Chen Hung-Yi, Yang Fu-Yu, Jong Gwo-Ping, Liou Yi-Sheng
Institute of Pharmacy, China Medical University, Taichung, Taiwan.
Department of Pharmacy, China Medical University Beigang Hospital, Beigang, Taiwan.
Eur J Clin Invest. 2017 May;47(5):388-393. doi: 10.1111/eci.12754. Epub 2017 Apr 19.
Antihyperglycemic drugs have been linked to new-onset atrial fibrillation (NAF). However, the effect of the different classes of antihyperglycemic drugs on the development of NAF in elderly patients has not been well studied. In this study, we investigated the association between different classes of antihyperglycemic drugs and NAF in elderly patients.
This was a nested case-control study performed using the database of National Health Insurance programme in Taiwan. Each participant aged 65 years and older who were NAF from 2005 to 2012 were assigned to the NAF group, whereas case was sex-, age-, diabetes duration-, index date-matched, and Charlson Comorbidity Index score-matched randomly selected participant without NAF were assigned to the non-NAF group. Multivariable logistic regression model was used for the estimation of odds ratios (ORs) and 95% confidence intervals (CIs) of NAF associated with use of different classes of antihyperglycemic agents. Nonusers served as the reference group.
We identified 1958 cases and 7832 controls. The risk of NAF after adjusting for sex, age, comorbidities and concurrent medication was higher among the users of insulin than among the nonusers (OR, 1·58; 95% CI, 1·37-1·82). Patients who took dipeptidyl peptidase 4 inhibitors were at lower risk of developing NAF than the nonusers (OR, 0·65; 95% CI, 0·45-0·93).
In this population, use of dipeptidyl peptidase 4 inhibitor was associated with a low risk of NAF. Insulin use was associated with a significant increase in the risk of NAF during the long-term follow-up.
降糖药物与新发房颤（NAF）有关。然而，不同类别的降糖药物对老年患者NAF发生的影响尚未得到充分研究。在本研究中，我们调查了不同类别降糖药物与老年患者NAF之间的关联。
这是一项利用台湾国民健康保险计划数据库进行的巢式病例对照研究。将2005年至2012年期间发生NAF的65岁及以上参与者纳入NAF组，而对照则是通过性别、年龄、糖尿病病程、索引日期匹配，以及Charlson合并症指数评分匹配后随机选择的无NAF参与者，纳入非NAF组。采用多变量逻辑回归模型估计与使用不同类别降糖药物相关的NAF的比值比（OR）和95%置信区间（CI）。未使用者作为参照组。
我们确定了1958例病例和7832例对照。在调整性别、年龄、合并症和同时使用的药物后，胰岛素使用者发生NAF的风险高于未使用者（OR，1.58；95%CI，1.37 - 1.82）。服用二肽基肽酶4抑制剂的患者发生NAF的风险低于未使用者（OR，0.65；95%CI，0.45 - 0.93）。
在该人群中，使用二肽基肽酶4抑制剂与NAF低风险相关。在长期随访中，使用胰岛素与NAF风险显著增加相关。
