Serum GFAP and UCH-L1 in selected older emergency department patients: positivity beyond documented head injury

Abstract Background Older adults frequently present to emergency departments (EDs) after falls or suspected head injury, often with multimorbidity, cognitive impairment, and uncertain injury timing. Glial fibrillary acidic protein (GFAP) and ubiquitin C-terminal hydrolase L1 (UCH-L1) may support imaging decisions after mild traumatic brain injury, but age- and disease-related elevations may reduce specificity in older patients. Methods We conducted a single-center exploratory observational pilot study with prospective serum collection and retrospective clinical and imaging data abstraction in Poland from October 2023 to April 2024. Consecutive eligible patients aged ≥ 65 years presenting during the primary investigator’s clinical shifts were enrolled. Participants were classified as head injury sampled ≤ 24 h ( n = 47), head injury sampled > 24 h ( n = 7), or selected non-trauma presentations without documented head injury ( n = 41). GFAP and UCH-L1 were measured on the Abbott ARCHITECT platform. We assessed positivity at manufacturer decision limits (GFAP ≥ 35 pg/mL; UCH-L1 ≥ 400 pg/mL). All patients in the ≤ 24-h subgroup underwent head CT; the composite endpoint included acute traumatic intracranial hemorrhage or cerebral contusion and/or acute fractures of the skull, facial skeleton, or upper cervical spine. Results GFAP positivity occurred in 91.5% (43/47), 100.0% (7/7; descriptive only), and 78.0% (32/41) of the three groups, respectively. UCH-L1 positivity occurred in 55.3% (26/47), 42.9% (3/7), and 26.8% (11/41). Nine of 47 patients (19.1%) in the ≤ 24-h subgroup had at least one CT-positive acute traumatic finding. GFAP showed a preliminary discriminatory signal (AUC 0.868; 95% CI 0.728–1.000), whereas UCH-L1 showed no apparent discrimination (AUC 0.488; 95% CI 0.222–0.755). A combined model yielded an AUC of 0.874 (95% CI 0.747–1.000), only minimally higher than GFAP alone. At the manufacturer GFAP decision limit, sensitivity was 100.0% (9/9; 95% CI 66.4–100.0) and specificity was 10.5% (4/38; 95% CI 2.9–24.8). Conclusions Among selected older ED patients, positivity at manufacturer decision limits was frequent beyond documented head injury. GFAP showed a preliminary signal for the composite acute traumatic CT endpoint, but estimates were imprecise and specificity was very low; UCH-L1 added little diagnostic information. These findings support cautious, age- and context-aware interpretation and require external validation. They do not establish new clinical thresholds or effects on geriatric care pathways.