Ethnic differences in premature CAD incidence in 14.8 million adults in england, 2004-2024: a retrospective cohort study

Introduction Premature CAD (onset ≤45 years) has disproportionate lifetime impact, with adverse outcomes and decades of treatment burden. South Asian populations are typically aggregated, yet Indian, Pakistani, and Bangladeshi communities experience markedly different CAD burden. We hypothesised aggregation masks clinically important heterogeneity and examined 14.8 million English adults to quantify differences. Methods We conducted a retrospective cohort study using Clinical Practice Research Datalink (CPRD) primary care electronic health records linked to hospital admission and mortality data (2004–2024), including 14,821,240 adults (≥18yrs, ≤45yrs) free of CAD at baseline. Ethnicity was defined using CPRD records, supplemented by hospital records when missing. CAD was defined as first MI or revascularisation (PCI/CABG) at age ≤45 using validated codes. Follow-up ran from registration until incident CAD, death, transfer out of practice, age 45, or study end. Cox models estimated associations by sex and ethnicity (including subtypes), adjusting for age and socioeconomic deprivation. Results Among 14,821,240 adults, 15,989 experienced CAD ≤45 over 80.9 million person-years at risk (median follow-up 4.2 years [IQR 2.0–8.1]). South Asian men had a two-fold higher premature CAD risk than White men (HR [95% CI]: 2.07 [1.96, 2.18]), and Black men a lower risk (0.61 [0.55, 0.68]). Within South Asian subgroups, risks were higher for Bangladeshi (3.13 [2.83, 3.47]) and Pakistani (2.54 [2.35, 2.75]) than Indian men (1.46 [1.35, 1.59]). South Asian women had a 22% higher premature CAD risk compared to White women (1.22 [1.07, 1.38]), and Black women a 28% lower risk (0.72 [0.61, 0.86]). Within South Asian subgroups, risk was much higher for Pakistani (1.81 [1.52, 2.17]) and Bangladeshi women (1.38 [1.02, 1.88]), whilst Indian women had a lower risk than White women (0.81 [0.66, 0.99]). Deprivation partially explained excess risk in Bangladeshi and Pakistani populations (entirely accounted for Bangladeshi women's excess risk), though marked heterogeneity between South Asian sub-groups persisted. Ethnic patterns differed significantly by sex (p<0.001 for interaction). This was marked for Indians, where men had 42% higher risk than White men, while Indian women had 21% lower risk than White women. Conclusion Disaggregating South Asian populations reveals marked heterogeneity in premature CAD risk, with significant sex differences. Bangladeshi and Pakistani men face more than two-fold excess risk, while Pakistani women remain moderately elevated. Indian men show modest elevation, but Indian women have lower risk than White women. Black ethnicity showed 50% lower risk in both sexes. Deprivation partially explains excess risk in Pakistani and Bangladeshi groups. This support targeted prevention incorporating South Asian subgroup, sex, and socioeconomic context rather than broad ethnic groupings.