IRDA/IRTA at rest: what slow EEG events “turn on” and “turn off” in BOLD — a BrainLatam commentary on Feige et al. (BPS: Global Open Science, 2025)

1) The scientific question

The core question is: when IRDA/IRTA events (intermittent rhythmic delta/theta activity) occur in resting-state EEG, which brain regions show time-locked BOLD changes—and what is shared versus diagnosis/etiology-specific across autoimmune-mediated vs primary psychiatric groups?

In practical terms: is IRDA/IRTA merely “nonspecific slowing,” or does it reflect a network-level excitability/compensation dynamic with a mappable functional signature?

IRDA IRTA at rest - what slow EEG events turn on and turn off in BOLD

2) The experiment

The authors use a large transdiagnostic resting-state EEG-fMRI design:

• 135 high-quality EEG-fMRI datasets

• APS (suspected autoimmune psychiatric syndromes): 33

• BPD (borderline personality disorder): 59

• healthy controls: 43

They acquire fMRI with ultra-fast MREG (TR ~100 ms), chosen to better sample and control fast physiological components. In EEG, they:

• correct gradient/BCG/EOG artifacts and apply ICA

• detect IRDA/IRTA events algorithmically

• convert detected events into an fMRI regressor convolved with a canonical HRF to map event-related BOLD correlates

Analysis proceeds in two layers:

1. transdiagnostic consensus areas (significant across all groups)

2. group-specific additional clusters (APS- or BPD-specific, non-overlapping with consensus)

3) Why this experiment answers the question

It answers the question because it directly tests:

1. A phenomenon-first logic: start from IRDA/IRTA as the anchor and ask “what network changes co-occur,” rather than starting from a diagnosis.

2. Event-linked coupling: it is not just “slow EEG overall,” but BOLD modulation time-locked to detected IRDA/IRTA ranges.

3. Common vs specific separation: consensus clusters suggest what belongs to the phenomenon; additional clusters suggest modulation by etiology and clinical organization.

4) What the results show (objective synthesis)

They identify 11 consensus regions associated with IRDA/IRTA, including both increases and decreases in BOLD activity.

• Increases (5 regions) include sensorimotor/temporal/cingulate-related areas (e.g., BA2, BA4, BA43, right BA18, right BA26/29/30).

• Decreases (6 regions) include bilateral BA10 and associative parietal/occipital/cingulate regions (e.g., BA39 left, BA23 left, BA19 left, left BA18).

Additional findings:

• APS shows five extra clusters, all deactivations, suggesting a broader pattern of reduced activity beyond the consensus set.

• BPD shows one additional activation in left BA17 (primary visual cortex).

Exploratory analyses also report associations between IRDA/IRTA density and selected symptom measures, suggesting potential clinical relevance.

5) BrainLatam reading — APUS (extended proprioception)

We interpret the consensus “activations” in sensorimotor and integration-related regions as a sign that IRDA/IRTA is not simply “sleep-like slowing,” but a state in which the system attempts to stabilize embodied readiness. In BrainLatam terms, APUS (extended proprioception) may be pulled into a stabilizing regime—maintaining coherence of action and bodily mapping while network dynamics shift.

The key point is the pattern: sensorimotor engagement co-occurs with regions often linked to salience/integration, reinforcing the view that IRDA/IRTA behaves like a network-level phenomenon, not an isolated rhythm.

6) BrainLatam reading — Tekoha (extended interoception)

From a Tekoha perspective, IRDA/IRTA can be read as an internal regulatory mode where local excitation and compensatory inhibition coexist. The mixed pattern of “turning some regions up” while “turning others down” fits a regulation logic: the organism rebalances system-level stability when a network enters a dysregulated excitability state.

The authors’ framing via the LANI hypothesis (local homeostatic inhibition following hyperexcitability) maps well onto this interpretation: excitation and compensation co-occur, and the clinical impact depends on which networks lose functional bandwidth.

7) Limits that define the next experiment

• Causality: this is event–BOLD correlation, not a causal test of mechanisms.

• Transdiagnostic heterogeneity: APS and BPD differ in sex distribution, age, and medication exposure, all of which can modulate both IRDA/IRTA occurrence and BOLD coupling.

• EEG-in-scanner detection: despite strong preprocessing, in-scanner EEG remains noisy; detector sensitivity/specificity is always critical.

• Canonical HRF: IRDA/IRTA–hemodynamic coupling may vary; more flexible HRF or model-based approaches could capture temporal differences.

8) BrainLatam translation to the organic world

BrainLatam translation to the organic world: this study shows that a clinically recognizable EEG phenomenon (IRDA/IRTA) has a measurable whole-brain functional footprint in BOLD—combining a transdiagnostic core with etiology-specific extensions. That supports using IRDA/IRTA not as a diagnosis label, but as a marker of a network state: a mode of excitability/compensation that may shape attention, speech, perception, and affect regulation.

9) Open BrainLatam question

If IRDA/IRTA reflects a network-level “on/off” rebalancing, what is the next decisive experiment:

• test whether interventions (anticonvulsants, immunotherapy, neuromodulation, autonomic/respiratory regulation) reduce IRDA/IRTA and normalize these BOLD clusters, and

• evaluate whether network normalization predicts symptom improvement better than IRDA/IRTA presence alone?

The body does not need belief to function.
It needs space, movement, and regulation.

Ref.:

Feige, B., Zedtwitz, K. von, Matteit, I., Coenen, V. A., Nickel, K., Runge, K., Harald Prüss, Rau, A., Reisert, M., Matthies, S., Maier, S. J., & Elst, van. (2025). Functional brain activity associated with intermittent rhythmic delta/theta activity: A transdiagnostic EEG-fMRI resting state study. Biological Psychiatry Global Open Science, 100661–100661. https://doi.org/10.1016/j.bpsgos.2025.100661