Restoration of spinopelvic alignment is a primary goal in ASD surgery, with the L1PA serving as a reliable global alignment parameter [1, 2]. Although L1PA can be modified intraoperatively, the direct surgical maneuvers to change the L1PA may not be intuitive, as the femoral heads and L1 vertebrae cannot be directly moved. Moreover, understanding how and where to apply segmental lordosis may help achieve L1PA goals. As such, to inform surgical correction strategies, the current study evaluated the influence of segmental lordosis changes at individual lumbar levels on postoperative L1PA. In our cohort, overall alignment improved following surgery, but only a minority of patients reached the “target” L1PA. Importantly, only changes at L4/5 and L5/S1 were significantly associated with reductions in L1PA, with L4/5 showing the strongest influence. In contrast, changes at more cranial lumbar segments had little measurable effect on L1PA. These findings highlight the lower lumbar spine as the most effective intraoperative target for optimizing sagittal alignment.

Our study demonstrated that changes in lower lumbar lordosis, particularly L4/5 and L5/S1, were the only levels significantly associated with changes in L1PA. This finding is not unexpected, as the majority of lumbar lordosis in a normally aligned spine resides within the L4–S1 region, which exerts the greatest influence on global sagittal balance [14]. L4/5 and L5/S1 together contribute an average of 35–40° of the lumbar lordosis, which remains constant across individuals and does not vary with PI [15], indicating that most of a spine’s necessary lordosis is “anchored” in the lower segments [15, 16]. Increase in the lower lumbar lordosis (L4-S1) shifts the L1 vertebra posteriorly relative to the pelvis, which reduces the angle between L1 and the femoral heads, i.e. the L1PA [1, 16]. Greater lordosis often coincides with a decrease in PT, reinforcing this effect [17].

Among the two levels contributing to the lower lumbar lordosis, correction at L4/5 had a slightly greater influence on decreasing L1PA than L5/S1. Although the difference was not statistically significant, this finding was somewhat surprising, as we expected that changes at the lowest level, L5/S1, would theoretically lead to the largest decrease/improvement in L1PA. Several biomechanical and surgical factors may explain this trend. The study may be limited by the minimal degree of correction achieved at L5/S1 (mean 0.7°), which likely reduced the statistical power to detect an association between lordosis change at that level and L1PA. Achieving substantial lordosis at L5/S1 is often constrained by sacral morphology and the inherent stiffness of the lumbosacral junction [18]. As a transitional segment, L5/S1 frequently exhibits reduced facet joint dimensions and increased coronal orientation, both of which limit segmental mobility and contribute to its rigidity [18]. Furthermore, studies have shown that adding an L5/S1 ALIF in long constructs often yields minimal additional sagittal correction [19]. Notably, L5/S1 demonstrated the smallest increase in lordosis overall, with more than half of patients exhibiting a decrease at this level, underscoring the difficulty of achieving correction at L5/S1. In contrast, greater increases in lordosis were observed at L4/5, possibly indicating greater segmental mobility, which enabled more reliable detection of its contribution to changes in L1PA.

L1PA is a global alignment parameter that cannot be directly adjusted intraoperatively once a long construct is instrumented [2, 16]. While achieving the target L1PA remains a central objective in contemporary ASD surgery, surgeons must recognize that this global parameter is directly influenced by segmental lumbar lordosis, particularly within the lower lumbar spine [2, 16]. Strategic correction at mobile segments such as L4/5 and L5/S1 can meaningfully shift L1PA, underscoring the importance of individualized, metric-driven alignment planning. Lower lumbar lordosis can be optimized through ALIF, TLIF, osteotomies, and rod contouring, each tailored to anatomy and fusion status. Techniques like cage placement, posterior release, and segmental compression at L4–S1 offer powerful tools to influence global alignment. Intraoperative long-cassette radiographs help verify L1PA, and if off-target, targeted adjustments at mobile segments can recalibrate alignment [7].

Despite the clinical relevance of our findings, several limitations merit consideration. First, the retrospective, single-institution nature of the study may limit the generalizability of findings. Second, although our regression model controlled for preoperative segmental lordosis as well as changes at each lumbar level, unmeasured confounders such as variability in surgical technique, instrumentation-related factors, and pelvic compensation may have influenced final postoperative alignment. Third, the proportion of patients achieving the target L1PA was relatively low, which could reflect the heterogeneity of ASD severity in our cohort and the multi-year nature of our database. Likewise, although we consulted with several statisticians and experts in the field, we certainly acknowledge that our statistical methodology has inherent flaws. Although multivariable linear regression including all segmental levels in a single model allowed for direct comparison of level-specific effects while accounting for interdependence between adjacent segments, this approach assumes linear and additive relationships that may oversimplify the complex, nonlinear biomechanics of spinal alignment. Additionally, inclusion of correlated segmental variables introduces the potential for multicollinearity, which may attenuate or obscure the true effect size of individual levels. Alternative modeling strategies, such as hierarchical or biomechanical modeling frameworks, could have been performed to better capture these relationships. Nonetheless, the selected approach was chosen for its interpretability and clinical applicability, though these tradeoffs should be considered when interpreting the findings.

It should also be noted that L1PA is influenced not only by angular changes across the L1-S1 segment, but also by variations in intervertebral height, which cumulatively alter the spatial relationship between L1, S1, and the hip axis. Accordingly, surgeons can modify the L1PA not only through angular correction, but also height restoration. Although intervertebral height was not captured in our analysis, future work should incorporate segment-specific disc height measurements to further refine the mechanistic understanding of L1PA modulation. In particular, distinguishing the relative contributions of height restoration versus angular correction may help guide selection of interbody techniques and implant strategies aimed at optimizing global sagittal alignment. Furthermore, inclusion of patients with mixed sagittal and coronal deformity represents a source of heterogeneity that may confound interpretation of sagittal alignment–specific effects on L1PA. Finally, while L1PA is strongly associated with clinical outcomes, our study did not directly assess the impact of segmental correction on patient-reported outcomes, which warrants future investigation. While L1PA has demonstrated promising associations with postoperative outcomes, further validation across diverse patient populations and surgical techniques is needed before routine incorporation into clinical decision-making.