Prevalence and associated factors of resting electrocardiogram abnormalities among systemic lupus erythematosus patients without cardiovascular disease

It is well established that SLE patients have increased mortality secondary to cardiovascular events even in the absence of traditional risk factors for CVD [15]. In this study of SLE patients, the prevalence of the ECG-4, which have been shown by Chou et al. to be predictive of CVD [14], is 21.4% and the prevalence of ECG-5 is 24.2%. The prevalence of ECG-CVD among SLE patients is greater than the general population (prevalence 3.6–17%) as shown in Table 4. Chou et al. showed that ECG-4 predicts subsequent CVD events [16] and that pathological Q-wave was also a predictor of future CVD [14]. In this study, ECG-CVD (ECG-4 and ECG-5) was more common in patients aged 50.14?±?14.1 years compared to the group without ECG-CVD (age 44.8?±?12.9 years); however, this is still younger than the general population.

In our analysis, we found an association between ECG-CVD (ECG-4 and ECG-5) and lupus disease activity (AMS 2 years prior to ECG) and damage (SDI). Some studies have found an association of azathioprine with atherosclerosis [17, 18]. Although there was a trend in our study for an association with immunosuppressive drugs (azathioprine, mycophenolate mofetil, cyclophosphamide, cyclosporine, and methotrexate), this was not statistically significant. Cumulative glucocorticoid therapy has been postulated as a potential risk factor for atherosclerosis in patients with SLE [2, 19, 20], but not all studies have confirmed this association [21]. In this study, cumulative glucocorticoid therapy and antimalarial drugs were significantly associated with ECG-4 and ECG-5 in the univariate analysis, but did not remain in the multivariate analysis. Statin therapy ever for hyperlipidemia was protective against ECG-CVD in the multivariate analysis.

While some previous studies showed an association between antiphospholipid antibodies and clinical CVD and/or events [20, 22], other studies did not [17, 21, 23]. In the present study, antiphospholipid antibodies were not associated with ECG-4 or ECG-5. More recently, Andrade et al. showed that patients with antiphospholipid syndrome do not develop premature atherosclerosis [24].

Smoking, hypertension, and DM, all recognized as traditional risk factors for CVD in SLE patients [2, 23], were not significantly associated with ECG-CVD this study. Several factors could explain the lack of association with traditional risk factors for CVD in our study. First, smoking ever was reported in 29% of each of the non-CVD and ECG-4 and -5 groups, and other studies have demonstrated the importance of studying the dose-effect (pack-years) of smoking [25]. Second, in CAD, the ECG is the reflection of damage to the myocardium that usually occurs at the end of the process of the development of CAD and its clinical implications. Third, with respect to hypertension, not only is LVH a reflection of long-term inadequate control of blood pressure, but ECG has a poor sensitivity for LVH (slightly lower in females than in males). Fourth, the number of patients with DM was small (only four patients had DM in ECG-4 and -5).

In the current study, the most common element of ECG-4 was ST-segment abnormalities and/or T-wave abnormalities (70%), followed by LVH in 32%, LAD or LAFB in 20%, and LBBB or RBBB in 11%. In ECG-5, Q-wave was found in 18% of the patients. The pooled adjusted HR (for CVD) of ST-segment abnormalities in resting ECG in the normal healthy population from several studies was found to be 1.9 [26–30]. De Bacquer et al. showed that ST?depression is predictive of CVD [31]. In a recent systematic review, both LVH LAD on resting ECG were associated with a similar increased risk for subsequent CVD events, with a pooled adjusted HR of 1.6 (95% CI, 1.4 to 1.8) [8, 16, 29–32]. LBBB was found in 3% of ECG-CVD, and it has been previously reported that, in the presence of LBBB on resting ECG, there is an increased incidence of ischemic heart disease events and/or death due to cardiovascular disease compared to control subjects [8, 33]. Interestingly, pathological Q-waves were noted in 11% of the patients with ECG-5 without a previous history of ischemic heart disease.

A limitation of this study is the lack of a control group to better evaluate the prevalence of CVD-ECG-4 abnormalities in patients without SLE. In order to evaluate the prevalence of CVD-ECG-4 abnormalities, a baseline ECG should have been done and this is also a limitation of this study. In addition, the low level of correlation among variables in the multivariate analyses is a limitation of this study; although we avoided using variables with moderate to high correlation in the multivariate analyses, some variables were inter-related to each other at a low level, e.g., older age and hyperlipidemia (correlation coefficient?=?0.30), and age and dsDNA antibodies (correlation coefficient?=?–0.22).