Abstract
Follistatin-like 1 (FSTL1) participates in neuroinflammation. This study was conducted to investigate whether serum FSTL1 levels are associated with cognitive impairment following acute intracerebral hemorrhage (ICH). In this prospective cohort study of supratentorial ICH, 309 patients were randomly allocated to study group (206 cases) and validation group (103 cases) according to 2:1 ratio. Serum samples were obtained at admission of patients so as to measure FSTL1 levels. The Montreal Cognitive Assessment Scale was applied for assessing cognitive status at poststroke 3 months, with the score of <26 signifying cognitive impairment. An independent correlation was confirmed between serum FSTL1 levels and Montreal Cognitive Assessment Scale scores (β, -0.184; 95% confidence interval (CI), -0.314-0.054; variance inflation factor, 1.453; P = .006). Serum FSTL1 levels were substantially higher in patients with cognitive impairment than in the remainders (median, 9.8 vs 6.2 ng/mL; P < .001). Serum FSTL1 levels, in linear relation to cognitive impairment likelihood (P = .105), were independently predictive of cognitive impairment (odds ratio, 1.115; 95% confidence interval, 1.029-1.208; P = .008). Association of serum FSTL1 levels with cognitive impairment was negligibly affected by age, sex, drinking, and more (all P for interaction > .05). The model, which encompassed serum FSTL1, National Institutes of Health Stroke Scale scores and hematoma volume, performed well under the receiver operating characteristic curve, calibration curve and decision curve, and using the Hosmer-Lemeshow test. This model was validated in the validation group. Admission serum FSTL1 levels following ICH are intimately associated with cognitive impairment, suggesting that serum FSTL1 may be an appealing predictive factor of cognitive impairment subsequent to ICH.