The number of older adults with chronic kidney disease (CKD) across the country is rapidly growing, driven by increased lifespans as well as higher incidences of diabetes and vascular disease. In the world of geriatric nephrology, more physicians are questioning the assumption that all older adults with low estimated glomerular filtration rate (eGFR) have renal pathology. Because of the way we estimate and classify renal function, some older patients labeled as having “CKD” may have relatively normal renal function for their age.

Part of this discrepancy relates to the tools we use to estimate GFR: the Modification of Diet in Renal Disease (MDRD) and Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equations. Both rely on serum creatinine, a reasonably accurate but imperfect biomarker, to estimate renal function. However, healthy individuals have more muscle mass than those with CKD, and therefore have a higher GFR at the same creatinine level. As a result, the MDRD and CKD-EPI equations tend to underestimate GFR in healthy adults (in one analysis, by up to 25% and 16%, respectively).¹

In 2012, the Kidney Disease Improving Global Outcomes (KDIGO) group released their practice guidelines for the evaluation of CKD.² By these guidelines, any adult with an eGFR of less than 60 persisting for three months is determined to have CKD, even absent other evidence of kidney damage. These eGFR cutoffs do not vary based on age; thus a 30-year-old and a 90-year-old with the same eGFR’s of 58 are lumped into the same CKD3a bucket.

The KDIGO guidelines hold 90-year-old nephrons to the same standards as their 30-year-old counterparts. Yet nephrons, like all functional units in the body, undergo a predictable process of aging called senescence, even in the absence of true pathology or disease. At the cellular level, glomerular capillary walls deteriorate, podocytes degrade, and tubular function slows, leading to fewer functional glomeruli. Even in the absence of comorbidities, renal function declines an estimated 1 ml/min/1.73 m2 (unit of GFR) per year starting at age 40.³ For example, a 30-year-old with a relatively average GFR of 105 may undergo a reduction of nearly 50% to a GFR of 55 by age 90.

When taken together, equations that underestimate eGFR in healthy adults, age-nonspecific CKD diagnostic criteria, and age-related renal senescence mean that CKD is likely over-diagnosed in older adults.

Notably, isolated reduced eGFR in older adults does not always portend a poor prognosis. The CKD Prognosis Consortium, one of the largest studies to examine this, initially used age-nonspecific cutoffs of eGFR to determine a threshold that was independently associated with adverse outcomes, including cardiovascular events and all-cause mortality. The study found that adverse outcomes increased significantly below an eGFR of 60, driving the KDIGO 2012 guidelines’ definition of CKD as an eGFR of <60. Years later, the same data from the Consortium was reanalyzed using age-specific eGFR cutoffs⁴ (the normal range of eGFR was >105 for ages 18-54, 90-104 for ages 55-64, and 75-89 for ages 65+). Among older individuals, researchers found that mortality did not significantly increase until an eGFR of <45. Older adults with an eGFR of 45-59 had the same rates of cardiovascular events and all-cause mortality as those with an eGFR of 60 and above. Furthermore, for older patients with an eGFR of 45-59 without albuminuria, the risk of progression to ESRD is rare (<1% risk in 5 years).⁵

Overdiagnosis of CKD in older adults is harmful. It can lead to unnecessary distress for patients and their families, result in subtherapeutic drug dosing, expose older adults to excessive interventions and side effects, overwhelm specialist resources, and elevate the cost of care. We can have more nuanced discussions with patients about management of mild to moderate CKD by recognizing the context around eGFR equations and CKD definitions. This may mean closely monitoring patients with stable CKD3a rather than pursuing more advanced testing or referrals. Providers must interpret eGFR in clinical context. Certain scenarios warrant a nephrology referral and urgent work-up for renal pathology, including rapidly declining renal function, active urine sediment, significant albuminuria, etc. If there are no other signs or sequelae of kidney damage present, a stable eGFR between 45–59 might be considered typical for the patient’s age.

As with many issues in geriatrics, our collective misunderstanding of what constitutes “normal” kidney function in older adults is rooted in a lack of inclusion of older adults in research, age-insensitive cutoffs in diagnostic criteria and clinical algorithms, and normative attitudes that the aging process as inherently pathological. SGIM members must understand the limitations of GFR estimation and CKD diagnostic criteria to provide better care for older adults with possible renal dysfunction.