Research Support Center, Shizuoka General Hospital, Shizuoka, Japan.
Department of Health Policy and Management, School of Medicine, Keio University, 35 Shinanomachi, Shinjuku-ku, Tokyo, 160-8582, Japan.
BMC Public Health. 2021 Oct 18;21(1):1880. doi: 10.1186/s12889-021-11933-z.
Hemoglobin A1c (HbA1c) levels are routinely measured during health check-ups and are used as an indicator of glycemic control in Japan. However, only a few studies have followed up individuals to assess the risk of diabetes development and worsening based on HbA1c screening results. This study evaluated the relationship between HbA1c screening results and the risk of diabetes development and worsening.
Data were collected from the Shizuoka Kokuho Database, a Japanese administrative claims database of insured individuals aged > 40 years. We included individuals available for follow-up from April 2012 to March 2018 who had not received any diabetes treatment before March 2014. HbA1c screening results were categorized into 4 groups based on the HbA1c levels at the 2012 and 2013 health check-ups: group A, those whose HbA1c levels were < 6.5% in 2012 and 2013; group B, those whose HbA1c levels > 6.5% in 2012 but < 6.5% in 2013; group C, those whose HbA1c levels were > 6.5% in 2012 and 2013; and group D, those whose HbA1c levels were < 6.5% in 2012 and > 6.5% in 2013. Logistic regression models were used to analyze diabetes development and worsening, defined as the initiation of diabetes treatment by March 2018 and the use of injection drugs by participants who initiated diabetes treatment by March 2018.
Overall, 137,852 individuals were analyzed. After adjusting for covariates, compared with group A, group B was more likely to initiate treatment within 4 years (odds ratio: 22.64; 95% confidence interval: 14.66-34.99). In patients who initiated diabetes treatment by March 2018, injection drugs were less likely used by group D than by group A (odds ratio: 0.28; 95% confidence interval: 0.12-0.61).
Our study suggests that although HbA1c levels measured during health check-ups were correlated with the risk of diabetes development and worsening, HbA1c levels in a single year may not necessarily provide sufficient information to consider these future risks.
在日本，血红蛋白 A1c（HbA1c）水平在健康检查期间通常会被测量，并作为血糖控制的指标。然而，只有少数研究对个体进行了随访，以根据 HbA1c 筛查结果评估糖尿病发展和恶化的风险。本研究评估了 HbA1c 筛查结果与糖尿病发展和恶化风险之间的关系。
数据来自 Shizuoka Kokuho 数据库，这是一个日本的医疗保险参保人年龄大于 40 岁的行政索赔数据库。我们纳入了 2012 年 4 月至 2018 年 3 月期间有随访可能的个体，这些个体在 2014 年 3 月之前没有接受过任何糖尿病治疗。根据 2012 年和 2013 年健康检查时的 HbA1c 水平，将 HbA1c 筛查结果分为 4 组：A 组，HbA1c 水平在 2012 年和 2013 年均<6.5%；B 组，HbA1c 水平在 2012 年>6.5%但在 2013 年<6.5%；C 组，HbA1c 水平在 2012 年和 2013 年均>6.5%；D 组，HbA1c 水平在 2012 年<6.5%但在 2013 年>6.5%。使用逻辑回归模型分析糖尿病的发展和恶化，糖尿病的发展和恶化定义为到 2018 年 3 月开始接受糖尿病治疗和到 2018 年 3 月开始接受糖尿病治疗的参与者使用注射药物。
总体上，分析了 137852 名个体。在校正了协变量后，与 A 组相比，B 组在 4 年内更有可能开始治疗（比值比：22.64；95%置信区间：14.66-34.99）。在 2018 年 3 月前开始接受糖尿病治疗的患者中，与 A 组相比，D 组使用注射药物的可能性较低（比值比：0.28；95%置信区间：0.12-0.61）。
我们的研究表明，尽管健康检查期间测量的 HbA1c 水平与糖尿病发展和恶化的风险相关，但单一年度的 HbA1c 水平不一定能提供足够的信息来考虑这些未来的风险。
