Association of Generalized Anxiety Disorder With Autonomic Hypersensitivity and Blunted Ventromedial Prefrontal Cortex Activity During Peripheral Adrenergic Stimulation: A Randomized Clinical Trial

Abstract

Importance: β-Adrenergic stimulation elicits heart palpitations and dyspnea, key features of acute anxiety and sympathetic arousal, yet no neuroimaging studies have examined how the pharmacologic modulation of interoceptive signals is associated with fear-related neurocircuitry in individuals with generalized anxiety disorder (GAD).

Objective: To examine the neural circuitry underlying autonomic arousal induced via isoproterenol, a rapidly acting, peripheral β-adrenergic agonist akin to adrenaline.

Design, setting, and participants: This crossover randomized clinical trial of 58 women with artifact-free data was conducted from January 1, 2017, to November 31, 2019, at the Laureate Institute for Brain Research in Tulsa, Oklahoma.

Exposures: Functional magnetic resonance imaging was used to assess neural responses during randomized intravenous bolus infusions of isoproterenol (0.5 and 2.0 μg) and saline, each administered twice in a double-blind fashion.

Main outcomes and measures: Blood oxygen level-dependent responses across the whole brain during isoproterenol administration in patients with GAD vs healthy comparators. Cardiac and respiratory responses, as well as interoceptive awareness and anxiety, were also measured during the infusion protocol.

Results: Of the 58 female study participants, 29 had GAD (mean [SD] age, 26.9 [6.8] years) and 29 were matched healthy comparators (mean [SD] age, 24.4 [5.0] years). During the 0.5-μg dose of isoproterenol, the GAD group exhibited higher heart rate responses (b = 5.34; 95% CI, 2.06-8.61; P = .002), higher intensity ratings of cardiorespiratory sensations (b = 8.38; 95% CI, 2.05-14.71; P = .01), higher levels of self-reported anxiety (b = 1.04; 95% CI, 0.33-1.76; P = .005), and significant hypoactivation in the ventromedial prefrontal cortex (vmPFC) that was evident throughout peak response (Cohen d = 1.55; P < .001) and early recovery (Cohen d = 1.52; P < .001) periods. Correlational analysis of physiological and subjective indexes and percentage of signal change extracted during the 0.5-μg dose revealed that vmPFC hypoactivation was inversely correlated with heart rate (r56 = -0.51, adjusted P = .001) and retrospective intensity of both heartbeat (r56 = -0.50, adjusted P = .002) and breathing (r56 = -0.44, adjusted P = .01) sensations. Ventromedial prefrontal cortex hypoactivation correlated inversely with continuous dial ratings at a trend level (r56 = -0.38, adjusted P = .051), whereas anxiety (r56 = -0.28, adjusted P = .27) and chronotropic dose 25 (r56 = -0.14, adjusted P = .72) showed no such association.

Conclusions and relevance: In this crossover randomized clinical trial, women with GAD exhibited autonomic hypersensitivity during low levels of adrenergic stimulation characterized by elevated heart rate, heightened interoceptive awareness, increased anxiety, and a blunted neural response localized to the vmPFC. These findings support the notion that autonomic hyperarousal may be associated with regulatory dysfunctions in the vmPFC, which could serve as a treatment target to help patients with GAD more appropriately appraise and regulate signals of sympathetic arousal.

Trial registration: ClinicalTrials.gov Identifier: NCT02615119.

Conflict of interest statement

Conflict of Interest Disclosures: Dr Ajijola reported having ownership in NeuCures outside the submitted work. Dr Kuplicki reported receiving grants from the National Institute of General Medical Sciences during the conduct of the study. Dr Paulus reported receiving general institutional support from the William K. Warren Foundation and grants from the National Institute of General Medical Sciences during the conduct of the study. No other disclosures were reported.

Figures

Figure 1.. CONSORT Diagram

BMI indicates body mass index; EEG, electroencephalography; GAD, generalized anxiety disorder; HC, healthy comparator; and MRI, magnetic resonance imaging.

Figure 2.. Physiological, Perceptual, and Affective Responses During Isoproterenol and Saline Infusions

Box plots are shown representing median (thick line) and 25th and 75th quartiles (thin lines) with whiskers extending to 1.5 × the IQR. Chronotropic dose 25 (CD25) values reflect the dose required to elevate the heart rate by 25/min. GAD indicates general anxiety disorder; HCs, healthy comparators.

Figure 3.. Group Differences in Ventromedial Prefrontal Cortex Response From Whole-Brain Analyses

A, Significant group activation differences during the peak response period were only found during administration of the 0.5-μg dose at a voxelwise threshold of P < .001 and a 95% false-positive rate cluster correction. B, Significant group activation differences during the early recovery period were only found during administration of the 0.5-μg dose. Right insets show the percentage of signal change above baseline for each dose. C, Mean activitation time course for the significant ventromedial prefrontal cortex (vmPFC) region (blue cluster from peak period) identified during the 0.5-μg infusion. Shaded lines indicate SEMs. BOLD indicates blood oxygenation level–dependent response; GAD, generalized anxiety disorder; and HC, healthy comparator.

Figure 4.. Insular Cortex Responses During Isoproterenol and Saline Infusion

A, Increased activity localized within bilateral clusters of ventral insular cortex was observed across all participants via voxelwise, whole-brain analysis during the peak effect epoch of the 2.0-μg isoproterenol infusion. B, No significant differences were found between groups within the ventral insular cortex clusters at each dose. The 2.0-μg infusion elicited a significantly larger insula response than the saline and the 0.5-μg infusion. The clusters identified at 2.0 μg were used as masks for extracting signal change for each dose and the saline infusions based on prior studies, in which the 2.0-μg dose proved to be sensitive at eliciting insula activity in healthy individuals. C, Cohen d effect sizes when comparing each isoproterenol dose with saline. Error bars indicate SEMs. GAD indicates generalized anxiety disorder; HC, healthy comparator.