Day: October 9, 2026

Spine (Phila Pa 1976). 2026;51(19):1354-1362

Body mass index is readily available during preoperative assessment, but its implications for cervical myelopathy surgery are not straightforward. A technically more demanding operation does not necessarily imply less neurological recovery, and similar average outcomes do not establish equal risk for every patient. Okubo and colleagues address this distinction in a prospective Japanese multicenter cohort, relating BMI to operative variables, cervical alignment and two-year clinical outcomes. Their findings associate obesity with longer procedures and greater blood loss, while showing no statistically significant differences in neurological improvement across BMI categories. The principal relevance is to individualized planning and counseling rather than selection of a specific surgical technique.

Objective

To assess the association between BMI and operative demands, perioperative complications, cervical alignment, neurological recovery and patient-reported outcomes after surgery for degenerative cervical myelopathy.

Methods

The study drew on a prospective cohort from ten high-volume spine centers in Japan between 2018 and 2022. Of 1,105 patients identified with degenerative cervical myelopathy, 935 entered the final analysis after exclusions for previous cervical surgery, missing essential information, loss to follow-up or death during observation.

Patients were classified as underweight (BMI below 18.5; n=35), normal weight (18.5-24.9; n=522), overweight (25.0-29.9; n=292), or having obesity (30 or above; n=86). Posterior decompression predominated in every group; anterior and posterior fusion procedures were less frequent. Complications were defined as adverse events within 30 days of surgery.

Assessment included Japanese Orthopaedic Association (JOA) scores, pain ratings, SF-36 physical and mental component scores, neuropathic pain symptoms, and cervical radiographs. Clinical outcomes at two years were compared using general linear models with adjustment for demographic, comorbidity and imaging variables. Models for outcome changes also incorporated operative time, blood loss and postoperative alignment and motion measurements.

Main results

Patients with obesity had mean operative time of 128.3 minutes and blood loss of 91.2 mL, versus 102.1 minutes and 38.1 mL in normal-weight patients; global comparisons across the four groups were significant. OPLL prevalence was 47.7% versus 16.9%. Adjusted JOA improvement was 3.5, 3.2, 2.6 and 3.1 points across ascending BMI categories (P=.129). Reported complication categories did not differ significantly. Obesity remained associated with less cervical lordosis and motion at two years. Between-group differences in pain, SF-36 and neuropathic pain changes were not statistically significant; this does not demonstrate equivalent outcomes.

Interpretation

The decision informed by this study is how BMI contributes to preparation and expectations when clinical and imaging findings already support surgery. Greater operative demands justify attention to associated comorbidities and cervical anatomy, particularly the higher frequency of OPLL, rather than assuming that neurological recovery will be poor. However, the study does not compare approach-selection algorithms or validate a BMI-specific exposure, fixation or closure technique. Persistent differences in alignment and motion also should not be described as surgical failure: they coexisted with neurological improvement. Conversely, favorable average recovery should not be interpreted as eliminating individual perioperative risk.

Limitations

This observational cohort cannot establish the independent causal effect of BMI. Patients with obesity were younger and more frequently had OPLL, diabetes and hypertension. Statistical adjustment reduces some measured confounding but does not resolve unmeasured differences. Moreover, adjusting for operative demands and postoperative radiographic findings may account for intermediate pathways through which BMI could affect outcomes, rather than estimating its total association with recovery.

The analysis included patients selected for surgery and excluded those who died or were lost during follow-up. Consequently, its complication comparisons cannot be generalized to all patients initially considered for surgery or used to establish mortality safety. Care at high-volume Japanese centers and the predominance of posterior decompression further limit extrapolation to other settings or procedure mixes.

Only 35 patients were underweight and 86 had obesity, without a separate analysis of severe obesity classes. Rare complications and modest clinical differences could therefore have been missed. The authors’ post hoc analysis indicated limited power for JOA and mental-component changes; nonsignificant findings are not evidence of equivalence. BMI also does not characterize muscle mass, nutritional status or visceral adiposity. Two years is insufficient to settle long-term adjacent-segment disease, reoperation or progressive alignment changes. Finally, inconsistencies in some tabulated variability labels and secondary outcome intervals warrant caution when interpreting precise estimates.

Clinical takeaway

In this selected surgical cohort, a higher BMI was associated with greater operative demands but not a statistically significant reduction in two-year neurological improvement. The findings argue against treating BMI alone as a reliable predictor of poor recovery. They do not establish a universal BMI threshold for proceeding with surgery, prove equal complication risk, or test whether preoperative weight reduction changes outcomes. Clinical severity, anatomy, comorbidities and the planned operation remain central to individualized assessment.