Previous studies aimed at identifying predictors of clinical outcomes following surgical decompression for degenerative cervical myelopathy (DCM) are limited by multicollinearity among predictors, whereby the high degree of correlation between covariates precludes detection of potentially significant findings. We apply partial least squares (PLS), a data-driven approach, to model multi-dimensional variance and dissociate patient phenotypes associated with functional, disability, and quality of life (QOL) outcomes in DCM. This was a post-hoc analysis of DCM patients enrolled in the prospective, multi-center AOSpine CSM-NA/CSM-I studies. Baseline clinical covariates evaluated as predictors included demographic (e.g., age, sex), clinical presentation (e.g., signs and symptoms), and treatment (e.g., surgical approach) characteristics. Outcomes evaluated included change in functional status (∆mJOA), disability (∆NDI), and QOL (∆SF-36) at 2 years. PLS was used to derive latent variables (LVs) relating specific clinical covariates with specific outcomes. Statistical significance was estimated using bootstrapping. Four hundred and seventy-eight patients met eligibility criteria. PLS identified 3 significant LVs. LV1 indicated an association between presentation with hand muscle atrophy, treatment by an approach other than laminectomy alone, and greater improvement in physical health-related QOL outcomes (e.g., SF-36 Physical Component Summary). LV2 suggested the presence of comorbidities (respiratory, rheumatologic, psychological) was associated with lesser improvements in functional status post-operatively (i.e., mJOA score). Finally, LV3 reflected an association between more severe myelopathy presenting with gait impairment and poorer mental health-related QOL outcomes (e.g., SF-36 Mental Component Summary). Using PLS, this analysis uncovered several novel insights pertaining to patients undergoing surgical decompression for DCM that warrant further investigation: (1) comorbid status and frailty heavily impact functional outcome; (2) presentation with hand muscle atrophy is associated with better physical QOL outcomes; and (3) more severe myelopathy with gait impairment is associated with poorer mental QOL outcomes.

先前旨在确定退行性脊髓型颈椎病 (DCM) 手术减压后临床结果的预测因子的研究受到预测因子之间多重共线性的限制，协变量之间的高度相关性排除了潜在重要发现的检测。我们应用偏最小二乘法 (PLS)，一种数据驱动的方法，对多维方差建模，并分离与 DCM 中的功能、残疾和生活质量 (QOL) 结果相关的患者表型。这是对参加前瞻性、多中心 AOSpine CSM-NA/CSM-I 研究的 DCM 患者的事后分析。作为预测因子评估的基线临床协变量包括人口统计学（例如，年龄、性别）、临床表现（例如，体征和症状）和治疗（例如，手术方法）特征。评估的结果包括 2 年时功能状态 (∆mJOA)、残疾 (∆NDI) 和 QOL (∆SF-36) 的变化。PLS 用于推导出与特定结果相关的特定临床协变量的潜在变量 (LV)。使用自举法估计统计显着性。478 名患者符合资格标准。PLS 确定了 3 个重要的 LV。LV1 表明出现手部肌肉萎缩、采用除单纯椎板切除术以外的方法进行治疗以及与身体健康相关的 QOL 结果（例如，SF-36 身体成分总结）的更大改善之间存在关联。LV2 表明合并症（呼吸道、风湿病、心理）的存在与术后功能状态（即 mJOA 评分）的改善较小有关。最后，LV3 反映了表现为步态障碍的更严重的脊髓病与较差的心理健康相关 QOL 结果之间的关联（例如，SF-36 心理成分总结）。使用 PLS，该分析揭示了一些与 DCM 手术减压患者相关的新见解，这些见解值得进一步研究：(1) 共病状态和虚弱严重影响功能结果；(2)手部肌肉萎缩的表现与更好的身体QOL结果相关；(3) 更严重的脊髓病伴步态障碍与较差的心理 QOL 结果相关。该分析揭示了一些与 DCM 手术减压患者相关的新见解，这些见解值得进一步调查：(1) 合并症状态和虚弱严重影响功能结果；(2)手部肌肉萎缩的表现与更好的身体QOL结果相关；(3) 更严重的脊髓病伴步态障碍与较差的心理 QOL 结果相关。该分析揭示了一些与 DCM 手术减压患者相关的新见解，这些见解值得进一步调查：(1) 合并症状态和虚弱严重影响功能结果；(2)手部肌肉萎缩的表现与更好的身体QOL结果相关；(3) 更严重的脊髓病伴步态障碍与较差的心理 QOL 结果相关。