Real Clear Journal Club

Science Figures Interpreted and Analyzed by RealClearScience

Guys, What's Worse than Erectile Dysfunction?

By Alex B. Berezow

Death. Definitely death.

In the graph, the X-axis classifies men into four groups according to severity of erectile dysfunction (ED): None, mild, moderate, and severe. The y-axis represents the number of deaths from all causes per 1,000 person-years. (A "person-year" is equivalent to one person being studied for one year; therefore, 10 person-years could be 10 people followed for 1 year, or 5 people followed for 2 years, or 2 people followed for 5 years, etc.) The men are further divided into two groups: Those who previously had cardiovascular disease (CVD), and those who have not had it.

The data shows a very clear trend: The more difficult it is for a man to "get it up," the likelier he is to die... and not from embarrassment. ED is a marker of cardiovascular disease (CVD). Therefore, the worse the ED, the worse the CVD. And the worse the CVD, the likelier a man is to die.

Here's the data again, presented another way (with some statistical analysis added) for men who have never had CVD:

Even mild or moderate ED is linked to a higher risk of death. (Although, the confidence interval crosses 1.00, meaning that the results for "mild" and "moderate" are not statistically significant.) Notably, severe ED is linked to roughly a doubling of the risk of death. (Note: RR* is the relative risk adjusted for age only, while RR# is the relative risk adjusted for age, smoking, alcohol, BMI, and various other confounders.)

The moral of the story: If you're not as awesome as you once were in the sack, that could be a sign of a more serious health problem. Go to the doctor.

Or, in the words of Rob Grenfell, listen to that "canary in your trousers."

Source: Banks E, Joshy G, Abhayaratna WP, Kritharides L, Macdonald PS, et al. (2013) Erectile Dysfunction Severity as a Risk Marker for Cardiovascular Disease Hospitalisation and All-Cause Mortality: A Prospective Cohort Study. PLoS Med 10(1): e1001372. doi:10.1371/journal.pmed.1001372


Recent Journal Club

Journal Club Archives