Cognitive and neural mechanisms of cognitive-behavioral therapy for avoidant/restrictive food intake disorder

摘要 避免/限制性食物摄入障碍（ARFID）影响3%的儿童和青少年， 营养缺乏、补充剂依赖和心理社会损害。ARFID遵循一种长期的 当然，没有循证治疗。ARFID的标志性特征是避免食物，这可能 维持极端水平的食物新恐惧症和/或过度激活的恐惧电路，以应对 食物线索我们的团队开发了一种手动认知行为疗法（CBT-AR）， 食物新恐惧症（认知机制）和恐惧回路（神经机制），以减少食物回避 （临床结果）。根据NIMH的实验治疗方法，在这个探索性/发展性的研究中， 分阶段R61/R33，我们将利用多学科专家团队治疗ARFID，神经 食物动机的机制，并对临床试验进行统计分析，以进行机械随机 CBT-AR的对照试验。首先，为了建立目标参与度，我们将随机抽取50名青少年（年龄10- 18岁） 与ARFID以1：1的比例进行15次每周一次的CBT-AR与营养咨询，以建立 瞄准目标我们选择营养咨询作为我们的积极控制，因为它不包括 关键的CBT-AR干预暴露，因此不太可能参与我们的目标机制。我们 假设与营养咨询相比，随机接受CBT-AR的患者将显示出显著的 食物新恐惧症（认知机制）和恐惧回路（神经机制）的下降幅度更大， 在标准化fMRI食物提示范例中对食物提示的反应（主要ROI：前扣带皮层 [ACC]次级ROI：杏仁核、眶额皮质[OFC]）。我们还将检查每周食物的变化 新恐惧症（认知机制），以确定进一步受益停止的会话数量，并使用 R33的CBT-AR最佳剂量。如果我们能够证明R33 CBT-AR组从治疗前到治疗后至少减少d=.40，并且治疗后 CBT-AR与营养咨询在食物新恐惧症（认知）中至少d=.40的组差异 或恐惧回路（Go/No-Go ROI：ACC;神经机制）。接下来，我们将随机抽取70名青年 (ages 10- 18岁）与ARFID以1：1的方式与CBT-AR的优化剂量或相同的疗程数进行比较 的营养咨询，以复制目标参与和建立目标验证。我们假设， 与营养咨询相比，随机接受CBT-AR的患者将表现出更大的减少， 在食物回避，这些减少食物回避将介导的减少食物 新恐惧症（认知机制）和恐惧回路（神经机制）激活。如果成功， 拟议的干预措施（CBT-AR）可以填补ARFID患者的一个重要的未满足需求， 进行更大规模的功效和有效性试验。

ABSTRACT Avoidant/restrictive food intake disorder (ARFID) affects 3% of children and adolescents and results in nutritional deficiencies, supplement dependence, and psychosocial impairment. ARFID follows a chronic course, and has no evidence-based treatment. The hallmark feature of ARFID is food avoidance, which may be maintained by extreme levels of food neophobia and/or hyperactivation of fear circuitry in response to food cues. Our team has developed a manualized cognitive-behavioral therapy (CBT-AR) that directly targets both food neophobia (cognitive mechanism) and fear circuitry (neural mechanism) to reduce food avoidance (clinical outcome). In line with NIMH’s experimental therapeutics approach, in this exploratory/developmental phased R61/R33, we will leverage a multidisciplinary team of experts in the treatment of ARFID, neural mechanisms of food motivation, and statistical analysis of clinical trials to conduct a mechanistic randomized controlled trial of CBT-AR. First, to establish target engagement, we will randomize 50 youth (ages 10-18yo) with ARFID in a 1:1 ratio to 15 weekly sessions (via telehealth) of CBT-AR vs. nutrition counseling to establish target engagement. We chose nutrition counseling as our active control because it does not include the crucial CBT-AR intervention of exposure, and is therefore unlikely to engage our target mechanisms. We hypothesize that, compared to nutrition counseling, patients randomized to CBT-AR will show significantly greater decreases in food neophobia (cognitive mechanism) and fear circuitry (neural mechanism) in response to food cues during a standardized fMRI food cue paradigm (primary ROI: anterior cingulate cortex [ACC]; secondary ROIs: amygdala, orbitofrontal cortex [OFC]). We will also examine weekly change in food neophobia (cognitive mechanism) to identify the number of sessions at which further benefit ceases, and use this optimized dose of CBT-AR for the R33. We will move on to the R33 if we are able to demonstrate a reduction of at least d=.40 in the CBT-AR group from pre- to post-treatment AND a post-treatment between- group difference of at least d=.40 in CBT-AR vs. nutrition counseling in either food neophobia (cognitive mechanism) OR fear circuitry (Go/No-Go ROI: ACC; neural mechanism). Next, we will randomize 70 youth (ages 10-18yo) with ARFID in a 1:1 fashion to the optimized dose of CBT-AR or the same number of sessions of nutrition counseling to replicate target engagement and establish target validation. We hypothesize that, compared to nutrition counseling, patients randomized to CBT-AR will exhibit significantly greater reductions in food avoidance, and that these reductions in food avoidance will be mediated by reductions in food neophobia (cognitive mechanism) and fear circuitry (neural mechanism) activation. If successful, the proposed intervention (CBT-AR) could fill an important unmet need for those living with ARFID and pave the way for larger-scale efficacy and effectiveness trials.