糖尿病家族史和睡眠时间对2型糖尿病罹患风险的交互作用

Interaction between family history of diabetes and sleep duration on prevalence of type 2 diabetes：a cohort study

目的：探讨糖尿病家族史和睡眠时间对罹患2型糖尿病（T2DM）风险的交互作用。方法于2013年5月采用多阶段随机抽样，对徐州市18～75岁的常住居民进行基线调查，在此基础上，收集未确诊为T2DM的人群对其开展为期5年的随访，并对随访人群糖尿病家族史和睡眠时间与罹患T2DM风险之交互作用进行评估。采用非条件Logistic回归分析糖尿病家族史和睡眠时间与T2DM之间的关系。结果15939名居民纳入随访研究，5年后共随访到11842名（74.3%）。受试者平均每晚睡眠时间为（7.2±1.1） h，T2DM发病率为3.10%。调整混杂因素后，与睡眠时间6～8 h/晚相比，睡眠时间不足（8 h/晚）未增加罹患T2DM的风险[风险比（RR）=1.31，95%可信区间（CI）：0.96～1.74，P〉0.05]。有糖尿病家族史者罹患T2DM的风险是无家族史者的4.35倍（RR=4.35，95%CI：2.01～7.14，P8 h/晚）合并糖尿病家族史者与罹患T2DM相关联（RR=1.12，95%CI：0.88～1.59，P〉0.01）；调整潜在的混杂因素后，睡眠时间不足（〈6 h/晚）和糖尿病家族史的交互作用的相对超额危险度比（RERI）、归因比（AP）和交互作用指数（S）分别为3.51（1.41～6.77）、0.46（0.26～0.62）和1.98（1.24～3.13），糖尿病家族史和睡眠时间过长交互作用的RERI、AP和S分别为0.07（-0.18～0.31）、0.02（-0.15～0.10）和0.95（-0.04～2.01）。结论糖尿病家族史和睡眠时间不足对增加T2DM罹患风险存在交互作用，糖尿病家族史和睡眠时间过长对T2DM罹患风险不存在交互作用。

Objective To explore the interaction between family history of diabetes and sleep duration on prevalence of type 2 diabetes mellitus （T2DM）. Methods A cohort study was conducted among permanent residents aged from 18 to 75 years in Xuzhou in May 2013. On the basis of this, a followed-up study of five years was conducted among a collection of participants without T2DM, and the interaction between family history of diabetes and sleep duration on prevalence of T2DM in followed-up participants was evaluated. Non-conditional logistic regression analysis was used to analyse the association between family history of diabetes and sleep duration on prevalence of T2DM. Results Total of 15 939 participants were involved inthe followed-up study, and after five years, finally 11 842 participants were still in the cohort. The average sleep duration of total participants was （7.2 ± 1.1） hours per night. The accumulative rate of T2DM among the followed-up participants was 3.10%.After adjusting for the confounding factors, compared with＆nbsp;that participants with sleep duration of 6-8 h/night, individuals with sleep duration8 hours per night （RR=1.31, 95%CI：0.96-1.74, P〉0.05）. Individuals with family history of diabetes were more likely to suffer from T2DM than those without（RR=4.35, 95%CI：2.01-7.14, P8 hours per night didn′t increase significantly（RR=1.12, 95%CI：0.88-1.59, P〉0.05） when compared with participants with sleep duration of 6-8 hours per night without family history of diabetes. After adjusted for potential confounding factors, values of the relative excess risk of interaction （RERI）,the attributable proportion（AP） ,and the synergy index（S） for the additive interaction betweenthe family history of diabetes and sleep duration8 hours per night. Conclusions Additive interactions exist between family history of diabetes and sleep duration8 hours per night at the same time.