Brown Denver M Y, Burkart Sarah, Groves Claire I, Balbim Guilherme Moraes, Pfledderer Christopher D, Porter Carah D, Laurent Christine St, Johnson Emily K, Kracht Chelsea L
Kansas State University, 1105 Sunset Ave, Manhattan, KS 66502 USA.
University of South Carolina, Arnold School of Public Health, 921 Assembly St, Columbia, SC 29208 USA.
J Act Sedentary Sleep Behav. 2024;3(1):23. doi: 10.1186/s44167-024-00062-8. Epub 2024 Oct 2.
Compositional data analysis (CoDA) techniques are well suited for examining associations between 24-h movement behaviors (i.e., sleep, sedentary behavior, physical activity) and indicators of health given they recognize these behaviors are co-dependent, representing relative parts that make up a whole day. Accordingly, CoDA techniques have seen increased adoption in the past decade, however, heterogeneity in research reporting practices may hinder efforts to synthesize and quantify these relationships via meta-analysis. This systematic review described reporting practices in studies that used CoDA techniques to investigate associations between 24-h movement behaviors and indicators of health.
A systematic search of eight databases was conducted, in addition to supplementary searches (e.g., forward/backward citations, expert consultation). Observational studies that used CoDA techniques involving log-ratio transformation of behavioral data to examine associations between time-based estimates of 24-h movement behaviors and indicators of health were included. Reporting practices were extracted and classified into seven areas: (1) methodological justification, (2) behavioral measurement and data handling strategies, (3) composition construction, (4) analytic plan, (5) composition-specific descriptive statistics, (6) model results, and (7) auxiliary information. Study quality and risk of bias were assessed by the National Institutes of Health Quality Assessment Tool for Observational Cohort and Cross-sectional Studies.
102 studies met our inclusion criteria. Reporting practices varied considerably across areas, with most achieving high standards in methodological justification, but inconsistent reporting across all other domains. Some items were reported in all studies (e.g., how many parts the daily composition was partitioned into), whereas others seldom reported (e.g., definition of a day: midnight-to-midnight versus wake-to-wake). Study quality and risk of bias was fair in most studies (85%).
Current studies generally demonstrate inconsistent reporting practices. Consistent, clear and detailed reporting practices are evidently needed moving forward as the field of time-use epidemiology aims to accurately capture and analyze movement behavior data in relation to health outcomes, facilitate comparisons across studies, and inform public health interventions and policy decisions. Achieving consensus regarding reporting recommendations is a key next step.
The online version contains supplementary material available at 10.1186/s44167-024-00062-8.
成分数据分析（CoDA）技术非常适合用于研究24小时运动行为（即睡眠、久坐行为、身体活动）与健康指标之间的关联，因为这些技术认识到这些行为是相互依赖的，代表了构成一整天的相对部分。因此，在过去十年中，CoDA技术的应用越来越广泛，然而，研究报告方法的异质性可能会阻碍通过荟萃分析来综合和量化这些关系的努力。本系统评价描述了使用CoDA技术研究24小时运动行为与健康指标之间关联的研究中的报告方法。
除了补充检索（如向前/向后引文、专家咨询）外，还对八个数据库进行了系统检索。纳入了使用CoDA技术的观察性研究，这些研究涉及对行为数据进行对数比转换，以检验基于时间的24小时运动行为估计值与健康指标之间的关联。提取报告方法并分为七个领域：（1）方法学依据；（2）行为测量和数据处理策略；（3）成分构建；（4）分析计划；（5）特定成分描述性统计；（6）模型结果；（7）辅助信息。采用美国国立卫生研究院观察性队列研究和横断面研究质量评估工具评估研究质量和偏倚风险。
102项研究符合我们的纳入标准。各领域的报告方法差异很大，大多数在方法学依据方面达到了高标准，但在所有其他领域的报告不一致。有些项目在所有研究中都有报告（如每日成分分为多少部分），而其他项目很少报告（如一天的定义：午夜至午夜还是醒至醒）。大多数研究（85%）的研究质量和偏倚风险为中等。
当前研究普遍表明报告方法不一致。随着时间利用流行病学领域旨在准确捕捉和分析与健康结果相关的运动行为数据、促进跨研究比较并为公共卫生干预和政策决策提供信息，显然需要一致、清晰和详细的报告方法。就报告建议达成共识是下一步的关键。
在线版本包含可在10.1186/s44167-024-00062-8获取的补充材料。
