nucleation

LIT-996: It’s not the thought that counts: Allostasis at the core of brain function

2 min read

Properties5
StatusActive — read 2026-10-14 (NOTE-761) in full; worth reading as the Barrett group's current case that the brain's core function is allostasis: an intrinsic system of visceromotor and interoceptive regions that matches tract-traced anatomy and contains the default-mode and salience hubs, a limbic-to-sensory laminar gradient read as a prediction hierarchy, interoceptive rhythms shaping brain-wide signalling, and a hypothesis that cognitive decline in Alzheimer's disease is an allostatic trade-off. Read it knowing that it is a review of mostly the group's own work, that the authors do not claim the allostasis-first hypothesis is correct, and that it says nothing about why or whether allostatic states are felt.
Version2
Tagsneuroscience, emotion-and-affect
Cited as “paper” byNOTE-761 — Theriault et al. — Allostasis at the core of brain function
Cited as “source” byTHEORY-291 — The brain's visceromotor and interoceptive regions form one intrinsic system that reproduces their tract-traced anatomy and contains the hubs of the default-mode and salience networks

Graph View

Jordan E. Theriault, Yuta Katsumi, Henning M. Reimann, Jiahe Zhang, Philip Deming, Bradford C. Dickerson, Karen S. Quigley, Lisa Feldman Barrett (2025), Neuron 113(24), 4107–4133.

Key takeaways

From the reading, NOTE-761.

  • One allostatic system. 21 seeds chosen from non-human tract-tracing (14 subcortical, 7 cortical visceromotor and interoceptive regions). Resting-state connectivity at 7 T (N = 90), replicating 3 T (N = 550), reproduces 100%, 94% and 96% of the monosynaptic cortico-cortical, subcortico-cortical and subcortico-subcortical connections.
  • It contains the named networks’ hubs. The system overlaps the central autonomic, skeletomotor and salience networks and a pain signature; its two motifs, default-mode-like and salience-like, meet in the rich-club hubs. The paper concludes that assigning psychological functions to these networks is “destined to fail”.
  • A laminar gradient, read as prediction. From sensory to limbic cortex neurons grow larger, sparser and more connected and layer 4 disappears. Limbic signals are read as low-dimensional allostatic predictions decompressing outwards, sensory signals as errors compressed inwards.
  • The body shapes brain-wide signalling. Breathing couples oscillations across regions; cardiac phase gates visual sampling; vagal stimulation affects memory and mood.
  • Stress and mood. Stress is an anticipated large energetic outlay, value-neutral, felt as stressful when prediction and need mismatch. Mood is restated as “a low-dimensional allostatic barometer”.
  • Alzheimer’s disease. Falling glucose uptake after amyloid but before cognitive decline may protect pH and waste clearance, so glucose-restoring therapies might rescue cognition while worsening the disease.
  • The thesis is disclaimed. The authors aim “to justify asking scientific questions … in a different way”, not to claim allostasis-first is correct.

Standing in the record

Filed on 2026-10-12 from the owner’s papers-feed for QUESTION-059, as the bodily-mechanism option: “a feature a functional organization might lack”. Read on 2026-10-14 in NOTE-761, which does not bear that out. The paper is about what the brain is for, stated in functional terms, and says nothing about what makes a state felt or whether an organization could lack it. Its affect claims restate Barrett’s, already held in THEORY-045. THEORY-291, filed from the reading with Barrett (2017, LIT-542), holds the anatomical core of its case at the strength it is reported; the allostasis-first synthesis and the Alzheimer’s hypothesis are not filed.

Backlinks