Letter to the Editor: Clinical Outcomes and Mechanical Complications After Adult Spinal Deformity Surgery in Patients Aged 80 Years or Older: A Propensity Score-Matched Comparative Study

To the Editor:

We read with interest the study by Inoue et al, which compares mechanical outcomes after long-segment adult spinal deformity (ASD) reconstruction between patients aged <80 and ≥80 years. ¹ The authors should be commended for focusing on a clinically important and increasingly common population, and for highlighting the relevance of vertebral bone quality and postoperative mechanical alignment. However, several issues warrant caution.

First, the central causal interpretation remains uncertain. The propensity score model adjusted for upper instrumented vertebra (UIV) Hounsfield units (HU), modified frailty index score, preoperative sagittal vertical axis, and lumbar lordosis. These variables are not merely baseline covariates; they may represent manifestations or downstream consequences of biological aging. Thus, matching on them addresses whether an 80-year-old and a younger patient with similar frailty, local bone quality, and deformity have comparable risk, rather than whether age contributes to risk in the real-world ASD population. This is a clinically valid conditional question, but it differs from the broader inference that age itself is unimportant. The attenuation of between-group differences after matching may partly reflect removal of pathways through which aging increases mechanical vulnerability.

Second, the study’s definition of “bone quality” is incomplete from a skeletal-health perspective. Femoral-neck dual-energy X-ray absorptiometry T-scores and UIV HU were reported, but neither measure fully characterizes skeletal fragility in patients with ASD. Degenerative spinal changes, vertebral rotation, vascular calcification, and prior fractures can compromise conventional bone-density interpretation; moreover, HU is a surrogate for local trabecular attenuation rather than a comprehensive assessment of fracture resistance. Bone quality in these patients also depends on prevalent vertebral fractures, bone turnover, vitamin-D status, sarcopenia, fall risk, and, critically, the timing and effectiveness of osteoporosis therapy. Teriparatide has been associated with lower proximal junctional failure after ASD surgery in osteoporotic patients, ² whereas poor bone quality remains a recognized contributor to proximal junctional kyphosis/failure. ³

Third, the use of an ROC-derived HU value near 102 as a potential high-risk range requires restraint. HU thresholds vary with scanner type, acquisition protocol, reconstruction kernel, region-of-interest placement, and vertebral level. More importantly, a threshold derived from a relatively small octogenarian subset with a limited number of PJFx events risks optimism bias. Previous studies have linked low CT attenuation with proximal junctional complications, ⁴ but no universally transportable UIV HU cutoff has been established. Rather than implying a clinical threshold, the data support incorporating HU into a multivariable, externally validated fracture-risk model that also includes age-adjusted alignment, construct characteristics, osteoporosis treatment, and muscle status.

Fourth, postoperative sacral slope should not be interpreted as a modifiable isolated target. Sacral slope is constrained by pelvic incidence and is interdependent with lumbar lordosis, pelvic tilt, thoracic compensation, and global alignment. A low postoperative sacral slope may be a marker of persistent compensation or under-correction, but it may also reflect patient-specific morphology. In older adults, pursuing conventional alignment targets can itself increase proximal junctional stress if correction exceeds age-adjusted alignment needs. Age-adjusted alignment concepts have been proposed specifically to avoid overcorrection-related mechanical complications. ⁵ Therefore, the relationship between postoperative sacral slope and PJFx should be evaluated alongside age-adjusted proportional alignment and distribution of lordosis, rather than as an independent correction objective.

In summary, this study appropriately shifts attention from chronological age alone toward skeletal and biomechanical vulnerability. However, its findings should be viewed as hypothesis-generating: optimal care of octogenarians undergoing ASD surgery requires integrated assessment of bone fragility, muscle function, individualized alignment targets, and perioperative bone-active treatment—not age or a single HU threshold in isolation.

ORCID iD

Lingyu Wang https://orcid.org/0009-0003-1263-8642