Date of Award

Spring 2020

Document Type

Open Access Dissertation


Exercise Science

First Advisor

Mark A. Sarzynski


Purpose: To compare baseline CVD risk factor profiles among groups of individuals with discordant levels of high- and low-density lipoprotein cholesterol (HDL-C and LDL-C respectively) compared to lipoprotein particle concentration (HDL-P and LDL-P respectively) and examine the associations between lipoprotein discordance and incident type 2 diabetes mellitus (T2D) and metabolic syndrome (MetS). Furthermore, to examine the association between lipoprotein discordance and exercise induced changes in CVD risk factors.

Methods: Standard lipid panels as well as lipoprotein subclass profiles via nuclear magnetic resonance spectroscopy were measured among three previously completed studies (CARDIA cohort, HERITAGE Family Study, and HART-D study) spanning over 4,000 male and female participants. 1) Aerobic training only (AT), 2) Resistance training only (RT), 3) Combined AT/RT. Four exclusive HDL and LDL discordance groups were created based on baseline median lipoprotein cholesterol/particle concentrations as follows: 1) low/low (< median for both cholesterol and particle), 2) low/high (< median cholesterol, ≥ median particle), 3) high/low, and 4) high/high. Cross-sectional associations between baseline discordance group and CVD risk factors were assessed via multivariable linear regression. Continuous discordance levels were created by subtracting individual particle percentile ranking from cholesterol percentile ranking (e.g. HDL-C – HDL-P). The association between continuous lipoprotein discordance and cross-sectional CVD risk factors was assessed correlational analysis, while the association of incident T2D or MetS with lipoprotein discordance was assessed via multivariable proportional hazards regression. The relationship between lipoprotein discordance and exercise induced changes in CVD risk factors was assessed via multivariable linear regression.

Results: Cross sectional analyses revealed that for HDL discordance, more favorable risk profiles (e.g. lower BMI, larger mean HDL-P and LDL-P size) were associated with higher HDL-C, while in LDL discordance groups more favorable risk factors were associated with lower LDL-P. In women only, both HDL and LDL discordance were associated with T2D and MetS risk with higher risk being associated with lower HDL-C and higher LDL-P. Associations of lipoprotein discordance with exercise induced changes in CVD risk factors were primarily limited to lipoprotein specific variables (e.g. HDL discordance significantly associated with changes in mean HDL-P size).

Conclusions: Lipoprotein discordance is associated with cross-sectional differences in CVD risk factors with more favorable risk factor profiles being found in those with high HDL-C and low LDL-P regardless of discordance status. Lipoprotein concordant/discordant status is also a significant risk factor for both T2D and MetS in women only, although racial disparities may be present. Furthermore, lipoprotein discordance status is primarily associated with exercise training induced changes in lipoprotein subclass measures. Thus, while lipoprotein discordance is minimally predictive of exercise training responses outside of lipoprotein subclass measures, our findings support the regular measurement of lipoprotein particle concentrations alongside of cholesterol measures in chronic disease management and prevention.