Department of Psychiatry and Addiction Center, University of Michigan, 4250 Plymouth Road, Ann Arbor, MI, 48109, USA.
Cogn Affect Behav Neurosci. 2021 Oct;21(5):1101-1114. doi: 10.3758/s13415-021-00907-8. Epub 2021 May 11.
The present study identified subgroups based on inhibitory and reward activation, two key neural functions involved in risk-taking behavior, and then tested the extent to which subgroup differences varied by age, sex, behavioral and familial risk, and substance use. Participants were 145 young adults (18-21 years old; 40.0% female) from the Michigan Longitudinal Study. Latent profile analysis (LPA) was used to establish subgroups using task-based brain activations. Demographic and substance use differences between subgroups were then examined in logistic regression analyses. Whole-brain task activations during a functional magnetic resonance imaging go/no-go task and monetary incentive delay task were used to identify beta weights as input for LPA modeling. A four-class model showed the best fit with the data. Subgroups were categorized as: (1) low inhibitory activation/moderate reward activation (39.7%), (2) moderate inhibitory activation/low reward activation (22.7%), (3) moderate inhibitory activation/high reward activation (25.2%), and (4) high inhibitory activation/high reward activation (12.4%). Compared with the other subgroups, Class 2 was older, less likely to have parental alcohol use disorder, and had less alcohol use. Class 4 was the youngest and had greater marijuana use. Classes 1 and 3 did not differ significantly from the other subgroups. These findings demonstrate that LPA applied to brain activations can be used to identify distinct neural profiles that may explain heterogeneity in substance use outcomes and may inform more targeted substance use prevention and intervention efforts.
本研究基于抑制和奖励激活两个关键的神经功能确定了亚组,然后测试了亚组差异在多大程度上因年龄、性别、行为和家族风险以及物质使用而有所不同。参与者是来自密歇根纵向研究的 145 名年轻成年人(18-21 岁;40.0%为女性)。使用潜在剖面分析(LPA)根据基于任务的大脑激活来建立亚组。然后在逻辑回归分析中检查亚组之间的人口统计学和物质使用差异。在功能磁共振成像 Go/No-Go 任务和货币奖励延迟任务期间的全脑任务激活被用来识别 beta 权重作为 LPA 建模的输入。四分类模型与数据拟合最好。亚组分为:(1)低抑制激活/中等奖励激活(39.7%),(2)中等抑制激活/低奖励激活(22.7%),(3)中等抑制激活/高奖励激活(25.2%),和(4)高抑制激活/高奖励激活(12.4%)。与其他亚组相比,第 2 类年龄较大,父母酒精使用障碍的可能性较小,酒精使用量较少。第 4 类年龄最小,大麻使用量最大。第 1 类和第 3 类与其他亚组没有显著差异。这些发现表明,应用于大脑激活的潜在剖面分析可以用于识别可能解释物质使用结果异质性的独特神经特征,并为更有针对性的物质使用预防和干预措施提供信息。