Increased EEG gamma power under exposure to drug-related cues: a translational index for cue-elicited craving in METH-dependent individuals

Methamphetamine (METH) dependence is a crucial public health concern. In the past 10 years, the number of illicit METH users in China has rapidly increased [1]. For drug users, one significant symptom is the craving elicited by drug-related cues [2]. Exposure to such cues may trigger drug-seeking and drug-taking behaviors, which could influence relapse [3‐7]. Thus, improving our understanding of cue-induced cravings and developing effective interventions are necessary.

Over the past few decades, neuroimaging techniques have been significant in cue reactivity research. Functional magnetic resonance imaging (fMRI) studies have revealed several processes associated with cue-induced cravings [8, 9]. For example, the central executive network, which is linked with cognitive control and executive functions, is engaged during appetitive cue reactivity [8]. Reward and emotional processing could also influence cue-induced craving [8, 10, 11].

Electroencephalogram (EEG) studies have revealed the brain-mind relationship in cue-induced reactivities based on oscillatory brain dynamics. Previous studies have reported increased activity of beta oscillations due to alcohol or cocaine dependence [12‐14]. Heroin-dependent individuals have enhanced beta and gamma powers than healthy controls (HCs) [15]. Nicotine administration elicits EEG activity shifts from low (delta, theta, alpha-1) to high (alpha-2, beta) frequencies, indicating a state of arousal [16‐18]. Some researchers found that METH-dependent participants showed lower delta and alpha powers with higher beta and gamma powers compared to healthy participants [19]. Zhao et al.’s study of METH use disorder revealed that a cue-induced craving was higher after abstinence of 1–3 months, along with a decreased power spectrum for theta and alpha, and an increase in the beta frequency band [20]. In contrast, some evidence suggests that gamma activity in the Medial Prefrontal Cortex/Orbital Frontal Cortex and right Dorsolateral Prefrontal Cortex decreases after cue exposure [5].

Thus, EEG is an objective and precise method for detecting brain dysfunction in people with substance use disorder, and EEG abnormalities may reflect the underlying changes in brain function caused by chronic drug abuse. In most studies, cue-induced cravings were correlated with increased power in the beta frequency band, which was correlated with perceptual and motor processes [14, 19, 20]. However, only a few studies have investigated the correlation between cue-induced cravings and gamma oscillations. Some studies reported that low-frequency EEG signals are related to alertness and motor imagery, whereas high-frequency signals are related to memory, attention, and emotion [21‐25]. Thus, such high-frequency oscillations are not specific to a particular cognitive function, but represent global binding across various neurons and networks [26‐29].

Evidence from previous studies suggested that gamma oscillations represent activation-related markers of neural communication and are important in complex multisensory perceptual paradigms (e.g., watching a video) as well as in high-order cognitive processes such as attention, memory, and emotion [26, 28, 30‐34]. Since these functions influence cue-induced craving, extracting the features of gamma activity might help us understand how the brains of METH abusers react to drug-related cues.

In our previous study, we developed a METH-related virtual reality (VR) social environment and a VR counterconditioning procedure (VRCP) for cue-induced craving assessment and treatment. Patients with METH dependence showed a significant increase in craving scores when exposed to METH-related cues in the VR environment, with a corresponding increase in heart rate variability. However, participants who received VRCP showed a larger decrease in the craving score from baseline to follow-up assessments compared to those who did not receive VRCP. The current study explores the sensitivity of EEG gamma-band power as an index of cue-induced craving in METH-dependent individuals when they are exposed to a METH-related VR environment. Specifically, we recorded the EEG signals of participants under different conditions (cue-induced vs. rest) before and after receiving VRCP.

Based on previous studies [35], we hypothesized that (1) the VR environment might elicit larger gamma activity in the METH group than in HCs and (2) receiving VRCP might lead to a decrease in gamma frequency power in METH-dependent individuals during cue-induced conditions, compared to the data recorded before VRCP. Two studies were designed to test these hypotheses.

Our study aimed to evaluate how craving elicited by a METH-related VR environment modulates EEG activity in METH abusers. Previous studies suggested that gamma oscillations influence complex multisensory perceptual paradigms [28]; however, no study has investigated gamma power using the METH-related VR environment. In this study, we first measured the EEG spectrum power of METH users under different conditions: resting state and exposure to a METH-related VR environment. The VR environment elicited a significantly larger increase in gamma power in the METH group than in the HC group. This result supports our first hypothesis that EEG gamma power may represent a marker of cue-induced reactivity in METH-dependent disorders. In addition, Fig. 4 shows that gamma activity provoked by drug-related cues was mainly recorded by the frontal electrodes. This finding may suggest some spatial information of the craving process; however, the results should be interpreted with caution [36].

Previous studies emphasized that gamma-band oscillations are significant in object representation [37]. Subsequent studies demonstrated that gamma-band activity is related to bottom-up representation and top-down processes. Gamma-band responses are closely associated with selective attention [38]. Given the relationship between gamma activity and top-down representation, the attractive stimuli exhibited by VR may strongly activate the top-down processes of patients with METH dependence, enhancing their gamma-band activity. Aligning with our results, some authors have concluded that frontal gamma activity is linked to top-down processes [39, 40].

Furthermore, drug-related reactivity impacts emotional properties. The processing of rewards includes the components of “liking” and “wanting” [41‐43]. “Liking” refers to the hedonic impact or affective process within pleasure. Based on these models, the increase in METH-liking scores after watching VR videos represents a growing positive affect. Consistent with our previous study [35], our results showed that patients reported a significant decrease in the VAS score for METH-liking after receiving VRCP, which suggests that VRCP attenuated the affective valance of METH cues.

The power spectrum analysis after VRCP treatment revealed a similar trend in the gamma power. Our results showed that VRCP treatment decreased gamma power when participants were exposed to a METH-related VR environment, compared with their condition before VRCP. The association between gamma oscillations and emotional processing may explain this. An increase in the spectral power of the gamma range has been reported in positive emotional tasks [44]. Some studies have shown increased gamma activity in the right hemisphere for emotional pictures (pleasant and unpleasant) compared to neutral pictures [31, 45], although our results did not show right hemisphere dominance. Maffei et al. found that watching emotional movies elicited greater gamma activity than watching neutral clips [28].

Likewise, evidence for the possible role of gamma-band activity in learning and short-term memory has been reviewed [29, 46]. Some studies have demonstrated that fast oscillations reflect the synchronous activity of large ensembles of neurons [30]. A reasonable explanation is that such high-frequency oscillations are not specific to a particular cognitive process, but rather represent the functional communication between various neurons and networks [26‐29]. Together, both these processes could lead to the observed effects of gamma oscillation.

However, this study has some limitations. First, this study enrolled only male participants, and the sample size was small, making it difficult to generalize the results. Although no statistically significant difference was found in age between the two groups, the difference should not be ignored in terms of brain functioning. Future studies should investigate female METH dependence patients too and expand the scale. Second, no source analysis was performed because of the constraints of the EEG system. Therefore, the brain networks underlying the observed effects remain unclear. Third, VR devices and environments may disturb participants’ baseline condition. To rule out such bias, the control condition should be set up in a VR environment. Fourth, the control group did not respond to the VAS at the resting state. Even if they are not users, they may have relevant knowledge or experiences. Furthermore, participants were not evaluated with a more comprehensive craving questionnaire, leading to a lack of comprehensive information about their craving. In addition, abstinence may induce incubation of a cue-induced drug craving. A previous study [20] reported that a cue-induced craving of METH was higher after 1–3 months’ abstinence than after 3–6, 6–12, and 12–24 months. This effect on a cue-induced craving may impact the different responses to METH-related cues, and should be considered in future studies. Last and most important, the present study did not set up a control condition for VRCP. The observed changes in VAS and gamma-band power after VRCP might be related to other potential variables such as habituation to the stimuli presented in the videos, rather than be an effect of VRCP.

In conclusion, the present study found some preliminary results, suggesting that gamma power may be a marker of cue-induced reactivity in METH-dependent disorders.