Department of Surgical Sciences, Uppsala University, Uppsala, Sweden.
Section of Vascular Surgery, Department of Clinical Science and Education, Karolinska Institute at Södersjukhuset, Stockholm, Sweden.
Eur J Vasc Endovasc Surg. 2022 Aug-Sep;64(2-3):217-224. doi: 10.1016/j.ejvs.2022.05.001. Epub 2022 May 7.
The aim of this study was to estimate the prevalence and predictive accuracy for cardiovascular (CV) morbidity by using different ankle brachial index (ABI) calculation methods in the general population.
ABI measurements and questionnaire data were collected from 5 080 randomly selected citizens aged 60 - 90 years. A 10 year follow up with data from Swedish national health registries was carried out. ABI was calculated using as numerator the highest (ABI-HI) or the lowest (ABI-LO) ankle BP obtained in each leg. Subjects were defined as references or having peripheral arterial disease (PAD) based on ABI-LO (Group 1) or ABI-HI (Group 2). Prevalence, mortality, CV events and risk were then analysed for these three groups, and their predictive power by using the area under the curve (AUC).
A total of 4 909 inhabitants were included in the cohort (References: 83.8%, Group 1: 6.7% and Group 2: 9.6%). The prevalence of PAD was 16% using ABI-LO, and 9.6% using ABI-HI. The 10 year all cause mortality for references and Groups 1 and 2 was 27.6%, 48.8%, and 67.2%, respectively. The overall age adjusted hazard ratio (95% confidence interval) for the composite outcome of CV mortality and a non-fatal CV event was 1.25 (1.06 - 1.49) for Group 1 and 2.11 (1.85 - 2.39) for Group 2. The prediction accuracy for ABI < 0.9 in predicting CV event measured with AUC was 0.60 for ABI-HI and 0.62 for ABI-LO.
An ABI < 0.9 should be considered a strong risk marker for future CV morbidity. Applying the traditional ABI calculation method of using the highest measured ankle BP, a group of subjects with high CV risk may be overlooked for intervention, and this why the lowest ankle BP should be the preferred for risk stratification. However, as a single predictive tool an ABI < 0.9 cannot adequately discriminate which individual will have a future CV event regardless of calculation method used.
本研究旨在通过使用不同的踝臂指数（ABI）计算方法来评估普通人群的心血管（CV）发病率的患病率和预测准确性。
从 5080 名随机选择的 60-90 岁公民中收集 ABI 测量值和问卷调查数据。对来自瑞典国家健康登记处的数据进行了为期 10 年的随访。使用每个腿部获得的最高（ABI-HI）或最低（ABI-LO）踝部血压作为分子来计算 ABI。根据 ABI-LO（第 1 组）或 ABI-HI（第 2 组），将受试者定义为参考或患有外周动脉疾病（PAD）。然后，分析了这三组的患病率、死亡率、CV 事件和风险，并使用曲线下面积（AUC）分析了它们的预测能力。
共有 4909 名居民被纳入队列（参考：83.8%，第 1 组：6.7%，第 2 组：9.6%）。使用 ABI-LO 时 PAD 的患病率为 16%，使用 ABI-HI 时为 9.6%。参考组和第 1 组和第 2 组的 10 年全因死亡率分别为 27.6%、48.8%和 67.2%。第 1 组和第 2 组的 CV 死亡率和非致命性 CV 事件的综合结果的总体年龄调整危险比（95%置信区间）分别为 1.25（1.06-1.49）和 2.11（1.85-2.39）。使用 AUC 测量 CV 事件的 ABI<0.9 的预测准确性分别为 ABI-HI 的 0.60 和 ABI-LO 的 0.62。
ABI<0.9 应被视为未来 CV 发病率的强烈风险标志物。应用传统的使用最高测量踝部血压的 ABI 计算方法，一组高 CV 风险的受试者可能会被忽视进行干预，这就是为什么应首选最低的踝部血压进行风险分层。然而，作为单一预测工具，无论使用哪种计算方法，ABI<0.9 都不能充分区分哪些个体将发生未来的 CV 事件。
