Introduction. Electrocardiography (ECG) is a widely used non-invasive diagnostic method for assessment of patients with cardiovascular diseases. Numerous different electrocardiographic criteria exist for detection of left ventricular hypertrophy (LVH). LVH is an important risk factor in patients with coronary artery disease (CAD) or diabetes mellitus and its presence is associated with increased cardiovascular morbidity and mortality. The aim of this study was to evaluate the prevalence of the most frequently used electrocardiographic left ventricular hypertrophy (ECG-LVH) criteria among patients with CAD and diabetes.

Methods. A cross-sectional, multicenter study was conducted in outpatient clinics across Poland. Family physicians performed physical examinations and collected relevant information about: onset of CAD and diabetes, presence and onset of hypertension, dyslipidemia, heart failure, diabetic complications, history of acute coronary syndrome and pharmacotherapy. In order to detect LVH, we used seven ECG criteria: 1) the Sokolow-Lyon voltage, 2) the Gubner voltage, 3) the criterion of the R wave amplitude on the leads V5–V6 and 4) aVL, 5) the gender specific Cornell voltage and 6) product, and 7) the Romhilt-Estes point score. Centralized manual assessment of the obtained ECG tracings were performed.

Results. We enrolled 1001 patients (48.5% women, 51.5% men, mean age 65 ± 11 years) into the study. At least one ECG-LVH criterion was met in 20.0% (n = 200) of the study participants. The ECG-LVH diagnosis was the most common when using the Romhilt-Estes point score (n = 138; 13.8%). The corresponding prevalence rates for the Cornell voltage, the Cornell product, the R wave amplitude on the lead aVL, the Sokolow-Lyon voltage, the Gubner voltage and the R wave amplitude on the leads V5-V6 criteria were 5.5% (n = 55), 5.2% (n = 52), 3.2% (n = 32), 2.2% (n = 22), 1.9% (n = 19) and 1.3% (n = 13) respectively. Subsequently, the prevalence of the three most frequently used in clinical practice electrocardiographic criteria for LVH (the Sokolow-Lyon voltage, the Cornell voltage and the Romhilt-Estes point score) was analyzed. At least one of them was fulfilled in 185 ECGs. All three criteria at the same time were met only in 5 ECGs (2.7% of 185). Two and only one out of three criteria were fulfilled in 20 (10.8%) and 160 (86.5%) ECGs respectively.

Conclusions. The co-occurrence of all assessed ECG-LVH criteria, including the three most frequently applied in clinical practice, is very low in diabetic CAD patients. The Romhilt-Estes point score identifies the highest number of ECG-LVH cases in this setting. However, it seems reasonable to use routinely several ECG criteria for detection of LVH. Further studies are needed to compare diagnostic values of ECG-LVH criteria with imaging methods and to assess prognostic values of various ECG-LVH criteria.