We were impressed by the results of the Detection of Silent Myocardial Ischemia in Asymptomatic Diabetic Subjects (DIAD) study (1), which found that silent myocardial ischemia (SMI) was present in 22% of a large cohort of asymptomatic patients with type 2 diabetes and that the strongest predictor for abnormal cardiac tests was an abnormal Valsalva test. We have also found that autonomic function predicts SMI but that abnormal sympathetic, rather than parasympathic, tests are predictive.
We selected 12 patients with type 2 diabetes and 12 nondiabetic patients with documented exercise-induced electrocardiogram (ECG) changes (Bruce protocol) and angiographically proven coronary artery disease from the cardiac catheterization database of the Royal Infirmary of Edinburgh, Scotland. The two groups were similar in age (59 ± 2 vs. 62 ± 3 years), number of diseased coronary arteries (2.3 ± 0.2 vs. 2.1 ± 0.2), and exercise characteristics (exercise time 326 ± 24 vs. 308 ± 39 s, maximum ST depression [2.2 ± 0.2 vs. 2.4 ± 0.3 mm]). All subjects underwent five well-validated and standard autonomic function tests (2): R-R interval variation with respiration (parasympathetic), R-R interval variation with the Valsalva maneuver (parasympathetic), supine and erect blood pressure (sympathetic), supine and erect heart rate (parasympathetic), and blood pressure response to sustained handgrip (sympathetic). Each test was scored 0 (normal), 1 (borderline), or 2 (abnormal), and all five tests were then used to compile an autonomic score between 0 and 10.
Although all 24 patients developed ECG evidence of ischemia during treadmill testing, only 50% developed pain (angina pectoralis [AP] group). In the remaining 50%, exertional ischemia was painless (SMI group). Nine of the 12 patients in the SMI group had diabetes, whereas only 3 of the 12 patients in the AP group had diabetes (P = 0.013).
The SMI group had a significantly higher total autonomic score (4.5 ± 0.4 vs. 0.9 ± 0.4), which was largely driven by differences in the two measures of sympathetic autonomic function (supine and erect blood pressure −5.1 ± 3.3 vs. 6.3 vs. ± 4.6 mmHg, P < 0.05, blood pressure response to sustained hand grip 6.2 ± 1.8 vs. 14.9 ± 1.7 mmHg, P < 0.0003). Two nondiabetic subjects with SMI had clearly abnormal autonomic tests.
Although autonomic neuropathy makes SMI more likely in patients with diabetes, it also occurs in those with normal autonomic function (3). In nondiabetic subjects, autonomic neuropathy does not exhibit a clear associate with SMI, despite the findings in our small series (4,5).
We used a five-test battery of autonomic function tests and found that tests of sympathetic autonomic function were impaired in the SMI group. The DIAD study only used heart rate changes during deep breathing, the Valsalva maneuver, and standing, which tests predominantly parasympathetic function. Abnormal parasympathetic function has been reported in another cohort of patients with SMI and diabetes (4). We have no explanation for the difference between our finding and those of others.
In summary, we concur that SMI is common in subjects with diabetes and autonomic neuropathy. Both parasympathetic and sympathetic pathways may be involved.
References
J.D.W. has been on advisory panels for Novo Nordisk, GlaxoSmithKline, and Takeda and has received honoraria from Novo Nordisk, Takeda, Bristol-Myers Squibb, and GlaxoSmithKline.
