Hypophosphatemia as a possible biomarker for epileptic seizures at the emergency department

Abstract

Introduction: Hypophosphatemia seems to be temporally associated with seizures, despite not being considered a trigger. We aimed to evaluate hypophosphatemia as a biomarker for seizures.

Methods: Retrospective study, including all consecutive patients admitted at our central hospital's emergency department from 01/01-31/03/2021, screened as "altered consciousness/syncope" or "seizures", with available phosphate levels.

Results: 277 patients included, mostly male (61.7%), mean age 64.3 years. Final diagnosis was "seizure" in 34.7% and "other diagnosis" in 65.3%. Patients with seizures were younger (p<0.001), had more frequent epilepsy (p<0.001) and alcoholism (p=0.01). Patients with other diagnosis had more often renal failure (p<0.001) and statin (p=0.02) or diuretic (p=0.003) therapy. Time to blood collection (from the event and from admission) was similar between groups. Patients with seizures had lower mean phosphate levels and more frequent hypophosphatemia (<2.4mg/dL) (p<0.001). Mean CK levels were similar in both groups (p=0.25). HyperCK (>200U/L) was more frequent in the seizure group (p=0.04). Odds ratio (OR) of hypophosphatemia for seizures was 4.330 (CI 95% 2.170-8.640, p<0.001), persisting after correction for confounders. OR of hyperCK was 1.890 (CI 95% 1.060-3.371, p=0.03), losing significance when adjusted. Sensitivity was low for both. Hypophosphatemia was more specific (91.2% vs 79.9%).

Conclusions: Our findings support hypophosphatemia as a seizure biomarker. More studies are needed.

Keywords: Creatine kinase; Hypophosphatemia; Phosphate; Seizures; Transient altered level of consciousness.