Kyphotic deformity is an abnormal force line across the coronal and sagittal planes of the spine caused by congenital dysplasia, trauma, tuberculosis, ankylosing spondylitis and other reasons. It manifests as abnormal bending of the spine backward, which not only exerts a serious impact on patients’ physical appearance but may also induce multiple problems such as low back discomfort, nerve compression, cardiopulmonary dysfunction due to the disorder of spinal mechanical structure, and seriously reduces the quality of life of patients (1). The disease can be caused by congenital vertebral dysplasia (such as hemivertebrae, wedge-shaped vertebrae), spinal tuberculosis, ankylosing spondylitis, traumatic fracture or osteoporotic compression fracture and other etiologies (2). With the development of modern imaging and spinal surgery technology, posterior osteotomy incorporating bone graft fusion and internal fixation has become the mainstream method for the treatment of moderate to severe kyphosis. Its core goal is to correct structural deformities by osteotomy, promote bone fusion by bone graft, maintain the orthopedic effect and rebuild spinal stability by internal fixation system (3).

Within this surgical system, the selection of fixation segments—specifically long segment fixation (LSF) or short segment fixation (SSF)—stands as one of the critical factors influencing surgical outcomes. LSF usually spans multiple adjacent normal vertebral bodies superior and inferior to the deformity area, and theoretically can provide stronger biomechanical stability, effectively reduce focal stress accumulation at the junctional zone after operation, which helps decrease the risk of orthopedic loss and early radiographic proximal junctional kyphosis (PJK) (4). However, the cost is to sacrifice more spinal motion segments, increase surgical duration, intraoperative hemorrhage, and soft tissue parameters dissection range, which may cause the acceleration of adjacent segment degeneration (ASD) (5). In contrast, SSF only fixes the core segment of the lesion and its directly adjacent vertebral body, featuring the advantages of reduced surgical invasion, shorter procedural time, and superior retention of motor function, and is in line with the concept of minimally invasive spine surgery (6). However, its mechanical strength is relatively limited. When dealing with patients with severe deformities or osteoporosis, there are risks of internal fixation failure, reduction loss and long-term correction loss (7). Understanding the curative effect of long and short segment fixation, postoperative adverse effects and risk factors of kyphosis recurrence can improve our medical level, better avoid risks and reduce postoperative complications. At the same time, the hospital stay of patients was shortened, the treatment cost was reduced, and better postoperative effect was obtained.

In current clinical practice, there is still a lack of unified standards for the selection of LSF and SSF, which is controversial. Several retrospective studies have tried to compare the efficacy of the two, but most of them did not fully control the confounding bias, which affected the reliability of the results. For example, retrospective cohort analysis of 127 adult cases of degenerative scoliosis by Jiang et al. found that although LSF has more advantages in correcting spinal imbalance, its perioperative complication rate is significantly higher than that of SSF, suggesting that the benefits and risks need to be weighed (8). Similarly, Zhang et al.'s. study also pointed out that implementation of long segment internal fixation represents an independent risk factor contributing to medical complications following surgery in elderly patients with ADS (9). These studies highlight the importance of controlling for baseline differences when comparing between groups.

In order to overcome the limitations of traditional comparison methods, probability score matching (PSM), as an effective statistical tool, is widely utilized in spinal surgical studies to minimize selection bias (10). Through PSM to balance the distribution of key covariates such as age, sex, basic disease, preoperative deformity degree between the two groups, the comparison between SSF and L group can be closer to the level of randomized controlled trials, so as to improve the reliability of the conclusion. In addition, combined with multivariate linear regression or logistic regression analysis, it can further identify independent predictors of adverse outcomes such as postoperative orthopedic loss, internal fixation failure, and early radiographic PJK, such as low Bone Mineral Density (BMD), severe preoperative imbalance, improper selection of upper fixed vertebra (UIV), and no use of transverse devices (11, 12).

In view of this, this study used a retrospective clinical control study method to compare and analyze the perioperative indicators, imaging parameters, functional recovery and complications of the two groups of patients with kyphosis who were undergoing LSF or SSF, accordingly. The clinical application value of the two fixation methods was objectively evaluated, and the linear regression analysis was utilized to examine the influencing factors, thereby furnishing a scientific foundation for the individualized selection of clinical fixation strategies and the prevention and control of related risks.

Imaging findings from the present study demonstrated a significant improvement in both the Cobb angle and SVA of patients in both groups immediately postoperatively, as compared to their preoperative measurements, but the correction rate of the L group was higher immediately after operation, and the Cobb angle loss rate 12 months after operation was significantly lower than that of the S group, and the excellent rate of sagittal balance and the poor rate of vertebral body alignment were significantly better than those of the S group (P < 0.05). The core mechanism of this result lies in the load dispersion effect at the biomechanical level. LSF extends the fixation range to the spinal mechanical stability zone by spanning more than two intact vertebral bodies adjacent to the upper and lower segments of the deformity area, effectively dispersing the stress concentration at the two extremities of the fixed segment, avoiding the mechanical imbalance caused by “isolated fixation of deformity segments” during SSF (17). Previous studies have shown that long segment fusion can more effectively maintain the coronal and sagittal spinal balance, while conferring superior long-term stability in the correction of adult degenerative scoliosis (22). In addition, finite element analysis further confirmed that LSF can significantly reduce the peak stress of the fixed segment and adjacent segments, reduce the risk of vertebral body displacement and angulation, and provide a solid biomechanical foundation for the long-term maintenance of orthopedic effect (23). Specifically, long segment fusion improves the stiffness and stability of the overall internal fixation system by increasing the number of fixed segments, so as to better resist the non physiological activities of the spine under the postoperative dynamic load, and reduce the risk of internal fixation loosening, fracture or bone fusion failure caused by micro motion (22). In contrast, although short segment fusion has shorter operation time and less blood loss, it is difficult to fully correct serious structural deformities due to its limited fixation range, and it is more prone to the loss of Cobb angle and the progression of sagittal imbalance in long-term follow-up, which may be related to the increase of compensatory load of adjacent segments (24). Therefore, for patients with more complex adult degenerative scoliosis such as Lenke-Silva VI type, long segment fusion combined with osteotomy technology has clear advantages in obtaining and maintaining a good spinal force line, which is helpful to improve the clinical function prognosis and quality of life of patients, although the surgical trauma is relatively large (25). In addition, modern spine surgery increasingly emphasizes accurate surgical planning based on individualized anatomical parameters (such as pelvic incidence angle PI, sacral inclination angle SS, lumbar lordosis LL, etc.) to achieve optimal reconstruction of sagittal balance (23). The long segment fusion strategy is more conducive to the overall orthopedic design according to these parameters, ensuring the postoperative LL-PI matching, thus reducing the incidence of long-term adjacent segment disease (26, 27). It should be noted that the sagittal balance assessment in this study was restricted to SVA, without incorporating pelvic parameters such as PI, PT, and LL. Pelvic parameters are important for evaluating sagittal spinal-pelvic harmony, but the core focus of this study was the direct effect of fixation segment length on kyphotic deformity correction and global sagittal alignment restoration. SVA, as a widely recognized global sagittal balance indicator (17, 22), can effectively reflect the overall spinal force line improvement after osteotomy, which is sufficient to address the research objectives. However, the lack of pelvic parameter analysis may limit the comprehensive evaluation of sagittal spinal-pelvic coordination, which is a consideration for future studies. In conclusion, long segment fusion shows superior biomechanical stability and long-term orthopedic effect in the treatment of complex adult spinal deformities, and its clinical decision should comprehensively consider the severity of deformity, bone condition, age and general condition of patients.

Perioperative outcomes indicated that both the operative duration and intraoperative blood loss in Group S were significantly reduced compared with those in Group L (P < 0.05), which aligns with the research findings reported by Ma et al. (28) on thoracolumbar fractures. SSF achieved the advantage of “minimally invasive” by reducing the scope of soft tissue dissection, shortening the time of screw placement, and simplifying the intraoperative fluoroscopy steps. According to the concept of accelerated rehabilitation surgery (ERAS), trauma reduction can reduce the peak of postoperative inflammatory response (lower levels of C-reactive protein and interleukin-6), help reduce the systemic stress response, and thus promote early ambulation and functional recovery (29, 30). In this study, although group L experienced greater surgical trauma, the functional scores (ODI and VAS scores) 12 months after surgery were significantly better (P < 0.05), suggesting that patients’ tolerance to short-term trauma may exchange for better long-term spinal stability and functional benefits (22). Notably, no statistically significant difference was observed in the duration of hospital stay between the two groups (P > 0.05), indicating that LSF did not prolong the overall rehabilitation cycle, which may be related to better immediate postoperative stability, better pain control and lower complication rate in group L (31, 32). In addition, multimodal analgesic strategies under the ERAS pathway, especially opioid sparing regimens, have been shown to effectively manage postoperative pain while reducing adverse reactions, providing similar comfort base for patients with different fixation methods (33). Therefore, the results suggest that when making decisions on the treatment of spinal fractures, clinicians should not only focus on perioperative trauma indicators such as operation time and blood loss, but should comprehensively evaluate the individual situation, fracture type, expected stability needs and long-term functional prognosis of patients, in order to achieve individualized precise treatment in the true sense (29).

Complications are the key factors affecting the prognosis of kyphosis surgery, and early radiographic PJK and internal fixation loosening/fracture are the most common adverse events after surgery. Combined with previous studies and basic medical mechanisms, the occurrence of complications such as early radiographic PJK and loosening of internal fixation may be closely related to the disorder of tissue repair caused by local excessive inflammation after surgery, in addition to the difference in mechanical stability (34). Findings from the present study revealed that the incidence of early radiographic PJK, internal fixation loosening/fracture, and total complications in Group L was significantly lower than that in Group S (P < 0.05), suggesting that LSF has significant advantages in the prevention and control of complications. This result is different from some previous studies. For example, the study of Jiang et al (8) on adult patients with degenerative scoliosis believed that the perioperative complication rate of LSF was higher, but the PSM method was not used to control the baseline confounders in this study. Potential significant disparities in key indicators—including age, BMD, and deformity severity—between the two groups might have introduced bias into the results. Following PSM, the baseline characteristics of the two groups were balanced, which eliminated the interference of confounding factors and rendered the results more credible. A notable finding in our study was the lower complication rate in the LSF group compared to the SSF group, which contrasts with some historical cohorts reporting higher perioperative risks associated with longer constructs. This discrepancy is likely attributable to the PSM design of our study. In non-matched retrospective series, patients undergoing LSF often present with more severe deformities, older age, or worse comorbidities (selection bias), inherently predisposing them to complications. By strictly matching baseline characteristics—particularly the preoperative Cobb angle and age—we minimized this bias. Consequently, our results suggest that when patient baseline conditions are equivalent, the superior biomechanical stability of LSF may actually reduce implant-related complications (e.g., screw loosening, rod breakage) compared to the biomechanically more vulnerable SSF, without significantly increasing perioperative systemic risks.

Multivariate logistic regression analysis revealed that SSF was an independent risk factor for increased postoperative Cobb’ angle loss rate and a key risk factor for early radiographic PJK. As the primary independent risk factor for postoperative orthopedic loss and early radiographic PJK, the essence of SSF is the stress concentration effect caused by insufficient mechanical stability. After spinal correction, a mechanical interface is formed between the fixed segment and the adjacent unfixed segment. The interface fixed by the short segment is too close to the deformity area to effectively disperse and buffer the residual corrective torque after surgery, resulting in the high concentration of stress in the adjacent upper vertebra and its upper intervertebral disc structure (35). LSF transfers the high stress mechanical interface from the fragile deformity correction apex region to the higher spinal segment with relatively normal anatomy by extending the fusion range (36). This region has better vertebral bone conditions and more uniform mechanical conduction, thus reducing the risk of borderline kyphosis (37). In addition, the multi screw anchor system with LSF can provide stronger anti rotation and anti shear ability, especially in the weight-bearing state of the spine, which can effectively resist the biomechanics leading to orthopedic loss (38).

Based on our findings, we propose a preliminary clinical decision framework for determining the fusion length in ADS. LSF is strongly indicated for patients with larger significant sagittal imbalance, or osteoporosis, where a long lever arm is crucial to prevent correction loss and mechanical failure. Conversely, SSF might be cautiously considered for a select subgroup of patients: those with flexible curves, mild deformity, good bone quality, and significant medical comorbidities that prohibit prolonged surgical time. In these specific cases, SSF offers a ‘less is more’ approach, provided that the patient and surgeon accept a potentially higher risk of correction loss in exchange for reduced surgical trauma. Finally, future research should assess the cost-effectiveness of these strategies. While LSF involves higher upfront costs due to more implants and longer operating room time, it may offer long-term economic benefits by reducing the rate of revision surgeries and late-stage complications compared to SSF.

This study is a retrospective clinical controlled study. Although strict inclusion and exclusion criteria and independent researcher data collection and analysis ensure objectivity, there is still inherent selection bias in retrospective studies. The study is a single center with limited sample size and concentrated patient background, which may reduce the universality of the results. The 12-month follow-up is insufficient to assess long-term outcomes such as PJK and ASD. These complications are often related to long-term spinal biomechanical changes and degenerative processes, and their incidence and severity are more likely to be accurately reflected in follow-up periods of ≥2 years. The 12-month follow-up was insufficient to evaluate the long-term efficacy and long-term complication risk after spinal orthopedic surgery, and the confounding factors such as postoperative rehabilitation compliance and bone graft material type were not included in the regression model. In the future, it is suggested to carry out a multi center prospective randomized controlled study with large samples, extend the follow-up time, expand the range of independent variables, explore the biomechanical mechanism combined with basic and clinical research, and improve the reliability of the results and clinical reference value.

This study confirmed that LSF was significantly better than SSF in maintaining kyphosis, restoring sagittal balance, reducing early radiographic PJK, and internal fixation failure, but the operation time and blood loss increased; SSF has less trauma, but the risk of orthopedic loss is higher in older patients with osteoporosis and severe deformities. Regression analysis showed that SSF, preoperative Cobb ≥ 60°, and low BMD were independent risk factors for postoperative orthopedic loss, while SSF, low BMD, and advanced age were predictors of early radiographic PJK. Individualized decision-making is needed in clinic: long segment or bone cement reinforcement is preferred for patients with poor bone mass and severe deformity; In young patients with good bone mass, the short segment can be carefully selected and the follow-up can be strengthened. In the future, it is necessary to carry out prospective research, extend follow-up, explore AI aided decision-making system, and promote the development of orthopedic treatment precision and value medicine.