Pre-chronic kidney disease -- Serial creatinine tracks glomerular filtration rate decline above 60 mL/min

This paper proposes that monitoring serial serum creatinine levels relative to an individual's historical maximum is a more reliable and practical method than estimated GFR equations for detecting early glomerular filtration rate decline above 60 mL/min, enabling earlier intervention before chronic kidney disease stage 3.

The Gist

The Big Idea: Catching Kidney Trouble Before It's "Official" Trouble

Imagine your kidneys are like a high-performance sports car engine. For a long time, mechanics (doctors) have only checked the engine when the "Check Engine" light turns on. In medical terms, that light turns on when your kidney function drops below a certain level (called Stage 3 CKD).

The problem? By the time that light turns on, you've already lost a huge chunk of your engine's power. You've lost half your kidney function, and the damage is often hard to reverse.

This paper argues that we need to stop waiting for the "Check Engine" light. Instead, we should be watching the speedometer (your blood creatinine levels) to see if the car is slowing down while it's still running fast.

The Problem with the Old Map (eGFR)

Currently, doctors use a formula called eGFR (Estimated Glomerular Filtration Rate) to guess how well your kidneys are working. They take your blood creatinine level and plug it into a calculator that also factors in your age, sex, and race.

The Flaw:
Think of the eGFR like a weather forecast based on a population average.

• If the forecast says "It's 70°F today," that might be true for the whole city, but it might be 10 degrees hotter in your specific neighborhood.
• Similarly, the eGFR formula gives a "population average" guess. When your kidneys are working well (above 60 mL/min), this formula is very imprecise. It's like trying to measure the width of a hair using a ruler marked in inches. The margin of error is so big that a small, real drop in kidney function gets hidden in the "noise" of the math.
• Also, the old formulas used "race" as a variable, which the authors argue is a flawed and outdated way to estimate biology.

The New Solution: The "Personal Baseline"

The authors suggest a much simpler, more personal approach: Stop comparing yourself to the crowd; compare yourself to your past self.

They propose tracking your Serum Creatinine (sCr) over time. Creatinine is a waste product your muscles make. Your kidneys filter it out. If your kidneys slow down, creatinine builds up in your blood.

The Analogy: The "Personal Speed Limit"
Imagine you drive the same route to work every day.

• The Old Way (Population Average): A traffic report says, "The average speed on this road is 55 mph." If you are driving 54 mph, the report says you are fine. But maybe you usually drive 60 mph on this road. The report missed that you slowed down.
• The New Way (Personal Baseline): You know that for the last 10 years, you have always driven between 58 and 60 mph on this road. One day, you notice you are consistently driving at 56 mph. Even though 56 mph is still "legal" (within the normal range), your personal trend shows you are slowing down.

The paper calls this "Pre-CKD" (Pre-Chronic Kidney Disease). It's the stage where your kidneys are starting to fade, but you haven't hit the "disease" threshold yet.

How It Works in Real Life

The paper presents four patient stories to prove this works:

1. The Stable Driver (Patient N): This person had the same creatinine level for 21 years. Their "engine" was running perfectly. No changes, no worry.
2. The Step-Down (Patient 6): This patient was muscular and fit. Their creatinine was high (1.4 mg/dL), so a specialist said, "You're fine, it's just your muscles." But when they looked at the history, they saw the number had slowly crept up over 13 years. It wasn't just muscle; the kidneys were slowly failing. The "Check Engine" light hadn't turned on yet, but the speedometer was dropping.
3. The Family History (Patient 7): This patient had a family history of kidney issues and took painkillers (NSAIDs). Their creatinine slowly climbed. By tracking the trend, doctors could catch the decline early, even though the numbers were still technically "normal" for the general population.
4. The Medication Effect (Patient 10): This patient took a diuretic (water pill). When they stopped the pill, their kidney numbers improved. When they started it again, the numbers got worse. Tracking the change helped doctors realize the medication was stressing the kidneys, allowing them to adjust treatment before permanent damage occurred.

Why This is a Game-Changer

1. It's "Race-Free": Because you are comparing a person to their own past, you don't need to guess based on their race, age, or gender. A Black patient and a White patient are judged by their own unique history, not a statistical average.
2. It's Cheaper and Easier: You don't need expensive new tests. You just need to look at old blood test results that are already sitting in the computer.
3. It Saves Kidneys: If you catch the "slowdown" early (Pre-CKD), you can change your diet, stop certain medications, or treat high blood pressure to stop the engine from failing completely.

The Bottom Line

The paper argues that we are too focused on the absolute number (is it above or below the line?) and ignoring the story (is the number going up or down for this specific person?).

By treating your blood creatinine level like a personal fitness tracker rather than a static medical chart, doctors can spot kidney trouble years earlier, without needing complex formulas or outdated assumptions about race. It's about listening to the subtle whispers of your body before it starts screaming.

Technical Summary

1. Problem Statement

The paper addresses a critical gap in the early detection of Chronic Kidney Disease (CKD), specifically Stage 1 and Stage 2, where the Glomerular Filtration Rate (GFR) is above 60 mL/min.

• Limitations of Current Standards: Standard clinical practice relies on estimated GFR (eGFR) equations (e.g., CKD-EPI, MDRD) and population-based reference intervals for serum creatinine (sCr).
• The "Creatinine-Blind" Range: Above 60 mL/min, eGFR equations have wide error margins (P30 error), often failing to detect early GFR decline. Furthermore, the population reference range for sCr is too broad to detect subtle, clinically significant changes within an individual.
• Consequences: Most adults with early CKD remain undiagnosed until GFR drops below 60 mL/min (Stage 3), missing the window for primary care interventions (e.g., ACE inhibitors, GLP-1s) that can slow progression.
• Race and Bias: Current eGFR equations historically included race as a variable, introducing potential disparities and ethical concerns, though "race-free" equations still suffer from mathematical inaccuracies in the pre-CKD range.

2. Methodology

The authors propose a shift from population-based reference intervals to within-individual longitudinal tracking of serum creatinine.

• Conceptual Framework:

  • sCrMax (Maximum Serum Creatinine): Defined as the highest historical sCr value for a specific patient. This represents the patient's "minimum GFR" to date.
  • sCrRCV (Reference Change Value): The minimum meaningful change in sCr relative to sCrMax. The authors cite values of 13.3% for sedentary individuals and 26.8% for physically active individuals.
  • PreCKD Definition: A condition identified when a patient's sCr exceeds their historical sCrMax by more than the sCrRCV, indicating a true decline in GFR before reaching the arbitrary 60 mL/min threshold.

• Data Analysis & Re-evaluation:

  • Re-analysis of Shemesh et al. (1985): The authors re-examined a seminal 40-year-old study that suggested observing subtle sCr changes above 80 mL/min. They digitized the original data and performed curve fitting (hyperbolic regression) to validate that the relationship between sCr and GFR is continuous and sensitive down to 60 mL/min, provided individual baselines are used.
  • Statistical Modeling: They compared segmented models (grouping data by GFR ranges) against a continuous inverse function model, finding the continuous model statistically superior (lower AIC/BIC, significant $\chi^2$ test).

• Case Study Approach:

  • Cohort: 11 patients from a non-nephrology specialty practice (7 White, 3 Black, 1 Asian American-Native Hawaiian/Pacific Islander) with sCr $\ge$ 1.0 mg/dL but no prior kidney diagnosis.
  • Comparison: Included a "normal" control patient (Patient N) with 21 years of stable data.
  • Visualization: Created "Adult Serum Creatinine (ASC)" charts plotting longitudinal sCr against time, highlighting sCrMax and the sCrRCV thresholds.

3. Key Contributions

• Proof-of-Concept for PreCKD: Demonstrates that tracking serial sCr relative to an individual's historical maximum (sCrMax) can detect GFR decline in the "normal" population range (GFR > 60 mL/min).
• Mathematical Justification for sCr over eGFR: Argues that converting sCr to eGFR in the pre-CKD range introduces unnecessary mathematical error. Direct comparison of sCr values is mathematically cleaner and more sensitive to early changes.
• Race-Free Monitoring: By using within-individual baselines, the method renders race, age, and sex irrelevant for the detection of decline, as these factors are constant for the individual over time. This mitigates racial disparities in diagnosis.
• Re-evaluation of Tubular Secretion (TScr): Challenges the historical view that tubular secretion of creatinine significantly masks GFR decline above 80 mL/min. Re-analysis suggests fractional TScr remains low (~16%) in the pre-CKD range, supporting the sensitivity of sCr.
• Clinical Utility: Proposes a practical, low-cost workflow for primary care clinicians to identify high-risk patients who would otherwise be missed by standard eGFR reporting.

4. Results

• Case Study Findings:
  • Patient N (Control): Showed stable sCrMax over 21 years, confirming the stability of the method in healthy individuals.
  • Patient 6 (Black, Muscular): Despite a nephrologist's initial assessment of "normal function" due to high muscle mass, longitudinal data showed a stepwise increase in sCrMax (27.8% rise over 13 years), exceeding the sCrRCV for active individuals. This triggered re-evaluation and endocrinology referral.
  • Patient 7 (Black, NSAID use): Showed a progressive 33% rise in sCrMax, suggesting genetic nephropathy exacerbated by NSAIDs.
  • Patient 10 (Asian American): Demonstrated that medication (thiazide diuretics) could artificially elevate sCr. Stopping the drug revealed the true baseline, preventing a false "PreCKD" diagnosis.
• Statistical Validation:
  • The continuous inverse function model (hyperbolic) fit the Shemesh data significantly better than segmented models, validating the use of sCr as a continuous marker down to 60 mL/min.
  • Residual analysis showed random error distribution, supporting the reliability of the method.
• Diagnostic Sensitivity: The method successfully identified PreCKD in patients with sCr levels as low as 1.16–1.2 mg/dL, which are typically considered "normal" by population standards but represented significant declines from their personal baselines.

5. Significance and Implications

• Early Intervention: Enables primary care providers to initiate kidney-protective therapies (e.g., SGLT2 inhibitors, ACE inhibitors) years earlier than current guidelines allow, potentially preventing progression to kidney failure.
• Cost-Effectiveness: Utilizes existing, low-cost serum creatinine assays rather than expensive exogenous filtration markers (e.g., inulin, iohexol) or new biomarkers like cystatin C (which has its own confounders).
• Ethical Advancement: Provides a "race-free" diagnostic approach that relies on individual physiology rather than population averages, addressing systemic biases in kidney care.
• Paradigm Shift: Moves the clinical focus from "Is the patient in Stage 3?" to "Is the patient's kidney function declining relative to their own baseline?" This aligns kidney disease monitoring with the management of other chronic conditions like prediabetes and hypertension, where individual trends matter more than population cut-offs.

Conclusion: The paper argues that serial creatinine tracking is a superior, practical, and mathematically sound method for detecting early kidney disease (PreCKD) in the GFR > 60 mL/min range, offering a solution to the limitations of eGFR equations and the insensitivity of population reference intervals.