Table 3.

Association of low GNRI score with post-stroke myocardial injury

Analysis method	OR	95% CI	P value
Logistic regression analysis (n = 377)
Model 1 (univariate model)a	2.74	1.74–4.33	< 0.001
Model 2 (demographic and previous patient-related covariates adjusted)b	3.22	1.96–5.29	< 0.001
Model 3 (stroke-related covariates adjusted)c	3.10	1.93–4.97	0.005
Model 4 (laboratory indicators adjusted)d	2.88	1.76–4.70	0.002
Model 5 (fully adjusted)e	3.25	1.93–5.48	< 0.001
Propensity score analysis
Model PSM (n = 202)f	3.59	1.93–6.67	< 0.001

Abbreviations: GNRI Geriatric Nutritional Risk Index, OR Odds ratio, CI Confidence interval, PSM Propensity score matching

^aModel 1 was an univariate regression model

^bModel 2 included low GNRI score, age, male, height, weight, BMI, SBP, DBP, current smoking, alcohol, hypertension, diabetes mellitus, COPD, peripheral vascular disease, and renal dysfunction

^cModel 3 included low GNRI score, admission NIHSS score, admission GCS score, dysphagia, TOAST stroke subtype, OCSP stroke subtype, intravenous thrombolysis, and endovascular treatment during hospitalization

^dModel 4 included low GNRI score, hemoglobin, albumin, blood glucose, triglycerides, total cholesterol, HDL-C, and LDL-C

^eModel 5 was adjusted for all the potential confounders. Univariate and multivariate results are shown in Supplementary Table 4

^f202 patients were matched (1:1) using propensity score approach. Univariate result is shown in Supplementary Table 5