Physical Activity and PDE5 Inhibitors in the Treatment of ErectileDysfunction: Results of a Randomized Controlled Studyjsm_1783 2201..2208

Giuseppe Maio, MD,* Salim Saraeb, MD,† and Antonio Marchiori, MD*

*Policlinico Abano Terme, Andrological Unit, Padova, Italy; †S. Antonio Hospital, Padova, Italy

DOI: 10.1111/j.1743-6109.2010.01783.x

A B S T R A C T

Introduction. Physical activity (PhA) has proven to be a protective factor for normal erectile function in numerousepidemiological studies.Aim. The aim of this study was to establish if PhA could have a therapeutic role in the treatment of erectiledysfunction (ED).Methods. This was a randomized, open-label study. A total of 60 patients complaining of ED were studied. Patientswere assessed at baseline and after 3 months of study treatment. At baseline, patients were randomized to receivephosphodiesterase type 5 inhibitor (PDE5i) alone (group A) or PDE5i plus regular (�3 hours/week), aerobic,non-agonistic PhA (group B).Main Outcome Measures. All subjects completed the International Index of Erectile Function (IIEF-15) question-naire and performed total testosterone (TT).Results. Mean PhA was 3.4 hours/week in group B vs. 0.43 in group A; mean energy expenditure in group B was1,868 kcal/ week or 22.8 metabolic equivalent (MET)/week. IIEF restoration of ED occurred in 77.8% (interventiongroup) vs. 39.3% (control) (P < 0.004). The IIEF-15 score resulted in statistical improvement in intervention groupin all the domains but one (orgasm): erectile function 24.7 vs. 26.8 (P = 0.003); confidence (Q15) 3.53 vs. 4.07(P = 0.006); sexual desire 6.46 vs. 7.18 (P = 0.028); intercourse satisfaction 9.85 vs. 11.25 (P = 0.001); total satisfaction7.17 vs. 8.07 (P = 0.009); total score 56.2 vs. 61.07 (P = 0.007). TT was statistically similar in the two groups; separateanalysis in each group showed statistical increase in group B 4.24 vs. 4.55 (P = 0.012). At multivariate logisticregression analysis, PhA was the only independent variable for normal erection (P = 0.010) (95% confidence interval[CI] 0.036–0.643), higher sexual satisfaction (P = 0.022) (95% CI 0.084–0.821) and normal total IIEF-15 score(P = 0.023) (95% CI 0.85–0.837).Conclusion. In this randomized controlled pilot study, PDE5i plus PhA was more effective than PDE5i alone in thetreatment of ED. Maio G, Saraeb S, and Marchiori A. Physical activity and PDE5 inhibitors in the treatmentof erectile dysfunction: Results of a randomized controlled study. J Sex Med 2010;7:2201–2208.

Key Words. Physical Activity; Sport; Exercise; Fitness; Treatment; Therapy; Erectile Dysfunction; PDE5 Inhibitors;Sildenafil; Tadalafil; Vardenafil

Introduction

N umerous studies have reported the impor-tance of physical activity (PhA) in the pre-

vention of cardiovascular diseases [1–3]. TheNational Institutes of Health Consensus Develop-ment Panel on Physical Activity and Cardiovascu-lar Health recommends that all Americans shouldengage in regular PhA [4].

Regular exercise has also proven to be beneficialon erectile function in several epidemiologicalstudies [5–9]. Data from the Massachusetts MaleAging Study, including a sample of 1,156 men aged40–70 years followed for approximately 8.8 years,showed that the lowest risk for erectile dysfunction(ED) was among men physically active at bothbaseline and follow-up compared to those whowere sedentary (<200 kcal/day of PhA), probability

2201

© 2010 International Society for Sexual Medicine J Sex Med 2010;7:2201–2208

for ED (95% confidence interval [CI]) was 14% vs.27% (P = 0.013) [5].

Kratzik et al. in a cross-sectional study on 674men, aged 45–60 years, at their place of workreported a positive correlation between Interna-tional Index of Erectile Function (IIEF)-5 andPhA from 1,000 to 4,000 kcal/week (r = 0.164;P < 0.001); the risk of severe ED was decreased by82.9% with at least 3,000 kcal/week energy expen-diture (EE) (odds ratio [OR] = 0.171, P = 0.018)[6].

In the Health Professionals Follow-up Study, acohort study on 22,086 men, PhA was associatedwith lower risk for ED; the multivariate relativerisk was 0.7 (95% CI 0.6–0.7) for >32.6 metabolicequivalent of exercise per week; interestingly allincluded exercise types, ranging from walking,jogging, running, cycling, tennis and squash dem-onstrated significant benefits [7].

In a population study on 2,412 men aged 40–70years, from four different countries, Nicolosi et al.reported that ED was inversely associated withPhA (OR = 0.5) [8].

A meta-analysis by Cheng et al. including sevencross-sectional studies revealed that the presenceof ED was negatively associated with PhA; thestudy reported a dose–response relationshipbetween ED and PhA, with higher PhA conferringlower risks for ED (OR = 1 for low activity,OR = 0.63 for moderate activity, and OR = 0.42for high activity) [9].

Aim

The aim of the present study was to establish ifPhA could have a therapeutic role in associationwith phosphodiesterase type 5 inhibitors (PDE5i)in the treatment of ED. As far as we know, nostudy has been published to date on the associationof PhA with the standard PDE5i treatment in sub-jects with ED.

Methods

PatientsMales aged 40–60 years affected by ED of anygrade, who where PDE5i naïve and did not take anysignificant PhA (less than 2 hours/week), wererecruited for the study and underwent preliminaryandrological evaluation (T0) which includeddetailed medical history, physical examination,fasting glucose, total and high-density lipoprotein(HDL) cholesterol, total testosterone (TT), and

PDE5i prescription. Exclusion criteria were: (i) EDsecondary to radical pelvic surgery due to actual orpotential nerve damage, spinal trauma, or hypogo-nadism; (ii) contraindications to PhA; (iii) patientrefusal to be physically active; (iv) no responders toPDE5i; (v) contraindications to PDE5i. BetweenJune 2007 and September 2008, 75 patients exam-ined at our center were assessed for eligibility. After6 weeks, in which patients took PDE5i, they werereevaluated (baseline evaluation) in order to selecteligible patients. At this evaluation, 15 patientswere excluded and 60 were accepted and conse-quently randomly assigned to either the interven-tion or control group using a computer-generatedrandom number sequence (Figure 1). This projectwas approved by an Institutional Review Board; allsubjects provided written informed consent for vol-untary, unpaid participation.

Participants were randomized to receive PDE5ialone (group A, control) or PDE5i plus regular(�3 hours/week) aerobic non-agonistic PhA(group B, intervention) for 3 months. Patients inthe intervention group were given detailed infor-mation on the importance of PhA in improvingvascular apparatus and possibly penile vasculariza-tion and ED in improving risk factors for ED suchas diabetes, hypertension and dyslipidemia; it wasalso carefully explained that good results dependedon maintaining regular PhA (three times or more aweek) and exercise for at least 3 hours a week.Each patient in this group received accurate infor-mation on the type and intensity of PhA (seebelow) and on the importance of calculating heartrate as a measure of PhA intensity; thorough

Figure 1 Consort flow of patients progress through the trial.PDE5 = phosphodiesterase type 5; PhA = physical activity.

2202 Maio et al.

J Sex Med 2010;7:2201–2208

instruction on how to calculate heartbeat was alsogiven. These patients had monthly interviews toreinforce the importance of maintaining regularexercise. Patients in the control group wereinformed on the beneficial effects of PhA and weregenerically advised to take PhA; they had an inter-view after 6 weeks. Patients from both groups weregiven written information on the type of PhA (seebelow reported characteristics), informed of theirown maximum heart rate and requested to recordand detail PhA (duration, type, intensity, registerheartbeats every 5–10 minutes) of any single activ-ity together with drug type and dose. The PDE5idrug was chosen by all the patients on the basis ofpersonal preference.

The PhA program followed, according to theAmerican Sport College Recommendations forindividuals who are quite unfit, was [10]: (i) fre-quency of training: 3–5 days a week; (ii) intensityof training: moderate, corresponding to 55–64%maximum heart rate; (iii) duration of training: �3hours a week, each session lasting 20–60 minutesof continuous or intermittent activity (minimumof 10-minute bouts accumulated throughout theday); (iv) type of activity: any aerobic activity:walking–hiking, running–jogging, cycling, cross-country skiing, aerobic exercises, swimming,skating, and various endurance game activities;with regard to cycling, a duration of no longerthan 60 minutes, an accurate padding of perineumand the use of a large saddle where possible wereadvised. EE was calculated on the basis of thestandards reported in the Compendium of PhysicalActivities, expressed in kcal/week and metabolicequivalent (MET)/week [11]. MET, metabolicequivalent, is defined as the ratio of work meta-bolic rate to a standard resting metabolic rate of1.0 kcal/kg/hour; 1 MET is the resting metabolicrate obtained during quiet sitting [11]; the advan-tage of expressing EE in MET over kcal is that theresult depends solely on the exercise and is notinfluenced by body weight.

Main Outcome Measures

Treatment OutcomesIIEF-15 [12] questionnaire was completed by allsubjects at baseline and at the end of study. TTmeasured by radioimmunoassay, expressed inng/mL, was assessed at baseline and at the end ofstudy. ED was assessed with IIEF-erectile functiondomain (IIEF-EF) (questions 1–5, 15), the severityof ED was grouped according to the score in: mild(17–25), moderate (11–16), severe (<11), no ED

(26–30); answers to question 15 (“Over the lastmonth, how do you rate your confidence that youcan get and keep your erection?”) were analyzedseparately to estimate self-confidence in erection.IIEF-15 specific domains were used to assess otheraspects of sexual function: (i) intercourse satisfac-tion (Q6–8); (ii) orgasmic function (Q9–10); (iii)sexual desire (Q11–12); (iv) overall satisfaction(Q13–14); for these domains of sexual function,before data analysis, we empirically considered asnormal the score obtained from the sum of the twohighest scores for each question: 12–15 for inter-course satisfaction, 8–10 for orgasmic function,sexual desire, overall satisfaction, 62–75 forIIEF-15 total score.

Statistical AnalysisStatistical analysis was performed using commer-cial software SPSS 15.0 (Chicago, IL, USA). Dataare presented as mean � standard deviation.Outcome data were analyzed for the intention-to-treat analysis.

Univariate AnalysisChi-square or Mann–Whitney to compare cat-egorical or discrete variables, respectively, andt-test for paired or independent samples tocompare continuous variables were used. Logisticregression for multivariable analysis was per-formed. P < 0.05 was considered statisticallysignificant.

Results

The total number of patients enrolled was 60: 30patients randomized to receive PDE5i alone (groupA) and 30 patients randomized to receive PDE5iplus PhA. After randomization, five subjects, whonever received the assigned treatment, wereexcluded from analysis: three from the interventiongroup who never performed the PhA program andtwo from the control group who refused the drugfrom the beginning of the study as it did not meetcouple expectations. The remaining 55 subjectscompleted their treatment program and follow-upand were included in the analysis. Patients’ baselinecharacteristics are reported in Table 1: age, bodymass index (BMI), cigarette smoking, hypercholes-terolemia (�240 mg/dL) were similar in bothgroups; hypertension was statistically more fre-quent in group A and diabetes in group B; meanPhA before therapy was similar in both groups(0.31 vs. 0.35 hours); no other significant comor-bidities were present.

Physical Activity in the Treatment of Erectile Dysfunction 2203

J Sex Med 2010;7:2201–2208

At baseline, scores of IIEF-15, total and specificdomains, degree of ED severity and TT were sta-tistically similar in the two groups (Table 1).

During the study period, mean PhA was0.43 � 0.61 (group A) vs. 3.40 � 0.43 hours/week(group B); duration, type and EE are reported inTable 2. Preferred PDE5i and drug intakeexpressed as number of doses were similar in bothgroups (Table 3); PDE5i therapy was very effectivein both groups.

At the end of treatment, IIEF-15 scoreimproved significantly in both groups (Tables 1

and 4; Figure 2). Comparison of the resultsbetween the two groups of subjects, group A andgroup B, respectively, showed (Table 4): erectilefunction 24.7 vs. 26.8 (P = 0.003); confidence(Q15) 3.53 vs. 4.07 (P = 0.006); sexual desire 6.46vs. 7.18 (P = 0.028); intercourse satisfaction 9.85vs. 11.25 (P = 0.001); total satisfaction 7.17 vs. 8.07(P = 0.009); IIEF-15 total score 56.2 vs. 61.7(P = 0.007). On the basis of IIEF-15 specificdomain (1–5, 15), a normal erection was obtainedin 39.3% in group A vs. 77.8% in group B(P < 0.004). TT at the end of treatment was statis-tically similar in the two groups (Table 5); whenanalyzing modifications of TT levels in eachgroup, statistical significant increase was foundexclusively in group B (4.24 vs. 4.55) (P = 0.012;

Table 1 Patients baseline characteristcs

Total population PDE5i alone PDE5i + PhA P

Age 50.23 � 6.57 50.32 � 6.95 50.14 � 6.28 nsBMI 27 � 2.87 27.1 � 3.42 26.8 � 2.22 nsCigarette smoking 41.81% 42.9% 40.7% nsTotal cholesterol

Normal 40.0% 42.9% 37.0% nsBorderline 38.2% 35.7% 40.7% nsHigh 21.8% 21.4% 22.2% ns

Hypertension 36.4% 50% 22.2% 0.032Overweight 58.2% 57.1% 59.3% nsObesity 16.4% 17.9% 14.8% nsDiabetes 18.2% 7.1% 29.6% 0.031Physical activity (hours/week) 0.33 � 0.39 0.31 � 0.40 0.35 � 0.38 nsErectile function* 15.7 � 4.15 15.5 � 4.18 15.8 � 4.19 nsED severity

Mild 43.6% 42.9% 44.1% nsModerate 40.0% 39.3% 40.7% nsSevere 16.4% 17.8% 14.8% ns

Confidence* 2.22 � 0.81 2.1 � 0.96 2.3 � 1.04 nsSexual desire* 5.04 � 1.21 4.9 � 1.13 5.2 � 1.30 nsOrgasmic function* 5.64 � 1.20 5.5 � 1.13 5.8 � 1.27 nsIntercourse satisfact* 8.24 � 1.73 7.9 � 1.71 8.5 � 1.71 nsOverall satisfact* 5.04 � 1.36 4.7 � 1.41 5.4 � 1.21 nsIIEF-15 total score 39.6 � 9.14 38.5 � 9.06 40.8 � 9.23 nsTotal testost (ng/mL) 4.35 � 0.94 4.45 � 0.93 4.24 � 0.95 nsResponders to PDE5i during screening period

Partial 65.5% 64.3% 66.7% nsTotal 34.5% 35.7% 33.3% ns

*IIEF-15 specific domain.PDE5i = phosphodiesterase type 5 inhibitor; PhA = physical activity; BMI = body mass index; ED = erectile dysfunction; IIEF = International Index of ErectileFunction.

Table 2 Physical activity

Group A Group B

Physical activity (hours/week) 0.43 � 0.61 3.40 � 0.43Energy Expenditure (kcal/week) 266 � 355 1,868 � 382Energy Expenditure (MET/week) 2.95 � 3.94 22.8 � 3.87Sport activities

Jogging—running 15% 22%Aerobic exercise 20% 16%Cycling 20% 15%Stationary cycling 5% 9%Treadmill walking-running 7%Brisk walking (>3.5 mph) 10% 7%Soccer 10% 6%Tennis 8% 6%Swimming 5%Other Activities 12% 7%

Table 3 Various PDE5 inhibitors in the study populationsubgroups

Totalpopulation

PDE5ialone

PDE5i +PhA P

Sildenafil 30.9% 32.1% 29.6% nsTadalafil 40% 39.3% 40.7% nsVardenafil 29.1% 28.6% 29.7% nsNumber of doses

(per month)5.12 � 1.34 5.17 � 1.44 5.07 � 1.24 ns

PDE5i = phosphodiesterase type 5 inhibitor; PhA = physical activity.

2204 Maio et al.

J Sex Med 2010;7:2201–2208

95% CI = 0.072–0.549). In group B, comparisonof the exercise intensity, expressed as EE both inkcal and MET, with IIEF-15 scores, did not showany statistical significant correlation.

Grouping scores of IIEF erection domain,overall sexual satisfaction and total IIEF-15 in twogroups: normal (scores: �26; �8; �62, respec-tively) and not normal, and adjusting data for age,

BMI, cigarette smoking, hypercholesterolemia,hypertension and diabetes at logistic regressionanalysis performed in both groups, PhA (�3hours/week) resulted the only independent vari-able for normal erection (P = 0.010) (95% CI0.036–0.643), higher sexual satisfaction(P = 0.022) (95% CI 0.084–0.821) and normaltotal IIEF-15 score (P = 0.023) (95% CI 0.85–0.837).

Discussion

In our study, regular PhA, �3 hours/week, for 3months, in subjects previously not physicallyactive, significantly improved erectile function: inthe intervention group, restoration of EDoccurred in 77.8% vs. 39.3% in the control group(P < 0.004).

The beneficial effects of PhA may be mediatedin a number of ways, including increase inendothelium-derived nitric oxide (NO), penilecyclic guanosine monophosphate levels, andnumber of endothelial progenitor cells (EPCs);changes in lipid profile, fibrinogen, carbohydratemetabolism, neurohormonal release, vascularinflammatory markers decrease as well as effectson blood pressure [1–3,13–16].

Shear stress is a potent physiological stimulusfor NO release [17,18]; flow-mediated dilation(FMD) was demonstrated in 25 healthy youngmales after just 10 weeks of aerobic and anaerobic

Table 4 IIEF-15 at end of treatment

Total population PDE5i alone PDE5I + PhA P

Erectile function 25.7 � 2.62 24.7 � 2.62 26.7 � 2.18 0.003ED severity degree

Mild 41.8% 60.7% 22.2% 0.004Normal erection 59.2% 39.3% 77.8% 0.004

Confidence 3.80 � 0.73 3.53 � 0.69 4.07 � 0.68 0.006Sexual desire 6.82 � 1.21 6.46 � 1.10 7.18 � 1.24 0.028Orgasmic function 8.18 � 1.21 7.96 � 1.40 8.40 � 0.87 0.169Intercourse satisfaction 10.55 � 1.54 9.85 � 1.50 11.25 � 1.25 0.001Overall satisfacion 7.64 � 1.28 7.17 � 1.29 8.07 � 1.15 0.009IIEF-15 total score 58.9 � 7.31 56.2 � 7.56 61.7 � 5.99 0.007Frequency sexual intercourse (last 4 weeks) 7 � 1.33 6.96 � 1.37 7.04 � 1.31 ns

ED = erectile dysfunction; IIEF = International Index of Erectile Function; PDE5i = phosphodiesterase type 5 inhibitor; PhA = physical activity.

0 1 2 3

Baseline End of Study

10

15

20

25

30IIE

F-1

5E

rect

ion

Do

mai

n

Gr.AGr. B

Gr. A

Gr. B

Figure 2 International Index of Erectile Function (IIEF)-15erection domain.

Table 5 Total testosterone (ng/mL)

Total population PDE5i alone PDE5I + PhAP (t-test ind.sample)

Baseline 4.35 � 0.94 4.45 � 0.93 4.24 � 0.95 nsEnd of study 4.51 � 0.78 4.47 � 0.71 4.55 � 0.85 nsP (t-test paired) 0.035 (95% CI 0.011–0.31) ns 0.012 (95% CI 0.072–0.549)

PDE5i = phosphodiesterase type 5 inhibitor; PhA = physical activity; CI = confidence interval.

Physical Activity in the Treatment of Erectile Dysfunction 2205

J Sex Med 2010;7:2201–2208

exercise program: FMD improved by 77% from2.2% to 3.9% (P < 0.01) and the basis for thisimprovement seems related to FMD mediatedendothelial shear stress [17].

PhA increases the number of EPCs that stimu-late vasculogenesis [19]; their number is consid-ered a predictor in vascular function and isinversely correlated with risk factors for coronarydisease [20–22]: circulating EPCs increased nearlyfourfold in peripheral blood after acute exhaustivedynamic exercise in 22 healthy volunteers [23];patients with coronary artery disease after a4-week program of moderate exercise showed asignificant EPC number increase [24].

PhA has beneficial effects on the lipid profile:endurance athletes have serum HDL cholesterol(HDL-C) 40–50% higher, low-density lipoproteincholesterol 5–10% lower, and triglyceride levels20% lower than their sedentary counterparts[25–27]. Kodama et al. in a meta-analysis on 25studies estimated that the minimal weekly exercisevolume for increasing HDL-C level should be900 kcal of EE or 120 minutes of exercise; univari-ate regression analysis indicated that every10-minute prolongation of exercise per session wasassociated with an approximate 1.4 mg/dL increase[28].

A program of changing lifestyles: reducingbody weight, improving quality of diet andincreasing PhA, in a randomized study on 209overweight or obese subjects with ED or atincreasing risk for ED conducted by Espositoet al. obtained a 32% restoration of ED in theintervention group vs. 7% in the control group(P = 0.015) [29]. The same authors had previouslyreported about one-third restoration of ED andsignificant reduction of vascular inflammatorymarkers in 55 obese men after a 2-year programof PhA and weight loss [15]. A review article byHannan et al. confirmed a beneficial impact ofPhA and obesity interventions on erectile func-tion [30]. Significant benefit of PhA in ED treat-ment was reported in a controlled study on 43patients with ED and hypertension after an8-week exercise program (45–60 minute/day;60–79% maximal heart rate) [31]. In the LookAHEAD trial of 263 patients with type 2 diabetes,cardiorespiratory fitness was found to be protec-tive of ED in a multivariable analysis (OR = 0.61;P < 0.001) [32]. PDE5i therapy was very effectivein both groups in terms of erectile function(Tables 1 and 4); improvements obtained in thePDE5i only group were similar to those reportedin the literature [33]. Preferred PDE5i, number

of doses and frequency of sexual intercourse, weresimilar in the two groups (Table 3).

Considering the other domains of sexual func-tion, this study evidences significant improve-ments in the intervention group compared to thecontrol group, in particular, confidence (Q15) 3.53vs. 4.07 (P = 0.006); sexual desire 6.46 vs. 7.18(P = 0.028); intercourse satisfaction 9.85 vs. 11.25(P = 0.001); total satisfaction 7.17 vs. 8.07 (P =0.009) (Table 4).

These improvements could be explained by: (i)improvement in erectile function; (ii) psychologi-cal improvement; and (iii) TT increase (as laterdiscussed).

The scientific consensus linking PhA to mentalhealth has resulted in recommendations that exer-cise should be used for the promotion and main-tenance of mental health and in the managementof mental health problems [34–37]. Actual mecha-nisms by which PhA affects mental health are stilluncertain: with regard to anxiety, a meta-analysiscontaining 104 studies (N = 3,048 men andwomen) indicated that acute (effect size [ES] =0.23) and chronic (ES = 0.25) exercise behavior isassociated with a reduction in state anxiety [38];since this meta-analysis, research has continued toreport that exercise reduces anxiety state [39].Most studies investigating changes in mood beforeand after an acute bout of exercise have also indi-cated increased positive (i.e., vigor) and decreasednegative (i.e., tension, depression, anger, fatigue,and confusion) mood states from pre- to postexer-cise [40].

Our study also shows a significant increase ofTT from baseline solely in the intervention group(4.55 � 0.85 vs. 4.24 � 0.95; P = 0.012; 95% CI0.072–0.549) (Table 5). The literature correlationbetween PhA and TT reports conflicting results: across-sectional study on 696 men aged 20–>70years (the European Prospective Investigation intoCancer and Nutrition study) comparing a non-exercise group to a vigorous exercise group (�3hours/week) reported higher testosterone levels(11%) in the latter group [41]. A study on 531healthy Singaporean Chinese men aged 29–72years showed a positive impact of PhA on TT level(14% increase) associated with a BMI significantreduction [42]. A longitudinal study of 995healthy, young (24–32 years), white or black malesshowed no correlation of PhA with TT; however,in multivariable linear regression, average PhA wassignificantly correlated with TT in obese whitemen [43]. It is possible that the influence of PhAmight be stronger among overweight or obese

2206 Maio et al.

J Sex Med 2010;7:2201–2208

men [44], and our population of patients was over-weight. PhA was similar at baseline in the twogroups and only slightly increased at the end ofstudy in the control group; this confirms the all toocommon experience that a simple advice of beingphysically active is not effective.

In our opinion, it is not surprising that signifi-cant effects are evident after just 3 months of PhA,since other studies have reported significantresults, regarding other aspects, in similar orshorter periods of time [17,23,27].

Limitations of the present study are: (i) this is apilot study performed on a small number ofpatients; (ii) PhA was not objectively monitored bymedical personnel or by sport instructors; theactivity was managed and reported solely by thepatients themselves; (iii) subjects self-reportingtheir performed PhA could overestimate the actualquantity of PhA: the group that received moreinformation on the importance of PhA for healthwould be more likely to overreport the PhA theyperformed; conversely, the group that received lessinformation would tend to underestimate; and (iv)despite a marked increase in the percentage ofpatients who had their IIEF-EF normalized(77.8% vs. 39.3%), the actual absolute change inthe score was not as impressive as seems: physicallyactive patients had a 19.8% greater improvementin erection domain (10.9 vs. 9.1) and an 18.1%greater improvement in overall IIEF-15 score(20.9 vs. 17.7) from baseline. This could be linkedto the limited period of treatment: it is likely thatextending PhA for a longer period of time couldresult in a more relevant outcome.

Conclusion

This randomized controlled study demonstratesthat PhA in association with PDE5i, after only 3months, gives significant beneficial effects in thetreatment of ED and in restoring a normal sexuallife compared with PDE5i alone. The study alsoshows a moderate yet significant increase of TT inthe intervention group. PDE5 inhibitors are veryeffective and, at present, represent almost exclu-sively the treatment of this pathology. PhA shouldbe seriously considered, not only as a preventivebut also as a curative measure, in the managementof ED together with pharmacological treatment.

Corresponding Author: Giuseppe Maio, MD, AbanoTerme Private Hospital, Operative Unit of Andrology,via Montegrande 28, Abano Terme 35031, Italy. Tel:(39) 348-2614920; Fax: (39) 049-667441; E-mail:giusmai@tin.it

Conflict of Interest: None.

Statement of Authorship

Category 1(a) Conception and Design

Giuseppe Maio; Salim Saraeb(b) Acquisition of Data

Giuseppe Maio; Salim Saraeb(c) Analysis and Interpretation of Data

Giuseppe Maio; Salim Saraeb; Antonio Marchiori

Category 2(a) Drafting the Article

Giuseppe Maio(b) Revising It for Intellectual Content

Giuseppe Maio; Salim Saraeb; Antonio Marchiori

Category 3(a) Final Approval of the Completed Article

Giuseppe Maio; Salim Saraeb; Antonio Marchiori

References

1 Leon AS, Cornett J, Jacobs DR, Rauramara R. Leisure timephysical activity levels and risk of coronary heart disease anddeath: The Multiple Risk Factor Intervention Study. JAMA1987;273:2388–95.

2 Blair SN, Kohl HW, Barlow CE, Paffenbarger RS, GibbonsLW, Macera CA. Changes in physical fitness and all causemortality. A prospective study of healthy and unhealthy men.JAMA 1995;273:1093–8.

3 Lakka TA, Venalainen JM, Rauramara R, Salonen R, Tuom-ilehto J, Salonen JT. Relation of leisure-time physical activityand cardiorespiratory fitness to the risk of acute myocardialinfarction in men. N Engl J Med 1994;330:1549–54.

4 NIH Consensus Development Panel on Physical Activity andCardiovascular Health. Physical activity and cardiovascularhealth. JAMA 1996;276:241–6.

5 Derby CA, Mohr BA, Goldstein I, Feldman HA, JohannesCB, McKinlay JB. Modifiable risk factors and erectile dys-function can lifestyle changes modify risk? Urology 2000;56:302–6.

6 Kratzik CW, Lackner JE, Märk I, Rücklinger E, SchmidbauerJ, Lunglmayr G, Schatzl G. How much physical activity isneeded to maintain erectile function? Results of the AndroxVienna Municipality Study. Eur Urol 2009;55:509–17.

7 Bacon CG, Mittleman MA, Kawachi I, Giovannucci E, GlasserDB, Rimm EB. A prospective study of risk factors for erectiledysfunction. J Urol 2006;176:217–21.

8 Nicolosi A, Glasser DB, Moreira ED, Villa M. Prevalence oferectile dysfunction and associated factors among men withoutconcomitant diseases: A population study. Int J Impot Res2003;15:253–7.

9 Cheng JY, Ng EM, Chen RY, Ko JS. Alcohol consumption anderectile dysfunction: Meta-analysis of population-basedstudies. Int J Impot Res 2007;19:343–52.

10 Pollock ML, Gaesser GA, Butcher JD, Després JP, DishmanRK, Franklin BA, Garber CE. ACSM position stand: Therecommended quantity and quality of exercise for developingand maintaining cardiorespiratory and muscular fitness, andflexibility in healthy adults. Med Sci Sports Exerc 1998;30:975–91.

Physical Activity in the Treatment of Erectile Dysfunction 2207

J Sex Med 2010;7:2201–2208

11 Ainsworth BE, Haskell WL, Whitt MC, Irwin ML, SwartzNM, Strath SJ, O’Brien WL, Bassett DR Jr, Schmitz KH,Emplaincourt PO, Jacobs DR Jr, Leon AS. Compendium ofphysical activities: An update of activity codes and MET inten-sities. Med Sci Sports Exerc 2000;32(suppl):S498–516.

12 Rosen RC, Riley A, Wagner G, Osterloh IH, Kirkpatrick J,Mishra A. The International Index of Erectile Function (IIEF):A multidimensional scale for assessment of erectile dysfunc-tion. Urology 1997;49:822–30.

13 Cook JP, Tsao PS. Is NO an anti-atherogenic molecule? Arte-rioscler Thromb 1994;14:653–5.

14 Mulhall J, Teloken P, Brock G, Kim E. Obesity, dyslipidemiasand erectile dysfunction: A report of a subcommittee of thesexual medicine society of North America. J Sex Med 2006;3:778–86.

15 Esposito K, Giugliano F, Di Palo C, Giugliano G, Marfella R,D’Andrea F, D’Armiento M, Giugliano D. Effect of lifestylechanges on erectile dysfunction in obese men. A randomizedcontrolled trial. JAMA 2004;291:2978–84.

16 Musicki B, Liu T, Strong T, Jin L, Laughlin MH, Turk JR,Burnett AL. Low-fat diet and exercise preserve eNOS regula-tion and endothelial function in the penis of early atheroscle-rotic pigs: A molecular analysis. J Sex Med 2008;5:552–61.

17 Clarkson P, Montgomery HE, Mullen MJ, Donald AE, PoweAJ, Bull T, Jubb M, World M, Deanfield JE. Exercise trainingenhances endothelial function in young men. J Am CollCardiol 1999;33:1379–85.

18 Faulx MD, Wright AT, Hoit DB. Detection of endothelialdysfunction with brachial artery ultrasound scanning. AmHeart J 2003;145:943–51.

19 Asahara T, Murohara T, Sullivan A, Silver M, van der Zee R,Li T, Witzenbichler B, Schatteman G, Isner JM. Isolation ofputative progenitor endothelial cells for angiogenesis. Science1997;275:964–7.

20 Liew A, Barry A, O’Brien T. Endothelial progenitor cells:Diagnostic and therapeutic considerations. BioEssays 2006;28:261–70.

21 Vasa M, Fichtlscherer S, Aicher A, Adler K, Urbich C.Number and migratory activity of circulating endothelial pro-genitor cells inversely correlate with risk factors for coronaryartery disease. Circ Res 2001;89:E1–7.

22 Loomans CJ, de Koning EJ, Staal FJ, Rookmaaker MB, Ver-seyden C. Endothelial progenitor cell dysfunction: A novelconcept in the pathogenesis of vascular complications of type 1diabetes. Diabetes 2004;53:195–9.

23 Rehman J, Li J, Parvathaneni L, Karlsson G, Panchal VR,Temm CJ, Mahenthiran J, March KL. Exercise acutelyincreases circulating endothelial progenitor cells andmonocyte-/macrophage-derived angiogenic cells. J Am CollCardiol 2004;43:2314–8.

24 Laufs U, Werner N, Link A, Endres M, Wassmann S. Physicaltraining increases endothelial progenitor cells, inhibits neoin-tima formation, and enhances angiogenesis. Circulation2004;109:220–6.

25 Coggan AR, Raguso CA, Gastaldelli A, Sidossis LS, YeckelCW. Fat metabolism during high-intensity exercise inendurance-trained and untrained men. Metabolism 2000;49:122–8.

26 Achten J, Jeukendrup AJ. Optimizing fat oxidation throughexercise and diet. Nutrition 2004;20:716–27.

27 Schrauwen P, van Aggel-Leijssen DP, Hul G, WagenmakersAJ, Vidal H, Saris WH, van Baak MA. The effect of a 3-monthlow-intensity endurance training program on fat oxidation andacetyl-CoA carboxylase-2 expression. Diabetes 2002;51:2220–6.

28 Kodama S, Tanaka S, Saito K, Shu M, Sone Y, Onitake F,Suzuki E, Shimano H, Yamamoto S, Kondo K, Ohashi Y,Yamada N, Sone H. Effect of aerobic exercise training onserum levels of high-density lipoprotein cholesterol. A meta-analysis. Arch Intern Med 2007;167:999–1008.

29 Esposito K, Ciotola M, Giugliano F, Maiorino MI, AutorinoR, Giugliano G, Nicoletti G, D’Andrea F, D’Armiento M,Giugliano D. Effects of intensive lifestyles changes on erectyledysfunction in men. J Sex Med 2009;6:243–50.

30 Hannan JL, Maio MT, Komolova M, Adams MA. Beneficialimpact of exercise and obesity interventions on erectile func-tion and its risk factors. J Sex Med 2009;6(3 suppl):254–61.

31 Lamina S, Okoye CG, Dagogo TT. Therapeutic effect of aninterval exercise training program in the management of erec-tile dysfunction in hypertensive patients. J Clin Hypertens(Greenwich) 2009;11:125–9.

32 Rosen RC, Wing RR, Schneider S, Wadden TA, Foster GD,West DS, Kitabchi AE, Brancati FL, Maschak-Carey BJ,Bahnson JL, Lewis CE, Gendrano lii IN. Erectile dysfunctionin type 2 diabetic men: Relationship to exercise fitness andcardiovascular risk factors in the Look AHEAD trial. J SexMed 2009;6:1414–22.

33 Moore RA, Derry S, McQuay HJ. Indirect comparison ofinterventions using published randomised trials: Systematicreview of PDE-5 inhibitors for erectile dysfunction. BMCUrol 2005;5:18. doi:10.1186/1471-2490-5-18. Available athttp://www.biomedcentral.com/content/pdf/1471-2490-5-18.pdf (accessed March 18, 2010).

34 Fox KR, Boutcher SH, Faulkner GE, Biddle SJH. The case forexercise in the promotion of mental health. In: Biddle SJH,Fox KR, Boutcher SH, eds. Physical activity and psychologicalwell-being. London: Routledge; 2000:1–10.

35 Grant T, ed. Physical activity and mental health—Nationalconsensus statements and guidelines for practitioners. London:Health Education Authority; 2000.

36 Burbach FR. The efficacy of physical activity interventionswithin mental health services: Anxiety and depressive disor-ders. J Ment Health 1997;6:543–66.

37 Fontaine KR. Physical activity improves mental health. Physi-cian Sports Med 2000;28:83–4.

38 Petruzzello SJ, Landers DM, Hatfield BD, Kubitz KA, SalazarW. A meta-analysis on the anxiety-reducing effects of acuteand chronic exercise. Sports Med 1991;11:143–82.

39 Focht BC, Koltyn KF. Influences of resistance exercise ofdifferent intensities on state anxiety and blood pressure. MedSci Sports Exerc 1999;31:456–63.

40 Berger BG, Motl RW. Exercise and mood: A selective reviewand synthesis of research employing the profile of mood states.J Appl Sport Psychol 2000;12:69–92.

41 Allen NE, Appleby PN, Davey GK, Key TJ. Lifestyle andnutritional determinants of bioavailable androgens and relatedhormones in British men. Cancer Causes Control 2002;13:353–63.

42 Goh VHH, Tong TYY, Mok HPP, Said B. Interactions amongage, adiposity, bodyweight, lifestyle factors and sex steroidhormones in healthy Singaporean Chinese men. Asian JAndrol 2007;9:611–21.

43 Wolin KY, Colangelo LA, Liu K, Sternfeld B, Gapstur SM.Associations of androgens with physical activity and fitness inyoung black and white men: The CARDIA Male HormoneStudy. Prev Med 2007;44:426–31.

44 Mohr BA, Bhasin B, Link CL, O’Donnell AB, McKinlay JB.The effect of changes in adiposity on testosterone levels inolder men: Longitudinal results from the Massachusetts MaleAging Study. Eur J Endocrinol 2006;155:443–52.

2208 Maio et al.

J Sex Med 2010;7:2201–2208