Glomerular Filtration Rate: When to Measure and in Which Patients?

Rajiv Agarwal; Pierre Delanaye


Nephrol Dial Transplant. 2019;34(12):2001-2007. 

In This Article

Abstract and Introduction


Of the glomerular, tubular and endocrine functions of the kidney, nephrologists have mostly focused their attention on the glomerular functions—albuminuria and glomerular filtration rate (GFR)—to grade the severity of chronic kidney disease (CKD). Although both albuminuria and GFR are associated with renal and cardiovascular morbidity and mortality, the utility of measured GFR (mGFR) has been questioned. GFR when measured adequately is the most precise measure of glomerular function and can be useful to individualize therapy among patients with CKD. In situations where estimated GFR is known to provide imprecise estimates of glomerular function, for example, sarcopenia and advanced cirrhosis, the measurement of GFR may be especially important. We discuss several clinical situations where mGFR can potentially influence the quality of life or complications of therapy because of interventions based on imperfect knowledge of GFR. We reason that although large databases may not detect the benefits of mGFR at the population level, precision medicine requires that therapy be individualized based on the best estimate of GFR that can be obtained particularly when the risk of harm is increased. The recent standardization of mGFRs is a step in the right direction and may help in treating the individual patient with CKD with a lower risk of complications and a better quality of life. We call for research in these subgroups of patients where it is clinically felt that mGFR is useful for clinical decision-making.


The kidney functions can be classified into three broad groups—glomerular, tubular and endocrine (Figure 1). Measuring glomerular filtration rate (GFR) can be theoretically obtained by plasma or urinary clearances of exogenous makers that have the following characteristics: freely filtrated through the glomerulus, not bound to proteins, neither secreted nor reabsorbed by tubules, and only excreted by the kidneys. Different markers are available (inulin, 51Cr-EDTA, 99Tc- DTPA, iohexol and iothalamate), each having their own strengths and limitations.[1] For the estimation of GFR, the most widely available and used endogenous marker is serum creatinine.[2] We will not point out the limitations of serum creatinine as a filtration marker, which have been discussed extensively in the literature, except for some less recognized observations.[2–4] First, it is now well established that in early stages of kidney failure, serum creatinine and measured creatinine clearance may remain normal despite a falling GFR.[5] Since creatinine is both freely filtered and secreted, its levels may be normal despite measured GFR (mGFR) dropping because of increased tubular secretion or reduced creatinine generation. The ratio of measured creatinine clearance/inulin clearance averages 1.16 for GFR >80 mL/min/1.73 m2, 1.57 for GFR 40–80 mL/min/1.73 m2 and 1.92 for GFR <40 mL/min/1.73 m2.[6] Thus, tubular secretion may be increased more with falling GFR. However, the correlation between creatinine clearance/mGFR and the mGFR is weak (ρ −0.49), which makes it difficult to predict the magnitude of creatinine secretion accurately. Less recognized is that serum creatinine is a negative acute-phase reactant. Thus, inflammation can reduce the generation of creatinine.[7,8] Serum cystatin C is another emerging marker of glomerular filtration.[9–11] It too is influenced by inflammation in that the latter can increase the generation of cystatin C.[7,12]

Figure 1.

Conceptual framework for renal functions. The renal functions can be divided broadly as glomerular, tubular and endocrine functions. Nephrologists have largely focused on the glomerular functions. Some such as eGFR and albuminuria can be influenced by tubular functions also. The importance of tubular and endocrine functions is discussed in the text. Dotted arrows denote that tubular functions can modulate some functions thought to be purely glomerular in nature (see text for details).

Tubular function is another important domain of kidney health. Besides filtration of albumin by the kidney, the uptake of urinary albumin by cubulin and megalin pathways reflects an important contribution of the tubule in albumin handling.[13] Less recognized is that albumin can be synthesized by the proximal tubules in states of injury.[14] Thus, urinary albumin excretion rate likely reflects glomerular and tubular functions. The associations of serum phosphorus and bicarbonate with progression of chronic kidney disease (CKD) to end-stage renal disease (ESRD) likely reflect—at least in part—the independent assessment of tubular function by these markers.[15,16]

Endocrine functions of the kidney are many, but the best recognized are the production of erythropoietin and activation of vitamin D.[17] Blood hemoglobin concentration, which is much more commonly measured than serum erythropoietin levels, is influenced by inflammation and iron deficiency. Therefore, it is not surprising that hemoglobin is a strong marker of progression of CKD to ESRD.[18] Similarly, 25 hydroxy vitamin D, a precursor for 1,25 dihydroxy vitamin D, is influenced not only by the dietary intake but also by sunlight exposure—a proxy for overall health—and negatively by inflammation. Thus, the association of 25 hydroxy vitamin D with poor outcomes comes as no surprise.[19]

Despite rich tools available to us to measure renal functions, nephrologists worldwide have focused mostly on the estimation of GFR (eGFR) and albuminuria for the assessment of kidney function.[20,21] Given that both eGFR and urine albumin to creatinine ratio are to varying degrees reflect glomerular, tubular and non-renal domains (such as age, sex, race, sarcopenia, liver disease and inflammation), it is not surprising that strong associations between these measures and all-cause mortality and ESRD are seen. The question emerges of whether mGFR has any value.[22,23]