Introduction. The association between erectile dysfunction (ED) and cardiovascular risk factors is well established and ED can be considered an early marker for cardiovascular disease.
Aim. To generate a calculator to predict the risk of diabetes, hypertension, hyperlipidemia or angina in men with ED, based on an analysis of data from the Men's Attitudes to Life Events and Sexuality (MALES) 2004 study.
Main Outcome Measure. A logistic regression model using the variables overall health, ED severity, having/not having a sexual partner, and waist size.
Methods. The MALES was a multinational, population-based study conducted in 2001, in which the prevalence of ED and comorbid medical conditions was assessed in 27,839 men aged 20–75 years. In 2004, the cohort of men with ED (N = 1843) were recontacted and 919 (50%) agreed to participate in the MALES 2004 longitudinal study. Multistep analysis of data from 808 patients was performed, with 289 variables evaluated. Only those variables significantly correlated with outcome and those making clinical sense were retained. A logistic regression model was applied to 90% of the sample; results were validated in the remaining 10% with sensitivity and specificity testing.
Results. Of the 2004 cohort, 20.7% had been diagnosed with diabetes, 44.3% with hypertension, 42.5% with hyperlipidemia, and 25.7% with angina. The following modifiable factors affected the risk of comorbidities, and were therefore included in the risk calculator: health status, waist size, ED severity, and having or not having a sexual partner. Using these variables in the model resulted in a sensitivity of 86.2% and specificity of 54.5%. The primary limitation of the calculator is that it is not a prediction calculator.
Conclusion. Erectile dysfunction is a key factor in calculating the probability of major risks to men's health, such as diabetes, hypertension, hyperlipidemia, and angina. Shabsigh R, Shah M, and Sand M. Erectile dysfunction and men's health: Developing a comorbidity risk calculator. J Sex Med 2008;5:1237–1243.