Yin Xin, Bao Wei, Ley Sylvia H, Yang Jiaxi, Cuffe Sherri Babaeian, Yu Guoqi, Chavarro Jorge E, Liu Peipei, Zhou Juan Helen, Tobias Deirdre K, Hu Frank B, Zhang Cuilin
Global Centre for Asian Women's Health, Yong Loo Lin School of Medicine, National University of Singapore, Singapore.
Department of Obstetrics and Gynaecology, Yong Loo Lin School of Medicine, National University of Singapore, Singapore.
JAMA Netw Open. 2025 Mar 3;8(3):e250142. doi: 10.1001/jamanetworkopen.2025.0142.
Women with a history of gestational diabetes (GD) are at high risk for developing type 2 diabetes (T2D). Sleep is a crucial lifestyle factor associated with cardiometabolic health, yet studies on its role in the progression from GD to T2D are sparse.
To investigate the associations of sleep duration and quality with T2D risk and levels of glucose metabolism biomarkers in women with a history of GD.
DESIGN, SETTING, AND PARTICIPANTS: This cohort study used data from the Nurses' Health Study II, an ongoing longitudinal cohort that began in 1989 and initially included 116 429 female nurses with health status and lifestyle factors updated every 2 to 4 years. A subset of participants with a history of GD was followed up through June 2021. Sleep characteristics were assessed in the 2001 questionnaire (administered from June 2001 to June 2003), which served as the baseline for follow-up. Data were analyzed from November 2023 to August 2024.
Snoring frequency, sleep duration, and daytime sleepiness.
Physician-diagnosed incident T2D was ascertained biennially via questionnaires. Cox proportional hazards regression models were used to estimate the hazard ratios (HRs) and 95% CIs. Biomarkers (glycated hemoglobin [HbA1c], C-peptide, and insulin) were compared across sleep characteristic categories using multivariable-adjusted least-squares means (LSMs) and 95% CIs.
During a total of 42 155 person-years of follow-up among 2891 women with a history of GD (mean [SD] age, 45.3 [4.4] years), 563 women (19.5%) developed T2D. Compared with women who reported rarely snoring, those with occasional or regular snoring had significantly higher T2D risk, with adjusted HRs of 1.54 (95% CI, 1.18-2.02) and 1.61 (95% CI, 1.21-2.13), respectively. Compared with women who slept 7 to 8 hours per day, shorter sleep duration (≤6 hours per day) was significantly associated with a higher risk of T2D (HR, 1.32; 95% CI, 1.06-1.64). Women who slept 6 or fewer hours per day and snored regularly had the highest risk of developing T2D (HR, 2.06; 95% CI, 1.38-3.07). Additionally, more frequent snoring was associated with higher HbA1c (LSM, 5.89; 95% CI, 5.75-6.02), C-peptide (LSM, 4.30; 95% CI, 3.70-4.99), and insulin (LSM, 11.25; 95% CI, 8.75-14.40) levels in the full adjusted models (P = .01 for trend for all).
In this cohort study of women with a history of GD, shorter sleep duration and both occasional and regular snoring were significantly associated with an increased risk of T2D. These findings suggest that improving sleep health may be important to reduce T2D incidence in this high-risk population.
有妊娠期糖尿病(GD)病史的女性患2型糖尿病(T2D)的风险很高。睡眠是与心脏代谢健康相关的关键生活方式因素,但关于其在从GD进展到T2D过程中作用的研究却很少。
探讨睡眠时间和质量与有GD病史女性的T2D风险及葡萄糖代谢生物标志物水平之间的关联。
设计、背景和参与者:这项队列研究使用了护士健康研究II的数据,这是一项正在进行的纵向队列研究,始于1989年,最初纳入了116429名女性护士,她们的健康状况和生活方式因素每2至4年更新一次。对有GD病史的参与者子集随访至2021年6月。睡眠特征在2001年问卷(2001年6月至2003年6月发放)中进行评估,该问卷作为随访的基线。数据于2023年11月至2024年8月进行分析。
打鼾频率、睡眠时间和日间嗜睡情况。
通过问卷每两年确定一次医生诊断的新发T2D。使用Cox比例风险回归模型估计风险比(HR)和95%置信区间(CI)。使用多变量调整后的最小二乘均数(LSM)和95%CI比较不同睡眠特征类别中的生物标志物(糖化血红蛋白[HbA1c]、C肽和胰岛素)。
在2891名有GD病史的女性(平均[标准差]年龄,45.3[4.4]岁)总共42155人年的随访期间,563名女性(19.5%)患了T2D。与报告很少打鼾的女性相比,偶尔或经常打鼾的女性患T2D的风险显著更高,调整后的HR分别为1.54(95%CI,1.18 - 2.02)和1.61(95%CI,1.21 - 2.13)。与每天睡7至8小时的女性相比,较短的睡眠时间(≤6小时/天)与更高的T2D风险显著相关(HR,1.32;95%CI,1.06 - 1.64)。每天睡眠≤6小时且经常打鼾的女性患T2D的风险最高(HR,2.06;95%CI,1.38 - 3.07)。此外,在完全调整模型中,更频繁的打鼾与更高的HbA1c(LSM,5.89;95%CI,5.75 - 6.02)、C肽(LSM,4.30;95%CI,3.70 - 4.99)和胰岛素(LSM,11.25;9%CI,8.75 - 14.40)水平相关(所有趋势P = 0.01)。
在这项针对有GD病史女性的队列研究中,较短的睡眠时间以及偶尔和经常打鼾均与T2D风险增加显著相关。这些发现表明,改善睡眠健康对于降低这一高危人群的T2D发病率可能很重要。