Abstract 019: Modifiable Lifestyle Factors And Plasma Branched Chain Amino Acids: An Analysis Of N=19,472 US Women

Abstract

Background: Circulating branched-chain amino acids (BCAAs) are associated with insulin resistance and higher risk of incident type 2 diabetes (T2D); however, despite their potential as a target for T2D prevention, upstream determinants of plasma BCAAs are largely unknown. Therefore, we investigated modifiable lifestyle factors in relation to plasma BCAAs in the Women’s Health Study. Methods: We performed a cross-sectional analysis among N=19,472 women (mean age=54.9 [SD=7.2] years) free of T2D, cardiovascular disease, and cancer. BCAAs were quantified via NMR spectroscopy and log-transformed among women reporting ≥8 hours of fasting at baseline blood draw. Lifestyle factors were ascertained from self-reported questionnaires at study baseline: body mass index (BMI, kg/m 2 ), smoking, alcohol, physical activity [PA, MET-hrs/wk], and diet (Mediterranean diet pattern score [aMED], alternative Health Eating Index score [aHEI], dietary BCAA intake [g/d], and individual protein sources [servings/d]). We estimated % mean differences in BCAAs (95% CI) for each lifestyle factor categorically (linear regression models) and continuously to calculate R 2 from univariate cubic spline models. Multivariate models were mutually adjusted for the other lifestyle factors and other T2D risk factors. Results: Overall, we observed positive associations between BMI and dietary protein sources with plasma BCAAs, while alcohol and PA were inversely associated with BCAAs in multivariable-adjusted models. Smoking and aMED and aHEI dietary quality scores were not associated with BCAAs. Compared with women with BMI<25.0, BCAAs were 8.2% (7.5, 8.8), 14.6% (13.8, 15.5), and 20.3% (17.8, 22.9) higher for BMI groups 25.0-29.9, 30.0-39.9, and ≥40.0, respectively (p-trend<0.0001). Compared to never drinkers, women with moderate or high alcohol intake had -1.8% (-2.8, -0.9) and -2.7% (-3.9, -1.5) lower BCAA levels. There was a slightly lower (-1.2% [-2.0, -0.4]) BCAA level among highest vs. lowest quartiles of PA (p-trend=0.0002). Women with higher intakes of dietary BCAAs had modest incrementally higher plasma BCAAs (comparing Q5 vs. Q1: 2.9% [2.0, 3.7]; p-trend<0.0001). Similarly, we observed dose-response relationships for higher intakes of major US protein sources (red meat, poultry, fish, eggs, dairy products) and higher plasma BCAAs. In continuous models, BMI explained 11.5% of the variability of plasma BCAAs, while the other modifiable factors each explained 0-1.1%. Conclusions: Our analysis in a large cohort of US women confirms prior evidence for a compelling relationship between BMI and plasma BCAAs. We also observed novel associations for dietary protein intake, alcohol, and PA, although the magnitudes of their contribution to BCAAs were small. Future clinical research on interventions to lower BCAAs for T2D prevention should consider body weight as a primary target for modification.