The study was approved by Liaocheng People's Hospital. Written consent was waived since this is a retrospective study. This retrospective clinical study included 124 patients with adolescent femoral shaft fractures treated in our department. The patients were divided into two groups: 62 treated with AIN and 62 treated with locking plate fixation. The inclusion criteria were: age 11–16 years, body weight 49–59 kg, high-energy trauma as the cause of injury, unilateral femoral shaft fracture, postoperative follow-up of at least 1 year with complete clinical and imaging data from injury to bony union, and surgery performed within 7 days of injury. The exclusion criteria were: pathological fractures, open fractures, concurrent fractures in other parts of the limbs, associated vascular or nerve injuries in the lower limbs, and underlying conditions such as osteogenesis imperfecta or neuromuscular diseases. To minimize selection bias, our study utilized propensity score matching for patient selection. The matching criteria included age, body mass index (BMI), injury cause, and AO classification of fracture, with a caliper value of 0.05.

Data were expressed as n (percentage, %) or mean ± standard deviation (SD). The normality of the data before analysis was examined using Anderson–Darling test. The comparisons of data between the two groups were done by unpaired t test with Welch's correction, Chi-square test or Fisher's exact test. The significant statistical difference was determined when p values were less than 0.05.

The demographic and clinical characteristics of adolescents with femoral shaft fractures treated with either AIN or LP were summarized in Table 1. Propensity score matching was employed to minimize selection bias, using age, BMI, injury cause, and AO classification of fracture as matching criteria, with a caliper value of 0.05. The process of the subject selection was shown in Supplementary Figure S1. The comparison revealed no significant differences between the two groups in terms of age (AIN: 13.75 ± 1.24 years, LP: 13.69 ± 1.38 years, p = 0.582), BMI (AIN: 23.35 ± 0.49 kg/m², LP: 23.16 ± 0.57 kg/m², p = 0.184), gender distribution (boys: AIN 61.3%, LP 53.2%; girls: AIN 38.7%, LP 46.8%; p = 0.468), injury cause (accident: AIN 32.3%, LP 40.3%; fall injury: AIN 46.8%, LP 45.2%; sports injury: AIN 20.9%, LP 14.5%; p = 0.522), and AO classification of the fracture (A: AIN 61.3%, LP 58.2%; B: AIN 24.2%, LP 20.9%; C: AIN 14.5%, LP 20.9%; p = 0.630).

Figure 1 illustrated the progression of fracture healing and the stability of fixation methods in adolescents treated with either an AIN or a LP over time. The typical x-ray images of a 13-year-old boy with a left femoral shaft fracture treated with AIN are shown at different stages: pre-operation, post-operation, 1-month post-operation, 3 months post-operation, and at the last follow-up (Figure 1A). The corresponding x-ray images for a 14-year-old boy with a left femoral shaft fracture treated with LP are also shown at these stages (Figure 1B). The images in both panels highlight the sequential healing process and the effectiveness of the respective treatment methods in maintaining fracture alignment and promoting bone healing.

The comparisons between the AIN and LP groups revealed significant differences in several surgical and early postoperative parameters (Figure 2). The incision length was significantly shorter in the AIN group (p < 0.001), with a mean incision length of 6.8 cm compared to 8.5 cm in the LP group (Figure 2A). Intraoperative blood loss was also notably lower in the AIN group, averaging 120 ml compared to 170 ml in the LP group (p < 0.001, Figure 2B). Additionally, the time to walking with load post-operation was significantly shorter for the AIN group, with patients beginning weight-bearing activities at an average of 34 days post-surgery compared to 47 days in the LP group (p < 0.01, Figure 2C). These findings indicate that the AIN procedure is less invasive and facilitates earlier mobilization compared to the LP method.

At 3 months post-operation, significant differences were observed in the widest diameter of the femoral head and the NAHS between the two groups (Figure 3). The AIN group exhibited a larger widest diameter of the femoral head and neck (p < 0.05, Figures 3A,B) and higher NAHS (p < 0.001, Figure 3C), suggesting better early postoperative hip function and geometry. Specifically, the mean NAHS was 87 in the AIN group vs. 75 in the LP group. However, by the last follow-up, these differences were no longer significant, with both groups showing comparable outcomes in the widest diameter of the femoral neck and head, as well as in the NAHS (Figures 3D–F). This convergence indicates that while both treatment methods ultimately lead to satisfactory long-term recovery, the AIN group achieves faster initial improvements in hip function and femoral geometry.

We next compared the levels of serum inflammatory cytokines and pain mediators between the AIN and LP groups 1-week post-operation. To establish baseline comparability, the serum levels of these inflammatory markers and pain mediators pre-operation were assessed and the data confirm that there were no significant differences between the groups before surgery (Supplementary Figures S2A–E).

However, there were significant differences in the levels of inflammatory markers and pain mediators between the two groups. Specifically, the LP group had higher serum concentrations of IL-6 (Figure 4A), IL-1β (Figure 4B), IL-8 (Figure 4C), PGE2 (Figure 4D), and pain mediator P (Figure 4E) compared to the AIN group. These results indicate that the AIN method is associated with a lower inflammatory response and reduced pain levels 1-week after surgery, suggesting it is less traumatic and more effective in minimizing postoperative pain.

To compare the levels of bone formation and resorption markers between the AIN and LP groups 1-month post-operation, we assessed the bone metabolism indicators between the two groups. There were no significant differences in serum levels of BALP, BGP, and β-CTX between the groups before surgery (Supplementary Figures S3A–C). Interestingly, 1-month after operation, the AIN group showed higher levels of bone formation markers BALP (Figure 5A) and BGP (Figure 5B), and lower levels of the bone resorption marker β-CTX (Figure 5C) compared to the LP group. These findings suggest that the AIN treatment promotes better bone formation and less bone resorption 1 month after surgery, which correlates with faster fracture healing observed in the AIN group.

Our findings indicated that AIN was less invasive than LP, as evidenced by significantly shorter incision lengths and lower intraoperative blood loss. Specifically, the mean incision length in the AIN group was 6.8 cm, compared to 8.5 cm in the LP group, and the average blood loss was 120 vs. 170 ml, respectively. These differences suggest that the AIN procedure minimizes tissue damage and bleeding, leading to a more favorable surgical profile. Additionally, patients in the AIN group began weight-bearing activities on average 34 days post-surgery, significantly earlier than the 47 days observed in the LP group. Early mobilization is crucial for reducing the risk of postoperative complications such as deep vein thrombosis and muscle atrophy, and it contributes to overall better recovery outcomes (13–15). Similar trends were reported that minimally invasive approaches in femoral fracture repair were associated with reduced soft tissue disruption and faster rehabilitation (16).

The AIN group exhibited superior early postoperative hip function and proximal femoral geometry compared to the LP group at 3 months post-operation. The widest diameters of the femoral head and neck were significantly larger in the AIN group, and the NAHS was higher, with a mean score of 87 compared to 75 in the LP group. These metrics suggest that AIN better preserves the anatomical integrity of the femur and supports hip function during the initial recovery phase. However, by the last follow-up, the differences in proximal femoral geometry and NAHS between the two groups were no longer significant, indicating that both treatment methods achieve similar long-term recovery outcomes. This convergence implies that while AIN offers faster initial improvements, both methods are effective in restoring function and anatomy over time.

A previous study that retrospectively analyzed the functional outcomes of LP vs. AIN for extra-articular distal femoral fractures noted a faster, albeit not statistically significant, union rate for AIN compared to LP (17). Our results similarly indicated that AIN generally leads to quicker fracture healing, supporting its advantage in facilitating faster recovery. Additionally, a meta-analysis of 20 studies involving 1,384 patients with proximal humeral fractures found that intramedullary nails were superior to LP in terms of shorter incision length, less peri-operative bleeding, reduced operation time, and faster fracture healing (18). These findings are consistent with our observations. Another study reported that AIN resulted in significantly lower shoulder pain and higher median Constant and ASES scores compared to LP, with fewer complications and revision surgeries (19). Collectively, these studies reinforce that while both AIN and LP are effective in achieving long-term clinical outcomes, AIN offers distinct advantages in the early postoperative period and in reducing complications.

Furthermore, similar patterns have been observed in femoral neck fracture treatments, where minimally invasive intramedullary fixation techniques were associated with improved short-term hip function and preservation of proximal femoral alignment (20). Moreover, biomechanical analyses have demonstrated that intramedullary implants distribute load more efficiently along the femoral axis compared to lateral plates, which may explain the superior early outcomes observed in our study. For instance, a biomechanical comparison using synthetic bone models found that proximal femoral intramedullary nails sustained approximately 1.78-fold greater axial load before failure than locking proximal anatomic femoral plates, highlighting their superior load-sharing capacity under axial compression (21). Additionally, finite element analysis studies have shown that intramedullary nails bear higher von Mises stresses and transmit loads centrally through the bone's mechanical axis, reducing stress concentration in the implant—a biomechanical advantage that correlates with improved early fracture stabilization (22).

Intramedullary nailing can trigger a “second hit” effect on a patient physiology, which includes several inflammatory and physiological responses (23–25). This second hit is caused by increased pressures within the intramedullary canal, intravasation of fat particles, activation of coagulation, marrow embolization to the lungs or brain, and overheating of the endosteum. Femoral reaming is associated with increased blood loss from the fractured extremity.

Various biomarkers have been studied to understand the cumulative effects of these factors during AIN (26). Key biomarkers such as IL-6, IL-8, IL-10, and TNF-α have been identified. IL-6, in particular, is a significant marker of the early post-traumatic immune response and is elevated in both reamed and unreamed nailing. Studies have shown that the type of surgical procedure, initial trauma severity, specific injury combinations, and timing of surgical intervention are critical factors influencing the inflammatory response (27). The period between 48 h and the 5th day post-trauma is identified as suboptimal for major surgeries due to the heightened risk of systemic inflammatory response syndromeand complications (26, 28). Effective resuscitation and careful timing of surgical interventions are crucial to mitigating the second hit's impact. Here, we found that 1-week after operation, the LP group had significantly higher serum levels of inflammatory cytokines IL-6, IL-1β, and IL-8, as well as pain mediators PGE2 and pain mediator P, compared to the AIN group. These elevated levels indicate a higher inflammatory response and greater pain in the LP group, suggesting that the AIN method is less traumatic and more effective in minimizing postoperative pain. Reduced inflammation and pain are associated with faster recovery and improved patient comfort, underscoring the advantages of AIN in postoperative management (29).

BALP and BGP were used as bone formation markers, and β-CTX was employed as a bone resorption marker. BALP promotes bone mineralization (30), while BGP regulates bone matrix crystallization and reflects recent osteoblast activity (31, 32). β-CTX, released during increased osteoclast activity, is the gold standard for assessing bone resorption (33). Bone metabolism markers provide insight into the biological processes underlying fracture healing. One-month post-operation, the AIN group demonstrated higher levels of bone formation markers BALP and BGP and lower levels of the bone resorption marker β-CTX compared to the LP group. These findings suggest that AIN promotes better bone formation and less bone resorption, correlating with faster and more robust fracture healing. Enhanced bone formation is crucial for the stabilization and consolidation of fractures, while reduced bone resorption minimizes the risk of delayed union or nonunion. The absence of significant differences in these markers pre-operation further supports the postoperative benefits observed with AIN.

Our study comparing AIN and LP fixation for adolescent femoral shaft fractures revealed significant advantages of AIN in terms of surgical impact, early recovery, and bone metabolism. These findings align with broader research in pediatric orthopedics. Chin See et al. demonstrated the efficacy of rigid intramedullary nailing in children with metabolic bone diseases, supporting its application beyond typical cases (34). Another study explored bioabsorbable intramedullary nails for pediatric forearm fractures, highlighting potential benefits in eliminating secondary surgeries but also noting technique-dependent complications (35). For overweight pediatric patients with narrow medullary canals, Tang et al. proposed combining elastic stable intramedullary nails with temporary external fixators as an alternative to locking compression plates, showing improved outcomes in operation time, blood loss, and union rates (36). Collectively, these studies, including our own, underscore the importance of tailoring fixation methods to patient-specific factors such as age, weight, bone quality, and fracture characteristics. While AIN demonstrates clear advantages in our adolescent cohort, the evolving landscape of pediatric orthopedic fixation techniques suggests a nuanced approach is necessary. Future research should focus on refining selection criteria for various fixation methods, optimizing techniques to minimize complications, and conducting long-term follow-up studies to ensure sustained positive outcomes across diverse patient populations.