Hasan SULTANOĞLU, Tuba Erdem SULTANOĞLU, Aydanur AKBABA
Health Sciences Quarterly - 2026;6(4):739-749
This study evaluated the diagnostic value of serum calcium (Ca), magnesium (Mg), and the calcium-to-magnesium (Ca/Mg) ratio for imaging-confirmed stroke among emergency department (ED) patients presenting with suspected cerebrovascular events. This single-center retrospective observational study included adult ED patients evaluated for suspected cerebrovascular events who underwent brain CT and/or MRI between 2020 and 2025. Of 2,256 eligible patients, 1,681 with missing data were excluded; 575 patients with available Ca and Mg measurements were analyzed. The primary outcome was CT/MRI-confirmed stroke (ischemic or hemorrhagic). Albumin-corrected Ca was calculated as corrected Ca = measured Ca + 0.8x(4.0 - albumin [g/dL]); Ca/Mg ratio was calculated as total Ca/Mg. Groups were compared using appropriate univariate tests. Diagnostic performance was assessed by ROC-AUC and optimal cut-offs were determined using the Youden index. Associations with stroke were examined using univariable and multivariable logistic regression adjusted for age, sex, Glasgow Coma Scale (GCS), systolic blood pressure, glucose, creatinine/eGFR, and comorbidities. Stroke was confirmed in 330/575 patients (57.4%). Compared with the non-stroke group, the stroke group had lower Mg (median 1.70 vs 2.00 mg/dL) and higher total Ca and Ca/Mg ratio (all p<0.001). The highest discriminatory performance was observed for the Ca/Mg ratio (AUC 0.876), followed by Mg (AUC 0.844), corrected Ca (AUC 0.801), and total Ca (AUC 0.737). In multivariable models, Mg remained inversely associated with stroke, and the Ca/Mg ratio remained independently associated with stroke, whereas total Ca and corrected Ca were not independently associated. Using the Youden cut-off, Ca/Mg >=5.06 yielded 70.6% sensitivity and 88.2% specificity; Mg <=1.78 mg/dL yielded 60.0% sensitivity and 90.2% specificity. In ED patients with suspected cerebrovascular events undergoing CT/MRI, Mg and particularly the Ca/Mg ratio showed meaningful diagnostic performance for imaging-confirmed stroke. These biomarkers should be considered adjunctive to, rather than replacements for, urgent neuroimaging. External validation in prospective, multicenter cohorts is warranted.
