A Novel Cognition-guided Neurofeedback Protocol for Methamphetamine Abuse Treatment
Poster No:
530
Submission Type:
Abstract Submission
Authors:
Huixing Gou1, Junjie Bu2, Xiaochu Zhang1
Institutions:
1University of Science and Technology of China, Hefei, Anhui, 2Anhui Medical University, Hefei, Anhui
First Author:
Co-Author(s):
Introduction:
The World Drug Report 2023 reveals a global rise in methamphetamine (meth) production and usage. In China, meth predominant among illicit substance. Meth abusers (MAS) exhibit cognitive deficits, especially in response inhibition (Potvin et al., 2018). Intense craving and recurrent relapse is the key feature of MAS, often triggered by exposure to cues associated with addiction drug (Carter et al., 1999). Recent studies have explored methods as virtual reality cue exposure therapy (Wang et al., 2019) and electrical stimulation (Ekhtiari et al., 2022) to reduce meth cue reactivity (CR), but these suffer from lack of effective tracking or short tracking times.
Neurofeedback (NF) is seen as a promising avenue to ameliorate MA-related impairments (Paulus et al., 2020). Our previous work developed a cognition-guided NF protocol, successfully deactivated nicotine CR in nicotine-dependent participants with significant effects on craving and smoking (Bu et al., 2019, 2021). This study aims to further develop this cognition-guided NF protocol, assess its efficacy in MA and compare it with the current routine treatment (RT) in Chinese compulsory rehabilitation centers.
Neurofeedback (NF) is seen as a promising avenue to ameliorate MA-related impairments (Paulus et al., 2020). Our previous work developed a cognition-guided NF protocol, successfully deactivated nicotine CR in nicotine-dependent participants with significant effects on craving and smoking (Bu et al., 2019, 2021). This study aims to further develop this cognition-guided NF protocol, assess its efficacy in MA and compare it with the current routine treatment (RT) in Chinese compulsory rehabilitation centers.
Methods:
We collected two data samples (Sample I and II) who met DSM-V criteria. Sample I (investigate cognition-guided NF effects for MA), consisting of Real-Feedback Group I (Real-NFG I, n = 31) and Yoke-Feedback Group (Yoke-NFG, received feedback based on Real-NFG I participant brain activity patterns, n = 32). Sample II (compare with RT), including Real-Feedback Group II (Real-NFG II, n = 16) and Routine-Treatment Group (RT-Group, completed normal labor production as usual, n = 16). This study employed a randomized double-blind controlled design. See more details in Fig. 1.
Results:
Short-term effects
Real-NFG I had a significantly better NF performance than Yoke-NFG (Fig. 2A). A two-way mixed-design ANOVA showed a significant group*Visit interaction for both training-learning (F = 6.057, p = 0.017), transfer-learning (F = 7.144, p = 0.009) probabilistic score, d-prime of Go/No-go task (Fig. 2B). And negative correlation was observed between them in Real-NFG I (Fig. 2C). Craving was significantly reduced in Real-NFG than Yoke-NFG (t = -2.106, p = 0.041). Moreover, cluster-based ERP analyses yielded a significant increase in N2 for Real-NFG I (t = 2.285, p = 0.031) but not for Yoke-NFG (t = -0.610, p = 0.547).
Long-term effects
Craving in Real-NFG I significantly decreased compared with Yoke-NFG (Fig. 2D), and this reduction was correlated with NF performance in Real-NFG I (r = -0.380, p = 0.035) but not in Yoke-NFG (r = -0.108, p = 0.556). Moreover, the two groups exhibited significant differences in beta time-frequency (TF) power change (Fig. 2E), and this change correlated with NF performance (r = -0.430, p = 0.022) and craving (r = 0.437, p = 0.020) in Real-NFG I but not in Yoke-NFG (NF: r = 0.136, p = 0.472; craving: r = 0.259, p = 0.166). Furthermore, significant group differences emerged in changes of questionnaires of BDI (t = 2.164, p = 0.035) and BIS (t = 2.627, p = 0.011). Notably, the Real-NFG was more likely not to contact old friends who used meth (Fig. 2F).
Effects prediction
SVM predicts short-term effects with 79.31% accuracy (Fig. 2G) and long-term with 78.57% accuracy (Fig. 2H).
Effects replication
Real-NFG II significantly deactivated the NF probabilistic score (Fig. 2I), increased d-prime (Fig. 2J). And positive correlation between them was observed (r = -0.344, p = 0.192). Moreover, the effects prediction classifier built in Sample I has an accuracy of 75% in Sample II (Fig. 2K).
Real-NFG I had a significantly better NF performance than Yoke-NFG (Fig. 2A). A two-way mixed-design ANOVA showed a significant group*Visit interaction for both training-learning (F = 6.057, p = 0.017), transfer-learning (F = 7.144, p = 0.009) probabilistic score, d-prime of Go/No-go task (Fig. 2B). And negative correlation was observed between them in Real-NFG I (Fig. 2C). Craving was significantly reduced in Real-NFG than Yoke-NFG (t = -2.106, p = 0.041). Moreover, cluster-based ERP analyses yielded a significant increase in N2 for Real-NFG I (t = 2.285, p = 0.031) but not for Yoke-NFG (t = -0.610, p = 0.547).
Long-term effects
Craving in Real-NFG I significantly decreased compared with Yoke-NFG (Fig. 2D), and this reduction was correlated with NF performance in Real-NFG I (r = -0.380, p = 0.035) but not in Yoke-NFG (r = -0.108, p = 0.556). Moreover, the two groups exhibited significant differences in beta time-frequency (TF) power change (Fig. 2E), and this change correlated with NF performance (r = -0.430, p = 0.022) and craving (r = 0.437, p = 0.020) in Real-NFG I but not in Yoke-NFG (NF: r = 0.136, p = 0.472; craving: r = 0.259, p = 0.166). Furthermore, significant group differences emerged in changes of questionnaires of BDI (t = 2.164, p = 0.035) and BIS (t = 2.627, p = 0.011). Notably, the Real-NFG was more likely not to contact old friends who used meth (Fig. 2F).
Effects prediction
SVM predicts short-term effects with 79.31% accuracy (Fig. 2G) and long-term with 78.57% accuracy (Fig. 2H).
Effects replication
Real-NFG II significantly deactivated the NF probabilistic score (Fig. 2I), increased d-prime (Fig. 2J). And positive correlation between them was observed (r = -0.344, p = 0.192). Moreover, the effects prediction classifier built in Sample I has an accuracy of 75% in Sample II (Fig. 2K).
Conclusions:
We deactivated meth CR in MAS by cognition-guided NF, thereby enhancing their RI capability. Participants' baseline NF performance and information could predict the intervention effects. Furthermore, we achieved effects replication in a separate data sample. All of these results suggest that our NF protocol is a promising avenue for drug addiction.
Disorders of the Nervous System:
Psychiatric (eg. Depression, Anxiety, Schizophrenia) 1
Higher Cognitive Functions:
Executive Function, Cognitive Control and Decision Making 2
Keywords:
Addictions
Behavioral Therapy
Cognition
Computational Neuroscience
Electroencephaolography (EEG)
Psychiatric Disorders
Therapy
1|2Indicates the priority used for review
Provide references using author date format
Bu, J. (2019). 'Effect of deactivation of activity patterns related to smoking cue reactivity on nicotine addiction', Brain, vol. 142, no. 6, pp. 1827–1841
Carter, B. L. (1999). 'Meta-analysis of cue-reactivity in addiction research', Addiction, vol. 94, no. 3, pp. 327–340
Ekhtiari, H. (2022). 'Transcranial direct current stimulation to modulate fMRI drug cue reactivity in methamphetamine users: A randomized clinical trial', Human Brain Mapping, vol. 43, no. 17, pp. 5340–5357
Paulus, M. P. (2020). 'Neurobiology, Clinical Presentation, and Treatment of Methamphetamine Use Disorder: A Review', JAMA Psychiatry, vol. 77, no. 9, p. 959
Potvin, S. (2018). 'Cognitive deficits in individuals with methamphetamine use disorder: A meta-analysis', Addictive Behaviors, vol. 80, pp. 154–160
Wang, Y. (2019). 'A virtual reality counterconditioning procedure to reduce methamphetamine cue-induced craving', Journal of Psychiatric Research, vol. 116, pp. 88–94