- Email: [email protected]
PREVALENCE AND INDEPENDENT RISK FACTORS FOR ERECTILE DYSFUNCTION IN SPAIN: RESULTS OF THE EPIDEMIOLOGIA DE LA DISFUNCION ERECTIL MASCULINA STUDY
0022-5347/01/1662-0569/0 THE JOURNAL OF UROLOGY® Copyright © 2001 by AMERICAN UROLOGICAL ASSOCIATION, INC.®
Vol. 166, 569 –575, August 2001 Printed in U.S.A.
PREVALENCE AND INDEPENDENT RISK FACTORS FOR ERECTILE DYSFUNCTION IN SPAIN: RESULTS OF THE EPIDEMIOLOGIA DE LA DISFUNCION ERECTIL MASCULINA STUDY ˜ IGO SAENZ DE TEJADA,* ANTONIO MARTIN-MORALES,* JOSE J. SANCHEZ-CRUZ,* IN LUIS RODRIGUEZ-VELA,* J. FERNANDO JIMENEZ-CRUZ* AND RAFAEL BURGOS-RODRIGUEZ* From the Urology Department, Carlos Haya Hospital, Malaga, Andalusian School of Public Health, Granada and Fundacion para la Investigacion y Desarrollo en Andrologia and Spanish Urological Association, Madrid, Spain
ABSTRACT
Purpose: We determined the prevalence of and risks factors for erectile dysfunction in Spain in a cross-sectional study. Materials and Methods: A total of 2,476 noninstitutionalized Spanish men 25 to 70 years old were interviewed at home and answered a self-administered questionnaire of 71 items, including 2 instruments to define erectile dysfunction, a simple self-assessment question to estimate erectile function and the International Index of Erectile Function. Data on disease, medication and toxic habits were also obtained. Results: With an overall participation rate of 75% the prevalence of erectile dysfunction according to the simple question was 12.1%. According to the erectile function domain of the International Index of Erectile Function the overall prevalence was 18.9%. Several independent risk factors were significantly associated with the probability of erectile dysfunction. Some differences arose according to the tool used to define the condition. However, there was a strong relationship of patient age with frequency or severity no matter which instrument was used to define erectile dysfunction. Diabetes (age adjusted odds ratio 4), high blood pressure (odds ratio 1.58), high cholesterol (1.63), peripheral vascular disorder (2.63), lung disease (3.11), prostate disease (2.93), cardiac problems (1.79), rheumatism (2.37) and allergy (3.08) were significantly associated with erectile dysfunction. Drug intake, which respondents called medication for nerves and sleeping pills, correlated strongly (odds ratio 2.78 and 4.27, respectively), as did tobacco use (2.5) and alcohol consumption (1.53). Conclusions: This study provides data on the prevalence of and risks factors for erectile dysfunction in Spain. The relationship of erectile dysfunction with certain risk factors, such as cardiovascular risk factors and drugs intake, are well known and our study corroborates these associations. Other associations with erectile dysfunction, such as prostate disease, allergy and rheumatism, support findings in previous reports, although to our knowledge the pathophysiological mechanisms remain unclear. Estimating the strength of the association of erectile dysfunction with distinct risk factors in terms of odds ratios enabled us to identify the factors to pursue when seeking to prevent erectile dysfunction. Furthermore, the relationship of tobacco with erectile dysfunction, which has been controversial in previous series, was well characterized in our study. KEY WORDS: impotence, questionnaires, epidemiology, risk factors
Erectile dysfunction is a serious public health problem affecting millions of men and their quality of life.1, 2 The National Institutes of Health Consensus Development Panel on Impotence stressed the need to perform epidemiological studies to determine the prevalence of male erectile dysfunction, and its associated medical and psychological factors, particularly in the setting of possible racial, ethnic, socioeconomic and cultural differences.1 Several studies have been done in various countries worldwide to determine the prevalence of erectile dysfunction.3 These studies using various methodologies and target populations have prevalence rates that vary widely from 70.8% to 12.8%. Nevertheless, few series were population based and used probabilistic samples representative of the general population. These studies also differ in many relevant aspects, such as the age range of men
evaluated or the definition of erectile dysfunction in the population or group studied, which limits the possibility of extrapolating prevalence rates to the general population. Furthermore, erectile dysfunction prevalence may vary according to cultural, racial and health variables among countries or continents. Some health related correlates of erectile dysfunction are well known and have been confirmed in various studies, including diabetes, hypertension, coronary heart disease and depression. Additional correlates, including cholesterol, tobacco, alcohol and conditions such as prostate disease, are mentioned in some reports but not in others since their role is controversial at least. Thus, strong epidemiological data clarifying health related correlates, conditions, life-style, demographics and drug consumption are obviously valuable for a better understanding of erectile dysfunction and subsequently for appropriate service delivery and resource allocation models as well as for developing prevention strategies when modifiable risk factors are established. We determined
Accepted for publication March 2, 2001. * Financial interest and/or other relationship with Pfizer, Fundacion para la Investigacion y Desarrollo en Andrologia and UROAN XXI. 569
570
PREVALENCE AND INDEPENDENT RISK FACTORS FOR ERECTILE DYSFUNCTION IN SPAIN
the prevalence of erectile dysfunction in Spanish men 25 to 70 years old as well as the potential personal, psychosocial, sociocultural, drug and life-style factors associated with erectile dysfunction. METHODS
In our population based prevalence study the target population included Spanish men 25 to 70 years old who were not institutionalized and who resided in the Iberian peninsula. A probabilistic multistage sampling design was used, with stratification of the primary sampling units. The primary sampling units were census sections, the secondary sampling units were dwellings and the final sampling units were subjects. Stratification was based on the variables of age group with 4 levels, including ages 25 to 39, 40 to 49, 50 to 59 and 60 to 70 years, autonomous community with 15 levels corresponding to each peninsular autonomous community and population density with 4 levels, including fewer than 10,000, 10,001 to 50,000, 50,001 to 500,000 and more than 500,000 inhabitants or provincial capitals. The sample setting was proportional to the number of men included in each population stratum. The first stage units or censual sections were selected using simple random sampling without replacement within each stratum formed by combining the levels of the variables, autonomous community and population density. The secondary sampling units or dwellings were selected using the random route procedure. When more than 1 person from the study population resided in a dwelling, 1 was randomly selected using the Kish pattern. Sample sizes were determined for the 95% confidence interval (CI) with a design effect of 1.1. Overall the study population included 2,476 men 25 to 70 years old inclusively who were distributed into groups, including 1,003, 520, 481 and 471 who were 25 to 39, 40 to 49; 50 to 59 and 60 to 70 years old, respectively. Data collection. Information was collected from March 23, 1998 to March 6, 1999 in each of the 15 autonomous communities of the peninsula by interviewers in the home of each participant. A national company specializing in this type of health study was engaged for the fieldwork. A comprehensive self-administered questionnaire of 71 items grouped into 7 sections was used to obtain the data, which included information on sociodemographic aspects, general health on the RAND 36-Item Health Survey,4, 5 tobacco use, alcohol intake on the CAGE questionnaire,6 chronic disease, disabilities and illnesses, prostatic symptoms on the International Prostate Symptom Score (I-PSS) questionnaire,7 erection capacity, concern about erection problems and relationship with a stable partner. Measurement of erectile dysfunction. Section 5 of the questionnaire, which assessed erectile function/dysfunction in the participants, contained 18 questions (see Appendix). The initial 15 questions were the International Index of Erectile Function (IIEF),8 which addresses certain domains of male sexuality, including erection, libido, orgasm, satisfaction with intercourse and overall sexual satisfaction. Question 18 on overall subject self-assessment of erectile capacity, called the simple question, was used as the first defining criterion of the various degrees of erectile dysfunction. The second criterion defining the various degrees of erectile dysfunction was the score of the answers to questions 1 to 5 and 15 of the IIEF. These questions form the IIEF erectile function domain and are scored on a scale of 26 to 30 —no, 17 to 25—mild, 11 to 16 —moderate and 6 to 10 —severe/complete dysfunction. We also asked about the frequency of erection on awakening in question 16 and assessed the subject impression of partner satisfaction with sexual activity in question 17. Data analysis. After the sample was weighted by the specified design the percent or prevalence and population estimates of the degree of erectile dysfunction in the population and the corresponding standard errors were obtained to con-
struct the 95% CI. We statistically analyzed factors associated with erectile dysfunction using contingency table analysis, and the chi-square and Fisher exact tests to determine independence. The odds ratio with the crude 95% CI, and 95% CI adjusted for stratification variables and other confounding variables was determined by constructing the binary logistic regression model. The effect of factors and covariates on dysfunction and the potential effect of interactions among factors and covariates on dysfunction were also studied. We used commercially available computer software for data analysis and for entering, cleaning and coding the data. RESULTS
Participation. The total participation rate was 75%. Except for the Autonomous Community of Catalonia with 37% participation, all regions showed a participation rate ranging from 61% in Andalusia to 94% in Castile-La Mancha. Overall prevalence. Prevalence According to the Simple Question: The prevalence of any degree of erectile dysfunction in Spanish men 25 to 70 years old was 12.1% (95% CI 10.8 to 13.3, fig. 1). Figure 1 shows the distribution of minimal, moderate or severe/complete severity. Therefore, according to this criterion 6.9% of the population (95% CI 5.9 to 7.9) had moderate, severe or complete erectile dysfunction. Prevalence According to the IIEF Erectile Function Domain: Based on this criterion the prevalence of all degrees of erectile dysfunction in men 25 to 70 years old was 18.9% (95% CI 17.15 to 20.67, fig. 1). Figure 1 shows the distribution of minimal, moderate and severe/complete severity. Therefore, 2.7% of the population (95% CI 2 to 3.4) had moderate or severe/complete erectile dysfunction. There was a significant difference in erectile dysfunction prevalence depending on whether the IIEF erectile function domain or simple question was used to define erectile dysfunction. The overall prevalence was higher on the IIEF erectile function domain than on the simple question because a large number of individuals with mild erectile dysfunction were detected by the erectile function domain. The simple question, which was based on individual self-assessment of erectile function, detected a significantly larger number of subjects in the moderate and severe/complete erectile dysfunction categories. Prevalence by age groups. Prevalence According to the Simple Question: The prevalence of erectile dysfunction increased with age (linear chi-square 224.006, 1 degree of freedom, p ⬍0.001). Table 1 lists the distribution of erectile dysfunction prevalence by age group. This table shows data on the rate of erectile dysfunction prevalence by age groups as well as the 95% CI and number of inhabitants (population estimates) within the 95% CI per age group. Furthermore, the severity of erectile dysfunction was also age dependent (fig. 2). Men 60 to 70 years old had a 10-fold increase in overall erectile dysfunction prevalence compared
FIG. 1. Overall erectile dysfunction (ED) prevalence on simple question and erectile function (EF) domain of IIEF.
PREVALENCE AND INDEPENDENT RISK FACTORS FOR ERECTILE DYSFUNCTION IN SPAIN
571
TABLE 1. Distribution of erectile dysfunction prevalence Age Group (yrs.) 25–39 40–49 50–59 60–70 Overall
% Incidence (95% CI incidence/population) IIEF Erectile Function
Simple Question
8.48 (6.52–10.44/203,747–326,246) 13.72 (10.48–16.95/183,767–297,219) 25.5 (21.07–29.93/315,367–447,980) 48.25 (42.9–54.21/459,438–588,937) 18.9 (17.15–20.67/1,276,903–1,546,377)
3.92 (2.7–5.13/107,183–203,648) 6.32 (4.22–8.42/87,252–174,091) 15.9 (12.57–19.23/240,143–367,378) 32.24 (27.98–36.49/524,908–684,557) 12.1 (10.8–13.3/1,067,215–1,322,595)
FIG. 2. Prevalence according to severity and age group on simple question.
with the 25 to 39-year-old group (3.92% versus 32.3%). When considering severe/complete erectile dysfunction, the increase in prevalence was 20-fold in the younger versus older groups (0.3% versus 6%). Figure 3 shows the probability of impotence according to age group in terms of odds ratios, in which the reference category was men 25 to 39 years old. When 25-year-old men were considered the reference category, the increase in the odds ratio was 1.07 yearly. Prevalence According to the IIEF Erectile Function Domain: The relationship of age with erectile dysfunction according to the IIEF erectile function domain was similar to that according to simple question. The prevalence of erectile dysfunction increased with age (linear chi-square 206.228, 1 degree of freedom, p ⬍0.001). Table 1 lists the distribution of severity. As in the simple question, the severity of erectile dysfunction depended significantly on age group (Pearson’s chi-square 240.625, 9 degrees of freedom, p ⬍0.001). Relationship of erectile dysfunction with other independent variables. Due to the strong relationship of age with erectile dysfunction the analysis of variables was age adjusted. Although we consider that erectile dysfunction is essentially a question of individual perception and we identified different prevalences applying the 2 instruments used to define erectile dysfunction, we based our analysis on the answers to each tool. The simple overall assessment question enabled comparison of our data to those of previous studies using this question directly9 or indirectly.10 Sociodemographic characteristics, and analysis of diseases, medications and toxic habits. Table 2 shows the age adjusted odds ratios of the various physical and sociodemographic characteristics analyzed. All diseases, medications and toxic habits evaluated were significantly associated with an increased probability of some degree of erectile dysfunction
FIG. 3. Odds ratio of having erectile dysfunction on simple question according to age group.
when the overall self-assessment question was used as the measuring tool (table 3). However, when the IIEF erectile function domain was used, conditions such as heart disease and allergy were not significantly related to an increased probability of erectile dysfunction (95% CI 0.83 to 2.30 and 0.75 to 2.49, respectively). The disease with a higher adjusted odds ratio for erectile dysfunction was diabetes with a greater than 4-fold likelihood of impotence in men with than without diabetes. Medication for nerves, including selfreported psychotropic drug intake in a broad sense, and sleeping pills, including self-reported intake of this type of medication, were related to erectile dysfunction with sleeping pills most strongly associated. After analyzing treatment versus no treatment for a certain disease with and without erectile dysfunction we detected no statistically significant differences regardless of the instrument used to define dysfunction. DISCUSSION
Our study introduces innovations in sample design (range of age studied) and the instruments used to identify subjects with erectile dysfunction, that is the self-assessment or simple question and the multidimensional IIEF questionnaire scale. There were certain reasons for including a population with a wider age range. It is not uncommon for men younger than 40 years to present to physicians, mostly complaining of primary erectile dysfunction. Therefore, they are also a population at risk. Followup in these patients, who are presumed to be in maximum health, may provide us in the future with highly clarifying data on the development of various diseases, and the correlation of these conditions and/or their treatment with erectile dysfunction. To date epidemiological studies of erectile dysfunction have been based mainly on subject assessment with an overall simple question as the criterion for establishing erectile dysfunction. However, since this question involves individual perception, an assessment that applies a multidimensional scale would appear to detect erectile dysfunction more accurately. To this end the IIEF was developed, proposed and recommended as an internationally validated tool for measuring the therapeutic impact of various actions on erectile dysfunction and for epidemiological studies.8 However, we were unfamiliar with the performance of the questionnaire in this setting, and so we also included the simple question in our series. Our study shows a different prevalence rate of any degree of erectile dysfunction in the 25 to 70-year-old Spanish male population than in previous series. Others reported a higher prevalence of erectile dysfunction, for example 70.6% in 585 men 25 to 70 years old seeking primary health care in Pakistan,11 64.7% in 1,982 men 40 to 70 years old in Turkey,12 54.9% in 599 Egyptian men 30 to 70 years old,13 53.6% in 655 men older than 25 years old in Morocco,14 50.7% in 917 Nigerian men 35 to 70 years old15 and 48% in 2,128 men 50 to 70 years old in Finland.16 When compared with studies done in Southern Europe, the prevalence of erectile dysfunction was similar in a single study that indicated a prevalence rate of erectile dysfunction of 12.8% in 2,010 Italian men older than 18 years old.17 Nevertheless, in another erectile dysfunction series from Italy the prevalence was higher, that is 14% in men 50 to 54, 30% in those 55 to
572
PREVALENCE AND INDEPENDENT RISK FACTORS FOR ERECTILE DYSFUNCTION IN SPAIN TABLE 2. Relationship of erectile dysfunction with physical and demographic characteristics Characteristics
Ht. (cm.) Wt. (kg.) Body vol. Educational level: Cannot read or write No educational certificate Primary education (VTI) Secondary education (VTII) College studies Degree studies $ Income/mo. ($1 ⬵ 175 pts.): Less than 50,000 50,001–70,000 70,001–90,000 90,001–150,000 150,001–200,000 Greater than 200,000
Odds Ratio
Odds Ratio
Simple Question (95% CI)
IIEF Erectile Function (95% CI)
Crude
Age Adjusted
Crude
Age Adjusted
0.94 (0.93–0.96) 1.00 (0.99–1.01) 1.07 (1.04–1.11)
0.99 (0.97–1.01) 1.00 (0.99–1.01) 1.03 (0.99–1.07)
0.93 (0.92–0.95) 0.99 (0.98–1.01) 1.08 (1.05–1.12)
0.97 (0.95–0.99) 1.00 (0.99–1.01) 1.05 (1.01–1.09)
3.80 (1.48–9.75) 2.89 (1.68–4.99) 1.44 (0.84–2.45) 1.04 (0.57–1.89) 1.16 (0.59–2.29) 1
1.06 (0.39–2.87) 1.00 (0.56–1.81) 0.97 (0.55–1.70) 0.97 (0.52–1.82) 0.88 (0.43–1.80) 1
7.38 (2.17–24.99) 3.51 (2.10–5.87) 1.77 (1.08–2.89) 0.96 (0.55–1.68) 1.14 (0.60–2.16) 1
2.77 (0.77–9.94) 1.63 (0.94–2.83) 1.38 (0.83–2.31) 0.91 (0.51–1.62) 0.93 (0.48–1.83) 1
2.15 (1.09–4.23) 2.26 (1.37–3.74) 2.03 (1.28–3.21) 1.68 (1.16–2.44) 0.80 (0.51–1.27) 1
1.91 (0.92–3.95) 0.88 (0.51–1.52) 1.17 (0.72–1.91) 1.29 (0.87–1.91) 0.71 (0.44–1.14) 1
3.21 5.31 2.97 1.90 1.50 1
3.76 (1.72–8.22) 2.94 (1.70–5.09) 2.23 (1.36–3.65) 1.61 (1.10–2.36) 1.44 (0.95–2.19) 1
(1.53–6.72) (3.19–8.83) (1.87–4.71) (1.32–2.74) (1.00–2.25)
TABLE 3. Relationship of erectile dysfunction with disease, drugs and toxic habits
Disease: High blood pressure Diabetes Cardiac problems Lung problems Circulatory problems Rheumatic problems Cholesterol problems Allergy Prostatic disease Medication: For nerves Sleeping pills Toxic habits: Greater than 40 cigarettes/day Alcohol abuse (CAGE index)
Odds Ratio
Odds Ratio
Simple Question (95% CI)
IIEF Erectile Function (95% CI)
Crude
Age Adjusted
Crude
Age Adjusted
3.18 (2.30–4.41) 7.09 (4.60–10.93) 4.23 (2.86–6.26) 4.92 (2.99–8.09) 4.50 (3.09–6.55) 4.29 (3.07–5.99) 2.62 (1.77–3.89) 2.39 (1.50–3.81) 7.29 (4.74–11.20)
1.58 (1.11–2.24) 4.08 (2.57–6.49) 1.79 (1.18–2.71) 3.11 (1.82–5.34) 2.63 (1.76–3.93) 2.37 (1.66–3.39) 1.63 (1.07–2.49) 3.08 (1.83–5.16) 2.93 (1.86–4.61)
3.25 (2.26–4.66) 3.86 (2.21–6.74) 3.06 (1.91–4.90) 2.99 (1.54–5.84) 3.86 (2.48–6.01) 4.35 (2.98–6.35) 2.55 (1.69–3.84) 1.25 (0.72–2.17) 6.96 (4.01–12.11)
1.72 (1.16–2.55) 1.98 (1.09–3.60) 1.38 (0.83–2.30) 2.31 (1.12–4.74) 2.39 (1.48–3.85) 2.46 (1.64–3.71) 1.62 (1.04–2.52) 1.37 (0.75–2.49) 2.67 (1.48–4.80)
3.21 (1.90–5.42) 4.92 (2.99–8.09)
2.78 (1.57–4.93) 4.27 (2.47–7.39)
3.01 (1.66–5.45) 3.14 (1.60–6.16)
2.28 (1.19–4.34) 2.78 (1.33–5.81)
2.56 (1.73–3.79) 1.36 (1.03–1.79)
2.50 (1.64–3.80) 1.53 (1.15–2.09)
1.50 (0.99–2.26) 1.42 (1.07–1.86)
1.59 (1.02–2.48) 1.62 (1.20–2.18)
64 and 51% in those older than 65 years old.18 The criteria used to determine erectile dysfunction and participant selection varied in these studies. The difference in methodology may explain the difference in the prevalence rate among and within countries.17, 18 On the other hand, even when comparing studies that to a certain extent applied a similar methodology (same criteria to define erectile dysfunction, same age range and random population sample), a different prevalence was reported, as in the Massachusetts Male Aging Study (MMAS)10 and Cross-National Study (CNS),9 which used a simple question to asses erectile dysfunction, as we did. Country-to-country differences in erectile dysfunction prevalence were identified when comparing the MMAS, CNS and our Epidemiologia de la Disfuncion Erectil Masculina Study (EDEM) despite the similar methodology used. For example, the prevalence of moderate to severe/complete erectile dysfunction at ages 40 to 70 years was 39% in Japan, 34.8% in the United States on the MMAS, 21% in Italy, 16% in Malaysia, 15% in Brazil and 10.5% in Spain on the EDEM. When comparing the overall prevalence rate of some degree of erectile dysfunction in men 40 to 70 years old with that in men 25 to 39 years old who were excluded from analysis in our sample, the values were 52.1% in the United States on the MMAS10 and 17.7% in Spain on the EDEM. To our knowledge the reasons for such large differences are unclear but they may reflect various cultural perceptions of erectile dysfunction. Of the sociodemographic data analyzed age most strongly correlated with the likelihood of erectile dysfunction. Fur-
thermore, the severity of dysfunction also correlated strongly with age. These findings are completely consistent with the other epidemiological studies of this condition reported to date, whether population based or otherwise. The remainder of the sociodemographic parameters, such as occupational status, income or educational level, were not related to the likelihood of erectile dysfunction in our study. When adjusted for age according to the simple question marital status was also not associated, although this variable was significantly associated with erectile dysfunction in the National Health and Social Life Survey.19 Nevertheless, when the IIEF erectile function domain was used to define erectile dysfunction, body volume and monthly income appeared to correlate with the probability of dysfunction. Although the association was slight, it was statistically significant and in regard to monthly income it confirms previous findings on the relationship of sexual dysfunction with socioeconomic status in this case.19 The analysis of the diseases studied revealed that they were significantly associated with some degree of erectile dysfunction. This finding may be explained by the selection criteria of the diseases evaluated in our study. Such criteria included selecting risk factors for erectile dysfunction that were already recognized in other epidemiological studies as well as the clinical impression, which was confirmed by our study, of the possible association of erectile dysfunction with other diseases. However, in contrast to previously reported results,10 the treatment or absence of treatment for each disease did not significantly influence the likelihood of erec-
PREVALENCE AND INDEPENDENT RISK FACTORS FOR ERECTILE DYSFUNCTION IN SPAIN
tile dysfunction, as mentioned. The results of our analysis differed slightly depending on which instrument was used to define erectile dysfunction. The data obtained by the simple self-assessment question completely agree with those reported previously.10 However, when the IIEF erectile function domain was used as the criterion, conditions with a likely epidemiological association with dysfunction, such as heart disease, and other diseases with no such obvious relationship, such as allergy, correlated with the probability of erectile dysfunction, although not in a statistically significant manner. We studied the agreement of the 2 instruments. The value of the index after crossing the 2 measurements of erectile dysfunction on the key question and IIEF erectile function domain was 0.344 when each scale was dichotomized for no dysfunction versus dysfunction. When agreement of the original scales with their corresponding 4 levels (no, mild, moderate and severe dysfunction) was analyzed, the value of the weighted index of 0.54 was obtained as an agreement measurement. Of the various levels of erectile dysfunction mild dysfunction had the greatest agreement according to each measurement, followed by moderate dysfunction. Nevertheless, we think that the question of which of the 2 instruments may be more useful and accurate for large population studies remains controversial. It is true that the IIEF has been validated and the diagnostic capacity of the erectile function domain has been evaluated.20 Agreement of a simple self-assessment question with the IIEF erectile function domain has also been evaluated and correlated well. Each instrument applies samples obtained from subjects involved in clinical trials rather than population and control samples.21 A good correlation of each criterion in population samples has recently been reported.22 A single selfassessment direct question to evaluate erectile dysfunction was applied in the population based sample of the MMAS followup evaluation, in addition to the Brief Male Sexual Function Inventory (BMSFI) and IIEF. Prevalence was similar to that determined on the IIEF, agreement was moderate ( 0.56 to 0.58) and associations with previously identified risk factors were similar for each classification. The single question correlated well with these other measures (r ⫽ 0.71 to 0.78, p ⬍0.001). However the incidence of subjects not classified due to missing data was 9% on the MMAS, 8% on the BMSFI and 18% on the IIEF. Because the incidence of subjects not classified due to missing data was 9% on the MMAS, 8% on the BMSFI and 18% on the IIEF, we believe that a direct self-assessment question may be a practical tool for population based studies, in which detailed clinical measures of erectile dysfunction are impractical. Considering that our study was cross-sectional, no inference of causality of erectile dysfunction with the evaluated factors of disease, medication or toxic habits was done. The relationship of erectile dysfunction with certain factors is well established, for example cardiovascular disease,3 diabetes,23, 24 medication intake and substance abuse.25, 26 Other illnesses have been associated with erectile dysfunction in clinical settings. Our study provides population based epidemiological data on the association of erectile dysfunction with lung diseases and adds data on the relationship with rheumatism, allergy and benign prostatic hyperplasia, which have been reported previously.9, 10 For toxic habits the relationship is controversial. No significant relationship of cigarette smoking with erectile dysfunction was identified in 1 report,10 while such a relationship was detected in other series27, 28 as well as in our study. Alcohol consumption was not associated with erectile dysfunction in some series,17, 29 while it was related in the MMAS10 and in our study. Further research and a deeper insight into the toxic factors associated with erectile dysfunction are required.
573
CONCLUSIONS
Our study, which was based on a random sample representing the Spanish male population 25 to 70 years old, had a broad participation rate and involved exhaustive statistical analysis. It shows results that to a certain extent confirm previous observations in regard to different prevalences in different geographical and cultural areas. It also confirms risk factors already detected in previous studies, such as cardiovascular factors, diabetes, hypertension, heart disease, peripheral vascular problems and hypercholesterolemia. In addition, it shows the strength of the association of erectile dysfunction with the consumption of drugs, particularly psychotropic drugs (sleeping pills and medication for nerves). Our study establishes the association of diseases with erectile dysfunction, of which some are plausibly biologically related, such as lung disease, and others are not so clearly related from the pathophysiological point of view, such as prostate disease, rheumatism and allergy. The latter association was previously detected in the MMAS.10 Slight but significant differences were detected in erectile dysfunction prevalence and grade depending on the instrument used to define and classify the condition. Recently others have assessed the agreement of various instruments to establish which may be more practical, reliable and simple to use in large population studies. All findings have stressed the need for further research in this field and for study designs that may clarify the associations, some unexpected and newly evidenced, of which the results may help us to achieve better understanding and management of this important health problem.
APPENDIX: LIST OF QUESTIONS IN SECTION 5—ERECTION CAPACITY
1. Over the past 4 weeks, how often were you able to get an erection during sexual activity? 2. Over the past 4 weeks, when you had erections with sexual stimulation, how often were your erections hard enough for penetration? 3. Over the past 4 weeks, when you attempted sexual intercourse, how often were you able to penetrate your partner? 4. Over the past 4 weeks, during sexual intercourse, how often were you able to maintain your erection after you had penetrated your partner? 5. Over the past 4 weeks, during sexual intercourse, how difficult was it to maintain your erection to completion of intercourse? 6. Over the past 4 weeks, how many times have you attempted sexual intercourse? 7. Over the past 4 weeks, when you attempted sexual intercourse, how often was it satisfactory to you? 8. Over the past 4 weeks, how much have you enjoyed sexual intercourse? 9. Over the past 4 weeks, when you had sexual stimulation or intercourse how often did you ejaculate? 10. Over the past 4 weeks, when you had sexual stimulation or intercourse, how often did you have the feeling of orgasm or climax (with or without ejaculation)? 11. Over the past 4 weeks, how often have you felt sexual desire? 12. Over the past 4 weeks, how would you rate your level of sexual desire? 13. Over the past 4 weeks, how satisfied have you felt with your overall sex life? 14. Over the past 4 weeks, how satisfied have you been with your sexual relationship with your partner? 15. Over the past 4 weeks, how do you rate your confidence that you can get and keep your erection?
574
PREVALENCE AND INDEPENDENT RISK FACTORS FOR ERECTILE DYSFUNCTION IN SPAIN
16. Over the past 4 weeks, how often have you experienced a full erection on awaking? 17. Do you think your partner is . . . a) Very satisfied with sexual activity; b) Satisfied; c) Neither satisfied nor dissatisfied; d) Rather unsatisfied; e) Very unsatisfied. 18. Do you consider yourself a man . . . a) With no erection problem; b) With a minimum incapacity; c) With a moderate incapacity; d) With a severe/complete incapacity for erection. REFERENCES
1. NIH Consensus Conference: Impotence: NIH Development Panel on Impotence. JAMA, 270: 83, 1993 2. Ayta, I. A., McKinlay, J. B. and Krane, R. J.: The likely worldwide increase in erectile dysfunction between 1995 and 2025 and some possible policy consequences BJU Int, 84: 50, 1999 3. Benet, A. E. and Melman, A.: The epidemiology of erectile dysfunction. Urol Clin North Am, 22: 699, 1995 4. Manual de Puntuacio´n de la Versio´n Espan˜ola del Cuestionario de Salud SF-36. Barcelona, Spain: Unitat de Recerca en Serveis Sanitaris, Institut Municipal d’Investigacio´ Me`dica, 1998 5. Ware, J. E., Snow, K. K., Kosinski, M. et al: SF-36 Health Survey manual and interpretation guide. Boston: New England Medical Center, The Health Institute, 1993 6. Ewing, J. A.: Detecting alcoholism. The CAGE questionnaire. JAMA, 252: 1905, 1984 7. Batista, J. E., Regalado, R., Chechile, E. et al: Validation of the International Prostate Symptom Score in Spain. The 2nd International Consultation on Benign Prostatic Hyperplasia: Proceedings. Edited by A. T. K. Cockett, Y. Aso, C. Chatelain, C. et al. Jersey, Channel Islands: Scientific Communication International, p. 129, 1993 8. Rosen, R. C., Riley, A., Wagner, G. et al: The international index of erectile function (IIEF): a multidimensional scale for assessment of erectile dysfunction. Urology, 49: 822, 1997 9. Glasser, D. and Sweeney, M.: The prevalence of erectile dysfunction in four countries: Italy, Brazil, Malaysia and Japan. Cross-National Study Group. Presented at International Consultation on Erectile Dysfunction, Paris, France, July 1–3, 1999 10. Feldman, H. A., Goldstein, I., Hatzichristou, D. G. et al: Impotence and its medical and psychosocial correlates: results of the Massachusetts Male Aging Study. J Urol, 151: 54, 1994 11. Hussain, S. S., Siddiqui, S., Razzaq, M. A. et al: A patient based epidemiology study to find the prevalence and correlates of erectile dysfunction (ED) in two cities in Pakistan. Presented at European Society for Impotence Research Meeting, Barcelona, Spain, January 30 to February 2, 2000 12. Akkus, E., Kadioglu, A., Esen, A. et al: Prevalence of erectile dysfunction among urban and rural population in Turkey. Presented at ESIR Meeting, Barcelona, Spain, XXX, 2000 13. Mahmoud, K. Z., Azim, S. A., Gadallah, M. et al: A patient based epidemiology study to find the prevalence and correlates of erectile dysfunction (ED) in Egypt. Presented at ESIR Meeting, Barcelona, Spain, XXX, 2000 14. Kadiri, N., Berrada, S., Tahiri, S. et al: Prevalence of erectile dysfunction (ED) and its correlates in Morocco: a population based epidemiological study. Presented at ESIR Meeting, Barcelona, Spain, XXX, 2000 15. Dogunro, S., Osegbe, D. and Jaguste, V. S.: Epidemiology of erectile dysfunction (ED) and its correlates: a patient based study in Nigerian population. Presented at ESIR Meeting, Barcelona, Spain, XXX, 2000 16. Koskimaki, K., Hakama, M. and Tammela, T. L. J.: Prevalence of erectile dysfunction in Finland. Presented at ESIR Meeting, Barcelona, Spain, XXX, 2000 17. La Pera, G., Taggi, F., Aloise, L. et al: Prevalence of the erectile dysfunction and of other genitourinary symptoms: preliminary results of the SIMONA study group. Presented at ESIR Meeting, Barcelona, Spain, XXX, 2000 18. Parazzini, F., Menchini-Fabris, F., Bortolotti, A. et al: Frequency and determinants of erectile dysfunction in Italy. Eur Urol, 37: 43, 2000 19. Laumann, E. O., Paik, A. and Rosen, R. C.: Sexual dysfunction in the United States. Prevalence and predictors. JAMA, 281: 537, 1999
20. Cappelleri, J. C., Rosen, R. C., Smith, M. D. et al: Diagnostic evaluation of the erectile function domain of the International Index of Erectile Function. Urology, 54: 346, 1999 21. Cappelleri, J. C., Siegel, R. L., Osterloh, I. H. et al: Relationship between patient self-assessment of erectile function and the erectile function domain of the International Index of Erectile Function. Urology, 56: 477, 2000 22. Derby, C. A., Araujo, A. B., Johannes, C. B. et al: Measurement of erectile dysfunction in population-based studies: the use of a single question self-assessment in the Massachusetts Male Aging Study. Int J Impot Res, 12: 197, 2000 23. Kayigil, O., Atahan, O. and Metin, A.: Multifactorial evaluation of diabetic erectile dysfunction. Int Urol Nephrol, 28: 717, 1996 24. Vibe-Petersen, J., Behrend, L., Jorgensen, T. et al: The prevalence of erectile dysfunction in a male diabetic population. Presented at ESIR Meeting, Barcelona, Spain, XXX, 2000 25. Wein, A. J. and Van Arsdalen, K. N.: Drug-induced male sexual dysfunction. Urol Clin North Am, 15: 23, 1988 26. Burns-Cox, N. and Gingel, C.: Erectile dysfunction: is medication to blame? Prescriber, 3: 77, 1997 27. Mannino, D. M., Klevens, R. M. and Flanders, W. D.: Cigarette smoking: an independent risk factor for impotence? Am J Epidemiol, 140: 1003, 1994 28. Condra, M., Morales, A., Owen, J. A. et al: Prevalence and significance of tobacco smoking in impotence. Urology, 27: 495, 1986 29. Schiavi, R. C.: Chronic alcoholism and male sexual dysfunction. J Sex Marital Ther, 16: 23, 1990 EDITORIAL COMMENTS The study of the prevalence and risk factors of erectile dysfunction on the Iberian Peninsula supplements the widely referenced MMAS from North America. The strength of this prevalence study is that it is population based rather than dependent on prevalence in men seeking urological care. This information combined with data on established risk factors as well as other correlative diseases allows urologists to arrive at a better understanding of erectile dysfunction. To society at large it may hopefully contribute to an improved recognition of erectile dysfunction and a more equitable delivery of health care to this needy subpopulation. The ramifications of confirming established risk factors and the confirmation of suspected or new ones, such as smoking, may help to address treatment or prevention strategies for risk factors that are not genetically or congenitally acquired. Of continued concern are the presumed cultural differences in erectile dysfunction prevalence compared to those in other well described population based studies. The variance of prevalence of moderate to severe erectile dysfunction ranges from 39% in Japan to 10.5% in Spain. Such large discrepancies are seemingly the only explanation of different cultural perceptions of erectile dysfunction. The methodology of measuring erectile dysfunction in this series is somewhat troublesome insofar as relates to use of the overall selfassessment in the so-called simple question. This use brings into question whether this particular questionnaire has been validated by the rigorous methods now commonly applied in population based studies of erectile dysfunction. It is interesting that the perception of erectile dysfunction differs substantially by age group and overall whether it was measured by the IIEF or the so-called simple question. However, the patterns of age related erectile dysfunction noted with either question are more consistent. This population study also provides epidemiological data on the association of erectile dysfunction with other suspected but more controversial topics, such as benign prostatic hyperplasia and cigarette smoking. The challenges were to convert such cross-sectional information into more longitudinal studies, thereby, strengthening the association of risk factors with disease. Kevin T. McVary Department of Urology Northwestern University Medical School Chicago, Illinois The authors report a population based study. Such studies are needed in various regions of the world to improve our understanding of the risk factors of erectile dysfunction and potentially generate hypotheses for preventing erectile dysfunction. Furthermore, this article is especially important because it not only confirmed the role
PREVALENCE AND INDEPENDENT RISK FACTORS FOR ERECTILE DYSFUNCTION IN SPAIN of previously known risk factors such as age, diabetes and cardiovascular disease, but also showed for the first time to my knowledge an association of erectile dysfunction with other medical conditions. The most exciting new finding to urologists is the association of erectile dysfunction with prostate disease. These conditions are common, chronic, slowly progressive and age related. As the baby boomer generation continues to age, the medical profession and specialty of urology are faced with the task of ensuring that this large, aging population remains healthy and vital in the face of extended life expectancy and increasing health care costs. This increase in life expectancy has brought much attention to the topic of geriatric medicine. Naturally urology as a specialty is concerned with many of these important medical conditions. This article focuses specifically on male population and the health problems that frequently plague men. Several major noncancer disease states adversely affect males as they age. These conditions include erectile dysfunction, prostate disease, cardiovascular disease and depression. The conditions often precipitate considerable morbidity and may even result in mortality. Furthermore, the presentation of 1 condition may correlate with the development of another. Available data on these correlations should be enough to warrant a multidisciplinary approach for research. This approach may help to maintain patients healthy and more importantly happy late into life, realizing the concept of successful aging. Ridwan Shabsigh Department of Urology Columbia Presbyterian Medical Center New York, New York REPLY BY AUTHORS In regard to the comment about problems (“troublesome”) that may arise from the use of the so-called simple question when meas-
575
uring erectile dysfunction, this question has been used in other studies for the same purpose (reference 9 in article). The MMAS indirectly (reference 10 in article) and directly (reference 22 in article) uses a global self-assessment question to classify the subjects participating in the study. It was precisely the “nonvalidation by the new current, rigorous methods of that particular questionnaire” which led us to use the IIEF as a tool to measure erectile dysfunction for the first time in a population-based epidemiological study. The uncertainty about its adequate behavior in this setting led us not to reject instruments already used. On the other hand, since the IIEF refers to the last 4 weeks, those subjects who have had no sexual activity during that period would have no data on erectile ability or disability. The use of both instruments allowed us to have 2 measures of the same phenomenon in the same population. Agreement between the 2 tools has been reported (references 21 and 22 in article). In our article we provided preliminary data on the agreement between the 2 instruments in our study sample. The comparison of the results obtained with one or the other tool allowed us to hypothesize that the IIEF could be a good instrument to detect the disease in its earliest stages (16.5% using the erectile function domain of the IIEF versus 5.2% obtained by the simple question regarding mild erectile dysfunction prevalence), since the simple question detects the patient when the degree of involvement is more severe (2.7% versus 6.9% for moderate and severe/complete erectile dysfunction using the erectile function domain and the simple question, respectively). The use of the IIEF would allow detection of the disease in its earliest stages when it could still be amenable to prevention, whereas the subject who answers the simple question classifying himself as having erectile dysfunction does so in the more severe categories. These could be the subjects who really are affected by this disease and, therefore, more prone to search for professional help.
Vol. 166, 569 –575, August 2001 Printed in U.S.A.
PREVALENCE AND INDEPENDENT RISK FACTORS FOR ERECTILE DYSFUNCTION IN SPAIN: RESULTS OF THE EPIDEMIOLOGIA DE LA DISFUNCION ERECTIL MASCULINA STUDY ˜ IGO SAENZ DE TEJADA,* ANTONIO MARTIN-MORALES,* JOSE J. SANCHEZ-CRUZ,* IN LUIS RODRIGUEZ-VELA,* J. FERNANDO JIMENEZ-CRUZ* AND RAFAEL BURGOS-RODRIGUEZ* From the Urology Department, Carlos Haya Hospital, Malaga, Andalusian School of Public Health, Granada and Fundacion para la Investigacion y Desarrollo en Andrologia and Spanish Urological Association, Madrid, Spain
ABSTRACT
Purpose: We determined the prevalence of and risks factors for erectile dysfunction in Spain in a cross-sectional study. Materials and Methods: A total of 2,476 noninstitutionalized Spanish men 25 to 70 years old were interviewed at home and answered a self-administered questionnaire of 71 items, including 2 instruments to define erectile dysfunction, a simple self-assessment question to estimate erectile function and the International Index of Erectile Function. Data on disease, medication and toxic habits were also obtained. Results: With an overall participation rate of 75% the prevalence of erectile dysfunction according to the simple question was 12.1%. According to the erectile function domain of the International Index of Erectile Function the overall prevalence was 18.9%. Several independent risk factors were significantly associated with the probability of erectile dysfunction. Some differences arose according to the tool used to define the condition. However, there was a strong relationship of patient age with frequency or severity no matter which instrument was used to define erectile dysfunction. Diabetes (age adjusted odds ratio 4), high blood pressure (odds ratio 1.58), high cholesterol (1.63), peripheral vascular disorder (2.63), lung disease (3.11), prostate disease (2.93), cardiac problems (1.79), rheumatism (2.37) and allergy (3.08) were significantly associated with erectile dysfunction. Drug intake, which respondents called medication for nerves and sleeping pills, correlated strongly (odds ratio 2.78 and 4.27, respectively), as did tobacco use (2.5) and alcohol consumption (1.53). Conclusions: This study provides data on the prevalence of and risks factors for erectile dysfunction in Spain. The relationship of erectile dysfunction with certain risk factors, such as cardiovascular risk factors and drugs intake, are well known and our study corroborates these associations. Other associations with erectile dysfunction, such as prostate disease, allergy and rheumatism, support findings in previous reports, although to our knowledge the pathophysiological mechanisms remain unclear. Estimating the strength of the association of erectile dysfunction with distinct risk factors in terms of odds ratios enabled us to identify the factors to pursue when seeking to prevent erectile dysfunction. Furthermore, the relationship of tobacco with erectile dysfunction, which has been controversial in previous series, was well characterized in our study. KEY WORDS: impotence, questionnaires, epidemiology, risk factors
Erectile dysfunction is a serious public health problem affecting millions of men and their quality of life.1, 2 The National Institutes of Health Consensus Development Panel on Impotence stressed the need to perform epidemiological studies to determine the prevalence of male erectile dysfunction, and its associated medical and psychological factors, particularly in the setting of possible racial, ethnic, socioeconomic and cultural differences.1 Several studies have been done in various countries worldwide to determine the prevalence of erectile dysfunction.3 These studies using various methodologies and target populations have prevalence rates that vary widely from 70.8% to 12.8%. Nevertheless, few series were population based and used probabilistic samples representative of the general population. These studies also differ in many relevant aspects, such as the age range of men
evaluated or the definition of erectile dysfunction in the population or group studied, which limits the possibility of extrapolating prevalence rates to the general population. Furthermore, erectile dysfunction prevalence may vary according to cultural, racial and health variables among countries or continents. Some health related correlates of erectile dysfunction are well known and have been confirmed in various studies, including diabetes, hypertension, coronary heart disease and depression. Additional correlates, including cholesterol, tobacco, alcohol and conditions such as prostate disease, are mentioned in some reports but not in others since their role is controversial at least. Thus, strong epidemiological data clarifying health related correlates, conditions, life-style, demographics and drug consumption are obviously valuable for a better understanding of erectile dysfunction and subsequently for appropriate service delivery and resource allocation models as well as for developing prevention strategies when modifiable risk factors are established. We determined
Accepted for publication March 2, 2001. * Financial interest and/or other relationship with Pfizer, Fundacion para la Investigacion y Desarrollo en Andrologia and UROAN XXI. 569
570
PREVALENCE AND INDEPENDENT RISK FACTORS FOR ERECTILE DYSFUNCTION IN SPAIN
the prevalence of erectile dysfunction in Spanish men 25 to 70 years old as well as the potential personal, psychosocial, sociocultural, drug and life-style factors associated with erectile dysfunction. METHODS
In our population based prevalence study the target population included Spanish men 25 to 70 years old who were not institutionalized and who resided in the Iberian peninsula. A probabilistic multistage sampling design was used, with stratification of the primary sampling units. The primary sampling units were census sections, the secondary sampling units were dwellings and the final sampling units were subjects. Stratification was based on the variables of age group with 4 levels, including ages 25 to 39, 40 to 49, 50 to 59 and 60 to 70 years, autonomous community with 15 levels corresponding to each peninsular autonomous community and population density with 4 levels, including fewer than 10,000, 10,001 to 50,000, 50,001 to 500,000 and more than 500,000 inhabitants or provincial capitals. The sample setting was proportional to the number of men included in each population stratum. The first stage units or censual sections were selected using simple random sampling without replacement within each stratum formed by combining the levels of the variables, autonomous community and population density. The secondary sampling units or dwellings were selected using the random route procedure. When more than 1 person from the study population resided in a dwelling, 1 was randomly selected using the Kish pattern. Sample sizes were determined for the 95% confidence interval (CI) with a design effect of 1.1. Overall the study population included 2,476 men 25 to 70 years old inclusively who were distributed into groups, including 1,003, 520, 481 and 471 who were 25 to 39, 40 to 49; 50 to 59 and 60 to 70 years old, respectively. Data collection. Information was collected from March 23, 1998 to March 6, 1999 in each of the 15 autonomous communities of the peninsula by interviewers in the home of each participant. A national company specializing in this type of health study was engaged for the fieldwork. A comprehensive self-administered questionnaire of 71 items grouped into 7 sections was used to obtain the data, which included information on sociodemographic aspects, general health on the RAND 36-Item Health Survey,4, 5 tobacco use, alcohol intake on the CAGE questionnaire,6 chronic disease, disabilities and illnesses, prostatic symptoms on the International Prostate Symptom Score (I-PSS) questionnaire,7 erection capacity, concern about erection problems and relationship with a stable partner. Measurement of erectile dysfunction. Section 5 of the questionnaire, which assessed erectile function/dysfunction in the participants, contained 18 questions (see Appendix). The initial 15 questions were the International Index of Erectile Function (IIEF),8 which addresses certain domains of male sexuality, including erection, libido, orgasm, satisfaction with intercourse and overall sexual satisfaction. Question 18 on overall subject self-assessment of erectile capacity, called the simple question, was used as the first defining criterion of the various degrees of erectile dysfunction. The second criterion defining the various degrees of erectile dysfunction was the score of the answers to questions 1 to 5 and 15 of the IIEF. These questions form the IIEF erectile function domain and are scored on a scale of 26 to 30 —no, 17 to 25—mild, 11 to 16 —moderate and 6 to 10 —severe/complete dysfunction. We also asked about the frequency of erection on awakening in question 16 and assessed the subject impression of partner satisfaction with sexual activity in question 17. Data analysis. After the sample was weighted by the specified design the percent or prevalence and population estimates of the degree of erectile dysfunction in the population and the corresponding standard errors were obtained to con-
struct the 95% CI. We statistically analyzed factors associated with erectile dysfunction using contingency table analysis, and the chi-square and Fisher exact tests to determine independence. The odds ratio with the crude 95% CI, and 95% CI adjusted for stratification variables and other confounding variables was determined by constructing the binary logistic regression model. The effect of factors and covariates on dysfunction and the potential effect of interactions among factors and covariates on dysfunction were also studied. We used commercially available computer software for data analysis and for entering, cleaning and coding the data. RESULTS
Participation. The total participation rate was 75%. Except for the Autonomous Community of Catalonia with 37% participation, all regions showed a participation rate ranging from 61% in Andalusia to 94% in Castile-La Mancha. Overall prevalence. Prevalence According to the Simple Question: The prevalence of any degree of erectile dysfunction in Spanish men 25 to 70 years old was 12.1% (95% CI 10.8 to 13.3, fig. 1). Figure 1 shows the distribution of minimal, moderate or severe/complete severity. Therefore, according to this criterion 6.9% of the population (95% CI 5.9 to 7.9) had moderate, severe or complete erectile dysfunction. Prevalence According to the IIEF Erectile Function Domain: Based on this criterion the prevalence of all degrees of erectile dysfunction in men 25 to 70 years old was 18.9% (95% CI 17.15 to 20.67, fig. 1). Figure 1 shows the distribution of minimal, moderate and severe/complete severity. Therefore, 2.7% of the population (95% CI 2 to 3.4) had moderate or severe/complete erectile dysfunction. There was a significant difference in erectile dysfunction prevalence depending on whether the IIEF erectile function domain or simple question was used to define erectile dysfunction. The overall prevalence was higher on the IIEF erectile function domain than on the simple question because a large number of individuals with mild erectile dysfunction were detected by the erectile function domain. The simple question, which was based on individual self-assessment of erectile function, detected a significantly larger number of subjects in the moderate and severe/complete erectile dysfunction categories. Prevalence by age groups. Prevalence According to the Simple Question: The prevalence of erectile dysfunction increased with age (linear chi-square 224.006, 1 degree of freedom, p ⬍0.001). Table 1 lists the distribution of erectile dysfunction prevalence by age group. This table shows data on the rate of erectile dysfunction prevalence by age groups as well as the 95% CI and number of inhabitants (population estimates) within the 95% CI per age group. Furthermore, the severity of erectile dysfunction was also age dependent (fig. 2). Men 60 to 70 years old had a 10-fold increase in overall erectile dysfunction prevalence compared
FIG. 1. Overall erectile dysfunction (ED) prevalence on simple question and erectile function (EF) domain of IIEF.
PREVALENCE AND INDEPENDENT RISK FACTORS FOR ERECTILE DYSFUNCTION IN SPAIN
571
TABLE 1. Distribution of erectile dysfunction prevalence Age Group (yrs.) 25–39 40–49 50–59 60–70 Overall
% Incidence (95% CI incidence/population) IIEF Erectile Function
Simple Question
8.48 (6.52–10.44/203,747–326,246) 13.72 (10.48–16.95/183,767–297,219) 25.5 (21.07–29.93/315,367–447,980) 48.25 (42.9–54.21/459,438–588,937) 18.9 (17.15–20.67/1,276,903–1,546,377)
3.92 (2.7–5.13/107,183–203,648) 6.32 (4.22–8.42/87,252–174,091) 15.9 (12.57–19.23/240,143–367,378) 32.24 (27.98–36.49/524,908–684,557) 12.1 (10.8–13.3/1,067,215–1,322,595)
FIG. 2. Prevalence according to severity and age group on simple question.
with the 25 to 39-year-old group (3.92% versus 32.3%). When considering severe/complete erectile dysfunction, the increase in prevalence was 20-fold in the younger versus older groups (0.3% versus 6%). Figure 3 shows the probability of impotence according to age group in terms of odds ratios, in which the reference category was men 25 to 39 years old. When 25-year-old men were considered the reference category, the increase in the odds ratio was 1.07 yearly. Prevalence According to the IIEF Erectile Function Domain: The relationship of age with erectile dysfunction according to the IIEF erectile function domain was similar to that according to simple question. The prevalence of erectile dysfunction increased with age (linear chi-square 206.228, 1 degree of freedom, p ⬍0.001). Table 1 lists the distribution of severity. As in the simple question, the severity of erectile dysfunction depended significantly on age group (Pearson’s chi-square 240.625, 9 degrees of freedom, p ⬍0.001). Relationship of erectile dysfunction with other independent variables. Due to the strong relationship of age with erectile dysfunction the analysis of variables was age adjusted. Although we consider that erectile dysfunction is essentially a question of individual perception and we identified different prevalences applying the 2 instruments used to define erectile dysfunction, we based our analysis on the answers to each tool. The simple overall assessment question enabled comparison of our data to those of previous studies using this question directly9 or indirectly.10 Sociodemographic characteristics, and analysis of diseases, medications and toxic habits. Table 2 shows the age adjusted odds ratios of the various physical and sociodemographic characteristics analyzed. All diseases, medications and toxic habits evaluated were significantly associated with an increased probability of some degree of erectile dysfunction
FIG. 3. Odds ratio of having erectile dysfunction on simple question according to age group.
when the overall self-assessment question was used as the measuring tool (table 3). However, when the IIEF erectile function domain was used, conditions such as heart disease and allergy were not significantly related to an increased probability of erectile dysfunction (95% CI 0.83 to 2.30 and 0.75 to 2.49, respectively). The disease with a higher adjusted odds ratio for erectile dysfunction was diabetes with a greater than 4-fold likelihood of impotence in men with than without diabetes. Medication for nerves, including selfreported psychotropic drug intake in a broad sense, and sleeping pills, including self-reported intake of this type of medication, were related to erectile dysfunction with sleeping pills most strongly associated. After analyzing treatment versus no treatment for a certain disease with and without erectile dysfunction we detected no statistically significant differences regardless of the instrument used to define dysfunction. DISCUSSION
Our study introduces innovations in sample design (range of age studied) and the instruments used to identify subjects with erectile dysfunction, that is the self-assessment or simple question and the multidimensional IIEF questionnaire scale. There were certain reasons for including a population with a wider age range. It is not uncommon for men younger than 40 years to present to physicians, mostly complaining of primary erectile dysfunction. Therefore, they are also a population at risk. Followup in these patients, who are presumed to be in maximum health, may provide us in the future with highly clarifying data on the development of various diseases, and the correlation of these conditions and/or their treatment with erectile dysfunction. To date epidemiological studies of erectile dysfunction have been based mainly on subject assessment with an overall simple question as the criterion for establishing erectile dysfunction. However, since this question involves individual perception, an assessment that applies a multidimensional scale would appear to detect erectile dysfunction more accurately. To this end the IIEF was developed, proposed and recommended as an internationally validated tool for measuring the therapeutic impact of various actions on erectile dysfunction and for epidemiological studies.8 However, we were unfamiliar with the performance of the questionnaire in this setting, and so we also included the simple question in our series. Our study shows a different prevalence rate of any degree of erectile dysfunction in the 25 to 70-year-old Spanish male population than in previous series. Others reported a higher prevalence of erectile dysfunction, for example 70.6% in 585 men 25 to 70 years old seeking primary health care in Pakistan,11 64.7% in 1,982 men 40 to 70 years old in Turkey,12 54.9% in 599 Egyptian men 30 to 70 years old,13 53.6% in 655 men older than 25 years old in Morocco,14 50.7% in 917 Nigerian men 35 to 70 years old15 and 48% in 2,128 men 50 to 70 years old in Finland.16 When compared with studies done in Southern Europe, the prevalence of erectile dysfunction was similar in a single study that indicated a prevalence rate of erectile dysfunction of 12.8% in 2,010 Italian men older than 18 years old.17 Nevertheless, in another erectile dysfunction series from Italy the prevalence was higher, that is 14% in men 50 to 54, 30% in those 55 to
572
PREVALENCE AND INDEPENDENT RISK FACTORS FOR ERECTILE DYSFUNCTION IN SPAIN TABLE 2. Relationship of erectile dysfunction with physical and demographic characteristics Characteristics
Ht. (cm.) Wt. (kg.) Body vol. Educational level: Cannot read or write No educational certificate Primary education (VTI) Secondary education (VTII) College studies Degree studies $ Income/mo. ($1 ⬵ 175 pts.): Less than 50,000 50,001–70,000 70,001–90,000 90,001–150,000 150,001–200,000 Greater than 200,000
Odds Ratio
Odds Ratio
Simple Question (95% CI)
IIEF Erectile Function (95% CI)
Crude
Age Adjusted
Crude
Age Adjusted
0.94 (0.93–0.96) 1.00 (0.99–1.01) 1.07 (1.04–1.11)
0.99 (0.97–1.01) 1.00 (0.99–1.01) 1.03 (0.99–1.07)
0.93 (0.92–0.95) 0.99 (0.98–1.01) 1.08 (1.05–1.12)
0.97 (0.95–0.99) 1.00 (0.99–1.01) 1.05 (1.01–1.09)
3.80 (1.48–9.75) 2.89 (1.68–4.99) 1.44 (0.84–2.45) 1.04 (0.57–1.89) 1.16 (0.59–2.29) 1
1.06 (0.39–2.87) 1.00 (0.56–1.81) 0.97 (0.55–1.70) 0.97 (0.52–1.82) 0.88 (0.43–1.80) 1
7.38 (2.17–24.99) 3.51 (2.10–5.87) 1.77 (1.08–2.89) 0.96 (0.55–1.68) 1.14 (0.60–2.16) 1
2.77 (0.77–9.94) 1.63 (0.94–2.83) 1.38 (0.83–2.31) 0.91 (0.51–1.62) 0.93 (0.48–1.83) 1
2.15 (1.09–4.23) 2.26 (1.37–3.74) 2.03 (1.28–3.21) 1.68 (1.16–2.44) 0.80 (0.51–1.27) 1
1.91 (0.92–3.95) 0.88 (0.51–1.52) 1.17 (0.72–1.91) 1.29 (0.87–1.91) 0.71 (0.44–1.14) 1
3.21 5.31 2.97 1.90 1.50 1
3.76 (1.72–8.22) 2.94 (1.70–5.09) 2.23 (1.36–3.65) 1.61 (1.10–2.36) 1.44 (0.95–2.19) 1
(1.53–6.72) (3.19–8.83) (1.87–4.71) (1.32–2.74) (1.00–2.25)
TABLE 3. Relationship of erectile dysfunction with disease, drugs and toxic habits
Disease: High blood pressure Diabetes Cardiac problems Lung problems Circulatory problems Rheumatic problems Cholesterol problems Allergy Prostatic disease Medication: For nerves Sleeping pills Toxic habits: Greater than 40 cigarettes/day Alcohol abuse (CAGE index)
Odds Ratio
Odds Ratio
Simple Question (95% CI)
IIEF Erectile Function (95% CI)
Crude
Age Adjusted
Crude
Age Adjusted
3.18 (2.30–4.41) 7.09 (4.60–10.93) 4.23 (2.86–6.26) 4.92 (2.99–8.09) 4.50 (3.09–6.55) 4.29 (3.07–5.99) 2.62 (1.77–3.89) 2.39 (1.50–3.81) 7.29 (4.74–11.20)
1.58 (1.11–2.24) 4.08 (2.57–6.49) 1.79 (1.18–2.71) 3.11 (1.82–5.34) 2.63 (1.76–3.93) 2.37 (1.66–3.39) 1.63 (1.07–2.49) 3.08 (1.83–5.16) 2.93 (1.86–4.61)
3.25 (2.26–4.66) 3.86 (2.21–6.74) 3.06 (1.91–4.90) 2.99 (1.54–5.84) 3.86 (2.48–6.01) 4.35 (2.98–6.35) 2.55 (1.69–3.84) 1.25 (0.72–2.17) 6.96 (4.01–12.11)
1.72 (1.16–2.55) 1.98 (1.09–3.60) 1.38 (0.83–2.30) 2.31 (1.12–4.74) 2.39 (1.48–3.85) 2.46 (1.64–3.71) 1.62 (1.04–2.52) 1.37 (0.75–2.49) 2.67 (1.48–4.80)
3.21 (1.90–5.42) 4.92 (2.99–8.09)
2.78 (1.57–4.93) 4.27 (2.47–7.39)
3.01 (1.66–5.45) 3.14 (1.60–6.16)
2.28 (1.19–4.34) 2.78 (1.33–5.81)
2.56 (1.73–3.79) 1.36 (1.03–1.79)
2.50 (1.64–3.80) 1.53 (1.15–2.09)
1.50 (0.99–2.26) 1.42 (1.07–1.86)
1.59 (1.02–2.48) 1.62 (1.20–2.18)
64 and 51% in those older than 65 years old.18 The criteria used to determine erectile dysfunction and participant selection varied in these studies. The difference in methodology may explain the difference in the prevalence rate among and within countries.17, 18 On the other hand, even when comparing studies that to a certain extent applied a similar methodology (same criteria to define erectile dysfunction, same age range and random population sample), a different prevalence was reported, as in the Massachusetts Male Aging Study (MMAS)10 and Cross-National Study (CNS),9 which used a simple question to asses erectile dysfunction, as we did. Country-to-country differences in erectile dysfunction prevalence were identified when comparing the MMAS, CNS and our Epidemiologia de la Disfuncion Erectil Masculina Study (EDEM) despite the similar methodology used. For example, the prevalence of moderate to severe/complete erectile dysfunction at ages 40 to 70 years was 39% in Japan, 34.8% in the United States on the MMAS, 21% in Italy, 16% in Malaysia, 15% in Brazil and 10.5% in Spain on the EDEM. When comparing the overall prevalence rate of some degree of erectile dysfunction in men 40 to 70 years old with that in men 25 to 39 years old who were excluded from analysis in our sample, the values were 52.1% in the United States on the MMAS10 and 17.7% in Spain on the EDEM. To our knowledge the reasons for such large differences are unclear but they may reflect various cultural perceptions of erectile dysfunction. Of the sociodemographic data analyzed age most strongly correlated with the likelihood of erectile dysfunction. Fur-
thermore, the severity of dysfunction also correlated strongly with age. These findings are completely consistent with the other epidemiological studies of this condition reported to date, whether population based or otherwise. The remainder of the sociodemographic parameters, such as occupational status, income or educational level, were not related to the likelihood of erectile dysfunction in our study. When adjusted for age according to the simple question marital status was also not associated, although this variable was significantly associated with erectile dysfunction in the National Health and Social Life Survey.19 Nevertheless, when the IIEF erectile function domain was used to define erectile dysfunction, body volume and monthly income appeared to correlate with the probability of dysfunction. Although the association was slight, it was statistically significant and in regard to monthly income it confirms previous findings on the relationship of sexual dysfunction with socioeconomic status in this case.19 The analysis of the diseases studied revealed that they were significantly associated with some degree of erectile dysfunction. This finding may be explained by the selection criteria of the diseases evaluated in our study. Such criteria included selecting risk factors for erectile dysfunction that were already recognized in other epidemiological studies as well as the clinical impression, which was confirmed by our study, of the possible association of erectile dysfunction with other diseases. However, in contrast to previously reported results,10 the treatment or absence of treatment for each disease did not significantly influence the likelihood of erec-
PREVALENCE AND INDEPENDENT RISK FACTORS FOR ERECTILE DYSFUNCTION IN SPAIN
tile dysfunction, as mentioned. The results of our analysis differed slightly depending on which instrument was used to define erectile dysfunction. The data obtained by the simple self-assessment question completely agree with those reported previously.10 However, when the IIEF erectile function domain was used as the criterion, conditions with a likely epidemiological association with dysfunction, such as heart disease, and other diseases with no such obvious relationship, such as allergy, correlated with the probability of erectile dysfunction, although not in a statistically significant manner. We studied the agreement of the 2 instruments. The value of the index after crossing the 2 measurements of erectile dysfunction on the key question and IIEF erectile function domain was 0.344 when each scale was dichotomized for no dysfunction versus dysfunction. When agreement of the original scales with their corresponding 4 levels (no, mild, moderate and severe dysfunction) was analyzed, the value of the weighted index of 0.54 was obtained as an agreement measurement. Of the various levels of erectile dysfunction mild dysfunction had the greatest agreement according to each measurement, followed by moderate dysfunction. Nevertheless, we think that the question of which of the 2 instruments may be more useful and accurate for large population studies remains controversial. It is true that the IIEF has been validated and the diagnostic capacity of the erectile function domain has been evaluated.20 Agreement of a simple self-assessment question with the IIEF erectile function domain has also been evaluated and correlated well. Each instrument applies samples obtained from subjects involved in clinical trials rather than population and control samples.21 A good correlation of each criterion in population samples has recently been reported.22 A single selfassessment direct question to evaluate erectile dysfunction was applied in the population based sample of the MMAS followup evaluation, in addition to the Brief Male Sexual Function Inventory (BMSFI) and IIEF. Prevalence was similar to that determined on the IIEF, agreement was moderate ( 0.56 to 0.58) and associations with previously identified risk factors were similar for each classification. The single question correlated well with these other measures (r ⫽ 0.71 to 0.78, p ⬍0.001). However the incidence of subjects not classified due to missing data was 9% on the MMAS, 8% on the BMSFI and 18% on the IIEF. Because the incidence of subjects not classified due to missing data was 9% on the MMAS, 8% on the BMSFI and 18% on the IIEF, we believe that a direct self-assessment question may be a practical tool for population based studies, in which detailed clinical measures of erectile dysfunction are impractical. Considering that our study was cross-sectional, no inference of causality of erectile dysfunction with the evaluated factors of disease, medication or toxic habits was done. The relationship of erectile dysfunction with certain factors is well established, for example cardiovascular disease,3 diabetes,23, 24 medication intake and substance abuse.25, 26 Other illnesses have been associated with erectile dysfunction in clinical settings. Our study provides population based epidemiological data on the association of erectile dysfunction with lung diseases and adds data on the relationship with rheumatism, allergy and benign prostatic hyperplasia, which have been reported previously.9, 10 For toxic habits the relationship is controversial. No significant relationship of cigarette smoking with erectile dysfunction was identified in 1 report,10 while such a relationship was detected in other series27, 28 as well as in our study. Alcohol consumption was not associated with erectile dysfunction in some series,17, 29 while it was related in the MMAS10 and in our study. Further research and a deeper insight into the toxic factors associated with erectile dysfunction are required.
573
CONCLUSIONS
Our study, which was based on a random sample representing the Spanish male population 25 to 70 years old, had a broad participation rate and involved exhaustive statistical analysis. It shows results that to a certain extent confirm previous observations in regard to different prevalences in different geographical and cultural areas. It also confirms risk factors already detected in previous studies, such as cardiovascular factors, diabetes, hypertension, heart disease, peripheral vascular problems and hypercholesterolemia. In addition, it shows the strength of the association of erectile dysfunction with the consumption of drugs, particularly psychotropic drugs (sleeping pills and medication for nerves). Our study establishes the association of diseases with erectile dysfunction, of which some are plausibly biologically related, such as lung disease, and others are not so clearly related from the pathophysiological point of view, such as prostate disease, rheumatism and allergy. The latter association was previously detected in the MMAS.10 Slight but significant differences were detected in erectile dysfunction prevalence and grade depending on the instrument used to define and classify the condition. Recently others have assessed the agreement of various instruments to establish which may be more practical, reliable and simple to use in large population studies. All findings have stressed the need for further research in this field and for study designs that may clarify the associations, some unexpected and newly evidenced, of which the results may help us to achieve better understanding and management of this important health problem.
APPENDIX: LIST OF QUESTIONS IN SECTION 5—ERECTION CAPACITY
1. Over the past 4 weeks, how often were you able to get an erection during sexual activity? 2. Over the past 4 weeks, when you had erections with sexual stimulation, how often were your erections hard enough for penetration? 3. Over the past 4 weeks, when you attempted sexual intercourse, how often were you able to penetrate your partner? 4. Over the past 4 weeks, during sexual intercourse, how often were you able to maintain your erection after you had penetrated your partner? 5. Over the past 4 weeks, during sexual intercourse, how difficult was it to maintain your erection to completion of intercourse? 6. Over the past 4 weeks, how many times have you attempted sexual intercourse? 7. Over the past 4 weeks, when you attempted sexual intercourse, how often was it satisfactory to you? 8. Over the past 4 weeks, how much have you enjoyed sexual intercourse? 9. Over the past 4 weeks, when you had sexual stimulation or intercourse how often did you ejaculate? 10. Over the past 4 weeks, when you had sexual stimulation or intercourse, how often did you have the feeling of orgasm or climax (with or without ejaculation)? 11. Over the past 4 weeks, how often have you felt sexual desire? 12. Over the past 4 weeks, how would you rate your level of sexual desire? 13. Over the past 4 weeks, how satisfied have you felt with your overall sex life? 14. Over the past 4 weeks, how satisfied have you been with your sexual relationship with your partner? 15. Over the past 4 weeks, how do you rate your confidence that you can get and keep your erection?
574
PREVALENCE AND INDEPENDENT RISK FACTORS FOR ERECTILE DYSFUNCTION IN SPAIN
16. Over the past 4 weeks, how often have you experienced a full erection on awaking? 17. Do you think your partner is . . . a) Very satisfied with sexual activity; b) Satisfied; c) Neither satisfied nor dissatisfied; d) Rather unsatisfied; e) Very unsatisfied. 18. Do you consider yourself a man . . . a) With no erection problem; b) With a minimum incapacity; c) With a moderate incapacity; d) With a severe/complete incapacity for erection. REFERENCES
1. NIH Consensus Conference: Impotence: NIH Development Panel on Impotence. JAMA, 270: 83, 1993 2. Ayta, I. A., McKinlay, J. B. and Krane, R. J.: The likely worldwide increase in erectile dysfunction between 1995 and 2025 and some possible policy consequences BJU Int, 84: 50, 1999 3. Benet, A. E. and Melman, A.: The epidemiology of erectile dysfunction. Urol Clin North Am, 22: 699, 1995 4. Manual de Puntuacio´n de la Versio´n Espan˜ola del Cuestionario de Salud SF-36. Barcelona, Spain: Unitat de Recerca en Serveis Sanitaris, Institut Municipal d’Investigacio´ Me`dica, 1998 5. Ware, J. E., Snow, K. K., Kosinski, M. et al: SF-36 Health Survey manual and interpretation guide. Boston: New England Medical Center, The Health Institute, 1993 6. Ewing, J. A.: Detecting alcoholism. The CAGE questionnaire. JAMA, 252: 1905, 1984 7. Batista, J. E., Regalado, R., Chechile, E. et al: Validation of the International Prostate Symptom Score in Spain. The 2nd International Consultation on Benign Prostatic Hyperplasia: Proceedings. Edited by A. T. K. Cockett, Y. Aso, C. Chatelain, C. et al. Jersey, Channel Islands: Scientific Communication International, p. 129, 1993 8. Rosen, R. C., Riley, A., Wagner, G. et al: The international index of erectile function (IIEF): a multidimensional scale for assessment of erectile dysfunction. Urology, 49: 822, 1997 9. Glasser, D. and Sweeney, M.: The prevalence of erectile dysfunction in four countries: Italy, Brazil, Malaysia and Japan. Cross-National Study Group. Presented at International Consultation on Erectile Dysfunction, Paris, France, July 1–3, 1999 10. Feldman, H. A., Goldstein, I., Hatzichristou, D. G. et al: Impotence and its medical and psychosocial correlates: results of the Massachusetts Male Aging Study. J Urol, 151: 54, 1994 11. Hussain, S. S., Siddiqui, S., Razzaq, M. A. et al: A patient based epidemiology study to find the prevalence and correlates of erectile dysfunction (ED) in two cities in Pakistan. Presented at European Society for Impotence Research Meeting, Barcelona, Spain, January 30 to February 2, 2000 12. Akkus, E., Kadioglu, A., Esen, A. et al: Prevalence of erectile dysfunction among urban and rural population in Turkey. Presented at ESIR Meeting, Barcelona, Spain, XXX, 2000 13. Mahmoud, K. Z., Azim, S. A., Gadallah, M. et al: A patient based epidemiology study to find the prevalence and correlates of erectile dysfunction (ED) in Egypt. Presented at ESIR Meeting, Barcelona, Spain, XXX, 2000 14. Kadiri, N., Berrada, S., Tahiri, S. et al: Prevalence of erectile dysfunction (ED) and its correlates in Morocco: a population based epidemiological study. Presented at ESIR Meeting, Barcelona, Spain, XXX, 2000 15. Dogunro, S., Osegbe, D. and Jaguste, V. S.: Epidemiology of erectile dysfunction (ED) and its correlates: a patient based study in Nigerian population. Presented at ESIR Meeting, Barcelona, Spain, XXX, 2000 16. Koskimaki, K., Hakama, M. and Tammela, T. L. J.: Prevalence of erectile dysfunction in Finland. Presented at ESIR Meeting, Barcelona, Spain, XXX, 2000 17. La Pera, G., Taggi, F., Aloise, L. et al: Prevalence of the erectile dysfunction and of other genitourinary symptoms: preliminary results of the SIMONA study group. Presented at ESIR Meeting, Barcelona, Spain, XXX, 2000 18. Parazzini, F., Menchini-Fabris, F., Bortolotti, A. et al: Frequency and determinants of erectile dysfunction in Italy. Eur Urol, 37: 43, 2000 19. Laumann, E. O., Paik, A. and Rosen, R. C.: Sexual dysfunction in the United States. Prevalence and predictors. JAMA, 281: 537, 1999
20. Cappelleri, J. C., Rosen, R. C., Smith, M. D. et al: Diagnostic evaluation of the erectile function domain of the International Index of Erectile Function. Urology, 54: 346, 1999 21. Cappelleri, J. C., Siegel, R. L., Osterloh, I. H. et al: Relationship between patient self-assessment of erectile function and the erectile function domain of the International Index of Erectile Function. Urology, 56: 477, 2000 22. Derby, C. A., Araujo, A. B., Johannes, C. B. et al: Measurement of erectile dysfunction in population-based studies: the use of a single question self-assessment in the Massachusetts Male Aging Study. Int J Impot Res, 12: 197, 2000 23. Kayigil, O., Atahan, O. and Metin, A.: Multifactorial evaluation of diabetic erectile dysfunction. Int Urol Nephrol, 28: 717, 1996 24. Vibe-Petersen, J., Behrend, L., Jorgensen, T. et al: The prevalence of erectile dysfunction in a male diabetic population. Presented at ESIR Meeting, Barcelona, Spain, XXX, 2000 25. Wein, A. J. and Van Arsdalen, K. N.: Drug-induced male sexual dysfunction. Urol Clin North Am, 15: 23, 1988 26. Burns-Cox, N. and Gingel, C.: Erectile dysfunction: is medication to blame? Prescriber, 3: 77, 1997 27. Mannino, D. M., Klevens, R. M. and Flanders, W. D.: Cigarette smoking: an independent risk factor for impotence? Am J Epidemiol, 140: 1003, 1994 28. Condra, M., Morales, A., Owen, J. A. et al: Prevalence and significance of tobacco smoking in impotence. Urology, 27: 495, 1986 29. Schiavi, R. C.: Chronic alcoholism and male sexual dysfunction. J Sex Marital Ther, 16: 23, 1990 EDITORIAL COMMENTS The study of the prevalence and risk factors of erectile dysfunction on the Iberian Peninsula supplements the widely referenced MMAS from North America. The strength of this prevalence study is that it is population based rather than dependent on prevalence in men seeking urological care. This information combined with data on established risk factors as well as other correlative diseases allows urologists to arrive at a better understanding of erectile dysfunction. To society at large it may hopefully contribute to an improved recognition of erectile dysfunction and a more equitable delivery of health care to this needy subpopulation. The ramifications of confirming established risk factors and the confirmation of suspected or new ones, such as smoking, may help to address treatment or prevention strategies for risk factors that are not genetically or congenitally acquired. Of continued concern are the presumed cultural differences in erectile dysfunction prevalence compared to those in other well described population based studies. The variance of prevalence of moderate to severe erectile dysfunction ranges from 39% in Japan to 10.5% in Spain. Such large discrepancies are seemingly the only explanation of different cultural perceptions of erectile dysfunction. The methodology of measuring erectile dysfunction in this series is somewhat troublesome insofar as relates to use of the overall selfassessment in the so-called simple question. This use brings into question whether this particular questionnaire has been validated by the rigorous methods now commonly applied in population based studies of erectile dysfunction. It is interesting that the perception of erectile dysfunction differs substantially by age group and overall whether it was measured by the IIEF or the so-called simple question. However, the patterns of age related erectile dysfunction noted with either question are more consistent. This population study also provides epidemiological data on the association of erectile dysfunction with other suspected but more controversial topics, such as benign prostatic hyperplasia and cigarette smoking. The challenges were to convert such cross-sectional information into more longitudinal studies, thereby, strengthening the association of risk factors with disease. Kevin T. McVary Department of Urology Northwestern University Medical School Chicago, Illinois The authors report a population based study. Such studies are needed in various regions of the world to improve our understanding of the risk factors of erectile dysfunction and potentially generate hypotheses for preventing erectile dysfunction. Furthermore, this article is especially important because it not only confirmed the role
PREVALENCE AND INDEPENDENT RISK FACTORS FOR ERECTILE DYSFUNCTION IN SPAIN of previously known risk factors such as age, diabetes and cardiovascular disease, but also showed for the first time to my knowledge an association of erectile dysfunction with other medical conditions. The most exciting new finding to urologists is the association of erectile dysfunction with prostate disease. These conditions are common, chronic, slowly progressive and age related. As the baby boomer generation continues to age, the medical profession and specialty of urology are faced with the task of ensuring that this large, aging population remains healthy and vital in the face of extended life expectancy and increasing health care costs. This increase in life expectancy has brought much attention to the topic of geriatric medicine. Naturally urology as a specialty is concerned with many of these important medical conditions. This article focuses specifically on male population and the health problems that frequently plague men. Several major noncancer disease states adversely affect males as they age. These conditions include erectile dysfunction, prostate disease, cardiovascular disease and depression. The conditions often precipitate considerable morbidity and may even result in mortality. Furthermore, the presentation of 1 condition may correlate with the development of another. Available data on these correlations should be enough to warrant a multidisciplinary approach for research. This approach may help to maintain patients healthy and more importantly happy late into life, realizing the concept of successful aging. Ridwan Shabsigh Department of Urology Columbia Presbyterian Medical Center New York, New York REPLY BY AUTHORS In regard to the comment about problems (“troublesome”) that may arise from the use of the so-called simple question when meas-
575
uring erectile dysfunction, this question has been used in other studies for the same purpose (reference 9 in article). The MMAS indirectly (reference 10 in article) and directly (reference 22 in article) uses a global self-assessment question to classify the subjects participating in the study. It was precisely the “nonvalidation by the new current, rigorous methods of that particular questionnaire” which led us to use the IIEF as a tool to measure erectile dysfunction for the first time in a population-based epidemiological study. The uncertainty about its adequate behavior in this setting led us not to reject instruments already used. On the other hand, since the IIEF refers to the last 4 weeks, those subjects who have had no sexual activity during that period would have no data on erectile ability or disability. The use of both instruments allowed us to have 2 measures of the same phenomenon in the same population. Agreement between the 2 tools has been reported (references 21 and 22 in article). In our article we provided preliminary data on the agreement between the 2 instruments in our study sample. The comparison of the results obtained with one or the other tool allowed us to hypothesize that the IIEF could be a good instrument to detect the disease in its earliest stages (16.5% using the erectile function domain of the IIEF versus 5.2% obtained by the simple question regarding mild erectile dysfunction prevalence), since the simple question detects the patient when the degree of involvement is more severe (2.7% versus 6.9% for moderate and severe/complete erectile dysfunction using the erectile function domain and the simple question, respectively). The use of the IIEF would allow detection of the disease in its earliest stages when it could still be amenable to prevention, whereas the subject who answers the simple question classifying himself as having erectile dysfunction does so in the more severe categories. These could be the subjects who really are affected by this disease and, therefore, more prone to search for professional help.