ABSTRACT
OBJECTIVE: To determine if new onset of low back pain in adults could be secondary to lumbar spondylolysis by establishing the age-related prevalence in the general population by examining patients undergoing computed tomography (CT) for reasons unrelated to back pain. MATERIALS AND METHODS: The records of 2,555 patients who had undergone abdominal and pelvic CT in 2008 were reviewed electronically. In order to determine a true representation of the general population, we reviewed all indications for CT, excluding patients with a primary complaint of low back pain as the primary indication for imaging. Equal numbers of patients were separated into age groups by decade to ensure an even distribution of ages for statistical analysis. Patients older than 70 years were grouped together to provide case numbers comparable to those of the other decades. Logistic regression analysis was performed to evaluate the significance of the results. Three board-certified radiologists, including two musculoskeletal fellows and a radiology resident, retrospectively evaluated CT scans for lumbar spondylolysis, including unilateral and bilateral defects. RESULTS: Of the 2,555 cases evaluated, there were 203 positive cases of defects of the lumbar pars interarticularis. This corresponded to an overall prevalence of 8.0%. Prevalence per decade was fairly evenly distributed and ranged from 7.0%( ages 30-39 years) to 9.2% (ages 70 years and above). Prevalence of ages 20-49 years was 7.9%, and that of ages 50 years and older was 8.0%. Male to female ratio was 1.5:1. Logistic regression showed no significant increase in spondylolysis based on age. CONCLUSION: No significant increase in the prevalence of lumbar spondylolysis was demonstrated in patients older than 20 years. This suggests that the development of symptomatic lumbar pars defects do not occur in this population and should not be considered as a rare but potentially treatable cause of new onset low back pain in adults. This study demonstrated an overall prevalence of pars defects of 8.0% in our population. As demonstrated in previous studies, the male to female ratio of 1.5:1 was a statistically significant difference.
