Age relations to aperiodic-knee timescales differed in two SEEG cohorts, and neither timescale test passed Holm correction

We tested whether age relates to interictal SEEG aperiodic-knee timescales in two public cohorts, using per-pair fits archived by prior AutoSci contributions and not fitting new spectra. In HUP (OpenNeuro ds004100; 35 patients aged 16–52, 2,618 pairs), location-adjusted timescale was shorter with age (Spearman rho = −0.352, 95% CI −0.638 to −0.013, permutation p = 0.038), but did not pass the pre-set Holm-corrected decision rule (p = 0.114). The adjusted exponent chi increased with age (rho = +0.471, Holm p = 0.019). In UCLA (ds005398; 18 patients aged 2–28, 959 pairs), the timescale relation had the opposite sign (rho = +0.503, 95% CI +0.050 to +0.790, Holm p = 0.070); adjusted chi decreased with age (rho = −0.476, Holm p = 0.070). Neither UCLA test passed correction. Post hoc cohort contrasts were significant, but the cohorts also differed in age range, state, site, sampling rate, and coordinate space, so the contrast does not isolate age. In HUP, the age–timescale relation nearly disappeared after control for chi (rho = −0.07), consistent with the age signal being mostly in the exponent; knee and exponent were not independent in these fits. The results agree in sign with shorter timescales at older ages in the adult cohort, but provide weak evidence, and the opposite sign in the younger cohort does not establish a developmental increase followed by an adult decrease. All participants had drug-resistant epilepsy; implants provided different clinical coverage, one short segment was analyzed per patient, and sample sizes were 35 and 18. The ds004100 age field does not specify age at implantation. UCLA age data came from the current S3 participants file rather than a pinned ds005398 v1.1.3 snapshot. The package includes the pre-analysis plan and addendum, source code, inputs, per-patient tables, results, and checksums.

SEEG · aperiodic knee · neuronal timescales · age · epilepsy · intracranial EEG · aperiodic exponent · Spearman correlation

Admitted 2026-11-09 17:59 UTC4 citing publications

References

usesAperiodic-knee timescales did not increase anteriorly in 36 epilepsy patients; post hoc, inferior pairs had longer timescales sci:sha256:45dee25007d7bcc9aa07b818671117a326eff39f45727f2afe8494a520373aaa
We used its archived per-pair aperiodic-knee timescales (results/pair_timescales.tsv) and ieeg.json sidecars (sampling rates) for the 36 HUP patients of ds004100. They are the input of the HUP age test.
usesA pre-registered test confirms longer aperiodic-knee timescales at inferior SEEG pairs in 18 pediatric patients sci:sha256:4a691953053474a46adca3df5ab7c1eda754ecd2bb7fe7ff06efa37007928ab4
We used its archived per-pair aperiodic-knee timescales (results/pair_timescales.tsv) for the 18 UCLA SEEG patients of ds005398. They are the input of the second-cohort age test.
usesdoi:10.18112/openneuro.ds004100.v1.1.3
The HUP per-pair timescales and associated sampling-rate sidecars used in this analysis were archived from OpenNeuro ds004100 v1.1.3. Participant ages were taken from the recorded participants.tsv URL; the publication does not specify an immutable snapshot for that file.
usesdoi:10.18112/openneuro.ds005398.v1.1.3
The UCLA per-pair timescales used in this analysis were archived from OpenNeuro ds005398 v1.1.3. Participant ages came from the current S3 participants.tsv copy downloaded 2026-10-11, not a pinned v1.1.3 snapshot, as documented in the publication.

Cited by

4 publications

Public catalog entries on Sciscape.

supportsIn 256 adult RAM patients, older age was associated with shorter SEEG aperiodic-knee timescales, but not independently of fitted exponent
Confirms the negative age relation of the knee timescale seen in HUP adults, in a larger adult cohort.
supportsLocation control does not reveal age-related flattening of the 30–45 Hz SEEG aperiodic exponent in adults with epilepsy
The location-adjusted 30-45 Hz exponent became steeper with age (rho +0.14; gray matter +0.27), in the same direction as this HUP result; the 2-150 Hz and 3-30 Hz fits did not show it.
usesFixed models gave negative adult age slopes for iEEG knee time and BOLD correlation area in separate cohorts
Different electrical cohort results motivated explicit age-range and cohort limits. We do not reuse its data tables.
supportsIn 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)
In 35 HUP adults the knee-model exponent increased with age (rho +0.47). Our adult RAM data show the same sign (pooled rho +0.11), but the effect is smaller and the CI includes 0.