Quick Fire Session

SCMR 22nd Annual Scientific Sessions

5 - Association of Prevalent Aortic Plaque and Measures of Kidney Function

Presenter(s)

• MC

  Michael Chuang, MD

  Asst. Director
  Beth Israel Deaconess Medical Center

Author Only(s)

• PG

  Philimon Gona, PhD

  Professor
  University of Massachusetts Boston
• NO

  Noriko Oyama-Manabe, MD

  Associate Professor
  Hokkaido University
• CS

  Carol Salton, BA

  Research Associate
  Beth Israel Deaconess Medical Center
• CO

  Christopher O'Donnell, MD

  Professor
  Boston VAMC
• WM

  Warren Manning, MD

  Professor
  Beth Israel Deaconess Medical Center

Background:

We sought to determine the association between prevalence of atherosclerotic aortic plaque (AAP) and two principal measures of kidney function, estimated glomerular filtration rate (eGFR) and urine albumin-to-creatinine ratio (UACR) in a community-dwelling adult population.

Methods:

1726 adults (aged 65±9y) from the Framingham Heart Study Offspring cohort underwent CMR imaging (ECG-gated, free breathing, fat suppressed T2W black blood turbo-spin echo, 1.03 x 0.64-mm in plane resolution, 5-mm slice thickness) of the descending aorta. AAP was identified, by a radiologist blinded to study-subject characteristics, as discrete luminal protrusions of at least 1-mm thickness. Fasting AM blood and urine were obtained at the adjacent cycle visit; eGFR used the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equations, UACR was calculated and log-transformed due to skewed distribution. LogUACR and eGFR were entered into separate age/sex-adjusted logistic models for prevalent AAP. Surviving (p‹0.10) factors were entered into a multivariable-adjusted model.

Results: Mean eGFR did not differ between men (n=820, eGFR= 84±17 ml/min/1.73m²), and women (n=920, eGFR=84±18). UACR was available in 564 (70.0%) men and 658 (71.5%) women. LogUACR was lower among men (1.59±1.08) than women (2.00±0.96). LogUACR (p=0.06) and eGFR (p‹0.01) each were associated with prevalent AAP in age/sex-adjusted logistic models. In the multivariable-adjusted model comprising eGFR, logUACR, and traditional cardiovascular disease risk factors, only eGFR, and not logUACR, was associated with AAP. Significant odds ratios (OR) for prevalent AAP were: eGFR: OR= 0.89 per 10 units (95% confidence intervals [CI] 0.81 – 0.98, p=0.014); smoking: OR=1.18 (CI 1.16 – 2.82, p=0.009); age: OR=1.36 per 10y (CI 1.15 – 1.62, p=0.0004); fasting glucose: OR=1.12 per 10 mg/dL (CI 1.03 – 1.22, p=0.01), LDL cholesterol: OR=1.05 per 10 mg/dL (CI 1.003 – 1.09, p=0.03) and BMI: OR=0.96 per kg/m² (CI 0.93 – 0.99, p=0.005).

Conclusion: In a community-dwelling cohort of adults, both UACR and eGFR, two common measures of kidney function, were associated with prevalent AAP in age/sex-adjusted models: greater UACR was weakly positively associated, while greater eGFR was inversely associated with AAP. On multivariate modeling, UACR was no longer associated with AAP, while eGFR remained inversely associated with AAP. In addition to eGFR, smoking, greater age, fasting glucose, and LDL cholesterol were positively associated with prevalent aortic plaque.