Abstract 18361: Lipidomic Profiles are Associated With Incident Coronary Heart Disease in Women
Background: Advances in lipidomics allow profiling of hundreds of lipid compounds, but prior studies have been limited by relatively small case numbers and scant data in women.
Methods: Four hundred confirmed, incident cases of MI and fatal CHD in the Women’s Health Initiative (WHI-OS) were frequency matched to 400 controls by age, race, hysterectomy, and time-period (mean age 62.7 years; no hormone use). Lipidomic analysis of 196 lipid species from 13 different lipid classes including cholesterol esters (CE), diacylglycerols (DAG),and triacylglycerols [TAG) was performed by liquid chromatography mass spectroscopy. Two independent modeling approaches were used: 1) Individual lipids were studied in logistic regression models and a false-discovery rate (FDR) correction was used to adjust for multiple comparisons (FDR corrected p-value of < 0.05); 2) Lipids were simultaneously considered using a least absolute shrinkage and selection operator (LASSO) algorithm. Metabolites selected by both approaches were included in combined multivariable logistic regression models.
Results: In multivariable-adjusted logistic models, 17 lipid metabolites were individually and significantly associated with CHD controlling for multiple comparisons, while in mutually adjusted LASSO regression, 8 lipids were significantly associated with CHD. Five lipids were associated with CHD by both methods (Table). When modeled together, only C18.1 CE and C34.3 DAG were associated with risk in fully adjusted models. Results will be validated in separate replication dataset of women from the WHI (396 cases/ 396 controls; to be presented).
Conclusions: Three long chain highly unsaturated triacylglycerols and one diacylglycerol were associated with decreased risk, while a short chain monounsaturated lipid CE (18.1) was associated with increased risk of CHD in women. Lipidomics may offer insights into biological mechanisms of CHD as well as identify potential markers of risk.
Author Disclosures: K.M. Rexrode: None. R. Balasubramanian: None. N. Paynter: None. S. Gopal: None. F. Guilianini: None. M. Allison: None. C.B. Eaton: None. L.W. Martin: None. B.V. Howard: None. L. Tinker: None. J.E. Manson: None. N. Cook: None. C.A. Albert: None. C. Clish: None.
- © 2015 by American Heart Association, Inc.