Division of Epidemiology and Community Health, University of Minnesota School of Public Health, 1300 S 2nd St Unit 300, Minneapolis, MN 55454, United States of America; Minnesota Population Center, University of Minnesota, 225 19th Ave S #50th, Minneapolis, MN 55455, United States of America.
Department of Psychology, University of Minnesota, 75 E River Rd, Minneapolis, MN 55455, United States of America.
Prev Med. 2021 May;146:106455. doi: 10.1016/j.ypmed.2021.106455. Epub 2021 Feb 23.
The aim of this paper was to better understand how child and adult adversities cluster together into classes, and how these classes relate to body weight and obesity. Analyses included 2015 and 2018 data from emerging adults (18-25 years old) who participated in a state surveillance system of 2- and 4-year college students in Minnesota (N = 7475 in 2015 and N = 6683 in 2018). Latent Class Analyses (LCA) of 12 child and adult adversities were run stratified by gender and replicated between 2015 and 2018. The distal outcome procedure and three-step Bolck-Croon-Hagenaars approach were used to estimate predicted BMI means and predicted probabilities of obesity for each class, adjusted for covariates. The LCA identified seven classes in women and 5 in men. In women, BMI ranged from 23.9 kg/m in the lowest-BMI class ("Adult Adversities and Childhood Household Dysfunction"; 95% CI: 22.6-25.1) to 27.3 kg/m in the highest-BMI class ("High Lifetime Adversities"; 95% CI: 25.9-28.7), a statistically significant difference of 3.4 kg/m. In men, the adjusted BMIs ranged from 24.6 kg/m ("Low Adversities"; 95% CI: 24.3-25.0) to 26.0 kg/m ("Childhood Household Mental Illness"; 95% CI: 25.1-26.9), a statistically significant difference of 1.4 kg/m. The pattern was similar for obesity. These results indicate that specific classes of child and adult adversities are strongly associated with BMI and obesity, particularly in women. A key contribution of LCA appeared to be identification of small classes at high risk for excess weight.
本文旨在深入了解儿童和成人逆境如何聚类为不同类别,以及这些类别与体重和肥胖的关系。分析数据来源于明尼苏达州的一项州级监测系统,该系统涵盖了 2 年制和 4 年制大学生(2015 年有 7475 人,2018 年有 6683 人)。对 12 种儿童和成人逆境进行潜在类别分析(LCA),按性别分层,并在 2015 年和 2018 年之间进行复制。采用远端结果程序和三步 Bolck-Croon-Hagenaars 方法,估计每个类别的预测 BMI 均值和肥胖预测概率,调整了协变量。LCA 在女性中识别出 7 个类别,在男性中识别出 5 个类别。在女性中,BMI 从最低 BMI 类别(“成人逆境和童年家庭功能障碍”;95%CI:22.6-25.1)的 23.9kg/m 到最高 BMI 类别(“终生高逆境”;95%CI:25.9-28.7),差异为 3.4kg/m,具有统计学意义。在男性中,调整后的 BMI 从“低逆境”(95%CI:24.3-25.0)到“童年家庭精神疾病”(95%CI:25.1-26.9),差异为 1.4kg/m,差异具有统计学意义。肥胖的模式与之类似。这些结果表明,特定类别的儿童和成人逆境与 BMI 和肥胖密切相关,尤其是在女性中。LCA 的一个重要贡献似乎是识别出了高风险的小类别。