MRI Shows Hypothalamic Atrophy in Narcolepsy Type 1, Sparing Subregion

The hypothalamus of an adult living with narcolepsy type 1 (NT1) measures smaller than that of a person without the disorder, and the loss concentrates toward the front of the structure rather than spreading evenly through it.¹ The pattern emerged from the largest MRI investigation of hypothalamic volume in NT1 to date, published in SLEEP.

Why NT1 1 Has Been Hard to See on MRI

NT1 is a chronic neurological sleep disorder marked by excessive daytime sleepiness, cataplexy, sleep paralysis, and disrupted nocturnal sleep, and it traces to the selective loss of the roughly 70,000 orexin-producing neurons housed in the hypothalamus. Symptoms usually begin between ages 7 and 25 years, and because they are frequently mistaken for psychiatric conditions, years can pass before a patient receives the correct diagnosis.²

That delay has complicated efforts to study the disease in living patients.¹ Postmortem tissue is typically collected decades after symptom onset, long after the process that destroys orexin neurons has run its course. Imaging offered a way closer to disease onset, but the hypothalamus occupies only about 4 cm³ and carries poor image contrast. Earlier voxel-based morphometry studies split evenly, 4 reporting gray matter reduction and 4 finding none, most of them on 1.5-T scanners.

Comparing 75 Patients With NT1 and 87 Healthy Controls

The cross-sectional, case-control analysis enrolled consecutive adults referred to a single narcolepsy center in Bologna, Italy, from September 2023 to November 2024. All participants underwent 3-T T1-weighted MRI, and an automated tool segmented the whole hypothalamus and its 5 subregions. Comparisons were adjusted for age, sex, and total intracranial volume; reported volumes are unadjusted means. Of 82 patients scanned, 7 were excluded for motion artifacts.

The main group was 52.0% male patients with a mean (SD) age of 33.1 (12.9) years compared with 87 age- and sex-matched controls who were 51.7% male patients and aged 32.5 (8.2) years. Mean disease duration reached 13.8 (9.7) years. Cataplexy was recorded in 96.0% of patients, 95.8% carried HLA-DQB1*0602, and cerebrospinal fluid orexin fell below 110 pg/mL in all patients tested. At scanning, 21.3% were untreated or had started treatment within the past year; the rest took wake-promoting agents (37.3%), sodium oxybate (37.3%), pitolisant (24.0%), or antidepressants (13.3%).

Anterior Subregions Showed the Steepest Volume Loss

Whole hypothalamic volume averaged 807.1 (76.1) mm³ in patients vs 847.6 (67.2) mm³ in controls (P = .009). The deficits ran deepest anteriorly: the anterior-inferior subregion measured 28.9 (8.9) mm³ vs 34.8 (7.1) mm³ in controls (P = .0008), and the anterior-superior subregion, 41.1 (9.3) mm³ vs 46.4 (8.2) mm³ (P = .003). Smaller but significant reductions appeared in the posterior (245.3 [29.8] vs 256.2 [21.4] mm³; P = .03) and tuberal-superior (221.6 [28.7] vs 237.8 [25.9] mm³; P = .04) subregions.

The tuberal-inferior subregion was the exception. Its volume measured 270.0 (31.1) mm³ in patients and 272.4 (28.3) mm³ in controls, a difference that did not reach significance, although it trended lower.

“Taken together, these findings provide evidence of persistent hypothalamic tissue alteration in long-standing NT1,” the researchers wrote.

Tuberal-Inferior Sparing and the Histamine Hypothesis

The sparing runs counter to what a map of orexin cell loss would predict, since the nuclei richest in orexin neurons sit in the tuberal-superior and tuberal-inferior subunits. The authors proposed that a compensatory increase in histaminergic neurons, or glial proliferation, could be masking atrophy there. Between 64% and 94% of the tuberomammillary nucleus, the brain's principal source of histamine, sits in the tuberal-inferior and posterior segments, and earlier work found sharp astrocyte increases across nuclei in that region.

Nothing about the atrophy tracked with the clinical picture. Hypothalamic volumes correlated with no demographic, clinical, or laboratory variable, including disease duration and orexin levels, and showed no relationship to treatment status or drug choice. The authors read the pattern as a residual structural scar established early in the disease course and followed by a long plateau, rather than as progressive neurodegeneration.

Study Limitations and What Imaging Could Add to NT1 Care

The cross-sectional design limits what the results can say about timing or causality, and the cohort was heterogeneous in disease duration and ongoing treatment. No children were enrolled, and no patient was scanned close to symptom onset, leaving open what happens while orexinergic damage is still active. The segmentation tool also works at gross anatomical resolution, not that of histology.

Because the volume loss was already established and moved with neither disease duration nor orexin levels, MRI appears to capture something that spinal fluid testing does not. The authors called for further quantitative imaging work to determine whether these measures could serve as a biomarker of NT1. In a disease where diagnosis still rests on lumbar puncture and multiple sleep latency testing and often arrives years late, a noninvasive marker would give clinicians and payers a faster path to confirmation.

References

1. Motta L, Venturi G, Biscarini F, et al. Hypothalamic volume reduction in adult patients with narcolepsy type 1. Sleep. 2026;49(6):zsag087. doi:10.1093/sleep/zsag087

2. Narcolepsy. National Institute of Neurological Disorders and Stroke. Accessed August 26, 2026. https://www.ninds.nih.gov/health-information/disorders/narcolepsy