Abstract
Background:
Previous research describes imaging via magnetic resonance imaging (MRI) as the gold standard for the diagnosis of spondylolysis after clinical examination. Existing literature on the accuracy of physical examination findings related to positive acute spondylolysis on MRI is limited.
Purpose:
To evaluate the diagnostic value of clinical examination maneuvers in assessing acute spondylolysis in adolescent athletes with low back pain related to positive radiographic findings.
Study Design:
Cohort study (Diagnosis); Level of evidence, 3.
Methods:
Data were abstracted from a sports medicine registry that prospectively collects data from the electronic health record (EHR) for patients seen for orthopaedic conditions across a regional health care network. Patients aged 8 to 18 years assessed for lumbar back pain via a standardized lumbar back pain assessment protocol and who had completed an MRI study between January 2019 and February 2024 were included. Patient information, pain duration, radiographic findings, and pain-associated physical assessment maneuvers were abstracted from the EHR.
Results:
Of 733 patients meeting study criteria, 260 (35.5%) had findings of acute spondylolysis on MRI. The mean age at initial evaluation was 14.8 years, with most patients (40.6%) presenting with 1 to 3 months of lumbar back pain. Overall, 94.2% of patients with acute spondylolysis on MRI had pain with hyperextension of the lumbar spine on physical examination, and 52.3% had pain with the single-leg hop maneuver. Those reporting pain with both hyperextension and single-leg hop had the highest prevalence of acute spondylolysis. Male patients, patients aged 13 to 14 years, and those presenting with 2 to 4 weeks of back pain had the highest prevalence of MRI-confirmed acute spondylolysis.
Conclusion:
This study evaluates the clinical accuracy of physical examination and patient characteristics associated with prevalence of acute spondylolysis in patients who present with lumbar back pain confirmed via MRI. A higher prevalence of acute spondylolysis was observed across several subgroups, including males, early to mid-adolescents, those with shorter symptom duration, and those demonstrating both hyperextension and single-leg hopping pain. Incorporating both maneuvers improved diagnostic accuracy for identifying acute spondylolysis. Findings demonstrate that both hyperextension and single-leg hopping pain may warrant advanced imaging to rule out acute spondylolysis as a cause of lumbar back pain.
Low back pain is a global health concern, affecting >619 million people in 2020 and is the leading cause of years lived with disability.8,15 In the adolescent athlete population, one of the more common and clinically relevant sources of low back pain is spondylolysis. 10 Spondylolysis is defined as a disruption in the pars interarticularis of a vertebra. 14 This occurs almost exclusively in the lumbar vertebrae and can be unilateral or bilateral. Spondylolysis can be seen in the context of a developmental defect or can occur as an acute injury. In the acute phase, early identification and rest may result in bony healing. If spondylolysis occurs as a developmental defect, then there is no chance of bony healing, even with early detection.
An acute spondylolysis is managed with rest, either with or without bracing, followed by physical therapy to address mechanical deficits in flexibility or strength or pelvic stabilization. Management can become surgical in those who fail nonoperative treatment.15,21 A chronic spondylolysis with no healing potential is treated with physical therapy and surgery if physical therapy fails.
Magnetic resonance imaging (MRI) is the preferred initial evaluation tool for acute spondylolysis diagnosis after a clinical examination due to its high sensitivity and specificity and lack of radiation.5,13 Computed tomography (CT) is advantageous for the visualization of bony anatomy; however, it is of secondary preference to MRI due to radiation concerns and the inability to distinguish between an acute fracture with bony healing potential and a chronic deficit.
Previous research has reported a wide range of spondylolysis prevalence among pediatric and adolescent athletes, from approximately 14% to nearly 40%, depending on study population and methodology.18,20,22,23,26,32 Previously identified risk factors for developing acute spondylolysis include being male, having tight hamstrings, having parents with the condition, and comorbidities like spina bifida occulta, Marfan syndrome, and other spine-related conditions.1,2,9,15,29,31 Sport type has also been implicated, particularly for sports with repetitive axial loading that may also have hyperextension and/or rotation of the lumbar spine.2,9,14,15 Although prior predictive models for spondylolysis have been proposed,2,28 no universally validated model currently exists to aid clinicians in diagnosing spondylolysis in youth.
The objective of this study was to evaluate the clinical effectiveness of physical examination when evaluating acute spondylolysis in adolescents. We aimed to evaluate the prevalence of acute spondylolysis in relation to certain patient characteristics, pain duration, examination findings, and vertebral level to aid clinicians in the determination of whether advanced imaging for suspected acute spondylolysis is indicated during patient visits.
Methods
Participants
Patients were part of a sports medicine registry that prospectively collects data from the electronic health record (EHR) for patients seen for various orthopaedic conditions across a regional health care network. Patients provided informed consent to have their personal health information included in this secure registry. This study was approved by the Children's Hospital of Philadelphia Institutional Review Board (#20-017341).
For this study, patients aged 8 to 18 years who were assessed for lumbar back pain and completed an MRI study between January 2019 and February 2024 were included. Spondylolysis was not required to be visible on plain radiographs for patients to undergo MRI. MRI studies were ordered at the discretion of pediatric sports medicine physicians for a range of clinical concerns, including but not limited to suspected spondylolysis, disk herniation, or nonspecific causes of lumbar back pain. All patients were evaluated by pediatric sports medicine specialists within our regional network.
Procedures
Patients were evaluated via a standardized lumbar back pain assessment protocol that included gathering a history and performing a physical examination to assess range of motion, strength, and neurological function. Sex; age; Child Opportunity Index (COI); duration of pain; mean hours per week of activity; pain with forward bending; pain with midline, right-sided, and left-sided hyperextension; pain with single-leg hop; and mean popliteal angle were abstracted from the registry. COI is a robust, nationally-normed score developed from census data as a composite measure of community-level infrastructure from resource availability and quality, and other geographic and neighborhood factors that contribute to childhood development. 6
All MRI studies were performed using standard lumbar spine protocols at our institution, which typically include sagittal T1-weighted and T2-weighted short tau inversion recovery sequences. MRI scans were classified as positive based on the presence of pars interarticularis edema with or without a frank defect, and as negative if no edema was present. All scans were interpreted by musculoskeletal radiologists and the ordering sports medicine pediatricians within our pediatric network, and findings were recorded as positive or negative for acute spondylolysis in the EHR. Clinical decision-making was made by consensus read of the MRI scans. Plain radiographs were consistently obtained in this cohort but were not included in our analysis since we were specifically interested in acute spondylolysis. Most clinicians obtained anteroposterior and lateral lumbar spine radiographs and not oblique views when evaluating for acute spondylolysis, since oblique views are more effective for visualizing chronic pars defects but are less useful in identifying early acute defects, which was the focus of this evaluation. For this study, MRI served as the reference standard: patients with MRI-confirmed acute spondylolysis were classified as positive, and all others were classified as negative. Images that demonstrated chronic pars defects without associated edema on MRI were interpreted as negative in our study.
Duration of pain was measured by the approximate number of days that the patient had been experiencing symptoms at the time of the visit. These were subsequently compiled into groups consisting of <2 weeks, 2 to 4 weeks, 1 to 3 months, 4 to <6 months, and ≥6 months. Pain with forward bending was assessed by having the patient stand with their feet together and having them bend forward and attempt to touch their toes to the best of their ability. Hyperextension pain was assessed with the patient in a similar position, but the clinician situated behind the patient with their hands on the patient's shoulders to gently guide them into midline, right-sided, and left-sided hyperextension. Unilateral hyperextension was achieved by rotating the trunk in one direction, and then gently guiding the patient into lumbar extension. The record for pain with forward bending and hyperextension pain was documented as a binary yes or no. Single-leg hop was performed by having the patient stand on one leg and hop up and down for 3 repetitions and then switching to the opposite leg. This was also recorded as a binary yes or no for ipsilateral pain.
Statistical Analysis
Descriptive statistics are presented as means and standard deviations for continuous variables and as counts and percentages for categorical variables. Sensitivity and specificity were calculated to evaluate the diagnostic performance of individual and combined physical examination maneuvers for identifying MRI-confirmed acute spondylolysis. Descriptive analysis was used to summarize demographic, physical examination, and radiographic data to evaluate the prevalence of acute spondylolysis in relation to patient characteristics and examination findings.
Results
A total of 733 patients were included, all of whom underwent MRI. Among these, 260 (35.5%) demonstrated MRI-confirmed acute spondylolysis, which is consistent with previously reported prevalence ranges for this population. From our overall study cohort, males had a 53.6% prevalence of acute findings compared with 22.7% in females. Demographic information on those with acute spondylolysis is provided in Table 1A. The mean age at initial evaluation was 14.8 years, with most patients (40.6%) presenting with 1 to 3 months of lumbar back pain. Unilateral spondylolysis (n = 149) was more prevalent than bilateral acute spondylolysis (n = 111), and the L5 vertebra was the most common level of acute findings, in both males (n = 105; 63.6%) and females (n = 60; 36.4%).
Demographics and Clinical Characteristics of Patients With Acute Spondylolysis a
All values are reported as mean ± SD or n (%). Child Opportunity Index (COI) is represented on a 0-100 scale.
The prevalence of acute spondylolysis with respect to sex, age, pain duration, and physical examination symptoms is summarized in Table 1B. In younger patients, females had more instances of acute spondylolysis compared with their male counterparts. The age group with the greatest proportion of acute spondylolysis observed on MRI was the 13- to 14-year age group, with 97 of 240 patients (40.4%) having acute spondylolysis. Patients presenting to the sports medicine clinic with a duration of 2 to 4 weeks of lumbar back pain who were referred for MRI had a 50% prevalence of subsequent acute spondylolysis findings (50/100 patients). Those reporting pain with both hyperextension and single-leg hop had the highest prevalence of acute spondylolysis findings, with 133 of 297 patients (44.8%) with positive findings. These results are also visually presented in Figure 1.
Prevalence of Acute Spondylolysis by Age, Pain Duration, and Physical Examination Findings a
All values are reported as n (%) or n.
Magnetic resonance imaging (MRI)–confirmed acute spondylolysis by physical examination maneuver.
Regarding sporting activity, patients playing rugby, soccer, football, basketball, and baseball who were referred for MRI had the highest prevalence of acute spondylolysis findings, all exceeding 50%. The distribution of sports participation among the overall cohort, along with sport-specific prevalence rates of acute spondylolysis, is summarized in Table 2.
Breakdown of Sports Participation and Acute Spondylolysis Prevalence a
All values are reported as n (%).
Of the 260 patients, 245 (94.2%) with acute spondylolysis on MRI had pain with hyperextension of the lumbar spine on physical examination. A total of 136 patients (52.3%) had single-leg hopping pain on physical examination, and 133 patients (51.2%) with acute spondylolysis reported pain with both hyperextension and single-leg hop on physical examination, compared to 297 of 733 patients (40.5%) from the overall study cohort reporting a combination of hyperextension and single-leg hopping pain. The prevalence of pain with the standardized lumbar back pain assessment protocol is summarized in Table 3 for both the entire study cohort and those with acute spondylolysis. Diagnostic performance values for each physical examination maneuver are also shown in Table 3, with hyperextension alone demonstrating 88.2% sensitivity and 18.8% specificity, single-leg hop alone demonstrating 2.4% sensitivity and 98.4% specificity, and the combination of both maneuvers demonstrating 91.7% sensitivity and 24.4% specificity.
Diagnostic Agreement Between Physical Examination Findings and MRI Results for Acute Spondylolysis a
All values are reported as n (%) unless otherwise indicated. Dashes indicate values that were not calculated due to lack of a true positive or true negative group for whole sample analyses.
Discussion
In our cohort, 35.5% of patients who received MRI exhibited positive findings of acute spondylolysis, which aligns with prior epidemiological findings in this population.18,20,22,23,26,32 We found that L5 was the most common location of acute findings, with 63.4% of acute spondylolysis at this level, followed by L4 at 27.7%. These results are consistent with recent findings from Kriz et al, 12 whose cohort demonstrated L5 and L4 spondylolysis prevalence rates of 65.3% and 23.6%, respectively.
Like Kriz and colleagues, 12 we also found soccer, basketball, and baseball to be among the sports associated with the highest prevalence of acute spondylolysis on MRI, along with football (18/28) and rugby (1/1), all with >50% prevalence in our cohort. These findings are consistent with prior evidence describing athletes in higher-load sports involving repetitive extension and/or rotation, such as wrestling, weightlifting, rugby, American football, gymnastics, diving, and cricket, as more likely to present with spondylolysis.4,7,15,17 In our cohort, gymnasts had the highest number of MRI studies ordered compared with athletes of other sports, although only 27 of 101 (26.7%) demonstrated acute spondylolysis on imaging. Rather than suggesting risk or incidence, these results highlight how certain sports may warrant a higher index of clinical suspicion and more tailored screening approaches for athletes presenting with lumbar back pain. Clinicians should be aware of the types of sports and activities that involve excessive loading of the lumbar spine and maintain spondylolysis in their differential diagnosis when evaluating athletes with back pain. It is important to note, however, that a multitude of factors are associated with evaluations of acute spondylolysis, including level of sport specialization, training volume and intensity, and strengthening and conditioning, rather than sport played alone, as emphasized by Kriz and colleagues. 12
We found that patients in the 13- to 14-year age group presenting to our clinic with MRI studies had the highest prevalence of acute spondylolysis (97/240; 40.4%), followed by patients ages 11 to 12 years (32/92; 34.8%). This finding partially aligns with previous research by Nitta et al, 18 who examined spondylolysis prevalence in school students with low back pain and reported a spondylolysis rate of 9.3% in elementary students, 59.3% in junior high school students, and 31.5% in high school students. Another group studying school-aged children found a high prevalence of nonunion (28%) in adolescents with acute spondylolysis and alarming rates for risk of progressive pathology in those patients, 27 which underscores the importance of appropriate and timely treatment. However, radiographic healing may not directly correlate with symptom improvement. Klein et al 11 conducted a meta-analysis demonstrating that while most patients treated nonoperatively achieved successful clinical outcomes, radiographic healing of pars defects occurred in only a minority, suggesting that clinical recovery is not dependent on bony union.
Of the 302 males who were included, 162 (53.6%) had acute spondylolysis on MRI, compared with just 22.7% of females (98/431), which is consistent with previous literature.3,15,24,30 In our data, younger males had lower rates of acute spondylolysis compared with females in the same age groups but had higher rates in mid-adolescence. This may be expected since males tend to have growth spurts 2 years later than females. Together, these findings may reflect a sex-specific vulnerability to bone integrity during adolescent development and pubescence. 19 In addition to biological factors, differences in sports participation between males and females may also contribute to this difference. Male athletes are more likely to engage in sports involving repetitive lumbar extension and rotation, such as football, baseball, and wrestling, which may play a role in explaining the higher prevalence observed in this group.
Our study identified patients presenting with 2 to 4 weeks of pain and receiving an MRI study to have the highest rate of acute spondylolysis findings, with 50% of this cohort (50/100 patients) exhibiting positive findings, with rates declining after 4 weeks. Previous literature has described longer durations of pain in athletes with pars/pedicle stress injuries at L4 and below. 11 As such, duration of pain should be taken in the context of the full complement of patient characteristics and symptoms when determining clinical concern for spondylolysis rather than an independent risk factor. Our results can be viewed in the context of prior efforts to develop clinical prediction models for spondylolysis.2,28 While our study did not attempt model generation, these earlier works highlight the ongoing need to refine and validate clinical criteria for this population.
When examining patient reports of pain with the various physical examination maneuvers associated with the standardized lumbar back pain examination, both the overall study cohort and those with confirmed acute spondylolysis demonstrated similarly high rates of hyperextension back pain (90.0% and 94.2%, respectively). With single-leg hopping pain, however, the discrepancy between those with acute spondylolysis compared to the overall study population was more apparent (52.3% vs 41.6%).
When evaluating these factors with respect to subsequent acute findings on MRI, among patients evaluated with hyperextension pain only, 30.9% (112/363) had MRI-confirmed acute spondylolysis. In contrast, when identifying those reporting a combination of both hyperextension pain and single-leg hopping pain, the rate of acute findings on subsequently obtained MRI rose to 44.8% (133/297 patients).
To further assess diagnostic accuracy, hyperextension was found to be highly sensitive (88.2%) but poorly specific (18.8%) for identifying acute spondylolysis. In contrast, the single-leg hop maneuver demonstrated very low sensitivity (2.4%) but high specificity (98.4%), while the combination of both maneuvers yielded the highest sensitivity (91.7%) with modest specificity (24.4%).
These findings suggest that the addition of hopping pain as a component in physicians’ clinical concern for acute spondylolysis can be useful in increasing the sensitivity of physical examination findings. Because of the axial loading of the spine that is produced with hopping, this can elicit a force similar to that which is created with hyperextension movements in sport that are most associated with spondylolysis. Because of this, hopping can be used to reproduce the patient's pain if spondylolysis is present. Furthermore, the use of a single-leg hop concentrates most of this force to only one side of the spine, allowing for a higher likelihood that painful symptoms are reproduced compared to distributing the load evenly throughout both pars interarticulares. Several different studies have linked different loading conditions with risk for spondylolysis, particularly in adolescent athletes.3,17,25
The key finding in this study is that when incorporating positive single-leg hopping pain on physical examination with hyperextension back pain, the sensitivity of predicting a positive acute spondylolysis on subsequent MRI is increased. The utility of these 2 physical examination maneuvers in tandem with other patient characteristics such as age, sex, and details of sports participation can allow clinicians to better identify when concern for spondylolysis should be high and warrants referral for MRI. While these findings provide useful clinical insight, they are exploratory in nature and should be validated prospectively before integration into diagnostic algorithms.
Limitations
In this case-control study, MRI studies were ordered by physicians based on their clinical evaluation. The decision to order an MRI study was based on history and physical examination findings. An MRI study was not ordered for every case of adolescent back pain, so the data reflect evaluation only when MRI was clinically indicated and do not imply incidences of adolescent back pain. MRI studies could be ordered for any clinical concern including spondylolysis, disk herniation, or unexplained lumbar back pain. In our study, 660 of 733 had hyperextension pain on physical examination. Spondylolysis would have been in the differential diagnosis for almost every patient, even if that was not the main clinical suspicion, and as such, we felt that including all MRI studies was clinically relevant. This study specifically looked at the clinical characteristics of those with positive acute spondylolysis MRI findings. Furthermore, we were interested in identifying acute pars injuries since clinical treatment would be different for acute versus chronic injuries. Patients with chronic pars defects and lumbar back pain who had an MRI study with no pars edema were included as negative MRI results, even though the pars defect may have been partially contributing to mechanical back pain. All MRI scans were interpreted by musculoskeletal radiologists and sports medicine pediatricians within our pediatric network; however, we did not perform an independent reliability assessment of radiological interpretation. Because this was a retrospective study using registry data not originally collected for this specific research purpose, the reliability of certain physical examination measures could not be independently verified. This registry includes patients evaluated within a single regional pediatric sports network and may not capture all cases of acute spondylolysis treated outside this setting.
Data were collected by retrospective review of a prospective sports medicine registry including multiple sports medicine physicians within a regional network. While the clinical evaluation and decision to order MRI will not be identical, the physical examination data collection template is the same, so the examination was done in the same manner by all clinicians, limiting bias in the physical examination findings. We did not conduct any developmental measurements, including measures such as Tanner staging; however, we did notice that females in the 8- to 10-year and 11- to 12-year age groups had a higher prevalence of acute spondylolysis compared with males, but males demonstrated a higher prevalence as a whole, so future work may help to clarify relationships between pubescence and the development of acute spondylolysis. Our findings are exploratory and require prospective validation before incorporation into clinical decision-making.
Also, lateral bending has been examined in previous studies but was not included for analysis within this population. Thus, we are unable to comment on the utility of lateral bending in association with acute spondylolysis in pediatric and adolescent patients.
Finally, the timing of the office evaluation may not have captured all acute physical examination findings. Some patients had been resting before their appointment and may have had resolution of pain at the time of the documented physical examination. An MRI study could have been ordered based on clinical history alone, such as recurrent back pain with activity in an athlete despite having no hyperextension pain on the day of evaluation. Some patients reported pain with sports activities but did not report pain in the office, which may underestimate the number of patients with positive examination findings with positive MRI findings.
Conclusion
Adolescent back pain has traditionally been associated with high spondylolysis rates, up to 47% in some studies, 16 particularly among athletes in extension-based sports such as gymnastics. In our cohort, hyperextension pain alone was highly sensitive (88.2%) but poorly specific (18.8%) for identifying acute spondylolysis and was associated with a 30.9% prevalence of MRI-confirmed findings. When both hyperextension and single-leg hop pain were present, the prevalence of acute spondylolysis increased to 44.8%, with combined maneuvers yielding the highest sensitivity (91.7%) and modest specificity (24.4%). These findings suggest that single-leg hopping pain should be incorporated into patients’ physical examination when concern for spondylolysis is present. While extension sports, such as gymnastics and dance, did have positive findings, many other sports had higher positive MRI rates. Therefore, clinical context as well as a physical examination which includes both extension and single-leg hopping can allow clinicians to more effectively determine when advanced imaging is warranted.
Footnotes
Acknowledgements
The authors acknowledge all the physicians who completed the clinical examinations and read the MRI scans. They also thank all the patients who sought care at the Children's Hospital of Philadelphia and consented to research.
Final revision submitted November 12, 2025; accepted November 21, 2025.
One or more of the authors has declared the following potential conflict of interest or source of funding: This study was funded by the Children's Hospital of Philadelphia Child Surgeons Association. AOSSM checks author disclosures against the Open Payments Database (OPD). AOSSM has not conducted an independent investigation on the OPD and disclaims any liability or responsibility relating thereto.
Ethical approval for this study was obtained from Children's Hospital of Philadelphia Institutional Review Board (#20-017341).
Data Accessibility Statement
The data can be made available on reasonable request by contacting the corresponding author.
