Department of Diabetes, Te Toka Tumai, Te Whatu Ora, New Zealand; University of Auckland, New Zealand.
Ivey Public Health Partners, LLC, New Zealand.
Appetite. 2023 Dec 1;191:107066. doi: 10.1016/j.appet.2023.107066. Epub 2023 Oct 16.
Individualised management of obesity remains challenging and, to date, most treatment is based on clinical judgement. This study aimed to develop and validate a novel questionnaire-based tool to identify three pre-defined eating behavior (EB) traits, emotional eating, reduced satiety (constant hunger) and reduced satiation (feasters) that may predict selective medication response given their targeted actions. We recruited 977 individuals from a tertiary academic diabetes clinic to participate in this two-phase validation study. Participants self-reported weight management activities and were asked to self-assess their EB characteristics. The initial questionnaire included 42 visual analogue scale questions. In Phase I, 729 participants completed the questionnaire, including Māori (11.8%) and Pacific peoples (19.3%). After random division of the study sample, Exploratory Factor Analysis (EFA) confirmed a three-factor model as the best fit. Stepwise removal of items with inadequate factor loading retained 27 of 42 items, which accounted for 96% of the variance. Confirmatory Factor Analysis (CFA), performed on the second half of the sample, demonstrated good model fit with the final 27-item questionnaire. Internal consistency was high for factor (α = 0.82-0.95) and demographic subgroups, and similar to those obtained in the EFA. Test-retest reliability in a subset of 399 participants who repeated the questionnaire after a four-week interval (Phase II) showed moderate to good reliability. Participants classified into one of three EB types based on the highest median score among the factors. Test-retest reliability was robust for emotional eaters (71.25%) and constant hunger (68.9%). The correlation between aggregate EB score (sum of three EB scores) and BMI was significant (Spearman rho = 0.314, P = .0005). The questionnaire reliably identified three distinct EB traits, which may be informative for precision medicine applications for obesity management.
个体化肥胖管理仍然具有挑战性，迄今为止，大多数治疗方法都是基于临床判断。本研究旨在开发和验证一种新的基于问卷的工具，以识别三种预先定义的饮食行为（EB）特征，即情绪性进食、饱腹感降低（持续饥饿）和饱食感降低（暴食者），这些特征可能因其靶向作用而预测选择性药物反应。我们从一家三级学术糖尿病诊所招募了 977 名参与者参与这项两阶段验证研究。参与者自我报告了体重管理活动，并被要求自我评估他们的 EB 特征。初始问卷包括 42 个视觉模拟量表问题。在第一阶段，729 名参与者完成了问卷，其中包括毛利人（11.8%）和太平洋岛民（19.3%）。在研究样本随机分组后，探索性因素分析（EFA）证实了三因素模型是最佳拟合。逐步去除因子负荷不足的项目后，保留了 42 个项目中的 27 个，占总方差的 96%。对样本的另一半进行验证性因素分析（CFA）表明，最终的 27 项问卷具有良好的模型拟合度。因子（α=0.82-0.95）和人口统计学亚组的内部一致性较高，与 EFA 中获得的结果相似。在随后的 399 名参与者中，有 399 名参与者在四周后（第二阶段）重复了问卷，测试-重测信度显示为中等至良好。根据三个因素中最高中位数得分将参与者分为三种 EB 类型之一。情绪性进食者（71.25%）和持续饥饿者（68.9%）的重测信度稳定。EB 总分（三个 EB 得分之和）与 BMI 之间的相关性显著（Spearman rho=0.314，P=0.0005）。该问卷可靠地识别了三种不同的 EB 特征，这可能对肥胖管理的精准医学应用具有信息性。
