Kidney Week

Abstract: TH-PO1136

Association of Serum Uromodulin with Mortality and Cardiovascular Disease in the Elderly – The Cardiovascular Health Study

Session Information

• CKD: Clinical, Outcomes, Trials - I
  October 25, 2018 | Location: Exhibit Hall, San Diego Convention Center
  Abstract Time: 10:00 AM - 12:00 PM

Category: CKD (Non-Dialysis)

• 1902 CKD (Non-Dialysis): Clinical, Outcomes, and Trials

Authors

• Steubl, Dominik, Tufts Medical Center, Tufts University, Boston, Massachusetts, United States

Dominik Steubl,

• Buzkova, Petra, UW, Seattle, Washington, United States

Petra Buzkova,

• Garimella, Pranav S., UCSD, San Diego, California, United States

Pranav S. Garimella,

• Ix, Joachim H., UCSD, San Diego, California, United States

Joachim H. Ix,

• Devarajan, Prasad, Cincinnati Children's Hospital, Cincinnati, Ohio, United States

Prasad Devarajan,

• Bennett, Michael R., Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio, United States

Michael R. Bennett,

• Chaves, Paulo H. M., Florida International University, Miami, Florida, United States

Paulo H. M. Chaves,

• Shlipak, Michael, San Francisco VA Medical Center, San Francisco, California, United States

Michael Shlipak,

• Sarnak, Mark J., Tufts Medical Center, Boston, Massachusetts, United States

Mark J. Sarnak,

Background

Uromodulin is released by renal tubular cells into the serum (sUMOD) and urine. Lower urine UMOD has been linked to mortality and cardiovascular disease but much less is known about sUMOD. We evaluated the association of sUMOD with those outcomes in community-dwelling older adults.

Methods

We measured sUMOD in a random subcohort of 933 participants enrolled in the Cardiovascular Health Study. The association of sUMOD with all-cause mortality, incident heart failure (HF) and incident CVD (myocardial infarction, stroke and mortality due to coronary disease or stroke) were evaluated using multivariable Cox regression, adjusting for study participants’ demographics, estimated glomerular filtration rate (eGFR), albuminuria and CVD risk factors. Generalized additive models with splines were used to address the functional form of sUMOD with outcomes. Due to non-linear associations of sUMOD with all outcomes, 2.5% of the values on either end of the sUMOD distribution were excluded from the analyses, limiting the range of sUMOD to 34.3 - 267.1 ng/ml.

Results

The mean age was 78±5 years, 40% were male, and 15% were non-white. Mean sUMOD level was 127±64 ng/ml, eGFR was 63 ml/min/1.73 m² and 42% had CKD defined as eGFR<60 ml/min/1.73 m². Patients in the lower sUMOD ranges had lower eGFR and higher albuminuria. During a median follow-up of 9.9 years, 805 patients died, 283 developed HF and 274 CVD. In multivariable analysis higher sUMOD was independently associated with lower mortality and CVD, and a trend for HF (Table).

Conclusion

Low sUMOD is independently associated with mortality and CVD in the elderly.

Funding

• Other NIH Support