Abstract
Introduction
The prevalence of spondylolysis reported from radiograph-based studies has been questioned in the recent computed tomography (CT)–based studies in adults; however, no new data are available in pediatric patients. Spina bifida occulta (SBO), which has been associated to spondylolysis, may be increasing its prevalence, according to the recent studies in adults in the last decades, but without new data in pediatric patients. We aimed to determine the prevalence of spondylolysis and SBO in pediatric patients using abdomen and pelvis CT as a screening tool.
Patients and Methods
We studied 228 patients, 4 to 15 years' old (107 males), who were evaluated with abdomen and pelvis CT scans for reasons not related to the spine. The entire lumbosacral spine was evaluated to detect the presence of spondylolysis and SBO. We compared the prevalence of spondylolysis in patients with and without SBO. A logistic regression analysis was performed to determine the effect of age and sex as independent predictors of spondylolysis and SBO.
Results
The mean age of all study patients was 11.2 ± 2.7 years. A total of 107 patients were males (46.9%) and 121 were females (53.1%). The prevalence of lumbar spondylolysis was 3.5%, 95% CI: 1.1 to 5.9% (eight patients). Spondylolysis was observed at L5 in seven patients and at L4 in one; seven patients exhibited bilateral spondylolysis, while one patient had a unilateral pars defect. Among the patients with spondylolysis, two exhibited olisthesis (0.9% of the sample), both corresponding to grade I slip. Males exhibited a 4.7% (95% CI: 0.6–8.7) prevalence of spondylolysis (five patients) compared with a 2.5% (95% CI: 0.3–5.3) prevalence in females (three patients), p = 0.48. The prevalence of SBO was 41.2%, 95% CI: 34.8–59.2% (94 patients). Overall, 6 patients had SBO at L5 (6.4%), 80 patients at S1 (85.1%), and 8 patients at S2 (8.5%). SBO was found in 51.4% of males (95% CI: 41.8–61) and in 32.2% of females (95% CI: 23.8–40.7), p < 0.01. The prevalence of SBO decreased with increasing age. The prevalence of spondylolysis was 5.3% in the patients with SBO and 2.2% in patients without SBO, p = 0.28. Logistic regression analysis revealed that male sex (OR = 2.1, 95% CI: 1.2–3.7; p < 0.01) and age (OR = 0.8, 95% CI: 0.7–0.9; p < 0.01) were significantly related to the presence of SBO. Conversely, neither sex (OR = 2, 95% CI: 0.5–8.5; p = 0.32) nor age (OR = 1, 95% CI: 0.8–1.4; p = 0.78) were significantly related to the presence of spondylolysis.
Conclusion
This is the first study using CT as a screening method to evaluate the prevalence of spondylolysis and SBO in a pediatric population. We found a 3.5% prevalence of lumbar spondylolysis and a 41.2% prevalence of SBO in this cohort. Most patients with spondylolysis had a pars defect at L5. We also demonstrated that SBO was independently influenced by male and younger age.