PAI-1, MMP-9, and NLR combined with NIHSS for predicting 90-day poor functional outcome in elderly acute ischemic stroke: a prospective observational cohort study

Objective To investigate whether a multi-marker panel comprising PAI-1, NLR, and MMP-9 enhances prognostication beyond the NIHSS score in elderly patients with acute ischemic stroke, and to develop a clinically applicable nomogram. Methods A total of 113 elderly AIS patients and 63 elderly non-AIS controls were prospectively enrolled. Fasting venous blood samples were collected at 06:00 on the first morning after admission (or at 06:00 on the day of admission for overnight admissions), and serum PAI-1, MMP-9, and NLR were measured. Clinical data and NIHSS scores within 24 h of admission were collected. Outcomes were assessed at 90-day follow-up using the modified Rankin Scale (mRS) (favorable outcome: mRS ≤ 2; poor outcome: mRS 2). Univariate and multivariate logistic regression analyses were performed to identify independent predictors. Three models were constructed: NIHSS alone, biomarkers alone, and their combination. Model performance was evaluated using ROC curves, calibration plots, decision curve analysis (DCA), and bootstrap internal validation. Results Serum levels of PAI-1, MMP-9, and NLR were significantly higher in AIS patients than in controls (all p 0.01). Among AIS patients, 50 (44.2%) had poor outcomes. In multivariable analysis, NIHSS score (OR, 2.24; 95% CI, 1.60–3.36; p 0.001), MMP-9 (OR, 1.71; 95% CI, 1.32–2.35; p 0.001), and NLR (OR, 1.42; 95% CI, 1.12–1.91; p = 0.011) were independently associated with poor outcome; PAI-1 showed a consistent effect direction but did not reach statistical significance ( p = 0.077). The combined model achieved an AUC of 0.889 (95% CI, 0.831–0.946), significantly outperforming both the NIHSS-only model (AUC, 0.781; DeLong p = 0.003) and the biomarker-only model (AUC, 0.791; DeLong p = 0.005). The combined model demonstrated excellent calibration (Brier score 0.135), good internal validity (optimism-corrected C-index 0.874), and positive net benefit across a wide threshold probability range (0.15–0.80) on DCA. Conclusion Elevated serum MMP-9 and NLR levels were independently associated with poor short-term prognosis in elderly AIS patients, while PAI-1 showed a consistent direction of association and contributed to the overall performance of the combined model. The nomogram combining these biomarkers with the NIHSS score provides improved risk stratification and may assist early clinical decision-making.