Neuroanatomical subtypes of tobacco use disorder and relationship with clinical and molecular features

Prog Neuropsychopharmacol Biol Psychiatry. 2024 Dec 26:111235. doi: 10.1016/j.pnpbp.2024.111235. Online ahead of print.

Abstract

Background: Individual neurobiological heterogeneity among patients with tobacco use disorder (TUD) hampers the identification of neuroimaging phenotypes.

Methods: The current study recruited 122 TUD individuals and 57 healthy controls, and obtained their 3D-T1 images. Heterogeneity through discriminative analysis (HYDRA) was applied to uncover the potential subtype of TUD where regional gray matter volume (GMV) was treated as the feature. Then we examined the clinical, neuroimaging and molecular characteristics of subtypes.

Results: Two distinct neuroanatomical subtypes were found. In subtype 1, TUD individuals showed decreased GMV in right orbitofrontal cortex (OFC), while subtype 2 exhibited distributed pattern of widely GMV increase. Moreover, subtype 1 showed older initial smoking age, longer duration of smoking than Subtype 2. Persistent smoking behavior in subtype 1 is more likely caused by substance dependence/addiction rather than psychosocial factors. GMV correlated negatively with cumulative tobacco exposure in Subtype 1 but not in Subtype 2. Besides, neuroanatomical aberrance in subtype 1 was mainly associated with dopamine system, while neuroanatomical abnormalities in subtype 2 were primarily associated with GABAa.

Conclusions: Overall, our results revealed two opposite neuroanatomical subtypes of TUD, which largely overlapped with their clinical and molecular features respectively. TUD subtypes taxonomy based on objective anatomy could help to facilitate the development of individualized treatment for TUD.

Keywords: Gray matter; Heterogeneity; Machine learning; Subtype; Tobacco use disorder.