During 16 years’ follow-up, we documented 200 incident Parkinson’s disease cases among men who provided information on their erectile function in the 2000 survey. Participants with erectile dysfunction defined as having poor or very poor erectile function prior to baseline were older and more likely to smoke, had a higher body mass index, had a higher prevalence of major chronic disease (cancer, stroke, hypertension, myocardial infarction, or diabetes), and consumed larger amounts of caffeine than did those who reported very good function ().
Baseline characteristics in 1986 according to reported erectile function before 1986 in the Health Professionals Follow-up Study*
As expected, the prevalence of erectile dysfunction increased with age (). However, men with Parkinson’s disease had a higher prevalence of erectile dysfunction relative to those without Parkinson’s disease in each age group. In 2000, 68.0 percent of Parkinson’s disease patients reported erectile dysfunction during the past 3 months, relative to 32.0 percent of participants without Parkinson’s disease (p < 0.0001, after adjusting for age, smoking status, and body mass index). Among participants without diabetes, the prevalences of erectile dysfunction were 68.8 percent and 31.2 percent for men with and without Parkinson’s disease, respectively, and for participants with diabetes, 61.9 percent and 38.1 percent, respectively.
Prevalence of erectile dysfunction in 2000 according to Parkinson’s disease (PD) status, Health Professionals Follow-up Study.
Subjects who reported erectile dysfunction before 1986 were 3.8 times more likely to develop Parkinson’s disease, relative to those who reported very good function (multivariate relative risk (RR) = 3.8, 95 percent confidence interval (CI): 2.4, 6.0; p < 0.0001) (). This association was stronger among men who reported erectile dysfunction at younger ages: The multivariate relative risks of Parkinson’s disease were 2.7, 3.7, and 4.0 (95 percent CI: 1.4, 11.1; p = 0.008) for men with first onset of erectile dysfunction (before 1986) at 60 or more, 50–59, and less than 50 years of age, respectively, relative to those without erectile dysfunction (). To test the robustness of the association, we conducted several sensitivity analyses that generated similarly significant results. The multivariate relative risk of Parkinson’s disease comparing men with erectile dysfunction with men with very good function was 3.9 (95 percent CI: 2.5, 6.2; p < 0.0001) after excluding men with prostate cancer at baseline and 4.0 (95 percent CI: 2.5, 6.5; p < 0.0001) after excluding men with cancer and stroke at baseline. The significant associations remained in the lag analyses that excluded the first several years of follow-up: The corresponding relative risks were 3.3 (95 percent CI: 2.0, 5.5; p < 0.0001) in the 4-year lag analyses and 3.2 (95 percent CI: 1.8, 5.8; p < 0.0001) in the 8-year lag analyses. A significant association with Parkinson’s disease risk was also observed for erectile function after 1986; adjusted relative risks were 1.3 and 2.9 (95 percent CI: 1.8, 4.7; p < 0.0001) for participants with good/fair and poor/very poor erectile function, respectively, in a comparison with those with very good function.
FIGURE 2 Relative risk (RR) and 95% confidence interval of Parkinson’s disease according to reported erectile function before 1986, Health Professionals Follow-up Study. Part A shows overall results; part B shows results according to age at first onset (more ...)
We further explored possible interactions of erectile function with age, body mass index, cigarette smoking, caffeine intake, and the presence of diabetes during follow-up (yes/no). None of these interactions was significant. The association between erectile function and risk of Parkinson’s disease was similar in men with diabetes (adjusted RR = 3.7) and those without (adjusted RR = 4.2). An association between erectile function and Parkinson’s disease risk also remained evident in subgroup analyses according to age, smoking status, body mass index, and caffeine intake at baseline (data not shown).