基于倾向性评分探讨睡眠时相提前综合征与2型糖尿病的关系

高绒绒; 高倩; 王彤

doi:10.16462/j.cnki.zhjbkz.2023.01.010

基于倾向性评分探讨睡眠时相提前综合征与2型糖尿病的关系

doi: 10.16462/j.cnki.zhjbkz.2023.01.010

030001 太原，山西医科大学公共卫生学院卫生统计学教研室

基金项目:

国家自然科学基金 82073674

山西省科技重大专项项目 202102130501003

山西省科技重大专项项目 202005D121008

详细信息

通讯作者:
王彤，E-mail：tongwang@sxmu.edu.cn

中图分类号: R181.2
计量
- 文章访问数: 253
- HTML全文浏览量: 124
- PDF下载量: 45
- 被引次数: 0
出版历程
- 收稿日期: 2022-03-28
- 修回日期: 2022-06-08
- 网络出版日期: 2023-02-09
- 刊出日期: 2023-01-10

Relationship between advanced sleep phase syndrome and type 2 diabetes mellitus based on the propensity score method

Department of Health Statistics, School of Public Health, Shanxi Medical University, Taiyuan 030001, China

Funds:

National Natural Science Foundation of China 82073674

The Grant of the Major Science and Technology Project of Shanxi Province 202102130501003

The Grant of the Major Science and Technology Project of Shanxi Province 202005D121008

More Information

Corresponding author: WANG Tong, E-mail: tongwang@sxmu.edu.cn

摘要

摘要: 目的探讨睡眠时相提前综合征(advanced sleep phase syndrome, ASPS)和2型糖尿病(type 2 diabetes mellitus, T2DM)间的关联性，为深入理解睡眠的健康效应及预防和干预T2DM提供理论依据。方法本研究数据来源于美国国家健康和营养调查(National Health and Nutrition Examination Survey, NHANES)(2017-2018年)数据库。ASPS由工作日的就寝时间和晨起时间定义。T2DM根据服用降糖药物来定义。使用基于倾向性评分的逆概率加权法(inverse probability treatment weight, IPTW)和重叠加权法(overlap weights, OW)尽可能地控制潜在混杂变量后，分析ASPS和T2DM的关联性。结果共有6 048名研究对象被纳入研究，其中759人患有T2DM，411人被诊断为ASPS，分别占12.5%、6.8%。T2DM在ASPS组中的患病率比非ASPS组高(20.0% vs. 12.0%, t=21.295, P < 0.001)。IPTW控制潜在混杂后，分析显示ASPS组患T2DM的风险是非ASPS组的1.540倍(OR=1.540, 95% CI: 1.184~2.002, P=0.001)；OW进一步完全均衡协变量后，分析显示二者差异有统计学意义(OR=1.362, 95% CI: 1.078~1.722, P=0.001)。结论 ASPS是T2DM独立风险因素，提示睡眠问题的管理在T2DM的防制中具有重要意义。
- 睡眠时相提前综合征 /
- 2型糖尿病 /
- 逆概率加权 /
- 重叠加权
Abstract: Objective This study aimed to analyze the correlation between advanced sleep phase syndrome (ASPS) and type 2 diabetes mellitus (T2DM), to better understand the health effects of sleep and provide a theoretical basis for prevention and intervention of T2DM. Methods The data for this study were obtained from the National Health and Nutrition Examination Survey (NHANES) (2017-2018) database. ASPS was defined by bedtime and morning rise time on weekdays. T2DM was defined as taking antidiabetic drugs. Associations between ASPS and T2DM were analyzed using the propensity score-based inverse probability treatment weights (IPTW) and overlap weights (OW) to control for potential confounding variables as much as possible. Results A total of 6 048 participants were included in the study, of whom 759 had T2DM and 411 were diagnosed with ASPS, accounting for 12.5% and 6.8%, respectively. The prevalence of T2DM was higher in the ASPS group than in the non-ASPS group (20.0% vs. 12.0%, t=21.295, P < 0.001). After controlling for potential confounding by IPTW, the analysis showed that the ASPS group had a 1.540-fold higher risk of developing T2DM than the non-ASPS group (OR=1.540, 95% CI: 1.184-2.002, P=0.001). After further fully balancing the covariates, OW analysis showed the significance of the two still existed (OR=1.362, 95% CI: 1.078-1.722, P=0.001). Conclusions ASPS is an independent risk factor for T2DM, suggesting that the management of sleep problems is of great significance in the prevention and treatment of T2DM.
- Advanced sleep phase syndrome /
- Type 2 diabetes /
- Inverse probability weighting /
- Overlap weighting

HTML全文

图 1 研究对象筛选流程图

Figure 1. Flow chart of research subjects selection

下载: 全尺寸图片幻灯片

图 2 协变量在处理组和对照组的边际密度及公共支撑

Figure 2. Marginal density and common support of covariates in treatment and control groups

下载: 全尺寸图片幻灯片

图 3 加权前及加权后两组个体各基线协变量均衡性情况

Figure 3. Baseline covariates balance of individuals in the two groups before and after weighting

下载: 全尺寸图片幻灯片

表 1 研究人群基本特征[n(%)]

Table 1. Characteristics of the study population [n(%)]

基本特征	总人数(N=6 048)	ASPS(n=411)	非ASPS(n=5 637)	t/χ²值	P值
睡眠时长(x±s, h)	7.68±1.61	7.71±0.94	7.67±1.65	0.505	0.613
糖尿病史				0.649	0.421
是	1 081(17.9)	80(19.5)	1 001(17.8)
否	4 967(82.1)	331(80.5)	4 636(82.2)
抑郁史				0.353	0.552
是	1 269(21.0)	81(19.7)	1 188(21.1)
否	4 779(79.0)	330(80.3)	4 449(78.9)
年龄(x±s, 岁)	48.14±19.69	52.06±16.34	47.86±19.89	4.185	< 0.001
性别				3.404	0.065
男	2 935(48.5)	218(53.0)	2 717(48.2)
女	3 113(51.5)	193(47.0)	2 920(51.8)
种族				27.972	< 0.001
墨西哥裔美国人	838(13.9)	86(20.9)	752(13.3)
非西班牙裔白人	2 093(34.6)	126(30.7)	1 967(34.9)
非西班牙裔黑人	1 914(31.6)	143(34.8)	1 771(31.4)
其他	1 203(19.9)	56(13.6)	1 147(20.3)
受教育水平				44.634	< 0.001
高中或以下	2 662(44.0)	229(55.7)	2 433(43.2)
大学	2 044(33.8)	142(34.5)	1 902(33.7)
大学毕业及以上	1 342(22.2)	40(9.7)	1 302(23.1)
婚姻状况				2.627	0.269
已婚	2 706(44.7)	195(47.4)	2 511(44.5)
与伴侣同住	742(12.3)	41(10.0)	701(12.4)
单身	2 600(43.0)	175(42.6)	2 425(43.0)
家庭收入与贫困比例				0.775	0.679
< 1.3	2 372(39.2)	166(40.4)	2 206(39.1)
1.3~3.5	2 139(35.4)	148(36.0)	1 991(35.3)
>3.5	1 537(25.4)	97(23.6)	1 440(25.5)
BMI(kg/m²)				20.510	< 0.001
< 18.5	541(8.9)	41(10.0)	500(8.9)
18.5~ < 24.0	1 174(19.4)	55(13.4)	1 119(19.9)
24.0~ < 28.0	1 414(23.4)	78(19.0)	1 336(23.7)
≥28.0	2 919(48.3)	237(57.7)	2 682(47.6)
饮酒状况				2.603	0.107
≤1杯/d	3 894(64.4)	249(60.6)	3 645(64.7)
>1杯/d	2 154(35.6)	162(39.4)	1 992(35.3)
吸烟状况				6.171	0.046
每天	920(15.2)	74(18.0)	846(15.0)
偶尔	2 711(44.8)	195(47.4)	2 516(44.6)
不吸	2 417(40.0)	142(34.5)	2 275(40.4)
高血压史				17.656	< 0.001
是	2 095(34.6)	182(44.3)	1 913(33.9)
否	3 953(65.4)	229(55.7)	3 724(66.1)
健康状况				33.404	< 0.001
非常好	1 985(32.8)	83(20.2)	1 902(33.7)
好或一般	3 904(64.6)	319(77.6)	3 585(63.6)
差	159(2.6)	9(2.2)	150(2.7)
身体活动度				1.178	0.555
剧烈	1 480(24.5)	93(22.6)	1 387(24.6)
适度	1 477(24.4)	98(23.8)	1 379(24.5)
无	3 091(51.1)	220(53.5)	2 871(50.9)
T2DM				21.295	< 0.001
否	5 289(87.5)	329(80.0)	4 960(88.0)
是	759(12.5)	82(20.0)	677(12.0)

下载: 导出CSV

表 2 ASPS与T2DM的相关性分析

Table 2. Analyses of the association between ASPS and T2DM

模型	OR (95% CI)值	χ²值	P值
单因素模型	1.826 (1.415~2.356)	4.630	< 0.001
模型1	1.619 (1.208~2.169)	3.222	0.001
模型2	1.540 (1.184~2.002)	3.221	0.001
模型3	1.362 (1.078~1.722)	2.586	0.001
注：单因素模型：未校正混杂因素；模型1、2、3：校正年龄、性别、种族、教育水平等在内的所有协变量。

下载: 导出CSV

参考文献(20)

[1]	Weng Z, Ou W, Huang J, et al. Circadian misalignment rather than sleep duration is associated with MAFLD: a population-based propensity score-matched study[J]. Nat Sci Sleep, 2021, 13: 103-111. DOI: 10.2147/NSS.S290465.
[2]	韩芳. 昼夜节律性睡眠障碍[J]. 生命科学, 2015, 27(11): 1448-1454. DOI: 10.13376/j.cbls/2015200. Han F. Circadian rhythm sleep disorder[J]. Life Science, 2015, 27(11): 1448-1454. DOI: 10.13376/j.cbls/2015200.
[3]	Sateia MJ. International classification of sleep disorders-third edition: highlights and modifications[J]. Chest, 2014, 146(5): 1387-1394. DOI: 10.1378/chest.14-0970.
[4]	Ando K, Kripke DF, Ancoli-Israel S. Delayed and advanced sleep phase symptoms[J]. Isr J Psychiatry Relat Sci, 2002, 39(1): 11-18. DOI: 10.1017/S1041610203008688.
[5]	Meng Q, Qi X, Fu Y, et al. Flavonoids extracted from mulberry (Morus alba L. ) leaf improve skeletal muscle mitochondrial function by activating AMPK in type 2 diabetes[J]. J Ethnopharmacol, 2020, 248: 112326. DOI: 10.1016/j.jep.2019.112326.
[6]	Wang L, Gao P, Zhang M, et al. Prevalence and ethnic pattern of diabetes and prediabetes in China in 2013[J]. JAMA, 2017, 317(24): 2515-2523. DOI: 10.1001/jama.2017.7596.
[7]	王文华. 睡眠障碍和2型糖尿病的关系: meta分析和孟德尔随机化研究[D]. 郑州: 郑州大学, 2018. Wang WH. The relationship between sleep disturbance and type 2 diabetes: meta-analysis and Mendelian randomization study[D]. Zhengzhou: Zhengzhou University, 2018.
[8]	Rosenbaum PR, Rubin DB. The central role of the propensity score in observational studies for causal effects[J]. Biometrika, 1983, 20(1): 41-55. DOI: 10.1093/biomet/70.1.41.
[9]	Cappuccio FP, D'Elia L, Strazzullo P, et al. Quantity and quality of sleep and incidence of type 2 diabetes: a systematic review and meta-analysis[J]. Diabetes Care, 2010, 33(2): 414-420. DOI: 10.2337/dc09-1124.
[10]	Rod NH, Vahtera J, Westerlund H, et al. Sleep disturbances and cause-specific mortality: results from the GAZEL cohort study[J]. Am J Epidemiol, 2011, 173(3): 300-309. DOI: 10.1093/aje/kwq371.
[11]	Austin PC, Stuart EA. Moving towards best practice when using inverse probability of treatment weighting (IPTW) using the propensity score to estimate causal treatment effects in observational studies[J]. Stat Med, 2015, 34(28): 3661-3679. DOI: 10.1002/sim.6607.
[12]	Johnson CL, Paulose-Ram R, Ogden CL, et al. National health and nutrition examination survey: analytic guidelines, 1999-2010[J]. Vital Health Stat 2, 2013, (161): 1-24.
[13]	DeShields SC, Cunningham TD. Comparison of osteoporosis in US adults with type 1 and type 2 diabetes mellitus[J]. J Endocrinol Invest, 2018, 41(9): 1051-1060. DOI: 10.1007/s40618-018-0828x.
[14]	Lee M, Lee S. Double robustness without weighting[J]. Stat Probab Lett, 2019, 146: 175-180. DOI: 10.1016/j.spl.2018.11.017.
[15]	Li F, Morgan KL, Zaslavsky AM. Balancing covariates via propensity score weighting[J]. J Am Stat Assoc, 2016, 113: 390-400. DOI: 10.1080/01621459.2016.1260466.
[16]	秦宇辰, 郭威, 阮一鸣, 等. 重叠加权法在医学研究混杂因素控制中的应用[J]. 中国卫生统计, 2020, 37(3): 363-366. Qin YC, Guo W, Ruan YM, et al. Application of overlapping weighting method in the control of confounding factors in medical research[J]. Chin J Health Statistics, 2020, 37(3): 363-366.
[17]	Patke A, Young MW, Axelrod S. Molecular mechanisms and physiological importance of circadian rhythms[J]. Nat Rev Mol Cell Biol, 2020, 21(2): 67-84. DOI: 10.1038/s41580-019-0179-2.
[18]	Toh KL, Jones CR, He Y, et al. An hPer2 phosphorylation site mutation in familial advanced sleep phase syndrome[J]. Science, 2001, 291(5506): 1040-1043. DOI: 10.1126/science.1057499.
[19]	Takahashi JS. Transcriptional architecture of the mammalian circadian clock[J]. Nat Rev Genet, 2017, 18(3): 164-179. DOI: 10.1038/nrg.2016.150.
[20]	陈筵明, 龚维坤, 孙英芬, 等. 昼夜节律紊乱与2型糖尿病关系的研究进展[J]. 中国细胞生物学学报, 2020, 42(9): 1684-1693. DOI: 10.11844/cjcb.2020.09.0023. Chen YM, Gong WK, Sun YF, et al. Research progress on the relationship between circadian rhythm disturbance and type 2 diabetes mellitus[J]. Chinese Journal of Cell Biology, 2020, 42(9): 1684-1693. DOI: 10.11844/cjcb.2020.09.0023.