Segmental kidney volumes measured by dynamic contrast-enhanced magnetic resonance imaging and their association with CKD in older people

Background: Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) is a potentially powerful tool for analysis of kidney structure and function. The ability to measure functional and hypofunctional tissues could provide important information in groups at risk for chronic kidney disease (CKD), such as the elderly. Study Design: Observational study with a cross-sectional design. Setting & Participants: 493 volunteers (aged 72-94 years; 278 women; mean estimated glomerular filtration rate [eGFR], 67 ± 15 mL/min/1.73 m₂; 40% with CKD) in the Age, Gene/Environment Susceptibility (AGES)-Reykjavik Study. Predictor: DCE-MRI kidney segmentation data. Outcomes & Measurements: eGFR, urine albumin-creatinine ratio (ACR), and risk factors for and complications of CKD. Results: After adjustment for age, sex, and height, eGFR was related to kidney volume (ΔR₂ = 0.19; P < 0.001), cortex volume (ΔR₂ = 0.14; P < 0.001), medulla volume (ΔR₂ = 0.18; P < 0.001), and volume percentages of fibrosis (ΔR₂ = 0.03; P < 0.001) and fat (ΔR₂ = 0.01; P = 0.03). In similarly adjusted models, log(ACR) was related to kidney volume (ΔR₂ = 0.02; P < 0.001) and fibrosis volume percentage (ΔR₂ = 0.03; P < 0.001). Using multivariable regression models adjusted for eGFR, ACR, age, sex, and height, kidney volume was related positively to body mass index (B = 29.9 ± 2.1 [SE] mL; P < 0.001), smoking (B = 19.7 ± 7.7 mL; P = 0.01), and diabetes mellitus (B = 14.8 ± 7.1 mL; P = 0.04) and negatively to hematocrit (B=-4.4 ± 2.1 mL; P = 0.04 [model R₂ = 0.72; P < 0.001]); relations were per 1-SD greater value of the variable. Fibrosis volume percentage was associated positively with body mass index (B = 0.28 ± 0.03; P < 0.001), cardiac output (B = 0.15 ± 0.03; P < 0.001), and heart rate (B = 0.08 ± 0.03; P = 0.01) and negatively with hematocrit (B=-0.07 ± 0.3; P = 0.02) and augmentation index (B=-0.06 ± 0.03; P = 0.04 [model R₂ = 0.49; P < 0.001]); again, relations are per 1-SD greater value of the variable. Limitations: Automatic segmentations were not validated by histology. The limited age range prevented meaningful interpretation of age effects on measured data or the automatic segmentation procedure. Conclusions: Kidney volume, cortex volume, and hypofunctional volume fraction assessed by DCE-MRI may provide information about CKD risk and prognosis beyond that provided by eGFR and urine ACR.