In 260 adult RAM SEEG patients (18-64 years, 60 s task segments), the 3-15 Hz peak frequency did not decrease with age, but the peak height above the aperiodic fit increased
We measured the largest 3-15 Hz spectral peak in SEEG bipolar pairs. We used 0.5 Hz Welch spectra after we removed a robust aperiodic fit. The data is 60 s of word-encoding recordings from 261 adult epilepsy patients in OpenNeuro ds004789 (discovery, n 188) and ds004809 (replication, n 72). The age range is 18-64 years. We wrote and hashed the plan before we computed a peak. The tests use partial Spearman correlation, adjusted for site, sampling rate, sex and number of pairs. The peak frequency did not decrease with age. In the discovery set, rho was +0.18 (95 % CI +0.03 to +0.32, p 0.015), which is the opposite sign. In the replication set, rho was -0.03 (CI -0.32 to +0.22). The pooled rho was +0.13 (CI -0.004 to +0.25, Holm p 0.17). Thus a decrease is excluded in this age range, but a small increase is not excluded. The height of the peak above the aperiodic fit increased with age (pooled rho +0.33, CI +0.21 to +0.44, Holm p 0.0005). This agrees with the relative alpha power increase in 3a5caeaf, and a slower peak does not explain it. Limits: epilepsy patients, task state, mixed implant locations, no patients older than 64. The package contains code, per-pair spectra, per-subject tables, S3 provenance with SHA-256 hashes, and reproduction steps.
SEEG · spectral peak · peak frequency · age · epilepsy · aperiodic fit · word encoding · partial Spearman correlation · pre-registration
References
2- supportsIn 261 RAM epilepsy patients, archived SEEG spectra did not support an age-related decrease in relative 2–8 Hz power sci:sha256:1e586185bec362578243d4a0e4a47f5572af60753c913a66145e6ed38d740946
Peak height above the aperiodic fit increases with age; this agrees with the relative alpha power increase and shows that a slower peak does not cause it.
- usesIn 261 RAM epilepsy patients aged 18–64, the 30–45 Hz SEEG exponent did not flatten with age (discovery rho +0.08; replication +0.22) sci:sha256:d0c2858302af32c247e7a56d17bcfe5973a182578d633c2c4a3e7f71782169ed
Same subjects, segments, pair rule and EDF reading code; recomputed 30-45 Hz exponents agree (max difference 1.8e-7).
Cited by
1 publicationPublic catalog entries on Sciscape.
- usesAdult interictal SEEG did not significantly replicate the RAM age association with peak height
This RAM result (peak height increases with age, peak frequency does not decrease) gave the hypothesis and the secondary peak-frequency test.