Abstract

Introduction

Chronic nicotine use perturbs multiple neurobiological systems involved with, for example, reward processing, emotion regulation, and cognitive control [1-3]. Alterations within specific neurocircuits may contribute to distinct nicotine withdrawal symptoms (e.g., anhedonia, anxiety, difficulty concentrating) [4] which make smoking cessation difficult. These aversive symptoms can be ameliorated to some degree by nicotine administration, which contributes to high relapse rates during early abstinence through negative reinforcement. Currently available pharmacological interventions (e.g., nicotine replacement, varenicline) vary in efficacy across individuals, highlighting the need to better characterize the underlying neurobiology linked with nicotine abstinence and administration. Brain regions linked with nicotine abstinence and administration may show alterations locally, in activity or connectivity, or remotely, in large-scale network integration.

Methods

In a repeated measures within-subject, double-blind, placebo-controlled study, overnight-abstinent smokers (n=20) and nonsmokers (n=20) completed 6 fMRI sessions during a nicotine and varenicline crossover drug administration regime. We first conducted two whole-brain voxel-wise complementary measures of intra-regional synchronization; regional homogeneity (ReHo), the local temporal synchronization of brain activity [5], and fractional amplitude of low frequency fluctuations (fALFF), the amplitude of spontaneous low frequency fluctuations [6]. A conjunction analysis of ReHo and fALFF outcomes comparing smokers vs. nonsmokers was conducted to identify convergent regions of altered local activity (pcorrected< 0.05). To further elucidate how intra-regional alterations were associated with inter-regional integration, areas of convergence identified above were then used as seed regions of interest in a resting-state functional connectivity (rsFC) analysis. For each subject and session, rsFC Z-maps from each seed region were submitted to group-level linear mixed effects modeling (pcorrected< 0.05). To enhance interpretations of rsFC outcomes, we utilized self-report metrics (affect, tobacco craving, nicotine withdrawal) in ancillary correlation analyses.

Results

Smokers (vs. nonsmokers) showed decreased ReHo in the thalamus, middle occipital gyrus (MOG), fusiform gyrus, and middle temporal gyrus (MTG), and increased ReHo in the ventromedial prefrontal cortex (vmPFC), medial frontal gyrus, and hippocampus. Smokers (vs. nonsmokers) also showed decreased fALFF in the thalamus, MTG, and cuneus, and increased fALFF in the vmPFC, hippocampus, cingulate, and globus pallidus. A conjunction of ReHo and fALFF outcomes identified convergence in the vmPFC, hippocampus, MTG, and cuneus (Fig. 1A). We then used convergent regions as seeds in a rsFC analysis to explore large-scale disruptions during nicotine administration. When considering the vmPFC seed, we observed a group (smoker vs. nonsmoker) by patch (nicotine vs. placebo) interaction, such that nicotine (vs. placebo) decreased rsFC with the insula, anterior cingulate cortex (ACC), frontal, parietal, and cerebellar regions among smokers, but not nonsmokers (Fig. 1B). Critically, rsFC between the vmPFC and insula, dorsolateral prefrontal cortex (dlPFC), and inferior parietal lobe was negatively correlated with positive affect (pcorrected< 0.05), whereas vmPFC-cerbellar rsFC was correlated with tobacco craving. When considering the hippocampal seed, we observed a patch effect, such that nicotine decreased rsFC with the fusiform gyrus, precuneus, and middle frontal gyrus. RsFC between the hippocampus and precuneus was positively correlated with self-reported withdrawal symptoms. Supporting Image: OHBMFigure_caption.jpg

Conclusions

Results provide support that alterations within specific neurocircuits may contribute to distinct aspects of nicotine withdrawal including affect, craving, and subjective withdrawal symptoms.