Neuropsychological practice mirrors the changing landscape of epilepsy: Temporal trends in a level-4 epilepsy center from 1986 to 2024.
PURPOSE Following reports of declining epilepsy surgery numbers and changing patient populations, this retrospective single-center study examined whether the patients referred for neuropsychological assessment at a level-4 epilepsy center changed over four decades, and whether neuropsychological practice mirrored these changes. METHODS We analyzed 9,877 first-time neuropsychological assessments performed at the University Epilepsy Center in Bonn between 1986 and 2024, grouped into four time-based quartiles, and tested trends in demographics, etiology, education, cognitive impairment, referral context, and assessment format (ANOVA, χ²). RESULTS Over time, referred patients became older, had later epilepsy onsets, and were more often assessed early in the disease course. The proportion aged >50 years rose from 12.0% to 33.6%, late-onset epilepsy (>50 years) from 3.2% to 20.7%, and new-onset cases from 5.3% to 25.7%. Presurgical referrals declined sharply from 82.0% to 17.9%. Localization remained predominantly temporal, whereas etiology shifted: among cases with pathology/etiology data (n = 3,362), hippocampal sclerosis declined from 51.4% to 28.3% and tumors decreased, while suspected limbic encephalitis rose from 1.8% to 23.0%. Estimated verbal IQ rose from 101.4 to 106.7, and ICCoDE/Bonn Code classifications showed a shift toward less widespread and less severe impairment profiles. Comprehensive batteries remained common, especially in surgical cases, while short, repeatable assessments of executive function and memory became the most frequent format (69.0% in 2024). CONCLUSION Over four decades, the referred population changed markedly: fewer surgical referrals, older patients, more new- and late-onset epilepsies, and more focused, less severe impairment. Neuropsychological practice shifted accordingly, from predominantly presurgical risk assessment toward targeted baseline testing and longitudinal monitoring of disease and treatment effects. The main future challenge is assessment that is both efficient and sensitive to dynamic cognitive change, including early neurodegeneration in late-onset epilepsy.