ApoA-I not superior to HDL-c to predict coronary heart disease
In the EPIC-Norfolk study and two validation cohorts, 4x4 interquartile comparisons did not show an association of ApoA-I levels with CHD risk within HDL-c quartiles. In contrast, risk was even increased with ApoA-I levels in some quartiles, as were prevalence of CHD risk factors.
Association of High-Density Lipoprotein-Cholesterol Versus Apolipoprotein A-I With Risk of Coronary Heart Disease: The European Prospective Investigation Into Cancer-Norfolk Prospective Population Study, the Atherosclerosis Risk in Communities Study, and the Women’s Health StudyLiterature - Van Capelleveen JC, Bochem AE, Boekholdt M, et al. - JAHA 2017, Epub ahead of print
Background
As mendelian studies cannot show a causal association between HDL-c levels and coronary heart disease (CHD) [1] and pharmacological HDL-c stimulation lacks prove of effect in humans [2-4], the hypothesis had been raised that HDL-c may not have a causal protective effect against atherosclerosis development itself and other parameters that reflect the physical structure and role of HDL may serve as a more relevant predictor of CHD risk.
For example, apolipoprotein A-I (apoA-I), which is the major constituent of HDL particles. However, prospective studies in which the association of apoA-I and HDL-c with CHD risk has been investigated, show conflicting data [5,6]. This may be due to the complex relationship between the two parameters, that is hard to dissect in conventional regression models.
Therefore, a new approach had been used in this study, in which HDL-c and apoA-I levels were categorized into quartiles in a 4x4 fashion. Subsequently, the association of apoA-I levels with CHD risk as well as with CHD risk factors, was assessed within the four HDL-c quartiles, and vice versa. This was done using a subset of the large European Prospective investigation into Cancer (EPIC) Norfolk cohort (United Kingdom, n=17,661) and a validation was executed with a subset of the Atherosclerosis Risk in Communities (ARIC) cohort (n=15,494) and Women’s Health Study (WHS, n=27,552). Risk factors: age, male sex, BMI, HbA1c, non-HDL-c, triglycerides, ApoB, systolic blood pressure (SBP), c-reactive protein (CRP), metabolic syndrome and alcohol intake.
Main results
- 12.6% of the EPIC-Norfolk cohort experienced CHD.
- HR for CHD events highest versus lowest HDL-c quartile was 0.34 (95% CI 0.30-0.39, P trend<0.001). Fully adjusted, this was 0.69 (95% CI 0.59-0.80, P trend <0.001).
- HR for CHD events highest versus lowest apoA-I quartile was 0.55 (95% CI 0.49-0.62, P trend<0.001). Fully adjusted, this was 0.75 (95% CI 0.66-0.86, P trend <0.001).
- In almost half of the patients, HDL-c quartile and apoA-I quartile were discordant.
- Within each apoA-I quartile, higher HDL-c levels were associated with lower CHD risk. In contrast, within each HDL-c quartile, we did not find a consistent association between apoA-I and CHD risk. However, a trend towards increased CHD risk with higher apoA-I levels was observed in the middle HDL-c quartiles.
- HDL-c levels within apo-I quartiles were inversely associated with CHD risk factors (all P<0.001). However, opposite was true for apoA-I levels within HDL-c quartiles: positive association with age, female sex, BMI, HbA1c, non-HDL-c, triglycerides, apoB, SBP and CRP (all P<0.001).
- Validation with the ARIC study (19.3% CHD events) showed comparable results regarding CHD risk. Significant trends were seen for the majority of HDL-c and apoA-I quartiles for risk factors: male percentage, BMI, non-HDL-c, triglyceride, ApoB, SBP and metabolic syndrome.
- Validation with the WHS study were done with tertiles because of the low number of CHD events (3.9%). Also here, similar results were obtained.
Conclusion
In the EPIC-Norfolk cohort, HDL-c and apoA-I were strongly and inversely associated with CHD risk. However, these associations were not interchangeable as apoA-I levels did not show a similar trend within HDL-c quartiles, while vice versa, this was. Therefore, apoA-I levels do not offer predictive information over and above HDL-c. In contrast, an unexpected trend was even observed towards increased risk in some apoA-I quartiles. This was supported by a positive association of apoA-I levels with CHD risk factors within all HDL-c quartiles. All observations were confirmed in two independent cohorts, suggesting biological relevant associations.
References
1. Voight BF, Peloso GM, Orho-Melander M, et al. Plasma HDL cholesterol and risk of myocardial infarction: a Mendelian randomisation study. Lancet. 2012;380:572–580.
2. Boden WE, Probstfield JL, Anderson T, Chaitman BR, Desvignes-Nickens P,
Koprowicz K, McBride R, Teo K, Weintraub W. Niacin in patients with low HDL cholesterol levels receiving intensive statin therapy. N Engl J Med. 2011;365:2255–2267.
3. Schwartz GG, Olsson AG, Abt M, Ballantyne CM, Barter PJ, Brumm J, Chaitman BR, Holme IM, Kallend D, Leiter LA, Leitersdorf E, McMurray JJ, Mundl H, Nicholls SJ, Shah PK, Tardif J-C, Wright RS. Effects of dalcetrapib in patients with a recent acute coronary syndrome. N Engl J Med. 2012;367:2089–2099.
4. Lincoff AM, Nicholls SJ, Riesmeyer JS, et al, Evacetrapib and cardiovascular outcomes in high-risk vascular disease. N Engl J Med. 2017;376:1933–1942.
5. Stampfer MJ, Sacks FM, Salvini S, Willett WC, Hennekens CH. A prospective study of cholesterol, apolipoproteins, and the risk of myocardial infarction. N Engl J Med. 1991;325:373–381.
6. Emerging Risk Factors Collaboration, Di Angelantonio E, Gao P, Pennells L, et al. Lipid-related markers and cardiovascular disease prediction. JAMA. 2012;307:2499–2506.