Characteristics of Lumbar Pars Interarticularis and Pedicle Stress Injuries by Sport in 902 Pediatric and Adolescent Athletes: A Retrospective Study

Background: Recent studies utilizing magnetic resonance imaging (MRI) for the evaluation of symptomatic lumbar spondylolysis in pediatric and adolescent athletes have indicated that upper level lumbar involvement has a higher incidence than previously reported. There has been a paucity of literature evaluating sport-specific patterns of lumbar spondylolysis, specifically upper versus lower level involvement. Purpose: To assess the potential risk factors for upper level stress injuries of the lumbar spine in pediatric and adolescent athletes. Study Design: Cross-sectional study; Level of evidence, 3. Methods: The medical records of 902 pediatric and adolescent athletes (364 female, 538 male; mean age, 14.5 ± 2.1 years) diagnosed with symptomatic pedicle and pars interarticularis stress injuries at 2 academic medical centers (July 2016 to June 2021) were reviewed. All patients had undergone MRI at the time of diagnosis. Only patients with pars/pedicle edema on MRI were included. Data regarding single-sport specialization, sport participation, sport category by biomechanics (axial rotation vs extension/axial loading), and vertebral level of injury over the 5-year period were analyzed. Stress reaction or active spondylolysis (SRAS) was the terminology used to designate grade 1, 2a, or 3 stress injuries according to the adapted Hollenberg classification system on MRI. Upper level vertebrae were defined as L3 or superior, whereas lower level vertebrae included L4 or inferior. Results: Of the 902 patients with SRAS injuries, most (n = 753 [83.5%]) had exclusively single-level lower stress injuries, while 67 (7.4%) had multilevel stress injuries. There were 82 athletes (9.1%) who had single-level upper stress injuries. Athletes with upper level pars/pedicle stress injuries were older at the time of diagnosis (15.8 ± 1.9 vs 14.3 ± 2.1 years, respectively; P < .001), had a shorter duration of low back pain before presentation (2.50 ± 2.70 vs 4.14 ± 6.73 months, respectively; P < .001), were more likely to specialize in a single sport (43.9% vs 32.3%, respectively; P = .046), and had a lower incidence of active spondylolysis on MRI at the time of diagnosis (42.7% vs 59.8%, respectively; P = .004) compared with athletes with lower level stress injuries. Athletes with lumbar stress injuries who specialized in a single sport had nearly twice the odds of having upper level involvement compared with multiple-sport athletes (adjusted odds ratio, 1.80 [95% CI, 1.06-3.04]; P = .03). Athletes with active spondylolysis on MRI at the time of diagnosis had nearly half the odds of having upper level involvement (adjusted odds ratio, 0.55 [95% CI, 0.33-0.91]; P = .02). Conclusion: Age at the time of diagnosis, duration of low back pain, single-sport specialization, and presence/absence of active spondylolysis on MRI at the time of diagnosis were primary predictors of whether an athlete’s lumbar stress injury was classified as either upper or lower level involvement. Overall, the variables included in multivariate analysis were modest predictors, explaining only 15.1% of the variance in the rates of lumbosacral stress injuries classified by spinal level. These specific biomechanical factors and other potential contributors to these findings warrant further investigation.