[Construction and efficacy verification of a risk prediction model for cognitive function changes in patients with...

Objective: To construct a risk prediction model for cognitive function changes in patients with vestibular migraine and to validate its effectiveness. Methods: 300 patients with vestibular migraine admitted to the First Affiliated Hospital of Zhengzhou University hospital from January 2022 to October 2024 were enrolled, with 54 male and 254 female, age from 22 to 77, with an average of 42.2 years old, who were randomly divided into a modeling group (to construct a risk prediction model for cognitive function changes, n=210) and a validation group (to validate the effectiveness of the model, n=90) using a random number table. All patients received routine treatment. During a 6-month following-up, the cognitive function of the subjects was evaluated using the Mini Mental State Examination (MMSE) scale. Based on this, the modeling group was divided into a cognitive function change group and a cognitive function normal group. The general information were compared between the cognitive function change group and cognitive function normal group. The influencing factors of cognitive function changes were explored through Logistic univariate and multivariate regression analysis. A nomogram was developed using the rms package in R. Calibration was assessed with calibration plots; discrimination was evaluated by the area under the receiver operating characteristic (ROC) curve. Decision curve analysis (DCA) was determined the benefits of the model. Results: There was no statistically significant difference in the general data between the modeling group and the validation group (P>0.05). The incidence of cognitive function changes was 39.73% (116/292), with modeling group and validation group being 39.71% (81/204) and 39.77% (35/88), respectively. Age [odds ratio (OR)=1.361, 95% confidence interval (95%CI): 1.256-1.475], duration of migraine attacks (OR=1.096, 95%CI: 1.052-1.142), migraine severity score (OR=3.629, 95%CI: 2.476-5.319), Dizziness Handicap Inventory (DHI) score (OR=1.020, 95%CI: 1.010-1.030), Pittsburgh sleep quality index (PSQI) score (OR=1.818, 95%CI: 1.460-2.265), and serum matrix metalloprotein-9 (MMP-9) level (OR=1.256, 95%CI: 1.182-1.335), serum 5-hydroxytryptamine (5-HT) level (OR=0.940, 95%CI: 0.920-0.960) and weekly exercise time (OR=0.955, 95%CI: 0.942~0.968) were all independent related factors for cognitive function changes (P<0.05). Based on the above factors, a risk prediction Nomogram model for cognitive function changes in patients was developed, which showed good discrimination and predictive performance. The model predicted that the modeling group and the validation group achieved net benefits when the threshold probabilities for cognitive function changes ranged from 0.04 to 0.97 and from 0.04 to 0.98, respectively. Conclusions: Age, duration of migraine attacks, migraine severity score, DHI score, PSQI score, and serum MMP-9, 5-HT levels and weekly exercise time are independent related factors for cognitive function changes in patients with vestibular migraine. The Nomogram model for cognitive function changes constructed by incorporating these factors has good efficacy.