Anthocyanin Intake and Physical Activity: Associations with the Lipid Profile of a US Working Population

Abstract

While growing evidence exists on the independent associations between anthocyanins and physical activity on cardiovascular disease (CVD) risk determinants, the possible interaction between these exposures has not yet been studied. We aimed to study the potential synergism between anthocyanin intake and physical activity on lipid profile measures. This cross-sectional study was conducted among 249 US career firefighters participating in the Feeding America's Bravest trial. Anthocyanin intake was calculated using a validated food frequency questionnaire (FFQ) and physical activity level by a validated questionnaire. Multivariable linear regression models determined the extent to which anthocyanin intake and physical activity predicted lipid parameters. Generalized linear models were used for joint effect and interaction analyses on the multiplicative and additive scales. Both anthocyanins and physical activity were independently inversely associated with total cholesterol:high density lipoprotein (HDL) cholesterol. Only physical activity was inversely associated with triglycerides, low density lipoprotein (LDL) cholesterol:HDL, and triglycerides (TG):HDL. Although the combined exposure of low anthocyanin intake and low physical activity was associated with lower (RR = 2.83; 95% CI: 1.42 to 5.67) HDL cholesterol <40 mg/dL, neither multiplicative (p = 0.72) nor additive interactions were detected (relative excess risk due to interaction (RERI): 0.02; 95% CI: -1.63 to 1.66; p = 0.98). Our findings provide insight on the potential synergism between anthocyanin intake and physical activity on the lipid profile.

Keywords: anthocyanins; cardiovascular disease; lipid profile; physical activity; working population.

Figure 1

Independent associations (β, 95% CI) of anthocyanin intake (SD) and physical activity level with lipid concentrations (mg/dL) and lipid ratios. * Physical activity was assessed using a scale of 0–7 representing levels of physical activity ranging from none to running > 10 miles/wk or spending > 3 hrs/wk in comparable physical activity (Table S2). MV1: multivariable model 1 adjusted for age, sex, BMI, smoking status, education level, marital status, prevalent hypertension, dyslipidemia, and type 2 diabetes. MV2: multivariable model 2 adjusted for age, sex, BMI, total energy intake, mMDS, smoking status, education level, marital status, multivitamin use, supplement use, sleep, prevalent hypertension, dyslipidemia, and type 2 diabetes. CI: confidence intervals, HDL: high density lipoprotein cholesterol, LDL: low density lipoprotein cholesterol, SD: standard deviation, TG: triglycerides, Total-c: total cholesterol. Table S1 shows β coefficients (95% CI) and p-vales of independent associations for all linear regression models.