Fig 1.
Top three panels. (1) A racially diverse group of 75 daily tobacco smokers was recruited from the community. A multi-stage screening process—including a baseline assessment of breath carbon monoxide (CO) levels—was used to assess eligibility. At the baseline laboratory session, participants completed questionnaire measures of tobacco use and dependence. (2) Participants were randomized to either the Smoke-as-Usual or 24-Hour Abstinence groups, stratified by age and sex. (3) At the neuroimaging session, protocol compliance was confirmed by breath CO. Prior to scanning, smoking urges and withdrawal symptoms were assessed. During scanning, participants completed a well-established threat-anticipation paradigm that encompassed measures of threat-evoked distress, physiological arousal, and brain function. Bottom two panels. Threat-Anticipation Paradigm. As shown schematically in the bottom-left panel, the threat-anticipation paradigm took the form of a 2 (Valence: Threat/Safety) × 2 (Temporal Certainty: Certain/Uncertain) repeated-measures, randomized event-related design. Subjects were completely informed about the task design and contingencies prior to scanning. On Certain Threat trials, subjects saw a descending stream of integers (‘count-down’) for 18.75 s. To ensure robust distress, the anticipation epoch culminated in the delivery of a noxious electric shock, unpleasant photographic image, and thematically related audio clip (e.g., scream). Uncertain Threat trials were similar, but the integer stream was randomized and presented for an uncertain and variable duration (8.75–30.00 s). Participants knew that something aversive was going to occur, but they had no way of knowing precisely when. Safety trials were similar, but terminated with the delivery of benign reinforcers (e.g., just-perceptible electrical stimulation). Mean duration of the anticipation epochs was identical across conditions. Subjects were periodically prompted to rate the intensity of fear/anxiety experienced during the anticipation period of the prior trial. Skin conductance was continuously acquired throughout. Neuroimaging Analyses. As shown schematically in the bottom-right panel, two approaches were used to test the impact of acute hour nicotine abstinence on neural reactivity to threat. Guided by mechanistic work in animal models, hypothesis testing focused on two well-established, anatomically defined EAc regions-of-interest (ROIs): the Ce (left) and BST (right). Because the ROIs (i.e., voxel-level measurements) were chosen a priori—on the basis of anatomy, rather than suprathreshold threat reactivity—this approach provides unbiased effect size estimates [48]. Standardized regression coefficients were extracted and averaged across voxels for each combination of ROI, threat type (Uncertain/Certain), and participant. Hypothesis testing used a standard mixed-effects general linear model (GLM). For illustrative purposes, 1-mm ROIs are shown. Analyses employed ROIs decimated to the 2-mm resolution of the fMRI data. As shown schematically in the bottom-right panel, exploratory whole-brain voxelwise analyses were also performed. Abbreviations—BST, bed nucleus of the stria terminalis; Ce, central nucleus of the amygdala; fMRI, functional magnetic resonance imaging; hrs., hours; M, mean; s, seconds.
Table 1.
Descriptive statistics for the complete sample, Smoke-as-Usual group, and 24-Hour Abstinence group.
Fig 2.
The impact of acute nicotine abstinence on subjective distress and objective physiological arousal elicited by the threat-anticipation paradigm.
Upper panels. Mean self-reported intensity of fear and anxiety experienced during the anticipation epoch of each condition for the Smoke-as-Usual (grey) and 24-hour Abstinence (red) groups. Participants were quasi-randomly prompted to rate each condition three times while completing the MTC paradigm. Lower panels. Mean SCL during the anticipation epochs of the MTC. (a) Consistent with prior work, distress was significantly elevated during the anticipation of Threat compared to Safety, and during the anticipation of temporally Uncertain compared to Certain reinforcers (ps<0.001). (b) The 24-Hour Abstinence group experienced significantly intensified fear and anxiety during the anticipation of Threat compared to Safety (Group × Valence, p = 0.002). (c) Psychophysiological arousal was also significantly increased during the anticipation of Threat compared to Safety, reinforcing the validity of the MTC paradigm (p<0.001). This increase was particularly evident during the anticipation of temporally Uncertain compared to Certain reinforcers (Valence × Certainty, p<0.001). On average, abstinent participants showed significantly greater arousal throughout the MTC paradigm (inclusive of all conditions) compared to those who smoked as usual (p = 0.008). (d) Echoing the subjective distress results, the 24-Hour Abstinence group showed a trend toward greater arousal during Threat anticipation (Group × Valence, p = 0.07). Bars indicate group means, whiskers indicate SEs, and dots indicate participant-level means.
Fig 3.
Uncertain and Certain Threat anticipation recruit a common cortico-subcortical network.
Key regions (cyan arrowheads) show significantly increased activation during the anticipation of both Uncertain Threat (left column) and Certain Threat (middle column) compared to their respective control conditions (FDR q<0.05, whole-brain corrected). Both threat conditions recruited a common neural circuit—including the BST and dorsal amygdala (Ce)—replicating prior work in university students [42]. Right column depicts the voxelwise overlap (Logical AND of the thresholded maps depicted in the left and middle columns). BST and dorsal amygdala images are masked to highlight the extended amygdala. Coronal insets show the corresponding statistical parametric maps without the extended amygdala mask. Abbreviations—Ant., anterior; dlPFC, Dorsolateral Prefrontal Cortex; FrO, Frontal Operculum; BST, Bed Nucleus of the Stria Terminalis; FDR, False discovery rate; PAG, Periaqueductal Gray; WB, whole-brain corrected.
Fig 4.
The impact of threat anticipation and acute nicotine abstinence on the EAc.
Figure depicts mean contrast coefficients (Threat—Safety) during the anticipation of temporally Certain or Uncertain Threat—relative to their respective control conditions—in the anatomically defined Ce (yellow) and BST (magenta) ROIs for the Smoke-as-Usual (grey) and 24-hour Abstinence (red) groups. The BST was more sensitive to threat—irrespective of temporal certainty—compared to the Ce (p = 0.003). On average, the two divisions of the EAc showed greater activation during Uncertain Threat anticipation (p = 0.02). Interpretation of these effects is tempered by a trend-level Region × Threat Certainty interaction (p = 0.10). The Ce showed similarly heightened activation during Certain and Uncertain Threat anticipation (p = 0.52), whereas the BST showed a clear preference for Uncertain Threat (p = 0.007). Other effects were nonsignificant (ps>0.33), indicating that acute nicotine abstinence had a negligible impact on EAC threat reactivity. Bars indicate group means, whiskers indicate SEs, and dots indicate participant-level means. Abbreviations—BST, bed nucleus of the stria terminalis; Ce, central nucleus of the amygdala; ROI, region of interest.
Fig 5.
Forest plot summarizing the consequences of acute nicotine abstinence across key outcomes.
To facilitate comparison, results are depicted as standardized group mean differences (dots; Cohen’s d). Whiskers indicate the precision of the standardized differences (95% confidence interval). Upper panel (light grey) shows group differences for subjective feelings of smoking urgency and withdrawal, both assessed just prior to scanning. Lower panel (dark grey) shows standardized group differences for the threat-reactivity (Threat minus Safety) measures acquired during fMRI scanning, including threat-evoked changes in subjective fear and anxiety, objective physiological arousal (SCL), and EAc activation. While all of the standardized group differences were in the expected positive direction (24-Hour Abstinence > Smoke-as-Usual), there were marked differences in magnitude across outcomes. Nicotine abstinence exerted a large to very large (red) impact on all of the self-report measures of subjective experience (ds = 0.75–2.33). In contrast, nicotine abstinence exerted a more moderate influence on threat-elicited arousal (green; d = 0.44) and a very small to nil impact on the fMRI measures of EAc threat reactivity (blue; ds = 0.01–0.11). Effect sizes are interpreted with reference to widely used benchmarks (Large, d = 0.80; Medium, d = 0.50; Small, d = 0.20) [138]. Abbreviations—BST, bed nucleus of the stria terminalis; Ce, central nucleus of the amygdala; CI, confidence interval; EAc, central extended amygdala; fMRI, functional magnetic resonance imaging.