Epidemiology of erectile dysfunction: A cross-sectional web-based survey conducted in an Indonesian national referral

Many epidemiological studies have demonstrated a high Background: prevalence of erectile dysfunction (ED) in different parts of the world. The objective of the present study was to establish the prevalence of ED in a healthy population from Indonesia and risk factors associated with ED. We performed a cross-sectional study to estimate the Methods: prevalence of erectile dysfunction in relatively healthy males in Jakarta that aged 20 to 80 years old. This cross-sectional study utilizes a web-based survey containing a translated version of The International Index of Erectile Function (IIEF-5) in Indonesian. Sexual domain functions in the IIEF-5 include 4 domains of erectile function starting from erection confidence, erection firmness, erection maintenance and sexual satisfaction. The mean age of respondents is 38.7±12.6 years old. Most of the Results: respondents were married (77.2%), had completed or undertaking tertiary education (66.3%) and worked in privately established companies (35.7%). The prevalence of ED was 35.6% (22.3% mild, 13.7% mild to moderate, 3.1% moderate and 0.8% severe). The prevalence of ED ranges from 6.5% in the 20-29 year old group, to as high as 88.0% in respondents aged 60 years old and above. Age, hypertension, stroke, history of heart disease, diabetes, kidney disease, history of prostate operations and interpersonal stress are significantly associated with ED ( = <0.001, <0.001, 0.015 p-value 0.000, 0.01, 0.002, <0.001 and 0.022 respectively). The prevalence of ED in Indonesia is about 35.6%. The Conclusion: prevalence of ED in this study ranges from 6.5% to as high as 88.0%. Age, hypertension, stroke, history of heart disease, diabetes, kidney disease, history of prostate operation and interpersonal stress are significantly associated with ED.


Introduction
Erectile dysfunction (ED) is principally defined as the inability to reach and maintain an adequate erection to achieve satisfying sexual intercourse 1 . Although not life-threatening, erectile dysfunction has a huge impact on the quality of life of the patient as well as their sexual partners with consequential fear, depression and loss of confidence or self-esteem.
Many epidemiological studies have demonstrated high prevalence of ED in different parts of the world. For instance, the Cologne Male Survey have found that the prevalence of erectile dysfunction could be as high as 19.2% amongst a group of male responder aging between 30 to 80 years old 2 . In the United states, about 18.4% of the adult male population (above the age of 20) suffered from sexual dysfunction 3 . Nikolosi et al. also explored sexual dysfunction in a number of Asian countries and found that approximately 15% of the whole Asian population suffers from ED 4 .
A simplified version of International Index of Erectile Function (IIEF-5) is a widely used self-administered questionnaires that has been translated and validated in many different languages. This questionnaire has a high degree of specificity and sensitivity to diagnose and evaluate the prevalence of ED 5 . In addition, this questionnaire is also commonly used for assessing ED severity before and after appropriate therapy 6-8 . One limitation however, is that IIEF cannot be used to distinguish the different pathophysiology's of ED 9 .
To date, the majority of studies looking at ED prevalence have focused on western countries or the Asian continent as a whole. No large population-based survey have been conducted to demonstrate the prevalence of ED in Indonesia specifically. Therefore, the objective of this present study was to establish prevalence of ED in a healthy population of males in Indonesia and risk factors for ED.

Methods
This was a cross-sectional study used to estimate the prevalence of erectile dysfunction in relatively healthy male population in Jakarta aged 20 to 80 years old. This cross-sectional study utilized a web-based survey containing a translated version of IIEF-5 in Indonesian (see extended data 10 ).

Study setting
This study was conducted via a web-based survey in Cipto Mangunkusumo Hospital, Jakarta in March 2019. First data entry began on 1 st March and last data entry was collected on 31 st March 2019.

Participants
The link to the web-based survey was put on display on a banner placed in the Urology Clinic of Cipto Mangunkusumo hospital and anyone who voluntarily opened the survey through the links provided are considered as potential respondents. We did not contact potential respondents individually because it may increase the risk of selection bias. The inclusion criteria were men aged 20-80 years old who completed the questionnaire within the enrollment period. All questionnaires filled beyond the enrollment period were excluded.

Ethical considerations and consent
This cross-sectional study has been approved by Cipto Mangunkusumo Hospital Research and Ethical Committee in February 2019 with an approval number of 0316/UN2.F1/ETIK/2018 before enrolling any participant for the study. Because this study incorporates web-based survey, written informed consents could not be obtained. Consent was taken electronically prior to participants completing the survey.

Data collection and management
We used Google Form to provide online questionnaire and collect respondent's data. We subsequently downloaded the completed questionnaire after the period of enrollment. Access to the completed questionnaire was considered confidential and only accessible to the authors. The online questionnaire was available here. Before being asked to fill in the questions, patients were first provided with a concise description of the study with confidentiality assurances. In order for the respondents to complete the survey, they had to provide their email address as well as their phone number to ensure that the same respondent would not take the survey more than once. We also imposed a mechanism in the survey that provided the respondents with a reminder that will pop up in their window if there are questions in the survey that have not been answered. This helps to assure that all respondents will complete all the questions provided in the survey and therefore minimizing missing data.
The translated IIEF-5 imbedded in the survey was used to determine the prevalence of. Sexual domain functions in the IIEF-5 include 4 domains of erectile function starting from erection confidence, erection firmness, erection maintenance and sexual satisfaction. These tools are useful in determining baseline erectile function as well in assess the outcome of a specific treatment modality 11,12 . Each IIEF-5 questions were provided with five response options, which enables the calculation of scores ranging from 5 to 25 for erectile function. The scores were then categorized into : severe ED (5-7), moderate ED (8-11), mild to moderate ED (12)(13)(14)(15)(16), mild ED (17)(18)(19)(20)(21), no ED (22)(23)(24)(25) 13 The IIEF-5 is limited by its inability to distinguish the etiologic basis for ED especially those of organic origins 9 . In order to address this limitation, some additional examinations should be performed, e.g. RigiScan, Doppler ultrasonography 9 . Unfortunately, additional examination was not possible due to the web-based nature of this study.
Only authors were able to access the completed surveys from a separate page under restricted access. Within this page, the authors are able to convert the data of the completed survey into a Microsoft Excel 2013 file which was later used for statistical analysis. This survey was planned to be maintained within one month period of time.

Statistical analysis
Data was analyzed using IBM SPSS Statistic 20. Demographic characteristics included in this study included the age group of respondents, marital status, educational level as well the occupation of respondents. The age group of the respondents was simplified into <40 years old and ≥ 40 years old groups to define age as a dichotomous variable to enable the subsequent analysis. Normality test using Kolmogorov-smirnov were performed for numerical data. Risk factors, including age, hypertension, stroke, history of heart disease, diabetes, kidney disease, history of prostate operation and interpersonal stress, and their association with ED in this study will be evaluated using chi-square test statistics. The result was considered significant when p < 0.05.

Results
Over 1 month a total of 255 respondents completed the webbased survey, the sociodemographic characteristics of the respondent are described in Table 1 (see underlying data 14 ). The mean age of respondents was 38.7±12.6 years old with the majority of the respondents (31.4%) in the 30-39 years old age group. Most of the respondents were married (197/255, 77.2%) and most had completed or were undertaking tertiary education (169/225, 66.3%). The majority of respondents were employees of private companies (91/255, 35.7%).
Erectile function of the respondents was been stratified into ED severity categorizations based on the established IIEF-5 definitions. The prevalence and severity of ED is described in Table 2 according to age groups. The prevalence of ED in this study was 35.6% (22.3% mild, 13.7% mild to moderate, 3.1% moderate and 0.8% severe). The prevalence of ED ranged from 6.5% in the 20-29 years old age group to as high as 88.0% in respondents aged 60 years and above.
The association of ED with various potential risk factors including hypertension, stroke, history of heart disease, diabetes, kidney disease, history of prostate operation and interpersonal stress were evaluated using Chi-square analysis. The age group of the respondents was simplified into <40 years old and ≥ 40 years old groups to define age as a dichotomous variable to enable the subsequent analysis. Table 3 summarizes the findings; age, hypertension, stroke, history of heart disease, Diabetes, kidney disease, history of prostate operation and interpersonal stress are significantly associated with ED (p-value = <0.001, <0.001, 0.015, <0.001, 0.01, 0.002, <0.001 and 0.022 respectively).

Discussion
The prevalence of ED has been widely evaluated in many different countries around the world, however, the comparison of prevalence between countries may prove difficult ordeal due to the various questionnaires used to assess ED in different studies. Additionally, heterogeneity of age as well as ethnic distribution of different groups of respondents further complicates matters. Two previous studies by Nikolosi et al. and Akkus et al. utilized one simple question "How would you describe yourself" to assess and classify ED severity into No ED, Mild ED, Moderate ED and Severe ED 15, 16 . Other more recent epidemiologic studies employed use of the self-reported IIEF-5 questionnaire, the most commonly used validated questionnaire that had been translated into various different languages worldwide 5, [17][18][19] . Routine use of IIEF-5 in the evaluation of ED prevalence helps in gathering epidemiologic data that are comparable, possibly reproducible, and standardized facilitating cross-cultural comparison.
The discussion of sexual issues, especially those of which are related to male potency may cause embarrassment and stress. This may result in reporting bias, general underestimation or even exaggeration of sexual problems, and even the unwillingness to participate in the study overall. We aimed to address this by using a web-based survey with adaptation of IIEF-5 that any potential respondents can voluntarily access and complete without any coercions, potentially limiting any reporting bias.  The prevalence of ED in this study ranges from 6.5% in the 20-29 age group to as high as 88.0% in respondents agrd 60 years old and above (Table 2). This consistent increase of ED prevalence with age was also demonstrated in a systematic review conducted by Prins et al. which found ED prevalence has been reported to be as high as 7% among men aged 18 to 29 years old, 2-9% among men aged 30-39 years old, 9-11% among men aged 40-49 years old, 16-18% among men aged 50-59 years old, 34% among men aged 60 to 69 years old, and 53% among men aged 70 to 80 years old. The rising trend in prevalence of ED as the population grows older can be explained by the presence of various comorbidities commonly found in the older population such as diabetes, heart disease, hypertension, and other vascular related diseases. These comorbidities are commonly associated with the development of ED 18 .
In this cross-sectional study, independent variables such as age, hypertension, history of heart disease, stroke, diabetes, kidney disease, history of prostate surgery and interpersonal stress are significantly associated with ED ((p-value = 0.000, 0.000, 0.015 0.000, 0.01, 0.002, 0.000 and 0.022 respectively). It is well known that the chances for developing ED increases with advancing age as mentioned by Prins et al. and Rhoden et al. and increases substantially from the age of 40 5,20 . Diabetes mellitus in itself is already an independent and strong predictor for ED development as this particular medical condition is strongly involved with microvascular changes, peripheral neuropathy and endothelial dysfunction 21 . The significant association between history of hypertension, heart disease and interpersonal stress with ED development in this study is also consistent with the findings in a another web-based study conducted by Shaeer et al., however, they did not find any significant association of renal disease with development of ED 17 . This contradictive result might be caused by the ambiguous definition of renal disease used in their survey that may have confused some respondents.
Interestingly, smoking which was generally thought to be associated with risk of ED, was not significantly associated with ED in the current study. This finding is consistent with studies conducted by Oyelade et al. and Shaeer et al. 17,19 . In contrast, an epidemiologic study conducted in Brazil by Moreira et al. found that heavy smoking (≥ 40 cigarettes/day) is a strong predicting risk factor for ED development 22 . The disparity in the results in between these studies is possibly due to the lack of information regarding the number of cigarettes smoked per day, as well as the duration of smoking that was not explored in this study.
Alcohol consumption was not found to be significantly associated with ED, consistent with Oyelade et al. and Shaeer et al. 17,19 . Interestingly, Moreira et al. found that moderate alcohol consumption was inversely related with the development of ED 22 . Low alcohol consumption was found to improve erections and increases libido, likely due to its vasodilatory effect and suppression of anxiety. Large amounts of alcohol consumption, however, may decrease libido due to central sedation 23 . To correctly evaluate the effect of alcohol consumption on development of ED, the amount and duration of alcohol consumption must also be considered, which was not explored in detail in this particular study.
In summary, incorporating IIEF-5 into a carefully designed and well-structured web-based survey assists in collecting epidemiological data on the prevalence of ED that is concise, standardized, and provides easily comparable data to the surveys conducted in other countries. Additionally, web-based survey also help to spare the potential respondents from embarrassment when addressing their sexual issues, which is hoped will provide more objective responses. Lastly, the web-based survey has proven to be cost efficient and is able to cover a large sample of the population at the country level and above.
One limitation of this web-based survey, however, is the difficulty in gathering responses from individuals of lower educational status or socioeconomic level which may have restricted access to internet connection may have resulted in sampling bias. Moreover, the web link to the survey in this study was only available to those attending the urology polyclinic in Cipto Mangunkusumo hospital. Future prevalence studies of ED in Indonesia may involve the distribution of the link to this web-based survey to all health institutions in Indonesia to provide a more accurate estimation of ED prevalence in the Indonesian population.

Conclusion
In conclusion, the prevalence of ED in our study was 35.6% (22.3% mild, 13.7% mild to moderate, 3.1% moderate and 0.8% severe). The prevalence of ED in this study ranges from 6.5% in the 20-29 years old group to 88.0% in respondents aged 60 years old and above. This shows the consistent increase in the prevalence of ED as the population is ages. Additionally, independent variables such as age, hypertension, history of heart disease, stroke, diabetes, kidney disease, prostate operation and interpersonal stress are significantly associated with ED. Data are available under the terms of the Creative Commons Zero "No rights reserved" data waiver (CC0 1.0 Public domain dedication).

Grant information
The author(s) declared that no grants were involved in supporting this work.