Coronary heart disease (CHD) remains a leading cause of death worldwide. It is a complex disorder stemming from interactions between multiple genetic and environmental risk factors. There is increasing evidence that one of the key factors generating and maintaining inflammation in the arterial intima is the complement system 1. Recently, the amount of complement factor H (CFH) was found to be much greater in the superficial layer of human coronary atherosclerotic lesions than in the deeper layers, indicating a potential role of CFH in atherogenesis 2.

CFH is encoded by a single gene (CFH) on human chromosome 1q32. Polymorphisms in this gene have been confirmed to be strongly associated with age-related macular degeneration (AMD) 34567. It is suggested that the variants in this gene influence the levels of CFH expression. For example, the Y402H (rs1061170) polymorphism, which is located within binding sites for heparin and C-reactive protein (CRP) 8, was found to change the binding properties and have functional implications 9, whilst the I62V (rs800292) polymorphism is located in exon 2, which contains a C3b binding site 6. These alterations may lead to complement related damage to blood vessels 4. CFH variants with more substantial effects may be implicated in the earlier age-of-onset of certain diseases 6. In 2006, Kardys and colleagues investigated the role of the Y402H polymorphism in the CFH gene and reported an increased risk for myocardial infarction in those who were homozygous for the His variant 10. However, three subsequent studies have not been able to replicate this result 111213.

In our study, we investigated the role of CFH gene polymorphisms as risk markers for early-onset CHD in a well-characterised Irish family-based study.

A total of 1494 individuals from 580 families were included (800 discordant sib-pairs and 64 parent-child trios). Table 1 shows the family structures of the participants. Of note, there are more male probands and more female siblings, which reflect the earlier onset of CHD in men compared with women (Table 2). Smoking and diabetes are more common in the probands; however, hypertension and elevated levels of total cholesterol were less common in the probands, probably reflecting pharmacological intervention in the patient group (Table 2). TDT/S-TDT results at the single SNP level are shown in Table 3. No transmission distortion was detected for any of the tested SNPs. Table 4 shows the association analysis of CFH gene haplotypes. No association between any of the haplotypes and early-onset CHD was found.

There is evidence that polymorphisms in the CFH gene may be involved in atherogenesis 19. The CFH gene is a member of the Regulator of Complement Activation (RCA) gene cluster 5. The activation of the complement system is an important link between inflammation and atherogenesis. It has been postulated that variants in the CFH gene may be associated with CHD through modulation of inflammatory pathways, as has been reported in age-related macular degeneration (AMD) patients 4.

The CFH gene has been found to contribute around 50% percent of the risk of AMD 5. AMD and atherosclerotic cardiovascular disease appear to share common pathogenic mechanisms: both are characterized by lipid deposition (drusen and plaques) and thickening of connective tissue (Bruch's membrane and arterial intima) 20. One possible pathogenic mechanism is that the normal role of CFH is to prevent uncontrolled complement activation and inflammation 6. Hence, mutations in CFH may increase inflammation and its pathological consequences. However, there is evidence that the mechanisms of atherogenesis may vary according to the size of arteries, which means that whilst CFH genetic variants may contribute to the atherogenesis in the tiny vessels in the eye, they may not have the same effect in the coronary arteries 21.

A recent study in Netherlands showed that the Y402H polymorphism was associated with increased risk of myocardial infarction. The HH homozygotes had a hazard ratio of 1.77 (95% confidence interval 1.23 to 2.55) 10. They first excluded the non-MI CHD patients (13% of the study population). In their case-control study, using samples from the Physicians' Health Study, Zee et al did not find any association between the CFH Y402H gene polymorphism and incident myocardial infarction, ischaemic stroke, or venous thromboembolism 11. Using samples from the Atherogene and ECTIM studies, Nicaud and colleagues were also unable to find any association between common polymorphisms in the CFH and coronary artery disease 12. Likewise, in their study of 1170 patients with angiographically confirmed coronary artery disease and 560 controls, Goverdhan and co-workers did not detect any association between the CFH Y402H gene variant and presence or severity of CAD 13. Such non-replication of findings is a common finding in the field of the genetics of complex disease 22.

One possible explanation for the lack of association in some of these reported studies is that CFH may be involved more in the formation of unstable plaques rather than the more stable atherosclerotic lesions. A significant characteristic of unstable lesions is the presence of numerous macrophages with matrix metalloproteinases, which can digest the fibrous cap part of the plaque. In the paper by Sivaprasad and colleagues, 3-fold increases of matrix metalloproteinase-9 were found in the AMD patients 23. This function may be mediated in part by C-Reactive Protein, which has been reported to be associated with unstable atherosclerotic lesions 24. Another supporting study showed that the activation of CFH is greater in the superficial rather than the deeper layer of the atherosclerotic plaque 2.

Whilst we have shown differences in the prevalence of conventional risk factors between the probands and siblings, family-based statistical analyses are, unfortunately, not suited to investigating gene-environment interactions since, unlike the subjects in a case-control study, families cannot be neatly divided into those with and those without the environmental risk factor of interest (such as smoking status or diabetes).

Although free from the problems of hidden population stratification, there is a relative lack of power with family based association studies. The number of families recruited was 580; however, the number of informative families ranged from 181 to 323 (Table 3). Of course, we cannot exclude the possibility that other polymorphisms in the CFH gene, which are not in linkage disequilibrium with any of these 6 SNPs, may play a role. However, we found no association between 6 common SNPs or 5 haplotypes of the CFH gene and early-onset CHD in this Irish family-based population.

In this family-based study, we found no evidence of an association between variants of the CFH gene and early-onset coronary heart disease.

The author(s) declare that they have no competing interests.

WM undertook the experimental work and wrote the manuscript; AH supervised the study; CCP performed statistical analysis; CB provided technical support; MSK and PGH collected patient samples and helped to draft the manuscript. FK and PPM provided overall direction to the project and revised the manuscript. All authors have read and approved the final manuscript.