Autor:innen:
M. Ritter (Lübeck, DE)
R. di Giuseppe (Kiel, DE)
M. Koch (Kiel, DE)
J. Borggrefe (DE)
H. Müller (Ulm, DE)
J. Kassubek (Ulm, DE)
G. Jacobs (Kiel, DE)
G. Jacobs (Kiel, DE)
M. Lutz (Kiel, DE)
U. Nöthlings (Bonn, DE)
W. Lieb (Kiel, DE)
W. Lieb (Kiel, DE)
N. Frey (Kiel, DE)
Background:
Osteopontin (OPN) emerged as a potential biomarker of cardiovascular diseases in experimental and clinical settings. However, little is known about clinical correlates of circulating OPN levels and their association with subclinical cardiometabolic traits in the general population. Therefore, we aimed to relate OPN to a broad spectrum of cardiovascular risk factors and cardiometabolic traits, including subcutaneous and visceral abdominal adipose tissue as well as liver fat content (LFC) in a community-based sample from Northern Germany.
Methods:
In this cross-sectional study, OPN concentrations were measured in 416 individuals from the general population (popgen control cohort; 1st follow-up visit; mean age 67±7 years; 42% women). Participants received anthropometric measurements and comprehensive physical examination, assessment of standard cardiovascular risk factors and whole body MRI measurements. Fasting blood samples were stored at -80°C until OPN was measured in plasma using ELISA. Visceral abdominal (VAT), subcutaneous abdominal (SAT) and subcutaneous trunk (STRAT) adipose tissue were quantified from a T1 weighted MRI sequence using ATLAS software and ARTIS algorithm. LFC was measured in a chemical shift imaging MRI sequence utilizing ImageJ software. Clinical correlates of OPN were identified by Spearman correlation coefficients and linear regression models with backward selection (candidate variables included age, sex, years of education (≥11; <11 years), BMI, systolic blood pressure, antihypertensive drugs, HbA1c, CRP, triglycerides, HDL, LFC, daily alcohol consumption, smoking (former; current smoker), pack years and physical activity). Linear and non-linear associations of OPN with SAT, STRAT, VAT and LFC were evaluated by means of restricted cubic spline regression models.
Results:
Median OPN concentrations were 394.2 ng/ml (325.3-500.4 ng/ml) in men and 380.5 ng/ml (297.0-481.3 ng/ml) in women. In multivariable-adjusted models, log-transformed OPN levels were independently and inversely associated with triglycerides (ß=-0.06 per 1-unit increase; p=0.02), HDL (ß=-0.10, p=0.03) and former smoking (ß=-0.08, p=0.03). The relation between LFC and OPN was J-shaped with both, very low as well as high biomarker levels being associated with higher liver fat values. The relationship between OPN and LFC was independent from potential confounders investigated in this study (ß=-0.02, p=0.02). OPN levels were not associated with SAT (ß=-0.001, p=0.36), STRAT (ß=-0.001, p=0.65) and VAT (ß=0.001, p=0.53).
Conclusion:
We identified triglycerides, HDL and former smoking as important correlates of circulating OPN in multivariable-adjusted models. In a restricted cubic spline regression model plasma OPN displayed a J-shaped association with LFC as determined by MRI. Further studies are needed to confirm our findings and to assess the clinical utility of OPN screening for cardiometabolic risk stratification.