Factors associated with gait recovery following surgery for degenerative cervical myelopathy.

OBJECTIVE: The objective was to quantify perioperative gait spatiotemporal parameters (STPs) in patients with degenerative cervical myelopathy (DCM) and to identify baseline predictors of gait recovery.

METHODS: This study evaluated a prospective observational cohort at a single academic center. Participants with DCM (n = 59) and non-DCM controls with cervical radiculopathy (n = 13) completed GAITRite walkway assessments at baseline, 6 months, and 12 months after standard-of-care treatments including surgery. STPs included gait speed, step/stride length, cadence, double support time, stride width, and Enhanced Gait Variability Index (eGVI). Relationships between STPs and common clinical scales, including the modified Japanese Orthopaedic Association lower extremity (mJOA-LE) subscore, Nurick grade, and Neck Disability Index (NDI), were evaluated with the Spearman ρ test. The Welch/Wilcoxon tests and linear mixed-effects models were used for between-group and longitudinal comparisons, respectively. Multivariable linear regression modeling was performed to evaluate independent predictors of gait recovery at last follow-up, including age, baseline STPs, mJOA-LE score, and Nurick grade.

RESULTS: At baseline, subjects with DCM demonstrated a distinct gait phenotype as compared to controls, characterized by significantly reduced gait speed (median [IQR] 83.1 [50.4-101.4] vs 105.1 [91.0-118.1] cm/sec, p = 0.003), shorter step length (51.7 [37.8-60.5] vs 59.1 [56.4-65.4] cm, p = 0.007), and shorter stride length (103.5 [75.1-120.3] vs 118.4 [113.0-130.4] cm, p = 0.006). Subjects with DCM also exhibited greater double support time (36.8% [32.7%-47.1%] vs 34.1% [30.8%-36.1%], p = 0.040) and lower cadence (95.1 [76.7-104.7] vs 103.7 [98.0-114.7] steps/min, p = 0.014). In contrast to NDI, mJOA-LE and Nurick scores showed significant correlations with STPs. After surgery, most STPs improved at 6 months and largely plateaued by 12 months. In multivariable models, the baseline value of each STP was the strongest predictor of change at last follow-up (all p < 0.01), with poorer baseline spatiotemporal performance predictive of greater interval recovery. Older age independently attenuated gains in gait speed and step/stride length (p ≤ 0.04). Baseline mJOA-LE and Nurick scores were not independent predictors of recovery after adjustment.

CONCLUSIONS: DCM is characterized by a slower and unstable gait with shorter stride lengths that significantly improves by 6 months after surgery and then plateaus. Baseline STPs track disease severity and postoperative gait recovery. Recovery magnitude depends primarily on baseline spatiotemporal performance and is attenuated by age, supporting the use of STPs and age-based stratification for prognostication.