Comment on: Hanssen et al. Associations between the ankle-brachial index and cardiovascular and all-cause mortality are similar in individuals without and with type 2 diabetes: nineteen-year follow-up of a population-based cohort study. Diabetes Care 2012;35:1731-1735.

In their article, Hanssen et al. (1) reported no associations between ankle-brachial index (ABI) and cardiovascular and all-cause mortality in individuals with and without diabetes in the Hoorn cohort. The results are considered important with regards to the long follow-up period in the study as well as some yet unidentified issues on the clinical role of ABI testing. However, I need to make several comments on the presented data.

In the Hoorn study, the ankle pressure was measured only from the posterior tibial artery. Whenever the operators found a normal ABI in the respective ankle, they took a further measurement from the dorsalis pedis artery or peroneal artery. In the last decade, however, the correct measurement and calculation method of the ABI was established in guidelines (2) that require measurement of both the posterior tibial artery and dorsalis pedis artery at the ankle and use of the “higher” one divided by the higher of two brachial pulses in the formula. Therefore, in the Hoorn study some individuals with a low ABI according to tibialis posterior pulse who actually had higher dorsalis pedis pressure than in the posterior tibial artery were probably misclassified to have peripheral arterial disease. Moreover, this might also have left some individuals with a high ABI undetected.

The authors noted that ABI >1.4 did not occur in their cohort. This is interesting for a sample of 624 individuals from the Netherlands because at least three previous surveys from the same country reported contradictory results. In the larger Rotterdam cohort and in the Netherlands Study of Depression and Anxiety, in which again only the posterior tibial artery at the ankles and only the right brachial pulse were recorded, the percentage of individuals with an ABI >1.4 was more than 3% (3,4). Moreover, in another study, which used the currently recommended ABI technique and definitions, the prevalence of a high ABI (>1.4) was 5.7% (5). Because a high ABI is a particular issue in diabetes (6) and is related to increased all-cause mortality (7), association of ABI with mortality in the diabetic individuals can only be investigated by combining the effects of both low (<0.9) and high (1.4) values.

Because the subjects with impaired glucose tolerance (IGT) did not differ significantly from individuals with normal glucose tolerance, the author combined individuals with IGT with individuals with normal glucose metabolism and compared them with individuals with diabetes. Many studies on different populations showed increased all-cause mortality in those with impaired fasting glucose and IGT (8–10). Interestingly, in the Hoorn cohort itself, even the subjects with impaired fasting glucose were previously reported to have increased cardiovascular mortality (11). Therefore, categorizing the subjects with IGT into normal in terms of mortality risk might have led to inaccurate distributions during statistical analysis.

ABI testing is a noninvasive technique to diagnose peripheral arterial disease with very high sensitivity and specificity. However, while arguing its possible roles in improving clinical decision making, correct use of the technique should be broadly applicable.