Introduction Social inequality in cardiovascular disease (CVD) is well-established. However, the causal paths driving this association are unclear. To disentangle the effect of education from underlying background factors we investigated the association between education and the risk of CVD using twin data to adjust for familial confounding.

Methods The study was based on data from the Danish Twin Registry linked to official registers in Statistics Denmark, including the National Inpatient Registry and Causes of Death Registry. A total of 12 240 monozygotic (MZ) and 20 822 dizygotic same sexed (DZSS) twins were analysed. Unpaired and intrapair analyses were compared.

Results In the unpaired analyses, an inverse educational gradient in CVD risk was observed, particularly in women. This association was not replicated in the intrapair analyses of female MZ twins, but it persisted among female DZSS twins. For men, the pattern was less clear.

Conclusions The attenuation of association in the intrapair analyses suggest that shared familial factors account for part of the observed association between education and CVD. The fact that it was primarily attenuated in MZ twins may point to genetic factors as an important source of confounding. However, since education was associated with CVD in the intrapair analysis of DZSS twins, there was still some evidence of effect of education. Finally, these Danish data suggest that the social inequality in CVD is larger for women than for men.