Division of Hospital Medicine, Emory University School of Medicine, Atlanta GA, United States.
Department of Biostatistics, University of Pennsylvania School of Veterinary Medicine, Kennett Square, PA, United States.
Front Endocrinol (Lausanne). 2022 Jun 2;13:871965. doi: 10.3389/fendo.2022.871965. eCollection 2022.
Approximately 50% of obese Black patients with unprovoked diabetic ketoacidosis (DKA) or severe hyperglycemia (SH) at new-onset diabetes achieve near-normoglycemia remission with intensive insulin treatment. Despite the initial near-normoglycemia remission, most DKA/SH individuals develop hyperglycemia relapse after insulin discontinuation. Traditional biomarkers such as normal glucose tolerance at the time of remission were not predictive of hyperglycemia relapse. We tested whether 1-h plasma glucose (1-h PG) at remission predicts hyperglycemia relapse in Black patients with DKA/SH.
Secondary analysis was performed of two prospective randomized controlled trials in 73 patients with DKA/SH at the safety net hospital with a median follow-up of 408 days. Patients with DKA/SH underwent a 5-point, 2-h 75-g oral glucose tolerance test after hyperglycemia remission. Hyperglycemia relapse is defined by fasting blood glucose (FBG) > 130 mg/dl, random blood glucose (BG) >180 mg/dl, or HbA1c > 7%.
During the median 408 (interquartile range: 110-602) days of follow-up, hyperglycemia relapse occurred in 28 (38.4%) participants. One-hour PG value ≥199 mg/dl discriminates hyperglycemia relapse (sensitivity: 64%; specificity: 71%). Elevated levels of 1-h PG (≥199 mg/dl) were independently associated with hyperglycemia relapse (adjusted hazard ratio: 2.40 [95% CI: 1.04, 5.56]). In a multivariable model with FBG, adding 1-h PG level enhanced the prediction of hyperglycemia relapse, with significant improvements in C-index (Δ: +0.05; p = 0.04), net reclassification improvement (NRI: 48.7%; p = 0.04), and integrated discrimination improvement (IDI: 7.8%; p = 0.02) as compared with the addition of 2-h PG (NRI: 20.2%; p = 0.42; IDI: 1.32%; p = 0.41) or HbA1c (NRI: 35.2%; p = 0.143; IDI: 5.8%; p = 0.04).
One-hour PG at the time of remission is a better predictor of hyperglycemia relapse than traditional glycemic markers among obese Black patients presenting with DKA/SH. Testing 1-h PG at insulin discontinuation identifies individuals at high risk of developing hyperglycemia relapse.
大约 50%的新发糖尿病伴无诱因糖尿病酮症酸中毒（DKA）或严重高血糖（SH）的肥胖黑人患者，通过强化胰岛素治疗可实现接近正常血糖缓解。尽管最初的接近正常血糖缓解，但大多数 DKA/SH 个体在胰岛素停用后会发生高血糖复发。传统的生物标志物，如缓解时的正常葡萄糖耐量，不能预测高血糖复发。我们检测了 DKA/SH 黑人患者缓解时 1 小时血浆葡萄糖（1-h PG）是否可预测高血糖复发。
对安全网医院的两项前瞻性随机对照试验中的 73 例 DKA/SH 患者进行二次分析，中位随访时间为 408 天。DKA/SH 患者在高血糖缓解后进行 5 点、2 小时 75g 口服葡萄糖耐量试验。高血糖复发定义为空腹血糖（FBG）＞130mg/dl、随机血糖（BG）＞180mg/dl 或 HbA1c＞7%。
在中位 408（四分位距：110-602）天的随访期间，28 名（38.4%）参与者发生高血糖复发。1 小时 PG 值≥199mg/dl 可区分高血糖复发（灵敏度：64%；特异性：71%）。升高的 1 小时 PG（≥199mg/dl）与高血糖复发独立相关（调整后的危险比：2.40[95%CI：1.04，5.56]）。在一个包含 FBG 的多变量模型中，添加 1 小时 PG 水平可提高高血糖复发的预测能力，C 指数显著提高（Δ：+0.05；p=0.04），净重新分类改善（NRI：48.7%；p=0.04）和综合鉴别改善（IDI：7.8%；p=0.02），优于添加 2 小时 PG（NRI：20.2%；p=0.42；IDI：1.32%；p=0.41）或 HbA1c（NRI：35.2%；p=0.143；IDI：5.8%；p=0.04）。
与传统的血糖标志物相比，肥胖黑人患者 DKA/SH 就诊时的 1 小时 PG 是高血糖复发的更好预测指标。在胰岛素停用后检测 1 小时 PG 可识别出发生高血糖复发风险较高的个体。
