Novo Nordisk , Princeton, NJ , USA.
Curr Med Res Opin. 2013 Dec;29(12):1609-15. doi: 10.1185/03007995.2013.830599. Epub 2013 Sep 23.
This retrospective study investigated the association between hypoglycemic events (HEs) and depression events (DEs) in patients with diabetes mellitus (type 1 and type 2).
Analyzed data were from health care claims for individuals with employer-sponsored primary or Medicare supplemental insurance from the Thomson Reuters Market Scan database during the years 2008 and 2009. A baseline period (January 2008 to December 2008) was used to identify eligible patients and collect baseline clinical and demographic characteristics. Eligible patients were aged ≥18 years with diabetes (ICD-9-CM codes: 250.00, 250.01, 250.02, 250.03) who had not experienced any HEs or DEs and were not on antidepressant therapy during the baseline period. We studied the relationships between the DEs and HEs before and after adjusting for the covariates.
Of the 923,024 patients meeting the inclusion criteria, 22,735 (2.46%) patients had HEs (ICD-9-CM coded: 251.0, 251.1, 251.2, 250.8) and 6164 (0.67%) patients had DEs (ICD-9-CM: 311) during the evaluation period. Patients reporting HEs had 78% higher odds of experiencing depression than patients without HEs before adjusting for the covariates. Similarly, after adjusting for the covariates, data indicated that patients with HEs had higher odds of experiencing depression (OR = 1.726; 95% CI = 1.52-1.96). Similar analyses in different age categories showed that the OR monotonically increases with age regardless of whether the other covariates are included in the model.
ICD-9-CM-coded HEs were independently associated with an increased risk of DEs in patients with diabetes, and this incidence increased with the patients' age.
A key limitation to this study is that only those HEs that resulted in health care provider contact and subsequent claims coding indicative of hypoglycemia were included. It is likely that many cases of mild hypoglycemia, particularly those not severe enough to warrant medical attention, were not captured in this study.
本回顾性研究调查了糖尿病患者（1 型和 2 型）低血糖事件（HEs）与抑郁事件（DEs）之间的关联。
分析数据来自于 Thomson Reuters Market Scan 数据库中，2008 年和 2009 年期间，有雇主赞助的主要医疗保险或补充医疗保险的个人的医疗保健索赔。使用基线期（2008 年 1 月至 2008 年 12 月）来确定合格患者，并收集基线临床和人口统计学特征。合格患者的年龄≥18 岁，患有糖尿病（ICD-9-CM 代码：250.00、250.01、250.02、250.03），在基线期内没有经历过任何 HEs 或 DEs，也没有接受抗抑郁治疗。我们研究了调整协变量前后 DEs 和 HEs 之间的关系。
在符合纳入标准的 923024 名患者中，22735 名（2.46%）患者发生 HEs（ICD-9-CM 编码：251.0、251.1、251.2、250.8），6164 名（0.67%）患者发生 DEs（ICD-9-CM：311）在评估期间。与未发生 HEs 的患者相比，报告发生 HEs 的患者发生抑郁的可能性高 78%，在调整协变量之前。同样，在调整协变量后，数据表明，与没有 HEs 的患者相比，患有 HEs 的患者发生抑郁的可能性更高（OR=1.726；95%CI=1.52-1.96）。在不同年龄组的类似分析表明，OR 随着年龄的增长而单调增加，无论模型中是否包含其他协变量。
ICD-9-CM 编码的 HEs 与糖尿病患者发生 DEs 的风险增加独立相关，且这种发生率随着患者年龄的增长而增加。
本研究的一个主要局限性是，仅包括导致医疗保健提供者接触并随后进行指示低血糖的索赔编码的那些 HEs。在这项研究中，很可能许多轻度低血糖症，特别是那些严重程度不足以需要医疗关注的病例，都没有被捕获。
