract symptoms; IIEF = International Index of Erectile Function; EF = erectile function; ED = erectile dysfunction; EjD = ejaculatory dysfunction; CPSI = Chronic Prostatitis Symptom Index; IPSS = international prostate symptom score.
25–29.9 vs. 24.9 or less 30 or more vs. 24.9 or less
Categorically coded BMI (kg/m )*
2
Logistic regression models
Table 2 The effect of categorically coded BMI on the rate of moderate to severe ED (IIEF-EF scale score <16), severe LUTS (IPSS score ⱖ19), the CPSI pain scale score >median, pain/discomfort on ejaculation, and decreased ejaculatory volume
The relationship between continuously coded BMI and the IIEF-EF scale score, the IPSS score, and the CPSI pain scale score was tested in age-adjusted linear regression models. The relationship between continuously coded BMI and the presence of pain/discomfort on ejaculation or of decreased ejaculate volume was tested in logistic regression models, due to the binary nature of these two end points. *Age-adjusted body mass index. †Regression coefficient, P value. ‡Odds ratio, P value. BMI = body mass index; IIEF = International Index of Erectile Function; EF = erectile function; ED = erectile dysfunction; EjD = ejaculatory dysfunction; CPSI = Chronic Prostatitis Symptom Index; IPSS = international prostate symptom score.
-0.04, 0.2†
Presence of pain/ discomfort on ejaculation
EjD (N = 590)
Continuously coded CPSI pain scale score
Logistic regression models
IIEF (N = 574)
CPSI (N = 500)
IPSS (N = 575)
Linear regression models
The effect of continuously coded BMI on the IIEF-EF scale score, the IPSS score, the CPSI pain scale score, and EjD
Continuously coded BMI (kg/m2)*
Table 1
Body Mass Index and Its Association with Genitourinary Disorders 2143
J Sex Med 2008;5:2141–2151
2144 such as BMI and the IIEF for example, the most bias-free assessment of the relationships between them should consist of a correlation or linear regression analysis, where univariable linear regression is synonymous with a correlation. Logistic regression represents an alternative to that method, where the outcome variable needs to be stratified. However, the predictor or predictors can assume either continuously coded or categorically coded formats. Both methods were used in our analysis. Both methods provide results that can be readily explained and easily understood. Alternatives to these modeling techniques include explanatory fractional analysis, where, for example, individual questionnaire items are used to explain the relationship in the outcome variable. However, such an approach is methodologically and substantially more complex, and the understanding of its results may be equally challenging. Therefore, we feel that the statistical methods used in this study provide an excellent trade-off between assessing of the underlying relationships and the ability to illustrate and understand the data. Hence, in our categorical analyses, BMI was stratified between non-obese (BMI < 24.9 kg/m2), overweight (BMI 25–29.9 kg/m2), and obese (BMI ⱖ 30 kg/m2). In our continuously coded analyses, no stratification was used. Continuously coded age was used for adjustment. All statistical tests were two-sided and P value < 0.05 was considered statistically significant. SPSS for Windows version 15.0 was used (SPSS Inc., Chicago, IL, USA). Results
Mean age was 54.1 years (median 54.0, range 30–83). BMI ranged from 19.1 to 48.8 kg/m2 (mean 26.6 kg/m2, median 26.2 kg/m2). Out of the total of 590 participants, 209 (35.4%) were non-obese, 296 (50.2%) were overweight and 85 (14.4%) were obese. Descriptive characteristics of the study cohort are shown in Table 3. The distribution of the IIEF-EF scale score (N = 574) ranged from 1 to 30 (mean 23.7, median 28.0). Out of the 574 participants, 111 (19.3%) had moderate to severe ED defined as IIEF-EF scale score <16. The distribution of the IPSS scores (N = 574) ranged from 0 to 34 (mean 6.9, median 5.0). Out of the 574 participants, 35 (6.1%) had severe LUTS (IPSS ⱖ 19). The CPSI pain scale scores ranged from 0 to 17 (2.2, 1). Pain/discomfort on ejaculation was reported by 183 (36.6%) and subjectively decreased ejaculate J Sex Med 2008;5:2141–2151
Bhojani et al. Table 3 Descriptive characteristics of the study cohort (N = 590) Variable
N (%)
Total number of patients (%) Age (years) Mean (median) Range BMI (kg/m2) Mean (median) Range WHO BMI (kg/m2) ⱕ24.9 (non-obese) 25–29.9 (overweight) 30–34.9 (obese) ⱖ35 (severely obese) IIEF-EF scale score (N = 574) Mean (median) Range Potent to moderate ED (ⱖ16) Moderate to severe ED (<16) Mean BMI (kg/m2) in participants with: Potent to moderate ED (ⱖ16) Moderate to severe ED (<16) IPSS score (N = 575) Mean (median) Range Mild to moderate LUTS (score < 19) Severe LUTS (score ⱖ 19) Mean BMI (kg/m2) in participants with: Mild to moderate LUTS (score < 19) Severe LUTS (score ⱖ 19) CPSI pain scale score (N = 500) Mean (median) Range CPSI pain scale score >1 (median) CPSI pain scale score ⱕ1 (median) Mean BMI (kg/m2) in participants with: Pain scale score >1 (median) Pain scale score ⱕ1 (median) EjD (N = 590) Pain/discomfort during ejaculation Low ejaculate volume Mean BMI (kg/m2) in participants with: Pain/discomfort during ejaculation Normal ejaculation Low ejaculate volume Normal ejaculate volume
590 (100.0) 54.1 (54.0) 30–83 26.6 (26.2) 19.1–48.8 209 296 74 11
(35.4) (50.2) (12.5) (1.9)
23.7 (28.0) 1–30 463 (80.7) 111 (19.3) 26.4 26.8 6.9 (5.0) 0–34 540 (93.9) 35 (6.1) 26.3 27.4 2.2 (1.0) 0–17 183 (36.6) 317 (63.4) 26.0 26.7 28 (4.7) 156 (26.4) 26.0 26.6 27.1 26.4
WHO = World Health Organization; BMI = body mass index; ED = erectile dysfunction; IIEF = International Index of Erectile Function; EF = erectile function; LUTS = lower urinary tract symptoms; EjD = ejaculatory dysfunction; IPSS = international prostate symptom score; CPSI = Chronic Prostatitis Symptom Index.
volume was recorded in 156 (26.4%). The distributions of continuously coded and categorized BMI are shown in Table 3. The distribution of categorized BMI according to the five examined categorical end points is shown in Table 4.
Effect of Continuously Coded BMI Table 1 shows the linear regression analyses that examine the effect of continuously coded BMI on three assessable end points that were coded on a continuous scale, namely, IIEF, IPSS, and CPSI.
7 (8.2%)
8 (2.7%)
ⱖ30
25–29.9
WHO = World Health Organization; BMI = body mass index; IIEF = International Index of Erectile Function; EF = erectile function; ED = erectile dysfunction; EjD = ejaculatory dysfunction; LUTS = lower urinary tract symptoms; IPSS = international prostate symptom score; CPSI = Chronic Prostatitis Symptom Index.
49 (23.4%) 73 (24.7%) 34 (40.0%) 160 (76.5%) 223 (75.3%) 51 (60.0%) <24.9
206 (93.6%) 264 (95.6%) 70 (88.6%) 43 (20.1%) 43 (15.1%) 25 (32.5%) 170 (79.8%) 241 (84.8%) 52 (67.5%)
14 (6.4%) 12 (4.3%) 9 (11.4%)
196 (93.8%) 288 (97.3%) 78 (91.8%) 91 (48.1%) 68 (28.2%) 24 (34.3%) 98 (51.8%) 173 (71.8%) 46 (65.7%)
13 (6.2%)
Normal ejaculate volume Normal ejaculation Higher than median (1)
Pain/discomfort on ejaculation
EjD (N = 590) CPSI pain scale score (N = 500)
Median and below (1) Moderate to severe ED (<16)
Severe LUTS (ⱖ19)
IPSS score (N = 575)
Mild to moderate LUTS (<19)
IIEF-EF scale score (N = 574)
Potent to moderate ED (ⱖ16) BMI (kg/m2)
Table 4
Tabulated data according to the WHO BMI categories and symptom severity
EjD (N = 590)
Decreased ejaculate volume
Body Mass Index and Its Association with Genitourinary Disorders
2145
The analyses demonstrated that elevated BMI is only statistically significantly related to IIEFEF scale score, as evidenced by an age-adjusted regression coefficient of 0.45 (P < 0.001). This relationship implies that for every one unit increase in BMI, the IIEF score decreased by 0.45 points. Conversely, BMI was unrelated to the IPSS score (regression coefficient: 0.02, P = 0.8) and CPSI pain scale score (regression coefficient: –0.04, P = 0.2). The effect of continuously coded BMI on the prevalence of the two binary coded end points, namely, pain/discomfort on ejaculation and on the prevalence of subjectively decreased ejaculate volume was examined in binary logistic regression models. After age adjustment, elevated BMI was statistically significantly related to lower rate of reported subjectively decreased ejaculate volume (odds ratio: 0.93, P = 0.02). This relationship implies that for every one unit increase in BMI, men were 7% less likely to report subjectively decreased ejaculate volume. Conversely, no statistically significant relationship was recorded between continuously coded BMI and pain/ discomfort on ejaculation (odds ratio: 0.91, P = 0.09).
Effect of Ccategorically Coded BMI Table 2 shows the effect of categorically coded BMI on the five examined end points. BMI was dichotomized between non-obese (BMI < 24.9 kg/ m2), overweight (BMI 25–29.9 kg/m2), and obese (BMI ⱖ 30 kg/m2). Non-obese (BMI < 24.9 kg/ m2) represented the reference category in data analyses. Since only 11 men (1.9%) were severely obese (BMI ⱖ 35 kg/m2), this category was combined with the obese (BMI ⱖ 30 kg/m2) category. In logistic regression analyses, overweight individuals were 1.9 times more likely to report moderate to severe ED (P = 0.01). Conversely, overweight individuals were 58% less likely to report CPP in excess of the median (P < 0.001), 64% were less likely to report pain/discomfort on ejaculation (P = 0.03), and 38% were less likely to report subjectively decreased ejaculate volume (P = 0.05). No statistically significant effect was seen between BMI ⱖ 30 kg/m2 and any of the five end points. Finally, categorized BMI was unrelated to IPSS score ⱖ19 (odds ratio: 0.6, P = 0.2). Discussion
Obesity represents one of the key features of the metabolic syndrome and is associated with J Sex Med 2008;5:2141–2151
2146 elevated levels of several proinflammatory cytokines, which contribute to a number of pathologies [26,27]. From the urologic perspective, two pathologies, namely, ED and LUTS, were identified as possible urological manifestations of the metabolic syndrome and indirectly of obesity [6,28]. Two recent reports demonstrated an association between the metabolic syndrome and ED [28,29]. In 2004 Gunduz et al. reported on 79 cardiology clinic outpatients with coronary artery disease and dyslipidemia. ED was diagnosed in 59 (74.7%) and all patients with metabolic syndrome (38) had ED [28]. Several cross-sectional studies demonstrated a link between obesity and ED [7,8], sedentary lifestyle [30], and smoking [7]. Kupelian et al. addressed the effect of BMI in the Massachusetts Male Aging Study, a population-based prospective cohort, where ED was assessed at three points during a 15-year period. In a low-risk subgroup of men with low BMI values (<25), ED was associated with a twofold higher risk of subsequent diagnosis of the metabolic syndrome (adjusted relative risk 2.09) [29]. The authors concluded that ED may represent an early warning sign in men otherwise considered at lower risk (low BMI) for the metabolic syndrome. In consequence, ED was proposed as a precursor of cardiovascular disease. Mueller et al. performed a detailed review of the association between cardiovascular disease, the metabolic syndrome, and ED. They demonstrated that endothelial damage represents the common pathway in cardiovascular disease, in the metabolic syndrome, and in ED [6]. Proinflammatory cytokines, such as interleukin-6 and 8 represent risk factors for endothelial dysfunction and possibly contribute to ED, as well as to other male disorders [31]. These may include LUTS, prostatitisrelated CPP, and EjD. Indeed, Rohrmann et al. demonstrated that obesity in young adulthood is associated with LUTS [32]. Their findings were derived from an observational longitudinal cohort of 2797 men. Interestingly, current BMI was unrelated to LUTS or prostate surgery for LUTS. Therefore, it does not appear to be a consensus regarding the role of BMI and LUTS. Interestingly, LUTS and ED frequently coexist [19,33–35]. The association between LUTS and ED was reported by MartinMorales et al. in a cross-sectional study of 2,476 Spanish men. In this study, the presence of LUTS was the strongest risk factor of ED (age-adjusted odds ratio of 2.67) [36]. Similarly, in 2003, Braun et al. reported a strong association between LUTS J Sex Med 2008;5:2141–2151
Bhojani et al. and ED in a German cohort of 4489 men [34]. The coexistence of ED and LUTS may indicate that BMI is implemented in LUTS pathogenesis [17,19,37,38]. Despite the interest in the link between obesity and ED and/or LUTS, few investigators addressed the association between obesity and other male pelvic disorders, such as EjD and/or CPP. Since male pelvic disorders are interrelated, EjD and/or CPP may also be affected by BMI distribution. For example, the prevalence of pain or discomfort during ejaculation occurs in 5–19% of men with LUTS [35,39–41] and in as many as 11% of men undergoing prostate cancer screening. However, to the best of our knowledge, few studies directly addressed the relationship between EjD and obesity [7]. Moreover, we found no English language medical literature addressing the link between obesity and CPP. Based on the potential importance of obesity in the pathogenesis of various male pathologies, we decided to systematically examine the association between obesity, defined as BMI ⱖ 30 kg/m2 and self-reported prevalence of ED, LUTS, CPP, and EjD. The effect of obesity may be assessed in two ways. BMI is commonly assessed in categories, according to the WHO definition. However, BMI, like weight and height, represents a continuously coded variable. Therefore, the effect of categorization may introduce a bias. The latter may either exaggerate or weaken the true effect of the continuously coded BMI [18]. Therefore it is imperative to examine the effect of BMI without converting this variable into strata. However, since many studies rely on categorically coded BMI, it is also imperative to examine the effect of categorically coded BMI, to allow comparisons with previous reports. To comply with these conditions, we performed two types of analyses. One relied on continuously coded BMI. The other relied on the WHO BMI categories. Since some of the end points of interest are also coded in continuous format, we applied the same logic to IIEF, IPSS, and CPSI scores. These three end points were first examined as continuous variables. Then, in order to be compatible with other studies that categorize these scale scores, we also categorized between generally accepted levels of ED, LUTS, and CPP [22,42,43]. In continuously coded BMI analyses, increasing BMI was related to lower IIEF-EF scale scores (P < 0.001). The regression coefficient of 0.45 indicates that for each one unit BMI increase, the IIEF scale score decreased by 0.45/30 units. This
Body Mass Index and Its Association with Genitourinary Disorders implies that after age adjustment a patient with a BMI of 29 would be expected to have a 3.6 unit (12%) lower IIEF scale score than his counterpart with a BMI of 21. The same effect was recorded in categorical BMI analyses, where overweight men (BMI 25–29.9 kg/m2) were nearly twice more likely (odds ratio: 1.9, P = 0.01) to have moderate to severe ED than their normal weight counterparts (BMI 24.9 or less). The rate of ED for obese men (BMI ⱖ 30 kg/m2) was not statistically significantly different from their non-obese counterparts (BMI < 24.9 kg/m2, [P = 0.3]), despite an odds ratio of 1.47. This finding may be related to lack of power, since the magnitude of the effect was similar to that recorded for overweight men. Taken together our data indicate that elevated BMI has a detrimental effect on EF. This finding is consistent with previous BMI vs. ED data, reported by Blanker et al. [7], Aytac et al. [8], and Walczak et al. [9], and with the general consensus that the metabolic syndrome predisposes to ED [6,44–46]. These findings are important—BMI may decrease the quality of life of men by virtue of predisposing to ED [13,47,48]. More importantly, elevated BMI is a precursor of several pathologies, including coronary artery disease [49]. Prevention of obesity may therefore not only reduce its urologic manifestations, but hopefully also other even more important systemic signs and symptoms. Our data addressing the relationship between both continuously coded or categorized BMI and the IPSS score failed to reveal a statistically significant association. Our data are consistent with several reports that demonstrated no association between BMI and LUTS or benign prostatic hyperplasia. For example a cross-sectional Austrian population-based study (N = 939) reported no association between BMI and IPSS [50]. A Turkish study (N = 286) found no relationship between serum lipids and IPSS [51]. Finally, a large-scale population-based study (N = 2797) from the united States, demonstrated that current BMI is unrelated to LUTS or to the history of prostatic surgery for LUTS [32]. It is noteworthy that in the same study, the authors also reported that elevated BMI in young adulthood predisposes to LUTS [32]. This is not the first study to report that a history of elevated BMI is a precursor of clinically relevant diseases many years later [52]. Therefore, BMI at specific time points may be more relevant than other circumstances. It is however noteworthy that several other studies suggested that BMI is associated with LUTS. For
2147
example, Hammarsten et al. (N = 158) demonstrated that obese men, defined according to elevated BMI, had larger prostates than their non-obese counterparts [15]. Ozden et al. (N = 78) demonstrated that men with biochemical profiles indicative of metabolic syndrome had enlarged prostates [16]. Paick et al. showed that metabolic profiles are related to IPSS data [18]. Finally, Giovannucci et al. demonstrated that abdominal obesity was related to the likelihood of prostate surgery for LUTS and among those that did not have prostate surgery for LUTS, abdominal obesity was associated with more frequent urinary symptoms [53]. Interestingly, BMI was unrelated to either outcome, which indicates that different definitions of obesity or different fat distributions may be related to different clinical manifestations. We identified no English language reports addressing the effect of BMI on CPP or EjD. Interestingly, our categorical analyses of BMI revealed that men with BMI 25–29.9 were 58% less likely to report CPSI pain scale scores above the median. This indicates that obesity, expressed as BMI, is related to lower rate of CPP. It is noteworthy that the analyses of continuously coded BMI on CPSI scale score failed to reveal any statistically significant relationship (P = 0.2). Therefore, the presence of a true effect between BMI and CPSI scale score should be interpreted with caution, since analyses of categorically coded data can exaggerate an existing relationship by introducing artificial cutoffs [18]. Similarly to CPSI scale score analyses, we found no effect between continuously coded BMI and pain/discomfort at ejaculation. However, the effect of categorized BMI revealed a 64% decrease in the rate of pain/discomfort at ejaculation (P = 0.03). This statistical significance of this effect needs to be interpreted with caution, as data categorization may introduce type I error and may artificially inflate the existing association. Therefore, further analyses of this association are certainly needed as the link between BMI and pain/discomfort at ejaculation has not received much attention previously. Elevated BMI was related to lower rate of self-reported subjectively decreased ejaculatory volume in categorically and continuously coded analyses. When coded as a continuous variable, every one unit BMI increase resulted in a 7% decrease in the rate of self-reported subjectively decreased ejaculate volume (P = 0.02). In categorized BMI analyses, overweight men were 38% J Sex Med 2008;5:2141–2151
2148 less likely to report subjectively decreased ejaculatory volume (P = 0.05). Therefore, elevated BMI appears to be associated with a lesser extent of ejaculatory dysfunction, at least in the current population. Several explanations may be proposed for what intuitively might appear as paradoxical. For example, it is possible that obese men are less affected by subjectively decreased ejaculate volume and tend to underreport it. Alternatively, it is also possible that indeed fewer obese men suffer of subjectively decreased ejaculate volume, especially that the protective association between BMI and subjectively decreased ejaculate was recorded in both analytic approaches (continuous and categorical analyses). No other study addressed the relationship between ejaculate volume and BMI. Therefore, our data cannot be compared with other studies. The same hypothesis might be proposed for CPP. However, the epidemiologic nature of the current study prohibits any valid attempts at elucidating the effect and mechanism of elevated BMI on either CPP or ED. This study has several strengths. First, it examined the prevalence of five genitourinary disorders and tested for presence and the strength of associations with BMI. Second, the study relied on two distinct statistical approaches. One uses categorical data and the other is based on continuously coded variables. This approach minimizes bias related to data transformations. Third, the study relied on a sample of healthy, populationdwelling men. Such individuals provide substantially more realistic estimates of the prevalence of urologic symptoms in the community, than hospital or clinic-based samples. Fourth, the study simultaneously examined several end points within the same population. This allows valid comparison of the coprevalence of the examined urologic disorders and of their suspected association with BMI. The present study also has several limitations. The sample size was relatively small. Moreover, BMI was not compared with other measures of overweightedness such as lean body mass and waist-to-hip ratio. Moreover, it is noteworthy that BMI and possibly other measures of overweightedness may not detect special patterns of lean body mass (muscle) distribution. For example men with elevated muscle mass may at times be quantified as overweight or obese if their physique is not specifically accounted for. The cross-sectional nature of our study only allows assessing the association between BMI and J Sex Med 2008;5:2141–2151
Bhojani et al. the targeted genitourinary disorders. Causality cannot be inferred. Another limitation of our study is that we did not have information on the comorbidities of our study population. This information may explain the high rate of ED in our study, as obesity is related to diabetes and consequently to ED. The questionnaire formats that were used also limited the amount of detail that can be extracted and interpreted. For example, the recorded prevalence of subjectively decreased ejaculate volume and/or the presence of pain/discomfort on ejaculation may be affected by the dichotomous type of input that was allowed (yes/no). It is plausible that a much finer definition of subjectively decreased ejaculatory volume or of pain/discomfort on ejaculation could have been established, if more answer choices were available. It is also important to notice that the logistic regression analysis, in its binary format, only allows two possible outcomes, which also may further reduce the variability of possible outcomes. This spectrum bias can be avoided in linear regression models, provided that continuously coded outcome is available. Finally, our population overwhelmingly consisted of white men. In consequence, race adjustment was not performed. Despite these limitations, our findings are noteworthy and deserve future attention.
Conclusions
Our findings indicate that elevated BMI is related to worse EF, which is consistent throughout the literature [7–9,30]. Moreover, it appears that current BMI does not exert a central role on IPSS score. Finally, elevated BMI may exert a protective effect on ejaculatory dysfunction and on associated pain, as well as on CPP. Our findings require corroboration, before a protective effect can be assigned between elevated BMI and CPP or ejaculatory dysfunction. However, it is highly noteworthy that in this relatively large cohort of patients, BMI failed to exert a uniformly adverse effect on several male pelvic pathologies. Corresponding Author: Pierre Karakiewicz, MD, Cancer Prognostics and Health Outcomes unit, University of Montreal Health Center, Campus St-Luc 1058, rue St-Denis, Montreal, Quebec, Canada H2X 3J4. Tel: 514 890 8000 ext. 35336; Fax: 514 227 5103; E-mail: [email protected] Conflict of Interest: None declared.
Body Mass Index and Its Association with Genitourinary Disorders Statement of Authorship
Category 1 (a) Conception and Design Pierre I. Karakiewicz; Georg Hutterer; Francesco Montorsi; Naeem Bhojani; Nazareno Suardi; Shahrokh Shariat (b) Acquisition of Data Pierre I. Karakiewicz; Claudio Jeldres; Hugues Widmer; Luc Valiquette; Paul Perrotte; Philippe Arjane; Francois Benard; Francois Peloquin (c) Analysis and Interpretation of Data Pierre I. Karakiewicz; Naeem Bhojani, Umberto Capitanio
Category 2 (a) Drafting the Article Naeem Bhojani; Claudio Jeldres; Pierre Karakiewicz, Umberto Capitanio (b) Revising It for Intellectual Content Pierre I. Karakiewicz, Umberto Capitanio
I.
Category 3 (a) Final Approval of the Completed Article Pierre I. Karakiewicz; Naeem Bhojani
References
1 Ogden CL, Carroll MD, Curtin LR, McDowell MA, Tabak CJ, Flegal KM. Prevalence of overweight and obesity in the united States, 1999–2004. JAMA 2006;295:1549–55. 2 Fontaine KR, Heo M, Allison DB. Obesity and prostate cancer screening in the USA. Public Health 2005;119:694–8. 3 Seidell JC. Obesity in Europe: Scaling an epidemic. Int J Obes Relat Metab Disord 1995;19(3 suppl): S1–4. 4 Pietinen P, Vartiainen E, Mannisto S. Trends in body mass index and obesity among adults in Finland from 1972 to 1992. Int J Obes Relat Metab Disord 1996;20:114–20. 5 Eckel RH, Grundy SM, Zimmet PZ. The metabolic syndrome. Lancet 2005;365:1415–28. 6 Muller A, Mulhall JP. Cardiovascular disease, metabolic syndrome and erectile dysfunction. Curr Opin Urol 2006;16:435–43. 7 Blanker MH, Bohnen AM, Groeneveld FP, Bernsen RM, Prins A, Thomas S, Bosch JL. Correlates for erectile and ejaculatory dysfunction in older Dutch men: A community-based study. J Am Geriatr Soc 2001;49:436–42. 8 Aytac IA, Araujo AB, Johannes CB, Kleinman KP, McKinlay JB. Socioeconomic factors and incidence of erectile dysfunction: Findings of the longitudinal Massachussetts Male Aging Study. Soc Sci Med 2000;51:771–8.
2149
9 Walczak MK, Lokhandwala N, Hodge MB, Guay AT. Prevalence of cardiovascular risk factors in erectile dysfunction. J Gend Specif Med 2002;5:19–24. 10 Ahn TY, Park JK, Lee SW, Hong JH, Park NC, Kim JJ, Park K, Park H, Hyun JS. Prevalence and risk factors for erectile dysfunction in Korean men: Results of an epidemiological study. J Sex Med 2007;4:1269–76. 11 Mulhall J, Teloken P, Brock G, Kim E. Obesity, dyslipidemias and erectile dysfunction: A report of a subcommittee of the sexual medicine society of North America. J Sex Med 2006;3:778–86. 12 El-Sakka AI, Hassoba HM, Sayed HM, Tayeb KA. Pattern of endocrinal changes in patients with sexual dysfunction. J Sex Med 2005;2:551–8. 13 Zohdy W, Kamal EE, Ibrahim Y. Androgen deficiency and abnormal penile duplex parameters in obese men with erectile dysfunction. J Sex Med 2007;4:797–808. 14 Wingard C, Fulton D, Husain S. Altered penile vascular reactivity and erection in the Zucker obesediabetic rat. J Sex Med 2007;4:348–62 (discussion 62–3). 15 Hammarsten J, Hogstedt B, Holthuis N, Mellstrom D. Components of the metabolic syndrome-risk factors for the development of benign prostatic hyperplasia. Prostate Cancer Prostatic Dis 1998;1: 157–62. 16 Ozden C, Ozdal OL, Urgancioglu G, Koyuncu H, Gokkaya S, Memis A. The correlation between metabolic syndrome and prostatic growth in patients with benign prostatic hyperplasia. Eur Urol 2007;51:199–203 (discussion 4–6). 17 Kasturi S, Russell S, McVary KT. Metabolic syndrome and lower urinary tract symptoms secondary to benign prostatic hyperplasia. Curr Urol Rep 2006;7:288–92. 18 Paick JS, Yang JH, Kim SW, Ku JH. Are age, anthropometry and components of metabolic syndrome-risk factors interrelated with lower urinary tract symptoms in patients with erectile dysfunction? A prospective study. Asian J Androl 2007; 9:213–20. 19 McVary K. Lower urinary tract symptoms and sexual dysfunction: Epidemiology and pathophysiology. BJU Int 2006;97(2 suppl):23–8 (discussion 44–5). 20 Guay A, Jacobson J. The relationship between testosterone levels, the metabolic syndrome (by two criteria), and insulin resistance in a population of men with organic erectile dysfunction. J Sex Med 2007;4:1046–55. 21 Corona G, Mannucci E, Petrone L, Balercia G, Paggi F, Fisher AD, Lotti F, Chiarini V, Fedele D, Forti G, Maggi M. NCEP-ATPIII-defined metabolic syndrome, type 2 diabetes mellitus, and prevalence of hypogonadism in male patients with sexual dysfunction. J Sex Med 2007;4:1038– 45. J Sex Med 2008;5:2141–2151
2150 22 Cappelleri JC, Rosen RC, Smith MD, Mishra A, Osterloh IH. Diagnostic evaluation of the erectile function domain of the International Index of Erectile Function. Urology 1999;54:346–51. 23 Barry MJ, Fowler FJ Jr., O’Leary, MP, Bruskewitz RC, Holtgrewe HL, Mebust WK, Cockett AT. The American Urological Association symptom index for benign prostatic hyperplasia. The Measurement Committee of the American Urological Association. J Urol 1992;148:1549–57 (discussion 64). 24 Karakiewicz PI, Perrotte P, Valiquette L, Benard F, McCormack M, Menard C, McNaughton Collins M, Nickel JC. French-Canadian linguistic validation of the NIH Chronic Prostatitis Symptom Index. Can J Urol 2005;12:2816–23. 25 Hald T, Nordling J, Andersen JT, Bilde T, Meyhoff HH, Walter S. A patient weighted symptom score system in the evaluation of uncomplicated benign prostatic hyperplasia. Scand J Urol Nephrol Suppl 1991;138:59–62. 26 Yudkin JS, Stehouwer CD, Emeis JJ, Coppack SW. C-reactive protein in healthy subjects: Associations with obesity, insulin resistance, and endothelial dysfunction: A potential role for cytokines originating from adipose tissue? Arterioscler Thromb Vasc Biol 1999;19:972–8. 27 Bastard JP, Jardel C, Bruckert E, Blondy P, Capeau J, Laville M, Vidal H, Hainque B. Elevated levels of interleukin 6 are reduced in serum and subcutaneous adipose tissue of obese women after weight loss. J Clin Endocrinol Metab 2000;85: 3338–42. 28 Gunduz MI, Gumus BH, Sekuri C. Relationship between metabolic syndrome and erectile dysfunction. Asian J Androl 2004;6:355–8. 29 Kupelian V, Shabsigh R, Araujo AB, O’Donnell AB, McKinlay JB. Erectile dysfunction as a predictor of the metabolic syndrome in aging men: Results from the Massachusetts Male Aging Study. J Urol 2006; 176:222–6. 30 Moreira ED Jr, Lisboa Lobo CF, Villa M, Nicolosi A, Glasser DB. Prevalence and correlates of erectile dysfunction in Salvador, northeastern Brazil: A population-based study. Int J Impot Res 2002;14(2 suppl):S3–9. 31 Ziccardi P, Nappo F, Giugliano G, Esposito K, Marfella R, Cioffi M, D’Andrea F, Molinari AM, Giugliano D. Reduction of inflammatory cytokine concentrations and improvement of endothelial functions in obese women after weight loss over one year. Circulation 2002;105:804–9. 32 Rohrmann S, Smit E, Giovannucci E, Platz EA. Associations of obesity with lower urinary tract symptoms and noncancer prostate surgery in the Third National Health and Nutrition Examination Survey. Am J Epidemiol 2004;159:390–7. 33 El-Sakka AI. Lower urinary tract symptoms in patients with erectile dysfunction: Analysis of risk factors. J Sex Med 2006;3:144–9. J Sex Med 2008;5:2141–2151
Bhojani et al. 34 Braun MH, Sommer F, Haupt G, Mathers MJ, Reifenrath B, Engelmann UH. Lower urinary tract symptoms and erectile dysfunction: Co-morbidity or typical “Aging Male” symptoms? Results of the “Cologne Male Survey”. Eur Urol 2003;44:588–94. 35 Frankel SJ, Donovan JL, Peters TI, Abrams P, Dabhoiwala NF, Osawa D, Lin AT. Sexual dysfunction in men with lower urinary tract symptoms. J Clin Epidemiol 1998;51:677–85. 36 Martin-Morales A, Sanchez-Cruz JJ, Saenz de Tejada I, Rodriguez-Vela L, Jimenez-Cruz JF, Burgos-Rodriguez R. Prevalence and independent risk factors for erectile dysfunction in Spain: Results of the Epidemiologia de la Disfuncion Erectil Masculina Study. J Urol 2001;166:569–74 (discussion 74–5). 37 Russell S, McVary KT. Lower urinary tract symptoms and erectile dysfunction: Epidemiology and treatment in the aging man. Curr Urol Rep 2005; 6:445–53. 38 McVary KT. BPH: Epidemiology and comorbidities. Am J Manag Care 2006;12:S122–8. 39 Nickel JC, Elhilali M, Vallancien G. Benign prostatic hyperplasia (BPH) and prostatitis: Prevalence of painful ejaculation in men with clinical BPH. BJU Int 2005;95:571–4. 40 Rosen R, Altwein J, Boyle P, Kirby RS, Lukacs B, Meuleman E, O’Leary MP, Puppo P, Robertson C, Giuliano F. Lower urinary tract symptoms and male sexual dysfunction: the multinational survey of the aging male (MSAM-7). Eur Urol 2003;44:637– 49. 41 Walz J, Perrotte P, Gallina A, Benard F, Valiquette L, McCormack M, Montorsi F, Karakiewicz PI. Ejaculatory disorders may affect screening for prostate cancer. J Urol 2007;178:232–7 (discussion 7– 8). 42 Litwin MS, McNaughton-Collins M, Fowler FJ Jr., Nickel, JC, Calhoun EA, Pontari MA, Alexander RB, Farrar JT, O’Leary, MP. The National Institutes of Health chronic prostatitis symptom index: development and validation of a new outcome measure Chronic Prostatitis Collaborative Research Network. J Urol 1999;162:369–75. 43 van Haarst EP, Heldeweg EA, Newling DW, Schlatmann TJ. A cross-sectional study of the international prostate symptom scores related to age and gender in Dutch adults reporting no voiding complaints. Eur Urol 2005;47:334–9. 44 Roumeguere T, Wespes E, Carpentier Y, Hoffmann P, Schulman CC. Erectile dysfunction is associated with a high prevalence of hyperlipidemia and coronary heart disease risk. Eur Urol 2003;44:355–9. 45 Bal K, Oder M, Sahin AS, Karatas CT, Demir O, Can E, Gumus BH, Ozer K, Sahin O, Esen AA. Prevalence of metabolic syndrome and its association with erectile dysfunction among urologic patients: Metabolic backgrounds of erectile dysfunction. Urology 2007;69:356–60.
Body Mass Index and Its Association with Genitourinary Disorders 46 Vlachopoulos C, Rokkas K, Ioakeimidis N, Stefanadis C. Inflammation, metabolic syndrome, erectile dysfunction, coronary artery disease: Common links. Eur Urol 2007 Dec;52(6):1590–600. 47 Haro JM, Beardsworth A, Casariego J, Gavart S, Hatzichristou D, Martin-Morales A, Schmitt H, Mirone V, Needs N, Riley A, Varanese L, von Keitz A, Kontodimas S. Treatment-seeking behavior of erectile dysfunction patients in Europe: Results of the Erectile Dysfunction Observational Study. J Sex Med 2006;3:530–40. 48 Kratzik CW, Schatzl G, Lunglmayr G, Rucklinger E, Huber J. The impact of age, body mass index and testosterone on erectile dysfunction. J Urol 2005; 174:240–3. 49 Grover SA, Lowensteyn I, Kaouache M, Marchand S, Coupal L, DeCarolis E, Zoccoli J, Defoy I. The prevalence of erectile dysfunction in the primary care setting: importance of risk factors for diabetes and
50
51 52
53
2151
vascular disease. Arch Intern Med 2006;166:213– 9. Haidinger G, Madersbacher S, Waldhoer T, Lunglmayr G, Vutuc C. The prevalence of lower urinary tract symptoms in Austrian males and associations with sociodemographic variables. Eur J Epidemiol 1999;15:717–22. Lekili M, Muezzinoglu T, Uyanik BS, Buyuksu C. Serum lipid levels in benign prostatic hyperplasia. World J Urol 2006;24:210–3. Morrison JA, Friedman LA, Gray-McGuire C. Metabolic syndrome in childhood predicts adult cardiovascular disease 25 years later: The Princeton Lipid Research Clinics Follow-up Study. Pediatrics 2007;120:340–5. Giovannucci E, Rimm EB, Chute CG, Kawachi I, Colditz GA, Stampfer MJ, Willett WC. Obesity and benign prostatic hyperplasia. Am J Epidemiol 1994;140:989–1002.
J Sex Med 2008;5:2141–2151
Body Mass Index and its Association with Genitourinary Disorders in Men Undergoing Prostate Cancer Screening Naeem Bhojani, MD,*† Paul Perrotte, MD,* Georg Hutterer, MD,‡ Nazareno Suardi, MD,*§ Claudio Jeldres, MD,*† Shahrokh F. Shariat, MD,¶ Umberto Capitanio, MD,* Philippe Arjane, MD,† Hugues Widmer, MD,† Francois Benard, MD,† Francois Peloquin, MD,† Francesco Montorsi, MD,§ and Pierre Karakiewicz, MD* *Cancer Prognostics and Health Outcomes unit, University of Montreal Health Center, Montreal, Canada; †University of Montreal—Urology, Montreal, Canada; ‡Graz Medical University—Urology, Graz, Austria; §Vita-Salute University San Raffaele—Urology, Milan, Italy; ¶University of Texas Southwestern—Urology, Dallas, TX, USA DOI: 10.1111/j.1743-6109.2008.00811.x
ABSTRACT
Introduction. Elevated body mass index (BMI) may predispose to several pelvic pathologies. Aims. We tested the association between BMI and five end points, namely, (i) erectile dysfunction (ED); (ii) lower urinary tract symptoms (LUTS); (iii) chronic prostatitis-associated pain (CPP); and ejaculatory dysfunction that is subdivided between (iv) pain/discomfort on ejaculation; and (v) subjectively decreased ejaculate volume. Methods. Age, height, and weight were prospectively recorded in a cohort of 590 consecutive healthy men undergoing prostate cancer screening. Continuously coded and categorized BMI (World Health Organization classification) were studied. Main Outcome Measures. Age-adjusted analyses relied on logistic and linear regression models, according to data type. Results. The average age was 54.1 years (range 30–83). Of all, 296 were overweight (50.2%, BMI 25–29.9 kg/m2) and 85 were obese (14.4%, BMI ⱖ 30 kg/m2). After age adjustment, elevated continuously coded BMI (P < 0.001) and elevated categorized BMI (P = 0.01) were associated with worse erectile function. Conversely, after age adjustment, elevated continuously coded BMI (P = 0.02) and elevated categorized BMI (P = 0.05) were associated with a lower rate of subjectively decreased ejaculate volume. Finally, after age adjustment, elevated categorically coded BMI was related to lower rates of CPP (P < 0.001) and to a lower rate of pain/discomfort on ejaculation (P = 0.03). Conclusions. In men undergoing prostate cancer screening, the effect of BMI on the five end points is not invariably detrimental. Elevated BMI may predispose to ED, but may also decrease the rate of pain/discomfort on ejaculation and may lower the reported rate of subjectively decreased ejaculate volume. Finally, it appeared to have no effect on LUTS. Bhojani N, Perrotte P, Hutterer G, Suardi N, Jeldres C, Shariat SF, Capitanio U, Arjane P, Widmer H, Benard F, Peloquin F, Montorsi F, and Karakiewicz P. Body mass index and its association with genitourinary disorders in men undergoing prostate cancer screening. J Sex Med 2008;5:2141–2151. Key Words. Body Mass Index; Erectile Dysfunction; Lower Urinary Tract Symptoms; Ejaculatory Pain; Chronic Prostatitis
Introduction
O
besity is common in Western countries [1]. According to the World Health Organization (WHO) criteria, more than 30% of adults in the united States are obese. Moreover, over 70% of Americans over 40 years of age are overweight (body mass index [BMI] ⱖ 25 kg/m2) [2]. Simi-
© 2008 International Society for Sexual Medicine
larly, obesity is a growing problem in Western European countries [3,4]. Central obesity represents one of several components of the metabolic syndrome [5]. The metabolic syndrome may predispose to several pelvic pathologies, such as erectile dysfunction (ED) and others [6]. Several cross-sectional studies have suggested an association between ED and obesity [7–14]. Similarly, a J Sex Med 2008;5:2141–2151
2142 link was reported between the metabolic syndrome and lower urinary tract symptoms (LUTS) [15–17]. Moreover, several investigators reported on the association of urological symptoms, ED, and the metabolic syndrome [18–21]. The link between elevated BMI and other male disorders might be equally strong. However, the presence and the strength of this link have not yet been quantified. Therefore, the purpose of the current study was to test for the presence and the strength of the association between established and the novel urological manifestations of the metabolic syndrome. In order to avoid the confounding effect of categorizing patients into diseased and nondiseased groups, we performed two types of analyses. One rested on categorized and the other on continuously coded data. Aims
We decided to assess the relationship between BMI and five separate end points, namely, (i) ED; (ii) LUTS; (iii) chronic prostatitis-associated pain (CPP); and ejaculatory dysfunction (EjD) that is subdivided between (iv) pain/discomfort on ejaculation; and (v) subjectively decreased (selfreported) ejaculate volume. Methods
The study cohort consisted of 590 consecutive men without evidence of prostate cancer, who participated in an annual prostate cancer screening event, the Prostate Cancer Awareness Days. Age, height, and weight were prospectively recorded in all participants. BMI was calculated and recorded as either a continuous variable or as defined according to the WHO classification (weight [kg]/height squared [m2]), as either nonobese (BMI < 24.9 kg/m2), overweight (BMI 25–29.9 kg/m2), or obese (BMI ⱖ 30 kg/m2). In all analyses, only three BMI categories were used, since very few (11 [1.9%]) severely obese men participated in the event. The study received Institutional Review Board (IRB) approval from University of Montreal Health Center I.R.B. Main Outcome Measures
The effect of either continuously coded or categorized BMI was tested in all analyses. ED represented the first end point. Its prevalence was assessed with the self-administered International J Sex Med 2008;5:2141–2151
Bhojani et al. Index of Erectile Function (IIEF)-erectile function (EF) scale. The relationship between IIEF was examined in linear regression models, where BMI and the IIEF scale score represented continuously coded variables (Table 1). Since IIEF-EF is also frequently categorized, we reexamined the effect of BMI on the IIEF-EF scale score of <16, which defines moderate to severe ED according to Cappelleri et al. [22]. Logistic regression models were used and the effect of categorized BMI was examined (Table 2). LUTS represented the second end point and its prevalence was assessed with the self-administered international prostate symptom score (IPSS). The relationship between the IPSS score was examined in linear regression models, where BMI and the IPSS scale score represented continuously coded variables (Table 1). Since the IPSS score can also be categorized according to LUTS severity, we reexamined the effect of BMI on the IPSS score ⱖ19, which defines severe LUTS [23]. Logistic regression models were used and the effect of categorized BMI was examined (Table 2). The pain scale score of the Chronic Prostatitis Symptom Index (CPSI) quantified CPP and represented the third end point [24]. The relationship between CPP score was examined in linear regression models, where BMI and the CPP scale score represented continuously coded variables (Table 1). We then categorized the CPP scores according to its median and reexamined the effect of BMI on the CPP score above median. Logistic regression models were used and the effect of categorized BMI was examined (Table 2). Subjectively decreased ejaculate volume and presence of pain/discomfort on ejaculation represented, respectively, the fourth and the fifth end points. The prevalence of these conditions that contribute to EjD was measured with items 8 and 10 of the Danish Prostate Symptom Score sexual function questionnaire [25]. Because of their inherent binary format (yes/no), these two end points could only be analyzed in logistic regression models. Two models were fitted. One relied on continuously coded BMI (Table 1). The second relied on categorized BMI (Table 2). Statistical analyses consisted of t-tests and of univariable and age-adjusted logistic and linear regression models. BMI represented the main predictor in all models. Continuously coded variables can be either kept in their original continuously coded format or can be categorized. Stratified data can only be examined in categorical format. When two continuously coded variables are examined,
Continuously coded IPSS score 0.02, 0.8†
Continuously coded IIEF-EF scale score 0.45, <0.001† 0.91, 0.09‡
0.93, 0.02‡
Presence of decreased ejaculate volume
IIEF-EF scale score Moderate to severe ED (<16) 1.88, 0.01† 1.47, 0.3†
IIEF (N = 574)
Presence of of pain/ discomfort on ejaculation 0.36, 0.03† 0.90, 0.9†
0.42, <0.001† 0.55, 0.07†
0.60, 0.2† 1.25, 0.6†
EjD (N = 590) CPSI pain scale score (>median)
CPSI (N = 500)
IPSS score (ⱖ19)
IPSS (N = 575)
0.62, 0.05† 0.54, 0.1†
Presence of decreased ejaculate volume
The relationship between categorically coded BMI and the IIEF-EF scale score, the IPSS score, the CPSI pain scale score and EjD was tested in age-adjusted logistic regression models. *Age-adjusted body mass index. †Odds ratio, P value. BMI = body mass index; LUTS = lower urinary tract symptoms; IIEF = International Index of Erectile Function; EF = erectile function; ED = erectile dysfunction; EjD = ejaculatory dysfunction; CPSI = Chronic Prostatitis Symptom Index; IPSS = international prostate symptom score.
25–29.9 vs. 24.9 or less 30 or more vs. 24.9 or less
Categorically coded BMI (kg/m )*
2
Logistic regression models
Table 2 The effect of categorically coded BMI on the rate of moderate to severe ED (IIEF-EF scale score <16), severe LUTS (IPSS score ⱖ19), the CPSI pain scale score >median, pain/discomfort on ejaculation, and decreased ejaculatory volume
The relationship between continuously coded BMI and the IIEF-EF scale score, the IPSS score, and the CPSI pain scale score was tested in age-adjusted linear regression models. The relationship between continuously coded BMI and the presence of pain/discomfort on ejaculation or of decreased ejaculate volume was tested in logistic regression models, due to the binary nature of these two end points. *Age-adjusted body mass index. †Regression coefficient, P value. ‡Odds ratio, P value. BMI = body mass index; IIEF = International Index of Erectile Function; EF = erectile function; ED = erectile dysfunction; EjD = ejaculatory dysfunction; CPSI = Chronic Prostatitis Symptom Index; IPSS = international prostate symptom score.
-0.04, 0.2†
Presence of pain/ discomfort on ejaculation
EjD (N = 590)
Continuously coded CPSI pain scale score
Logistic regression models
IIEF (N = 574)
CPSI (N = 500)
IPSS (N = 575)
Linear regression models
The effect of continuously coded BMI on the IIEF-EF scale score, the IPSS score, the CPSI pain scale score, and EjD
Continuously coded BMI (kg/m2)*
Table 1
Body Mass Index and Its Association with Genitourinary Disorders 2143
J Sex Med 2008;5:2141–2151
2144 such as BMI and the IIEF for example, the most bias-free assessment of the relationships between them should consist of a correlation or linear regression analysis, where univariable linear regression is synonymous with a correlation. Logistic regression represents an alternative to that method, where the outcome variable needs to be stratified. However, the predictor or predictors can assume either continuously coded or categorically coded formats. Both methods were used in our analysis. Both methods provide results that can be readily explained and easily understood. Alternatives to these modeling techniques include explanatory fractional analysis, where, for example, individual questionnaire items are used to explain the relationship in the outcome variable. However, such an approach is methodologically and substantially more complex, and the understanding of its results may be equally challenging. Therefore, we feel that the statistical methods used in this study provide an excellent trade-off between assessing of the underlying relationships and the ability to illustrate and understand the data. Hence, in our categorical analyses, BMI was stratified between non-obese (BMI < 24.9 kg/m2), overweight (BMI 25–29.9 kg/m2), and obese (BMI ⱖ 30 kg/m2). In our continuously coded analyses, no stratification was used. Continuously coded age was used for adjustment. All statistical tests were two-sided and P value < 0.05 was considered statistically significant. SPSS for Windows version 15.0 was used (SPSS Inc., Chicago, IL, USA). Results
Mean age was 54.1 years (median 54.0, range 30–83). BMI ranged from 19.1 to 48.8 kg/m2 (mean 26.6 kg/m2, median 26.2 kg/m2). Out of the total of 590 participants, 209 (35.4%) were non-obese, 296 (50.2%) were overweight and 85 (14.4%) were obese. Descriptive characteristics of the study cohort are shown in Table 3. The distribution of the IIEF-EF scale score (N = 574) ranged from 1 to 30 (mean 23.7, median 28.0). Out of the 574 participants, 111 (19.3%) had moderate to severe ED defined as IIEF-EF scale score <16. The distribution of the IPSS scores (N = 574) ranged from 0 to 34 (mean 6.9, median 5.0). Out of the 574 participants, 35 (6.1%) had severe LUTS (IPSS ⱖ 19). The CPSI pain scale scores ranged from 0 to 17 (2.2, 1). Pain/discomfort on ejaculation was reported by 183 (36.6%) and subjectively decreased ejaculate J Sex Med 2008;5:2141–2151
Bhojani et al. Table 3 Descriptive characteristics of the study cohort (N = 590) Variable
N (%)
Total number of patients (%) Age (years) Mean (median) Range BMI (kg/m2) Mean (median) Range WHO BMI (kg/m2) ⱕ24.9 (non-obese) 25–29.9 (overweight) 30–34.9 (obese) ⱖ35 (severely obese) IIEF-EF scale score (N = 574) Mean (median) Range Potent to moderate ED (ⱖ16) Moderate to severe ED (<16) Mean BMI (kg/m2) in participants with: Potent to moderate ED (ⱖ16) Moderate to severe ED (<16) IPSS score (N = 575) Mean (median) Range Mild to moderate LUTS (score < 19) Severe LUTS (score ⱖ 19) Mean BMI (kg/m2) in participants with: Mild to moderate LUTS (score < 19) Severe LUTS (score ⱖ 19) CPSI pain scale score (N = 500) Mean (median) Range CPSI pain scale score >1 (median) CPSI pain scale score ⱕ1 (median) Mean BMI (kg/m2) in participants with: Pain scale score >1 (median) Pain scale score ⱕ1 (median) EjD (N = 590) Pain/discomfort during ejaculation Low ejaculate volume Mean BMI (kg/m2) in participants with: Pain/discomfort during ejaculation Normal ejaculation Low ejaculate volume Normal ejaculate volume
590 (100.0) 54.1 (54.0) 30–83 26.6 (26.2) 19.1–48.8 209 296 74 11
(35.4) (50.2) (12.5) (1.9)
23.7 (28.0) 1–30 463 (80.7) 111 (19.3) 26.4 26.8 6.9 (5.0) 0–34 540 (93.9) 35 (6.1) 26.3 27.4 2.2 (1.0) 0–17 183 (36.6) 317 (63.4) 26.0 26.7 28 (4.7) 156 (26.4) 26.0 26.6 27.1 26.4
WHO = World Health Organization; BMI = body mass index; ED = erectile dysfunction; IIEF = International Index of Erectile Function; EF = erectile function; LUTS = lower urinary tract symptoms; EjD = ejaculatory dysfunction; IPSS = international prostate symptom score; CPSI = Chronic Prostatitis Symptom Index.
volume was recorded in 156 (26.4%). The distributions of continuously coded and categorized BMI are shown in Table 3. The distribution of categorized BMI according to the five examined categorical end points is shown in Table 4.
Effect of Continuously Coded BMI Table 1 shows the linear regression analyses that examine the effect of continuously coded BMI on three assessable end points that were coded on a continuous scale, namely, IIEF, IPSS, and CPSI.
7 (8.2%)
8 (2.7%)
ⱖ30
25–29.9
WHO = World Health Organization; BMI = body mass index; IIEF = International Index of Erectile Function; EF = erectile function; ED = erectile dysfunction; EjD = ejaculatory dysfunction; LUTS = lower urinary tract symptoms; IPSS = international prostate symptom score; CPSI = Chronic Prostatitis Symptom Index.
49 (23.4%) 73 (24.7%) 34 (40.0%) 160 (76.5%) 223 (75.3%) 51 (60.0%) <24.9
206 (93.6%) 264 (95.6%) 70 (88.6%) 43 (20.1%) 43 (15.1%) 25 (32.5%) 170 (79.8%) 241 (84.8%) 52 (67.5%)
14 (6.4%) 12 (4.3%) 9 (11.4%)
196 (93.8%) 288 (97.3%) 78 (91.8%) 91 (48.1%) 68 (28.2%) 24 (34.3%) 98 (51.8%) 173 (71.8%) 46 (65.7%)
13 (6.2%)
Normal ejaculate volume Normal ejaculation Higher than median (1)
Pain/discomfort on ejaculation
EjD (N = 590) CPSI pain scale score (N = 500)
Median and below (1) Moderate to severe ED (<16)
Severe LUTS (ⱖ19)
IPSS score (N = 575)
Mild to moderate LUTS (<19)
IIEF-EF scale score (N = 574)
Potent to moderate ED (ⱖ16) BMI (kg/m2)
Table 4
Tabulated data according to the WHO BMI categories and symptom severity
EjD (N = 590)
Decreased ejaculate volume
Body Mass Index and Its Association with Genitourinary Disorders
2145
The analyses demonstrated that elevated BMI is only statistically significantly related to IIEFEF scale score, as evidenced by an age-adjusted regression coefficient of 0.45 (P < 0.001). This relationship implies that for every one unit increase in BMI, the IIEF score decreased by 0.45 points. Conversely, BMI was unrelated to the IPSS score (regression coefficient: 0.02, P = 0.8) and CPSI pain scale score (regression coefficient: –0.04, P = 0.2). The effect of continuously coded BMI on the prevalence of the two binary coded end points, namely, pain/discomfort on ejaculation and on the prevalence of subjectively decreased ejaculate volume was examined in binary logistic regression models. After age adjustment, elevated BMI was statistically significantly related to lower rate of reported subjectively decreased ejaculate volume (odds ratio: 0.93, P = 0.02). This relationship implies that for every one unit increase in BMI, men were 7% less likely to report subjectively decreased ejaculate volume. Conversely, no statistically significant relationship was recorded between continuously coded BMI and pain/ discomfort on ejaculation (odds ratio: 0.91, P = 0.09).
Effect of Ccategorically Coded BMI Table 2 shows the effect of categorically coded BMI on the five examined end points. BMI was dichotomized between non-obese (BMI < 24.9 kg/ m2), overweight (BMI 25–29.9 kg/m2), and obese (BMI ⱖ 30 kg/m2). Non-obese (BMI < 24.9 kg/ m2) represented the reference category in data analyses. Since only 11 men (1.9%) were severely obese (BMI ⱖ 35 kg/m2), this category was combined with the obese (BMI ⱖ 30 kg/m2) category. In logistic regression analyses, overweight individuals were 1.9 times more likely to report moderate to severe ED (P = 0.01). Conversely, overweight individuals were 58% less likely to report CPP in excess of the median (P < 0.001), 64% were less likely to report pain/discomfort on ejaculation (P = 0.03), and 38% were less likely to report subjectively decreased ejaculate volume (P = 0.05). No statistically significant effect was seen between BMI ⱖ 30 kg/m2 and any of the five end points. Finally, categorized BMI was unrelated to IPSS score ⱖ19 (odds ratio: 0.6, P = 0.2). Discussion
Obesity represents one of the key features of the metabolic syndrome and is associated with J Sex Med 2008;5:2141–2151
2146 elevated levels of several proinflammatory cytokines, which contribute to a number of pathologies [26,27]. From the urologic perspective, two pathologies, namely, ED and LUTS, were identified as possible urological manifestations of the metabolic syndrome and indirectly of obesity [6,28]. Two recent reports demonstrated an association between the metabolic syndrome and ED [28,29]. In 2004 Gunduz et al. reported on 79 cardiology clinic outpatients with coronary artery disease and dyslipidemia. ED was diagnosed in 59 (74.7%) and all patients with metabolic syndrome (38) had ED [28]. Several cross-sectional studies demonstrated a link between obesity and ED [7,8], sedentary lifestyle [30], and smoking [7]. Kupelian et al. addressed the effect of BMI in the Massachusetts Male Aging Study, a population-based prospective cohort, where ED was assessed at three points during a 15-year period. In a low-risk subgroup of men with low BMI values (<25), ED was associated with a twofold higher risk of subsequent diagnosis of the metabolic syndrome (adjusted relative risk 2.09) [29]. The authors concluded that ED may represent an early warning sign in men otherwise considered at lower risk (low BMI) for the metabolic syndrome. In consequence, ED was proposed as a precursor of cardiovascular disease. Mueller et al. performed a detailed review of the association between cardiovascular disease, the metabolic syndrome, and ED. They demonstrated that endothelial damage represents the common pathway in cardiovascular disease, in the metabolic syndrome, and in ED [6]. Proinflammatory cytokines, such as interleukin-6 and 8 represent risk factors for endothelial dysfunction and possibly contribute to ED, as well as to other male disorders [31]. These may include LUTS, prostatitisrelated CPP, and EjD. Indeed, Rohrmann et al. demonstrated that obesity in young adulthood is associated with LUTS [32]. Their findings were derived from an observational longitudinal cohort of 2797 men. Interestingly, current BMI was unrelated to LUTS or prostate surgery for LUTS. Therefore, it does not appear to be a consensus regarding the role of BMI and LUTS. Interestingly, LUTS and ED frequently coexist [19,33–35]. The association between LUTS and ED was reported by MartinMorales et al. in a cross-sectional study of 2,476 Spanish men. In this study, the presence of LUTS was the strongest risk factor of ED (age-adjusted odds ratio of 2.67) [36]. Similarly, in 2003, Braun et al. reported a strong association between LUTS J Sex Med 2008;5:2141–2151
Bhojani et al. and ED in a German cohort of 4489 men [34]. The coexistence of ED and LUTS may indicate that BMI is implemented in LUTS pathogenesis [17,19,37,38]. Despite the interest in the link between obesity and ED and/or LUTS, few investigators addressed the association between obesity and other male pelvic disorders, such as EjD and/or CPP. Since male pelvic disorders are interrelated, EjD and/or CPP may also be affected by BMI distribution. For example, the prevalence of pain or discomfort during ejaculation occurs in 5–19% of men with LUTS [35,39–41] and in as many as 11% of men undergoing prostate cancer screening. However, to the best of our knowledge, few studies directly addressed the relationship between EjD and obesity [7]. Moreover, we found no English language medical literature addressing the link between obesity and CPP. Based on the potential importance of obesity in the pathogenesis of various male pathologies, we decided to systematically examine the association between obesity, defined as BMI ⱖ 30 kg/m2 and self-reported prevalence of ED, LUTS, CPP, and EjD. The effect of obesity may be assessed in two ways. BMI is commonly assessed in categories, according to the WHO definition. However, BMI, like weight and height, represents a continuously coded variable. Therefore, the effect of categorization may introduce a bias. The latter may either exaggerate or weaken the true effect of the continuously coded BMI [18]. Therefore it is imperative to examine the effect of BMI without converting this variable into strata. However, since many studies rely on categorically coded BMI, it is also imperative to examine the effect of categorically coded BMI, to allow comparisons with previous reports. To comply with these conditions, we performed two types of analyses. One relied on continuously coded BMI. The other relied on the WHO BMI categories. Since some of the end points of interest are also coded in continuous format, we applied the same logic to IIEF, IPSS, and CPSI scores. These three end points were first examined as continuous variables. Then, in order to be compatible with other studies that categorize these scale scores, we also categorized between generally accepted levels of ED, LUTS, and CPP [22,42,43]. In continuously coded BMI analyses, increasing BMI was related to lower IIEF-EF scale scores (P < 0.001). The regression coefficient of 0.45 indicates that for each one unit BMI increase, the IIEF scale score decreased by 0.45/30 units. This
Body Mass Index and Its Association with Genitourinary Disorders implies that after age adjustment a patient with a BMI of 29 would be expected to have a 3.6 unit (12%) lower IIEF scale score than his counterpart with a BMI of 21. The same effect was recorded in categorical BMI analyses, where overweight men (BMI 25–29.9 kg/m2) were nearly twice more likely (odds ratio: 1.9, P = 0.01) to have moderate to severe ED than their normal weight counterparts (BMI 24.9 or less). The rate of ED for obese men (BMI ⱖ 30 kg/m2) was not statistically significantly different from their non-obese counterparts (BMI < 24.9 kg/m2, [P = 0.3]), despite an odds ratio of 1.47. This finding may be related to lack of power, since the magnitude of the effect was similar to that recorded for overweight men. Taken together our data indicate that elevated BMI has a detrimental effect on EF. This finding is consistent with previous BMI vs. ED data, reported by Blanker et al. [7], Aytac et al. [8], and Walczak et al. [9], and with the general consensus that the metabolic syndrome predisposes to ED [6,44–46]. These findings are important—BMI may decrease the quality of life of men by virtue of predisposing to ED [13,47,48]. More importantly, elevated BMI is a precursor of several pathologies, including coronary artery disease [49]. Prevention of obesity may therefore not only reduce its urologic manifestations, but hopefully also other even more important systemic signs and symptoms. Our data addressing the relationship between both continuously coded or categorized BMI and the IPSS score failed to reveal a statistically significant association. Our data are consistent with several reports that demonstrated no association between BMI and LUTS or benign prostatic hyperplasia. For example a cross-sectional Austrian population-based study (N = 939) reported no association between BMI and IPSS [50]. A Turkish study (N = 286) found no relationship between serum lipids and IPSS [51]. Finally, a large-scale population-based study (N = 2797) from the united States, demonstrated that current BMI is unrelated to LUTS or to the history of prostatic surgery for LUTS [32]. It is noteworthy that in the same study, the authors also reported that elevated BMI in young adulthood predisposes to LUTS [32]. This is not the first study to report that a history of elevated BMI is a precursor of clinically relevant diseases many years later [52]. Therefore, BMI at specific time points may be more relevant than other circumstances. It is however noteworthy that several other studies suggested that BMI is associated with LUTS. For
2147
example, Hammarsten et al. (N = 158) demonstrated that obese men, defined according to elevated BMI, had larger prostates than their non-obese counterparts [15]. Ozden et al. (N = 78) demonstrated that men with biochemical profiles indicative of metabolic syndrome had enlarged prostates [16]. Paick et al. showed that metabolic profiles are related to IPSS data [18]. Finally, Giovannucci et al. demonstrated that abdominal obesity was related to the likelihood of prostate surgery for LUTS and among those that did not have prostate surgery for LUTS, abdominal obesity was associated with more frequent urinary symptoms [53]. Interestingly, BMI was unrelated to either outcome, which indicates that different definitions of obesity or different fat distributions may be related to different clinical manifestations. We identified no English language reports addressing the effect of BMI on CPP or EjD. Interestingly, our categorical analyses of BMI revealed that men with BMI 25–29.9 were 58% less likely to report CPSI pain scale scores above the median. This indicates that obesity, expressed as BMI, is related to lower rate of CPP. It is noteworthy that the analyses of continuously coded BMI on CPSI scale score failed to reveal any statistically significant relationship (P = 0.2). Therefore, the presence of a true effect between BMI and CPSI scale score should be interpreted with caution, since analyses of categorically coded data can exaggerate an existing relationship by introducing artificial cutoffs [18]. Similarly to CPSI scale score analyses, we found no effect between continuously coded BMI and pain/discomfort at ejaculation. However, the effect of categorized BMI revealed a 64% decrease in the rate of pain/discomfort at ejaculation (P = 0.03). This statistical significance of this effect needs to be interpreted with caution, as data categorization may introduce type I error and may artificially inflate the existing association. Therefore, further analyses of this association are certainly needed as the link between BMI and pain/discomfort at ejaculation has not received much attention previously. Elevated BMI was related to lower rate of self-reported subjectively decreased ejaculatory volume in categorically and continuously coded analyses. When coded as a continuous variable, every one unit BMI increase resulted in a 7% decrease in the rate of self-reported subjectively decreased ejaculate volume (P = 0.02). In categorized BMI analyses, overweight men were 38% J Sex Med 2008;5:2141–2151
2148 less likely to report subjectively decreased ejaculatory volume (P = 0.05). Therefore, elevated BMI appears to be associated with a lesser extent of ejaculatory dysfunction, at least in the current population. Several explanations may be proposed for what intuitively might appear as paradoxical. For example, it is possible that obese men are less affected by subjectively decreased ejaculate volume and tend to underreport it. Alternatively, it is also possible that indeed fewer obese men suffer of subjectively decreased ejaculate volume, especially that the protective association between BMI and subjectively decreased ejaculate was recorded in both analytic approaches (continuous and categorical analyses). No other study addressed the relationship between ejaculate volume and BMI. Therefore, our data cannot be compared with other studies. The same hypothesis might be proposed for CPP. However, the epidemiologic nature of the current study prohibits any valid attempts at elucidating the effect and mechanism of elevated BMI on either CPP or ED. This study has several strengths. First, it examined the prevalence of five genitourinary disorders and tested for presence and the strength of associations with BMI. Second, the study relied on two distinct statistical approaches. One uses categorical data and the other is based on continuously coded variables. This approach minimizes bias related to data transformations. Third, the study relied on a sample of healthy, populationdwelling men. Such individuals provide substantially more realistic estimates of the prevalence of urologic symptoms in the community, than hospital or clinic-based samples. Fourth, the study simultaneously examined several end points within the same population. This allows valid comparison of the coprevalence of the examined urologic disorders and of their suspected association with BMI. The present study also has several limitations. The sample size was relatively small. Moreover, BMI was not compared with other measures of overweightedness such as lean body mass and waist-to-hip ratio. Moreover, it is noteworthy that BMI and possibly other measures of overweightedness may not detect special patterns of lean body mass (muscle) distribution. For example men with elevated muscle mass may at times be quantified as overweight or obese if their physique is not specifically accounted for. The cross-sectional nature of our study only allows assessing the association between BMI and J Sex Med 2008;5:2141–2151
Bhojani et al. the targeted genitourinary disorders. Causality cannot be inferred. Another limitation of our study is that we did not have information on the comorbidities of our study population. This information may explain the high rate of ED in our study, as obesity is related to diabetes and consequently to ED. The questionnaire formats that were used also limited the amount of detail that can be extracted and interpreted. For example, the recorded prevalence of subjectively decreased ejaculate volume and/or the presence of pain/discomfort on ejaculation may be affected by the dichotomous type of input that was allowed (yes/no). It is plausible that a much finer definition of subjectively decreased ejaculatory volume or of pain/discomfort on ejaculation could have been established, if more answer choices were available. It is also important to notice that the logistic regression analysis, in its binary format, only allows two possible outcomes, which also may further reduce the variability of possible outcomes. This spectrum bias can be avoided in linear regression models, provided that continuously coded outcome is available. Finally, our population overwhelmingly consisted of white men. In consequence, race adjustment was not performed. Despite these limitations, our findings are noteworthy and deserve future attention.
Conclusions
Our findings indicate that elevated BMI is related to worse EF, which is consistent throughout the literature [7–9,30]. Moreover, it appears that current BMI does not exert a central role on IPSS score. Finally, elevated BMI may exert a protective effect on ejaculatory dysfunction and on associated pain, as well as on CPP. Our findings require corroboration, before a protective effect can be assigned between elevated BMI and CPP or ejaculatory dysfunction. However, it is highly noteworthy that in this relatively large cohort of patients, BMI failed to exert a uniformly adverse effect on several male pelvic pathologies. Corresponding Author: Pierre Karakiewicz, MD, Cancer Prognostics and Health Outcomes unit, University of Montreal Health Center, Campus St-Luc 1058, rue St-Denis, Montreal, Quebec, Canada H2X 3J4. Tel: 514 890 8000 ext. 35336; Fax: 514 227 5103; E-mail: [email protected] Conflict of Interest: None declared.
Body Mass Index and Its Association with Genitourinary Disorders Statement of Authorship
Category 1 (a) Conception and Design Pierre I. Karakiewicz; Georg Hutterer; Francesco Montorsi; Naeem Bhojani; Nazareno Suardi; Shahrokh Shariat (b) Acquisition of Data Pierre I. Karakiewicz; Claudio Jeldres; Hugues Widmer; Luc Valiquette; Paul Perrotte; Philippe Arjane; Francois Benard; Francois Peloquin (c) Analysis and Interpretation of Data Pierre I. Karakiewicz; Naeem Bhojani, Umberto Capitanio
Category 2 (a) Drafting the Article Naeem Bhojani; Claudio Jeldres; Pierre Karakiewicz, Umberto Capitanio (b) Revising It for Intellectual Content Pierre I. Karakiewicz, Umberto Capitanio
I.
Category 3 (a) Final Approval of the Completed Article Pierre I. Karakiewicz; Naeem Bhojani
References
1 Ogden CL, Carroll MD, Curtin LR, McDowell MA, Tabak CJ, Flegal KM. Prevalence of overweight and obesity in the united States, 1999–2004. JAMA 2006;295:1549–55. 2 Fontaine KR, Heo M, Allison DB. Obesity and prostate cancer screening in the USA. Public Health 2005;119:694–8. 3 Seidell JC. Obesity in Europe: Scaling an epidemic. Int J Obes Relat Metab Disord 1995;19(3 suppl): S1–4. 4 Pietinen P, Vartiainen E, Mannisto S. Trends in body mass index and obesity among adults in Finland from 1972 to 1992. Int J Obes Relat Metab Disord 1996;20:114–20. 5 Eckel RH, Grundy SM, Zimmet PZ. The metabolic syndrome. Lancet 2005;365:1415–28. 6 Muller A, Mulhall JP. Cardiovascular disease, metabolic syndrome and erectile dysfunction. Curr Opin Urol 2006;16:435–43. 7 Blanker MH, Bohnen AM, Groeneveld FP, Bernsen RM, Prins A, Thomas S, Bosch JL. Correlates for erectile and ejaculatory dysfunction in older Dutch men: A community-based study. J Am Geriatr Soc 2001;49:436–42. 8 Aytac IA, Araujo AB, Johannes CB, Kleinman KP, McKinlay JB. Socioeconomic factors and incidence of erectile dysfunction: Findings of the longitudinal Massachussetts Male Aging Study. Soc Sci Med 2000;51:771–8.
2149
9 Walczak MK, Lokhandwala N, Hodge MB, Guay AT. Prevalence of cardiovascular risk factors in erectile dysfunction. J Gend Specif Med 2002;5:19–24. 10 Ahn TY, Park JK, Lee SW, Hong JH, Park NC, Kim JJ, Park K, Park H, Hyun JS. Prevalence and risk factors for erectile dysfunction in Korean men: Results of an epidemiological study. J Sex Med 2007;4:1269–76. 11 Mulhall J, Teloken P, Brock G, Kim E. Obesity, dyslipidemias and erectile dysfunction: A report of a subcommittee of the sexual medicine society of North America. J Sex Med 2006;3:778–86. 12 El-Sakka AI, Hassoba HM, Sayed HM, Tayeb KA. Pattern of endocrinal changes in patients with sexual dysfunction. J Sex Med 2005;2:551–8. 13 Zohdy W, Kamal EE, Ibrahim Y. Androgen deficiency and abnormal penile duplex parameters in obese men with erectile dysfunction. J Sex Med 2007;4:797–808. 14 Wingard C, Fulton D, Husain S. Altered penile vascular reactivity and erection in the Zucker obesediabetic rat. J Sex Med 2007;4:348–62 (discussion 62–3). 15 Hammarsten J, Hogstedt B, Holthuis N, Mellstrom D. Components of the metabolic syndrome-risk factors for the development of benign prostatic hyperplasia. Prostate Cancer Prostatic Dis 1998;1: 157–62. 16 Ozden C, Ozdal OL, Urgancioglu G, Koyuncu H, Gokkaya S, Memis A. The correlation between metabolic syndrome and prostatic growth in patients with benign prostatic hyperplasia. Eur Urol 2007;51:199–203 (discussion 4–6). 17 Kasturi S, Russell S, McVary KT. Metabolic syndrome and lower urinary tract symptoms secondary to benign prostatic hyperplasia. Curr Urol Rep 2006;7:288–92. 18 Paick JS, Yang JH, Kim SW, Ku JH. Are age, anthropometry and components of metabolic syndrome-risk factors interrelated with lower urinary tract symptoms in patients with erectile dysfunction? A prospective study. Asian J Androl 2007; 9:213–20. 19 McVary K. Lower urinary tract symptoms and sexual dysfunction: Epidemiology and pathophysiology. BJU Int 2006;97(2 suppl):23–8 (discussion 44–5). 20 Guay A, Jacobson J. The relationship between testosterone levels, the metabolic syndrome (by two criteria), and insulin resistance in a population of men with organic erectile dysfunction. J Sex Med 2007;4:1046–55. 21 Corona G, Mannucci E, Petrone L, Balercia G, Paggi F, Fisher AD, Lotti F, Chiarini V, Fedele D, Forti G, Maggi M. NCEP-ATPIII-defined metabolic syndrome, type 2 diabetes mellitus, and prevalence of hypogonadism in male patients with sexual dysfunction. J Sex Med 2007;4:1038– 45. J Sex Med 2008;5:2141–2151
2150 22 Cappelleri JC, Rosen RC, Smith MD, Mishra A, Osterloh IH. Diagnostic evaluation of the erectile function domain of the International Index of Erectile Function. Urology 1999;54:346–51. 23 Barry MJ, Fowler FJ Jr., O’Leary, MP, Bruskewitz RC, Holtgrewe HL, Mebust WK, Cockett AT. The American Urological Association symptom index for benign prostatic hyperplasia. The Measurement Committee of the American Urological Association. J Urol 1992;148:1549–57 (discussion 64). 24 Karakiewicz PI, Perrotte P, Valiquette L, Benard F, McCormack M, Menard C, McNaughton Collins M, Nickel JC. French-Canadian linguistic validation of the NIH Chronic Prostatitis Symptom Index. Can J Urol 2005;12:2816–23. 25 Hald T, Nordling J, Andersen JT, Bilde T, Meyhoff HH, Walter S. A patient weighted symptom score system in the evaluation of uncomplicated benign prostatic hyperplasia. Scand J Urol Nephrol Suppl 1991;138:59–62. 26 Yudkin JS, Stehouwer CD, Emeis JJ, Coppack SW. C-reactive protein in healthy subjects: Associations with obesity, insulin resistance, and endothelial dysfunction: A potential role for cytokines originating from adipose tissue? Arterioscler Thromb Vasc Biol 1999;19:972–8. 27 Bastard JP, Jardel C, Bruckert E, Blondy P, Capeau J, Laville M, Vidal H, Hainque B. Elevated levels of interleukin 6 are reduced in serum and subcutaneous adipose tissue of obese women after weight loss. J Clin Endocrinol Metab 2000;85: 3338–42. 28 Gunduz MI, Gumus BH, Sekuri C. Relationship between metabolic syndrome and erectile dysfunction. Asian J Androl 2004;6:355–8. 29 Kupelian V, Shabsigh R, Araujo AB, O’Donnell AB, McKinlay JB. Erectile dysfunction as a predictor of the metabolic syndrome in aging men: Results from the Massachusetts Male Aging Study. J Urol 2006; 176:222–6. 30 Moreira ED Jr, Lisboa Lobo CF, Villa M, Nicolosi A, Glasser DB. Prevalence and correlates of erectile dysfunction in Salvador, northeastern Brazil: A population-based study. Int J Impot Res 2002;14(2 suppl):S3–9. 31 Ziccardi P, Nappo F, Giugliano G, Esposito K, Marfella R, Cioffi M, D’Andrea F, Molinari AM, Giugliano D. Reduction of inflammatory cytokine concentrations and improvement of endothelial functions in obese women after weight loss over one year. Circulation 2002;105:804–9. 32 Rohrmann S, Smit E, Giovannucci E, Platz EA. Associations of obesity with lower urinary tract symptoms and noncancer prostate surgery in the Third National Health and Nutrition Examination Survey. Am J Epidemiol 2004;159:390–7. 33 El-Sakka AI. Lower urinary tract symptoms in patients with erectile dysfunction: Analysis of risk factors. J Sex Med 2006;3:144–9. J Sex Med 2008;5:2141–2151
Bhojani et al. 34 Braun MH, Sommer F, Haupt G, Mathers MJ, Reifenrath B, Engelmann UH. Lower urinary tract symptoms and erectile dysfunction: Co-morbidity or typical “Aging Male” symptoms? Results of the “Cologne Male Survey”. Eur Urol 2003;44:588–94. 35 Frankel SJ, Donovan JL, Peters TI, Abrams P, Dabhoiwala NF, Osawa D, Lin AT. Sexual dysfunction in men with lower urinary tract symptoms. J Clin Epidemiol 1998;51:677–85. 36 Martin-Morales A, Sanchez-Cruz JJ, Saenz de Tejada I, Rodriguez-Vela L, Jimenez-Cruz JF, Burgos-Rodriguez R. Prevalence and independent risk factors for erectile dysfunction in Spain: Results of the Epidemiologia de la Disfuncion Erectil Masculina Study. J Urol 2001;166:569–74 (discussion 74–5). 37 Russell S, McVary KT. Lower urinary tract symptoms and erectile dysfunction: Epidemiology and treatment in the aging man. Curr Urol Rep 2005; 6:445–53. 38 McVary KT. BPH: Epidemiology and comorbidities. Am J Manag Care 2006;12:S122–8. 39 Nickel JC, Elhilali M, Vallancien G. Benign prostatic hyperplasia (BPH) and prostatitis: Prevalence of painful ejaculation in men with clinical BPH. BJU Int 2005;95:571–4. 40 Rosen R, Altwein J, Boyle P, Kirby RS, Lukacs B, Meuleman E, O’Leary MP, Puppo P, Robertson C, Giuliano F. Lower urinary tract symptoms and male sexual dysfunction: the multinational survey of the aging male (MSAM-7). Eur Urol 2003;44:637– 49. 41 Walz J, Perrotte P, Gallina A, Benard F, Valiquette L, McCormack M, Montorsi F, Karakiewicz PI. Ejaculatory disorders may affect screening for prostate cancer. J Urol 2007;178:232–7 (discussion 7– 8). 42 Litwin MS, McNaughton-Collins M, Fowler FJ Jr., Nickel, JC, Calhoun EA, Pontari MA, Alexander RB, Farrar JT, O’Leary, MP. The National Institutes of Health chronic prostatitis symptom index: development and validation of a new outcome measure Chronic Prostatitis Collaborative Research Network. J Urol 1999;162:369–75. 43 van Haarst EP, Heldeweg EA, Newling DW, Schlatmann TJ. A cross-sectional study of the international prostate symptom scores related to age and gender in Dutch adults reporting no voiding complaints. Eur Urol 2005;47:334–9. 44 Roumeguere T, Wespes E, Carpentier Y, Hoffmann P, Schulman CC. Erectile dysfunction is associated with a high prevalence of hyperlipidemia and coronary heart disease risk. Eur Urol 2003;44:355–9. 45 Bal K, Oder M, Sahin AS, Karatas CT, Demir O, Can E, Gumus BH, Ozer K, Sahin O, Esen AA. Prevalence of metabolic syndrome and its association with erectile dysfunction among urologic patients: Metabolic backgrounds of erectile dysfunction. Urology 2007;69:356–60.
Body Mass Index and Its Association with Genitourinary Disorders 46 Vlachopoulos C, Rokkas K, Ioakeimidis N, Stefanadis C. Inflammation, metabolic syndrome, erectile dysfunction, coronary artery disease: Common links. Eur Urol 2007 Dec;52(6):1590–600. 47 Haro JM, Beardsworth A, Casariego J, Gavart S, Hatzichristou D, Martin-Morales A, Schmitt H, Mirone V, Needs N, Riley A, Varanese L, von Keitz A, Kontodimas S. Treatment-seeking behavior of erectile dysfunction patients in Europe: Results of the Erectile Dysfunction Observational Study. J Sex Med 2006;3:530–40. 48 Kratzik CW, Schatzl G, Lunglmayr G, Rucklinger E, Huber J. The impact of age, body mass index and testosterone on erectile dysfunction. J Urol 2005; 174:240–3. 49 Grover SA, Lowensteyn I, Kaouache M, Marchand S, Coupal L, DeCarolis E, Zoccoli J, Defoy I. The prevalence of erectile dysfunction in the primary care setting: importance of risk factors for diabetes and
50
51 52
53
2151
vascular disease. Arch Intern Med 2006;166:213– 9. Haidinger G, Madersbacher S, Waldhoer T, Lunglmayr G, Vutuc C. The prevalence of lower urinary tract symptoms in Austrian males and associations with sociodemographic variables. Eur J Epidemiol 1999;15:717–22. Lekili M, Muezzinoglu T, Uyanik BS, Buyuksu C. Serum lipid levels in benign prostatic hyperplasia. World J Urol 2006;24:210–3. Morrison JA, Friedman LA, Gray-McGuire C. Metabolic syndrome in childhood predicts adult cardiovascular disease 25 years later: The Princeton Lipid Research Clinics Follow-up Study. Pediatrics 2007;120:340–5. Giovannucci E, Rimm EB, Chute CG, Kawachi I, Colditz GA, Stampfer MJ, Willett WC. Obesity and benign prostatic hyperplasia. Am J Epidemiol 1994;140:989–1002.
J Sex Med 2008;5:2141–2151