Effects of Acute Drug Administration on Emotion: a Review of Pharmacological MRI Studies

Emotional blunting is a recognised feature across several psychiatric conditions, including substance use disorders, and may both precede and follow problematic drug use. Previous work has suggested that psychoactive drugs can alter emotional experience by dampening responses to negative stimuli or by changing the salience of other rewards, but the neural mechanisms remain incompletely described. Pharmacological MRI (phMRI) — the pairing of acute drug administration with task-based functional MRI — provides a translational approach to examine how drugs acutely modulate brain circuits involved in emotion, reward and self-regulation, notably the amygdala, striatum, anterior cingulate and prefrontal cortex. Van Hedger and colleagues set out to review recent phMRI studies that tested the acute effects of commonly used recreational drugs on neural responses during emotion-related tasks in healthy human participants. The authors limited inclusion to studies published in the preceding 15 years that (a) used acute drug administration combined with fMRI, (b) tested healthy humans, and (c) required participants to complete a task involving emotion or mood during scanning. The search used online engines (for example PubMed and Google Scholar) and yielded 21 papers meeting these criteria, which are summarised in the review tables.

This article is a narrative review of phMRI studies rather than a systematic review with a detailed risk-of-bias assessment. The authors searched online databases (explicitly naming PubMed and Google Scholar) for articles published in the previous 15 years that met three inclusion criteria: healthy human participants, acute drug administration paired with fMRI (phMRI), and an in-scanner task involving emotion or mood. The extracted text indicates that 21 papers fulfilling these criteria were identified and are presented in the review tables. The review also summarises common methodological features of phMRI studies. Many phMRI experiments used a within-subjects design in which the same participants completed drug and placebo sessions, typically scheduled so scanning occurred at expected peak drug effects. For some drugs with intravenous administration (for example alcohol infusions in some studies, and ketamine), researchers infused the drug before or during fMRI scanning. Subjective and physiological measures were frequently collected during sessions to characterise acute drug effects, but the availability of direct correlations between subjective reports and neural measures varied across studies. The authors do not report a formal meta-analytic pooling of effect sizes or a standardised risk-of-bias appraisal in the extracted text. Because the extracted text contains study-level summaries rather than a methods appendix, details such as exact database search strings, screening procedures, and study quality ratings are not clearly reported in this extraction. The review therefore synthesises patterns across drug classes and tasks using the available study descriptions and reported outcomes.

Across the 21 reviewed phMRI papers, the authors report heterogeneous drug effects on neural responses to emotional stimuli, with some consistent patterns within drug classes and many exceptions. Alcohol: Multiple phMRI studies generally found that alcohol attenuates neural responses to negative emotional stimuli. Alcohol reduced activation in the amygdala, insula, parahippocampal gyrus and visual areas when participants viewed fearful faces. In social drinkers alcohol increased nucleus accumbens activation to neutral faces, and nucleus accumbens responses correlated positively with subjective intoxication ratings, suggesting enhanced reward processing in some contexts. By contrast, heavy drinkers (operationalised in one study as 20–40 drinks per week; other work classified heavy social drinkers as ≥10 drinks per week with 1–5 binge episodes) often showed blunted amygdala responses even in placebo sessions and did not exhibit alcohol-related increases in nucleus accumbens activation, consistent with tolerance or long-term changes in emotional reactivity. Some studies also reported alcohol-related reductions in insula activation to emotional faces regardless of valence; however, at least one study failed to detect alcohol effects on amygdala activation despite similar procedures. Cannabinoids: Findings for THC were inconsistent. Several studies reported THC-related attenuation of amygdala reactivity to fearful or angry faces and reductions in subgenual anterior cingulate activation during negative pictures, while other work reported THC-related increases in posterior regions (for example precuneus) or increases in amygdala responses linked to individual differences in CB1 receptor density. Functional connectivity results were mixed: one study found THC increased basolateral amygdala–prefrontal connectivity during social-threat faces, whereas another found no change in anterior cingulate–amygdala connectivity. Cannabidiol (CBD), in the smaller number of studies examined, tended to decrease activity in anterior and posterior cingulate and the amygdala while viewing fearful faces and to attenuate anterior cingulate–amygdala connectivity. Across cannabinoid studies, reported neural effects were frequently unrelated to subjective measures, and the authors note a need for further combined phMRI/PET work to probe mechanisms such as CB1 receptor density. Analgesics and ketamine: Low-dose μ-opioid agonist administration (for example 0.2 mg buprenorphine) reduced behavioural sensitivity or orienting to fearful faces in healthy volunteers, but phMRI data linking opioids to emotion circuitry are sparse. A phMRI study of oxycodone in low-risk volunteers did not find amygdala or nucleus accumbens differences versus placebo, though a higher oxycodone dose reduced medial orbitofrontal cortex activity while viewing happy faces. Acetaminophen has shown some evidence of blunting emotional ratings and reducing neural responses to social rejection in separate studies, but no studies of acetaminophen and facial-emotion phMRI were reported. Ketamine studies more consistently showed reduced neural responses to negative and neutral stimuli in amygdala and hippocampus, decreased pregenual anterior cingulate activation 24 hours after dosing in one study, and reductions in insula and dorsolateral prefrontal cortex during working memory tasks with negative affective words. Results suggest ketamine acutely reduces brain activation to negative stimuli, with some studies also reporting increased visual and striatal responses to neutral faces. Stimulants: Acute stimulant effects were variable. High-dose modafinil (three times a typical cognitive dose in one study) increased bilateral amygdala, anterior cingulate and striatal activation when viewing fearful faces, whereas methylphenidate did not alter amygdala activation in the same paradigm. Caffeine increased midbrain periaqueductal gray activation and decreased medial prefrontal activity to threat-related stimuli without affecting threat-related amygdala activation. MDMA showed a distinct profile: one study reported reduced amygdala activity to angry faces and increased ventral striatum activation to happy faces, while another MDMA study did not find neural differences despite some behavioural impairment in fear recognition. Psychedelics: Psilocybin reduced amygdala reactivity to negative and neutral pictures, with greater attenuation related to larger increases in subjective positive mood. Functional MRI after a 100 μg dose of LSD also showed decreased amygdala responses to fearful faces, and the magnitude of attenuation correlated with subjective drug effects. Very low ‘‘micro-dose’’ LSD (13 μg) altered amygdala seed-based functional connectivity at rest in one study. The authors note that most psychedelic phMRI work examined negative stimuli and that positive emotional processing under psychedelics remains less explored. Adverse events and subgroup analyses: The extracted text focuses on neural activation patterns and reports few systematic safety data. Several studies compared participants with differing histories of substance use (for example heavy versus social drinkers) and reported attenuated neural and subjective responses in heavier users, suggesting tolerance or long-term blunting effects in some chronic-user groups.

Van Hedger and colleagues interpret the collected evidence as broadly consistent with the idea that acute administration of several psychoactive drugs can reduce neural responsiveness to negative emotional stimuli in limbic regions such as the amygdala. Alcohol, for example, showed relatively consistent amygdala attenuation across studies and sometimes enhanced reward-region responses to neutral social stimuli. Ketamine also produced robust reductions in neural responses to negative stimuli across multiple regions including amygdala, hippocampus and insula, a pattern that the authors suggest could relate to ketamine’s therapeutic effects in depression. At the same time, the authors stress considerable heterogeneity. Cannabinoid effects (particularly for THC) were equivocal, with studies reporting both increases and decreases in amygdala activation and mixed connectivity results; CBD more commonly attenuated amygdala and cingulate responses but has been less extensively studied. Stimulant drugs showed divergent effects depending on the compound and dose: modafinil and caffeine tended to increase reactivity in limbic or midbrain regions, whereas MDMA attenuated amygdala reactivity to certain negative stimuli and enhanced ventral striatal responses to positive stimuli. Psychedelics consistently reduced amygdala responses to negative stimuli in the studies reviewed, but the literature on positive-emotion processing under psychedelics remains sparse. The authors acknowledge multiple limitations and uncertainties. Reduced BOLD signal does not necessarily equate to diminished subjective emotional experience, and many studies did not report robust correlations between neural and self-report measures. Technical and physiological confounds may also affect interpretation: some studies used older 1.5T scanners, and pharmacological effects on vascular tone can alter the BOLD signal independently of neural changes. The review highlights that many studies relied on static facial stimuli, which may limit ecological validity, and that most work focused only on acute administration rather than the consequences of repeated or chronic use. Important gaps include a paucity of recent phMRI studies on opioids and methamphetamine despite their public health relevance, and limited combined phMRI/PET work to probe receptor-level mechanisms (for example CB1 density effects with THC). To address these issues, the authors propose methodological improvements for future research: use of higher-field scanners and complementary imaging techniques (for example arterial spin labelling, PET), adoption of more dynamic and ecologically valid emotional stimuli (for example morphing faces or film clips), and closer examination of relationships between neural responses and subjective affect in both acute and chronic use contexts. They also recommend expanding phMRI study portfolios to include drugs and populations most relevant to current public health concerns.

The authors conclude that acute administration of many psychoactive drugs can reduce neural responses to emotional stimuli in limbic regions, a neural profile that may underlie reported emotional blunting and could influence drug-seeking behaviour. Alcohol and ketamine show relatively consistent dampening of amygdala and related responses to negative stimuli; cannabinoids produce mixed effects with CBD more likely to attenuate limbic activation, while stimulants vary by drug and dose (with MDMA showing attenuation of negative and enhancement of positive responses). Psychedelics reduced amygdala activation to negative stimuli in the studies reviewed, though positive-emotion effects require further study. Finally, the authors emphasise that methodological advances and targeted phMRI work are needed to relate neural changes to subjective affect, to disentangle vascular from neural effects on BOLD, and to investigate chronic use and understudied drugs such as opioids and methamphetamine. These steps, they argue, will help clarify how acute drug effects on emotion-related circuits contribute to both the appeal and the longer-term emotional consequences of drug use.