《Chinese Journal of Rehabilitation Theory and Practice》 ›› 2023, Vol. 29 ›› Issue (11): 1346-1352.doi: 10.3969/j.issn.1006-9771.2023.11.013

Previous Articles     Next Articles

Predictive efficacy of Geriatric Nutritional Risk Index on neurological outcome after stroke

ZHU Lielie, XIA Jianning, PU Xinyu, SHAO Xiangzhi, ZHANG Jiacheng, WU Dengchong()   

  1. Department of Rehabilitation, Wenzhou TCM Hospital of Zhejiang Chinese Medical University, Wenzhou, Zhejiang 325000, China
  • Received:2023-08-11 Revised:2023-10-16 Published:2023-11-25 Online:2023-11-30
  • Contact: WU Dengchong, E-mail: dcwuwzs@126.com
  • Supported by:
    Wenzhou Municipal Science and Technology Bureau Project(Y20220064);Wenzhou Association for Science and Technology Project(jczc153)


Objective To evaluate the efficacy of Geriatric Nutritional Risk Index (GNRI) on admission on predicting neurological outcomes for stroke patients at the convalescence stage (three months after stroke).

Methods A total of 277 inpatients with ischemic stroke in Wenzhou TCM Hospital of Zhejiang Chinese Medical University from November, 2022 to January, 2023 were divided into good outcome group (n = 195) and poor outcome group (n = 82) according to the score of modified Rankin Scale (mRS) three months after stroke. Their clinical data and GNRI were compared.

Results Compared with the good outcome group, the poor outcome group showed a lower level of GNRI (t = -9.569, P < 0.001), more proportion of patients with high nutritional risk (χ2 = 68.861, P < 0.001). More cases with poor outcome were found with higher nutritional risk (Z = 65.406, P< 0.001). After covariate adjustment, higher level of GNRI was an independent protective factor for poor outcome three months after stroke (OR = 0.895, 95%CI 0.864 to 0.927, P < 0.05), and the OR increased with the nutritional risk grade of GNRI (P < 0.05). The area under the curve (AUC) of GNRI, score of NIHSS, and BMI were 0.812 (Z = 11.576, P < 0.001), 0.759 (Z = 8.328, P < 0.001), and 0.594 (Z = 2.716, P = 0.007), respectively; while the combination of GNRI and NIHSS was more effective (AUC = 0.875, Z = 17.389, P < 0.001). The sensitivities of GNRI, NIHSS and the combination of the two in predicting neurological function were 65.85%, 76.83% and 79.21%, respectively, and the specificities were 81.54%, 60.51% and 82.95%, respectively.

Conclusion GNRI on admission is associated with neurological function three months after stroke. GNRI can be used alone or in combination with NIHSS score to predict neurological outcome in the rehabilitation period, which is a useful complement to traditional predictors.

Key words: stroke, elderly, nutrition, neurological function, outcome

CLC Number: