Abstract 16482: Proteomic Measures Associated With Leisure-Time Physical Activity and Heart Failure Risk in Late-Life - The Atherosclerosis Risk in Communities Study

Introduction: Leisure-time physical activity (LTPA) is associated with lower heart failure (HF) risk. However, the underlying molecular mechanisms by which LTPA impacts HF risk is unclear.

Methods: We evaluated participants at Atherosclerosis Risk in Communities study visits 3 (1993-95) and 5 (2011-13) with plasma proteomics data (Somascan v4) and without HF or stroke. LTPA was assessed by the Baecke questionnaire as metabolic equivalent (MET)-minutes/week. We performed cross-sectional associations of 4,955 proteins with LTPA at visit 3 and then at visit 5 LTPA using multivariable quantile regression models at Bonferroni-corrected significance (p<1e-5, p<0.0004, respectively). We related these proteins to incident HF after visit 3 and then visit 5 using multivariable Cox regression models at Bonferroni-corrected significance (p<0.00086, p<0.003, respectively).

Results: Thirty-nine proteins were associated with LTPA at visit 3 (n=10,402; age 60±6 yrs, 54% women; mean 688.6 MET-minutes/week, 773 incident HF events) and visit 5 (n=4,135; 75±5 yrs, 58% women; mean 834.8 MET-minutes/week, 362 events), of which 5 proteins were also associated with incident HF at both visits. Higher levels of THBS2, GABARAP, and BAGE2 were associated with lower LTPA and higher HF risk while higher levels of ART3 associated with higher LTPA and lower risk of incident HF (Figure). PTK7 demonstrated discordant associations with LTPA and HF risk. Levels of THBS2, GABARAP, BAGE2, and ART3 accounted for a significant proportion of the LTPA-HF association (p<0.05).

Conclusion: Large-scale proteomics can identify novel biomarkers potentially linking LTPA to HF risk. Further study is needed of THBS2, GABARAP, BAGE2, and ART3, which may link the association of LTPA on HF.