Login to Access Video or Poster Abstract: MP84-04
Sources of Funding: None

Introduction

Multiple disease states and medications have been correlated with erectile dysfunction (ED). There are currently no methods to predict future erectile function or to determine the impact of multiple comorbidities and medication use on loss of erectile function over time.

Methods

The Olmstead County Study of Urinary Symptoms and Health Status among Men represents prospective, ongoing study of men aged 40 and older and living in Olmsted County, Minnesota. The database includes detailed information on demographics, laboratory testing, medications, and comorbid conditions as well as assessments of ED using the Brief Male Sexual Inventory (BMSI) beginning in 2002. Subsequent BMSI questionnaires were completed in 2004, 2006, and 2009. Men without a sexual partner or those with prior pelvic surgery were excluded. Multivariable modeling was performed to determine factors associated with BMSI scores at the 2002 time point. Significant variables were then included in a weighted, step-wise scoring system to assess the predictability of loss of erectile function over time.

Results

937 patients met inclusion criteria, with a mean age of 61 years (SD 8.7). Mean BMSI scores in 2002, 2004, 2006, and 2009 were 7.4 (3.6), 6.9 (3.7), 6.6 (3.8), and 6.0 (3.9), respectively. Multivariate analysis revealed that increasing age (parameter estimate [PE] -0.2 / year over age 40), arterial disease (myocardial infarction, stroke, coronary artery disease; PE -0.7), diabetes (PE -0.7), insulin use (PE -2.0), and beta blocker (PE -0.6) were all associated with a decreased BMSI score (p<.05). The combination of significant factors into a predictive model demonstrated an overall r-squared value of 0.30, suggesting that only 30% of the variability in BMSI scores over time could be accounted for using these factors. In contrast, the use of the BMSI 2002 results demonstrated an r-squared value of 0.50, suggesting that baseline erectile function is a superior predictor for future ED compared to assessments of comorbid conditions and medication use.

Conclusions

While multiple comorbidities and medications are associated with ED, they fail to produce a predictive model superior to baseline assessment of erectile function. These findings suggest that despite the known association between select comorbid conditions, medications, and ED, the presence of these conditions does not result in a more rapid deterioration in erectile function over time compared to cases with no comorbid conditions. This information provides a significant addition to our current understanding of the natural history of ED progression.

Funding

None