Background: Increasing age as a risk factor for erectile dysfunction (ED) is in most studies assumed to be a linear function. If this is not the case the assumption could lead to bias, e.g. when men of different ages are compared in interventional studies on ED. Objective: To explore the risk of developing ED over time for men from different age groups. Materials and methods: A questionnaire was sent to a number of male residents in Gothenburg, Sweden, in 1992 (n = 10,458). Men were randomly selected according to year of birth to obtain several cohorts at 5-year intervals of ages 45, 50, 55 years, etc., up to the age of 85 or older. In 2003 an analogous, slightly expanded, questionnaire was sent to a random sample of men from the age cohorts 46, 51 years, etc. (n = 10,845). A total of 4072 men received both surveys, thereby constituting a group of men followed longitudinally for 11 years. The future risk of developing ED in the different age cohorts, adjusted for a number of ED risk factors, was then assessed. Results: A total of 3257 men responded to both questionnaires (response rate = 80%, age range = 56-103 years). The risk of having ED increased substantially with increasing age, both within each survey and longitudinally between the surveys. The adjusted risk of developing ED within the next 11 years increased with a factor of 10, from 1.8% at the age of 45 years at baseline to as much as 11.4% at the age of 65 years. Conclusion: Age as a risk-factor for ED is a non-linear function and should be adjusted as such to avoid bias when including men of different ages in interventional studies on ED.
Keywords: Longitudinal survey; age; epidemiology; erectile dysfunction; risk-factors.