The Characterisation and Risk Factors of Ischaemic Heart Disease in Patients With Coeliac Disease

L. Emilsson; R. Carlsson; M. Holmqvist; S. James; J. F. Ludvigsson


Aliment Pharmacol Ther. 2013;37(9):905-914. 

In This Article


The most important finding of this study is a more favourable IHD risk-factor profile in patients with CD, with the exception of high levels of biochemical markers in both CD patients and reference individuals. Hence, traditional IHD risk factors are unlikely to explain our previous observations of an increased risk of IHD in CD.[4]

Comparison With the Literature

Some studies have found an increased risk of IHD in CD[2–4] and others have not.[5, 6, 18] Most evidence suggests that CD patients have a cholesterol profile that, from a IHD risk perspective, is more beneficial[19] and that the introduction of a gluten-free diet does not alter cholesterol levels.[20] Worth mentioning is also that CD patients received statins to the same extent as reference individuals, despite having lower cholesterol levels. Moreover, BMI seems to be lower in CD patients,[7, 21, 22] with most studies indicating that smoking is less common in CD.[8, 19, 21, 23] In that way the findings of our study are consistent with previous data. Furthermore, West et al. reported less HT in individuals with undiagnosed CD.[6] Concerning the use of antihypertensive drugs, the evidence is less clear.[3,22] In the current study CD patients seem to have similar levels of antihypertensive treatment (apart from Ca-blockers) but fewer CD patients had a record of HT. However, all these studies, except for Lewis et al.,[20] examined cardiovascular risk factors in the general CD population and did not in detail examine characteristics of CD patients with IHD. Concerning the evidence of heart failure and better left ventricular ejection fraction seen in this study, this confirms similar findings including a different risk-factor profile and better ejection fraction seen in patients with other inflammatory or autoimmune conditions such as rheumatoid arthritis.[24]

Potential Mechanisms

Systemic inflammation may be one explanation for the increased risk of IHD in CD.[4] CRP in our study was 12% higher in CD patients with myocardial infarction than in reference individuals with myocardial infarction, but the difference was not statistically significant. The lack of substantial difference in CRP levels either implies that systemic inflammation is not important for IHD risk in CD or (more likely) that CRP is not an optimal marker of systemic inflammation in CD. Studies have shown elevated levels of various interleukins and tumour necrosis factor[25] in active CD, but elevated CRP has not been associated with CD in the few published studies on this topic.[26,27]Another autoimmune disease, systemic lupus erythematosus (SLE), has similarly been linked to an increased risk of atherosclerosis potentially due to chronic inflammation in SLE as these patients lack traditional IHD risk factors.[28] Assessments of carotid plaques have shown that atherosclerosis is both more common and positively associated with disease duration in both SLE and rheumatoid arthritis.[29] As of today we are not aware of any study examining disease severity in CD and cardiovascular pathology but one case report has described a patient with refractory CD whose cerebellum autopsy showed small vessel angiopathy, potentially due to circulating neoplastic clones of activated T-cells. In this patient the lymphocytes surrounding and infiltrating the vessel walls had the same phenotype as the intraepithelial lymphocytes.[30] Given that CD is an autoimmune disease (like SLE and RA) chronic inflammation may explain the adverse cardiovascular outcome seen in Swedish patients with CD. Some 55.4% of patients with both CD and MI had persistent villous atrophy implicating that chronic inflammation may be involved in the pathogenesis since this percentage is considerably higher than the overall percentage (43%) of coeliac patients with persistent VA at their second biopsy.[31]

Another potential explanation for the excess risk of IHD in CD is ascertainment bias (patients with CD may, to a larger extent than reference individuals, be investigated for any symptom, including cardiovascular symptoms). However, such systematic bias is unlikely for two reasons. First, CD patients with myocardial infarction had even higher rates of elevated biochemical markers of infarction, which argues that their myocardial infarctions were not diagnosed differently to myocardial infarctions in the average population. Second, in a previous study we found that patients with CD were at a 22% increased risk of death from IHD (95% CI = 1.06–1.40),[4] a finding that cannot be explained by ascertainment bias. Although the proportion of coeliac patients with any diabetes was similar to that in reference individuals, patients with CD had more type 1 diabetes suggesting a longer duration of diabetes since type 1 often has its onset in childhood. Also hyperhomocysteinaemia found to be associated to CD through vitamin deficiencies is a risk factor for cardiovascular disease[32] and might therefore contribute to the increased risk of MI in patients with CD.

Myocardial infarction with angiographically normal coronary arteries has lately attracted more attention.[33] Several mechanisms for the development of myocardial infarction in the absence of stenosis have been suggested, including inflammation.[33,34] Myocardial infarction with angiographically normal coronary arteries and stress-induced cardiomyopathy also seem to be more common than previously recognised. In fact, a recent study found that this subtype of myocardial infarction constitutes 7% of all myocardial infarction.[35]Myocardial infarction without stenosis is more prevalent in women[34,35] and may have other risk factors than stenotic coronary artery disease. Our study found that CD patients have elevated biochemical markers when diagnosed with myocardial infarction but more often a beneficial classical risk-factor profile and better results on coronary angiography. We therefore speculate that CD may be a risk factor specifically for myocardial infarction with less extensive coronary disease.

Strengths and Limitations

The major strength of this nationwide study is the large number of patients and its use of detailed clinical data through SWEDEHEART. All diagnoses and angiographies performed were standard procedures and therefore not subject to any particular study or inclusion bias. We required two independent records of IHD for study inclusion (national registers and the SWEDEHEART record) and individuals with CD were identified through small-intestinal biopsy yielding a high specificity.

A limitation is our lack of individual-based data on dietary adherence of CD patients. Earlier validation has shown that 17% of CD patients have signs of low dietary adherence according to patient charts.[11] Although our study was nationwide it was limited to Swedish patients with CD. We cannot rule out that CD patients in countries where no link between CD and cardiovascular disease has been shown have a different cardiac risk profile even in the presence of IHD.