Table3_Association between rest-activity rhythm and diabetic retinopathy among US middle-age and older diabetic adults.docx

BackgroundThe disruption of circadian rhythm has been reported to aggravate the progression of diabetic retinopathy (DR). Rest-activity rhythm (RAR) is a widely used method for measuring individual circadian time influencing behavior. In this study, we sought to explore the potential association between RAR and the risk of DR.
MethodsDiabetic participants aged over 40 from 2011-2014 National Health and Nutrition Examination Survey (NHANES) were enrolled. Data from the wearable device ActiGraph GT3X was used to generate RAR metrics, including interdaily stability (IS), intradaily variability (IV), most active 10-hour period (M10), least active 5-hour period (L5), and Relative amplitude (RA). Weighted multivariable logistic regression analysis and restricted cubic spline analysis were conducted to examine the association between RAR metrics and DR risk. Sensitivity analysis was also conducted to examine the robustness of the findings. An unsupervised K-means clustering analysis was conducted to identify patterns in IV and M10.
ResultsA total of 1,096 diabetic participants were enrolled, with a DR prevalence of 20.53%. The mean age of participants was 62.3 years, with 49.57% being male. After adjusting covariates, IV was positively associated with DR (β: 3.527, 95%CI: 1.371-9.073). Compared with the lowest quintile of IV, the highest quintile of IV had 136% higher odds of DR. In contrast, M10 was negatively associated with DR (β: 0.902, 95%CI: 0.828-0.982), with participants in the highest M10 quintile showing 48.8% lower odds of DR. Restricted cubic spline analysis confirmed that these associations were linear. Meanwhile, sensitivity analysis confirmed the robustness. K-means clustering identified three distinct clusters, with participants in Cluster C (high-IV, low-M10) had a significantly higher risk of DR comparing with Cluster A (low-IV, high-M10).
ConclusionA more fragmented rhythm and lower peak activity level might be associated with an increased risk of DR. These findings indicate that maintaining a more rhythmic sleep-activity behavior might mitigate the development of DR. Further research is necessary to establish causality and understand the underlying mechanisms, and focus on whether interventions designed to enhance daily rhythm stability and increase diurnal activity level can effectively mitigate the risk of progression of DR.

研究背景
有研究表明，昼夜节律紊乱可加重糖尿病视网膜病变（diabetic retinopathy, DR）的病情进展。休息-活动节律（rest-activity rhythm, RAR）是评估影响个体行为的昼夜节律时间的常用手段。本研究旨在探讨RAR与DR患病风险之间的潜在关联。
研究方法
本研究纳入了2011-2014年美国国家健康与营养检查调查（National Health and Nutrition Examination Survey, NHANES）中40岁以上的糖尿病参与者。研究采用可穿戴设备ActiGraph GT3X采集的数据生成RAR相关指标，包括日间稳定性（interdaily stability, IS）、日内变异性（intradaily variability, IV）、最活跃10小时时段（most active 10-hour period, M10）、最不活跃5小时时段（least active 5-hour period, L5）以及相对振幅（Relative amplitude, RA）。采用加权多变量logistic回归分析与限制性立方样条分析，探究RAR指标与DR患病风险之间的关联；同时开展敏感性分析以验证研究结果的稳健性。此外，通过无监督K-means聚类分析，识别IV与M10的特征模式。
研究结果
本研究共纳入1096名糖尿病参与者，DR患病率为20.53%；参与者平均年龄为62.3岁，男性占比49.57%。校正混杂因素后，IV与DR患病风险呈正相关（β=3.527，95%置信区间：1.371~9.073）。与IV最低五分位数组相比，IV最高五分位数组的DR患病比值比升高136%。与之相反，M10与DR患病风险呈负相关（β=0.902，95%置信区间：0.828~0.982）；M10最高五分位数组的DR患病比值比降低48.8%。限制性立方样条分析证实，上述关联均呈线性特征。同时，敏感性分析验证了结果的稳健性。K-means聚类分析共识别出3个特征迥异的聚类簇；与聚类簇A（低IV、高M10）相比，聚类簇C（高IV、低M10）的参与者DR患病风险显著更高。
研究结论
节律更碎片化以及活动峰值水平更低，或与DR患病风险升高相关。本研究结果提示，维持更规律的睡眠-活动行为或可延缓DR的病情进展。未来仍需开展进一步研究以明确其中的因果关系并阐释潜在机制，同时需聚焦于旨在提升日常节律稳定性与日间活动水平的干预措施能否有效降低DR的病情进展风险。