School of Medicine, University of Auckland, Room 12.085 A, Level 12, Auckland City Hospital, Auckland, 1142, New Zealand.
J Gastroenterol. 2020 Aug;55(8):775-788. doi: 10.1007/s00535-020-01682-y. Epub 2020 Jun 3.
New-onset diabetes is the most common sequela of acute pancreatitis (AP). Yet, prospective changes in glycaemia over time have never been investigated comprehensively in this study population. The primary aim was to determine the cumulative incidence of new-onset prediabetes and new-onset diabetes after AP over 24 months of follow-up in a prospective cohort study. The secondary aim was to identify trajectories of glycaemia during follow-up and their predictors at the time of hospitalisation.
Patients with a prospective diagnosis of AP and no diabetes based on the American Diabetes Association criteria were followed up every 6 months up to 24 months after hospital discharge. Incidence of new-onset prediabetes/diabetes over each follow-up period was calculated. Group-based trajectory modelling was used to identify common changes in glycaemia. Multinomial regression analyses were conducted to investigate the associations between a wide array of routinely available demographic, anthropometric, laboratory, imaging, and clinical factors and membership in the trajectory groups.
A total of 152 patients without diabetes were followed up. The cumulative incidence of new-onset prediabetes and diabetes was 20% at 6 months after hospitalisation and 43% over 24 months of follow-up (p trend < 0.001). Three discrete trajectories of glycaemia were identified: normal-stable glycaemia (32%), moderate-stable glycaemia (60%), and high-increasing glycaemia (8%). Waist circumference was a significant predictor of moderate-stable glycaemia. None of the studied predictors were significantly associated with high-increasing glycaemia.
This first prospective cohort study of changes in glycaemia (determined at structured time points in unselected AP patients) showed that at least one out of five patients develops new-onset prediabetes or diabetes at 6 months of follow-up and more than four out of ten-in the first 2 years. Changes in glycaemia after AP follow three discrete trajectories. This may inform prevention or early detection of critical changes in blood glucose metabolism following an attack of AP and, hence, reduce the burden of new-onset diabetes after acute pancreatitis.
新发糖尿病是急性胰腺炎(AP)最常见的后遗症。然而,在该研究人群中,从未全面研究过随时间推移血糖的变化。主要目的是在一项前瞻性队列研究中,确定在 AP 后 24 个月的随访中,新发糖尿病前期和新发糖尿病的累积发生率。次要目的是确定随访期间血糖的变化轨迹及其入院时的预测因素。
根据美国糖尿病协会标准,对前瞻性诊断为 AP 且无糖尿病的患者进行随访,在出院后 6 个月至 24 个月内每 6 个月随访一次。计算每个随访期间新发糖尿病前期/糖尿病的发生率。采用基于群组的轨迹建模方法识别血糖变化的常见模式。采用多项回归分析调查广泛的常规可用人口统计学、人体测量学、实验室、影像学和临床因素与轨迹组之间的关系。
共随访了 152 例无糖尿病患者。住院后 6 个月和 24 个月的新发糖尿病前期和糖尿病的累积发生率分别为 20%和 43%(p 趋势 < 0.001)。确定了三种不同的血糖变化轨迹:正常稳定血糖(32%)、中度稳定血糖(60%)和高升高血糖(8%)。腰围是中度稳定血糖的显著预测因子。研究中的预测因素均与高升高血糖无显著相关性。
这是首次前瞻性队列研究血糖变化(在未选择的 AP 患者中在结构化时间点测定),表明至少五分之一的患者在随访 6 个月时发生新发糖尿病前期或糖尿病,在头 2 年内超过十分之四。AP 后血糖变化遵循三个离散轨迹。这可能为 AP 发作后血糖代谢的关键变化提供预防或早期检测,并降低急性胰腺炎后新发糖尿病的负担。