ISSN 0439-755X
CN 11-1911/B

心理学报 ›› 2022, Vol. 54 ›› Issue (10): 1193-1205.doi: 10.3724/SP.J.1041.2022.01193

• 研究报告 • 上一篇    下一篇


曾红1(), 郑志灵2, 罗晓红1, 王鹏飞1, 王孟成1, 苏得权1, 杨文登1, 黄海娇1, 彭淑娜1   

  1. 1.广州大学教育学院心理系, 广州 510006
    2.广州市交通运输职业学校, 广州 510440
  • 收稿日期:2021-07-11 发布日期:2022-08-24 出版日期:2022-10-25
  • 通讯作者: 曾红
  • 基金资助:

Automatic drug use behavior: Characteristics of cue-induced reactivity and behavior extinction

ZENG Hong1(), ZHENG Zhiling2, LUO Xiaohong1, WANG Pengfei1, WANG Mengcheng1, SU Dequan1, YANG Wendeng1, HUANG Haijiao1, PENG Shuna1   

  1. 1. Department of Psychology, Guangzhou University, Guangzhou 510006, China
    2. Guangzhou Traffic and Transportation Vocational School, Guangzhou 510440, China
  • Received:2021-07-11 Online:2022-08-24 Published:2022-10-25
  • Contact: ZENG Hong


采用ERP技术考察海洛因成瘾者相关线索下的用药行为的性质和神经机制, 在此基础上结合虚拟现实技术(VR)和途径偏向矫正训练(AAT), 探究基于“相关线索-习惯性用药”联结的自动化用药行为的动作消退、替代训练的有效性。研究发现, (1)相比对照线索, 成瘾者在相关线索下的反应时更长, N1波幅更小, N1潜伏期更短; 相比对照组被试, 成瘾者在相关线索下的行为任务的正确率更低, 反应时更长, N1波幅更小; N1波幅的组间组内差异和潜伏期的组内差异主要集中在中央区。(2)结合VR技术的途径偏向矫正训练后, 训练组被试的生理指标和途径偏向系数显著降低, 且显著低于对照组被试; 追踪测试中, 训练组被试的分数显著低于对照组。这表明相关线索下成瘾者的用药行为具有自动化、无意识的性质, 感觉运动区N1的异常活动是其神经机制之一。VR提取用药动作记忆-AAT消退替代训练, 可以有效削弱或消退成瘾者的成瘾动作倾向, 而且训练效果持续性较好, 可达到减少或停止现实情境中的用药和复吸行为的目的。

关键词: 药物成瘾, 相关线索反应, 自动化用药行为, 途径偏向, 消退, ERP, VR


Under relevant cues, people who are addicted tend to use drugs with little attention, purpose or cognitive effort, conforming to an automated “habitual” response behavior. When the substance is not available, this habitual response behavior will be transformed into psychological craving.

A common goal of addiction treatment is to reduce cue-induced reactivity that has automatic and unconscious features. Based on previous research, this study used a self-report questionnaire, two behavioral experiments, and measures of Event-Related Potentials (ERPs) to explore the characteristics and neural mechanisms of cue-induced reactivity in heroin addicts under drug-related cues. On this basis, a training program using Virtual Reality (VR) technology and an Approach-Avoidance task (AAT) was designed to weaken the association between drug-related cues and automatic drug use behavior.

In Study 1, the participants (N= 38) were men who were addicted to heroin (n=19) and a matched sample of healthy controls (n = 19). The Visual Analog Craving Scale (VAS) was used to investigate cue-induced reactivity by self-report. We then assessed ERPs to explore the nature and neural mechanism of cue-induced reactivity. We hypothesized that compared to the healthy controls, the participants who were addicted to heroin would show more cue-induced reactivity in the form of automatic response outside of awareness.

Corroborating our hypotheses in Study 1, we proceeded to Study 2. N= 60 men who were addicted to heroin were equally divided into a treatment group (n= 30) and a no-treatment group (n= 30). The treatment was VR-AAT training for 10 sessions over the course of two months, with the goal of reducing or eliminating cue-induced reactivity. Before and after the treatment, VR-AAT and the Craving Automatized Scale-Substances (CAS-S) questionnaire were used to test the difference between the two groups.

The results showed that 1) In between-group comparisons, VAS scores were higher in the addicts than in the control group. And in Study 2, the within-group analyses showed that self-reports of drug use behavior being “unconscious” and “involuntary” were significantly correlated with the duration of addiction. 2) In within-group analyses, N1 amplitude was smaller, latency was shorter, and reaction time was longer in response to drug-related cues versus neutral cues. Other between-group analyses of the ERP data, addicts demonstrated smaller N1 amplitude and longer reaction time in response to drug-related cues compared to the healthy controls, and these differences were concentrated in the central regions of the brain (i.e., the parietal lobe). 3) After the VR-AAT training, the physiological index (based on temperature, diastolic pressure and systolic pressure) decreased significantly, and the coefficient of approach bias of AAT [(Avoid behavior RTs drug-related- Approach behavior RTs drug-related) - (Avoid behavior RTs neutral- Approach behavior RTs neutral)] was significantly smaller than before training. The results were the same at a two-month follow-up.

Our findings revealed that cue-induced reactivity showed characteristics of being automatic and unconscious, with greater N1 reactivity in sensorimotor and related brain areas. The VR-AAT training was effective in reducing cue-induced reactions among men addicted to heroin, suggesting that this method has potential applied value in treating heroin addiction and in designing relapse prevention programs. This research contributes to the addiction literature in two ways. First, AAT training has been used in the treatment of alcohol addiction, but not for other forms of addiction. This is the first research to use AAT training to treat heroin addiction. Second, the treatment in the current study is the first one to combine AAT with virtual reality technology. The VR-AAT method might also be a feasible approach for other substance abuse treatments, although more experimental evidence would be needed to support this. One limitation of this study was that the basis of automatic drug use behavior and automatic refusing drug use behavior cannot be distinguished. This issue can be investigated in future research.

Key words: drug addiction, cue induced reactivity, automatic drug use behavior, bias approach, extinguish, ERP, VR