Kidney Week

Abstract: FR-PO902

Gait Speed and Mortality in Older Adults with CKD: The Chronic Renal Insufficiency Cohort

Session Information

• Geriatric Nephrology
  November 03, 2023 | Location: Exhibit Hall, Pennsylvania Convention Center
  Abstract Time: 10:00 AM - 12:00 PM

Category: Geriatric Nephrology

• 1300 Geriatric Nephrology

Authors

• Liu, Christine, Stanford University, Stanford, California, United States

Christine Liu,

• Parvathinathan, Gomathy, Stanford University, Stanford, California, United States

Gomathy Parvathinathan,

• Stedman, Margaret R., Stanford University, Stanford, California, United States

Margaret R. Stedman,

• Seliger, Stephen L., University of Maryland School of Medicine, Baltimore, Maryland, United States

Stephen L. Seliger,

• Weiner, Daniel E., Tufts Medical Center, Boston, Massachusetts, United States

Daniel E. Weiner,

• Tamura, Manjula, Stanford University, Stanford, California, United States

Manjula Tamura,

Group or Team Name

• CRIC Study Investigators.

Background

Current models to predict mortality in older persons with CKD use only demographic, kidney, and medical history data. Yet prior work shows impaired physical function independently and strongly predicts death in older adults. ^,We assessed if a measure of physical function, gait (walking) speed, improves mortality prediction for older adults with CKD.

Methods

We included Chronic Renal Insufficiency Cohort participants who were ≥65 years with eGFR <60 mL/min/1.73m², not receiving kidney replacement therapy, non-missing UACR, and with at least 1 gait speed assessment. Gait speed at usual pace on a 4.57 meter course was timed and then categorized (≥ 0.84 meters/second, 0.83 to 0.65, 0.64 to 0.47, ≤ 0.46, or unable to do). We designated the visit with gait speed measurement as baseline. Our primary outcome was time to all-cause death. We used a flexible parametric modeling approach with kidney replacement therapy as a competing risk, adjusting for age, sex, race, eGFR, UACR, smoking, diabetes, heart failure and stroke. C statistics were used to compare models with and without gait speed as a predictor.

Results

Among 2,345 persons, mean age was 70.0±4.4 years; 43% were female and 41% Black. Mean eGFR was 43.3±12.5 mL/min/1.73m² and median UACR was 33.8 (IQR 9.3-283.4). At baseline, 80% had a gait speed ≤ 0.83 m/s. Over 5 years, 393 persons died, and 164 had kidney failure. For those that died, median survival time was 2.6 years. In time-to-event analyses, slower gait speeds correlated with greater mortality (Figure). After adjustment, the inclusion of gait speed as a predictor improved model performance with the c statistic improving from 0.625 (95% CI 0.624-0.627) to 0.736 (95% CI 0.735-0.737).

Conclusion

In older adults with CKD, we found that gait speed improves mortality prediction. Clinicians should consider assessing gait speed when discussing life expectancy and goals of care with older adults with CKD.

Figure: Kaplan Meier curve of gait speed and all-cause mortality, censoring for kidney failure

Funding

• NIDDK Support