Your Labs Are Interpreted INCORRECTLY
There is little to no scientific basis for lab "reference ranges." In fact, they are arbitrary and often completely incorrect.
FrugalDoc
11/14/20266 min read


Every type of doctor orders "labs" to determine the patient's health. They use reference ranges to determine whether lab values are "normal" or "abnormal." Where the patient falls with respect to reference ranges (intervals) determines whether the person is healthy or sick and requires treatment.
Most of my FrugalDoc clients tell me they are frustrated by the interpretation of their labs by their doctor because they are very often 'normal,' but the person does NOT feel well.
WHAT COULD BE WRONG?
Three important things to know about reference ranges:
A normal result in one lab may be abnormal in another: You must use the range supplied by the laboratory that performed your test to evaluate whether your results are “within normal limits.”
An abnormal result does not mean you are sick: A test result outside the reference range may or may not indicate a problem. Since many reference values are based on statistical ranges in healthy people, you may be a healthy person outside the statistical range, especially if your value is close to the expected reference range.
A normal result does not promise health: While having all test results within normal limits is certainly a good sign, it’s not a guarantee. For many tests, there is a lot of overlap between results from healthy people and those with diseases, so there is still a chance of an undetected problem.
Dissecting the "three important things" about labs and why this means the reference ranges are NOT science-based.
A normal result in one lab may be abnormal in the other. This is the 21st century, and lab testing instruments are highly standardized. If the reference ranges vary from lab to lab for a given biomarker, it could ONLY indicate that they are ARBITRARY- not science-based. White blood cell counts are a perfect example. The lower reference range for WBC varies from 3400 to 5000 cells/μL. This range is well outside of what could be expected from lab-to-lab variability.
An abnormal result does not mean you are sick. In this case, the reference range must be re-evaluated using the best available data, and the range you were evaluated against is most likely NOT VALID.
A normal result does not promise health. If a normal result does NOT promise HEALTH, then the reference range is either too broad, or the data supporting the reference range is flawed. "Total Cholesterol" is the perfect example. See the graphic below.


This image comes from a research paper titled "Total cholesterol and all-cause mortality by sex and age: a prospective cohort study among 12.8 million adults."
This is one of many studies showing similar results. The 12.8 million participants help ensure the data is accurate. It was published in the journal "Nature," one of the top scientific and medical journals.
Here is the reference link: https://www.nature.com/articles/s41598-018-38461-y
Here is the conclusion from the following request of the literature. "List all the scientific journal articles that show that a total cholesterol value above 200 mg/dL has lower all-cause mortality compared to values below 200 mg/dL."
In summary, the strongest and largest study supporting the pattern you asked about is the 2019 Scientific Reports paper by Yi et al. (12.8 million people). Several other high-quality papers show consistent U-shaped relationships in which TC levels around or modestly above 200 mg/dL are associated with lower all-cause mortality than levels below 200 mg/dL.
Summaries of the other papers evaluating total cholesterol and mortality are provided at the end of this blog.
Here are standard criticisms of the reference ranges currently being used and some suggested alternatives.
The reference ranges rest on standardized, consensus protocols (This means they are not science but are really driven by drug treatments), proper sample selection, controlled measurement, and transparent statistical calculation.
The 95% central interval is a pragmatic statistical convention (rooted in probability theory and the historical “normal range”), not a biologically derived optimal range grounded in clinical outcomes or physiology for every analyte.
“Healthy” is operationally defined and imperfect—many conditions are asymptomatic, selection criteria vary, and reference populations may not perfectly match the patients being tested (different age, ethnicity, lifestyle, geography, etc.)
Results outside the interval do not automatically equal disease, and results inside do not guarantee health. Individual biological variation, the index of individuality, measurement uncertainty, and clinical context matter more. (What does "clinical context" mean? Could it be medical liability, the influence of big pharma and health insurance money that demand drug treatment, or other motivations unrelated to patient health?)
For some tests, clinical decision limits (outcome-based thresholds linked to risk or treatment benefit, e.g., certain glucose cutoffs) are preferred over pure reference intervals. (Here, the standard of care is acknowledging that reference ranges are not adequate to ensure a patient's current or future good health.)
A better path forward. The approach FrugalDoc uses.
What is of utmost importance to you? I hope your answer is living a LONG and HEALTHY LIFE!
Then, the best way to establish science-based reference ranges is to determine the hazard ratio (interval) for early all-cause mortality risk. Fortunately, these "HRs" are WELL ESTABLISHED for blood (and other) biomarkers. Importantly, each biomarker has its own hazard ratio (HR) values across the various levels of that biomarker in your (and your neighbor's) blood!
I asked OpenEvidence medical AI to compare two (2) people of the same age with different biomarker hazard ratios. This is the standard-of-care medical AI, and here is what it determined - by looking at physiological evidence and the definition of HRs. I gave two (2) hazard ratios for early all-cause mortality: 1.0 and 1.2. The 1.2 HR means the person is at an approximately 20% higher risk of dying suddenly, thus dying younger. I also had the AI evaluate this based on two different ages, 50 and 70, and the comparison was for exact age matches.
The person with HR 1.2 has a higher instantaneous mortality rate at every age (under the proportional-hazards assumption), so their survival curve declines faster, and their remaining life expectancy is shorter. The absolute difference in expected remaining lifespan is larger when the comparison starts at age 50 than at age 70, because the elevated hazard compounds over more years.
Interpretation: all-cause mortality absolutely occurs earlier/more frequently on average.
Here is the "punch line."
The reference ranges derived from hazard ratios are very different from the current standard-of-care reference ranges.
The reference ranges are EVIDENCE-BASED; thus, the HR-based reference ranges address the "lab-to-lab variation" (ridiculous) issue.
In many instances, the HR-derived reference ranges fall within the standard-of-care reference ranges but with a MUCH NARROWER RANGE compared to the WIDE standard-of-care range. This addresses the "a normal result does not promise health" statement.
In some instances, the HR-derived reference ranges DO NOT EVEN OVERLAP with standard-of-care reference ranges. This addresses the "An abnormal result does not mean you are sick" statement. See the next section for an example of total cholesterol, where the HR-derived reference range DOES NOT OVERLAP with what your doctor is telling you!
Key references and summaries of studies showing that people with total cholesterol above 200 mg/dL have lower all-cause mortality than those below 200 mg/dL.
Key scientific journal articles supporting lower all-cause mortality near/above 200 mg/dL vs. lower levels
Here are the most directly relevant peer-reviewed studies:
Yi SW, Yi JJ, Ohrr H. Total cholesterol and all-cause mortality by sex and age: a prospective cohort study among 12.8 million adults. Scientific Reports. 2019;9:1596.
Massive Korean cohort (12.8 million adults). Clear U-curve.
In the range 50–199 mg/dL, each ~39 mg/dL higher TC was associated with 23% lower mortality.
Above 200 mg/dL the association reversed slightly upward.
Lowest mortality generally at 210–249 mg/dL (slightly lower optimal ranges in the youngest adults).
Explicitly shows lower mortality for levels around/above 200 than for levels well below 200.
Huang JY, Lo K, Huang YQ, Feng YQ. A nonlinear association of total cholesterol with all-cause and cause-specific mortality. Nutrition & Metabolism. 2021;18:25 (PMC7945313).
U.S. data (NHANES-related). U-curve for all-cause and cancer mortality.
Hazard ratios for all-cause mortality were elevated in the lowest groups (<120 and 120–159 mg/dL) relative to the 160–199 mg/dL reference and higher categories (including 200–239).
Strong inverse relationship in the range below 200 mg/dL.
Jiang Q et al. Low and Decreasing Cholesterol Levels and Risk of All-Cause and Cause-Specific Mortality: A Prospective and Longitudinal Cohort Study. Engineering. 2025.
Chinese cohorts + UK Biobank (untreated adults). Nonlinear associations.
Optimal TC for lowest all-cause mortality near 200 mg/dL (Chinese) to ~250 mg/dL (UK).
Low TC (<120 or 120–160 mg/dL) linked to higher all-cause and cancer mortality vs. the medium range around 160–200.
Additional supporting context
Many other large studies (Copenhagen General Population Study and various East Asian and older-adult cohorts) report analogous U- or L-shaped patterns for LDL-C or TC, with the lowest all-cause mortality risk often in ranges corresponding to TC roughly 200 - 265 mg/dL.
Older-adult-specific analyses (e.g., some Italian and other cohorts) sometimes show TC ≥200 mg/dL associated with lower mortality than <200 mg/dL, particularly in women.
Stay Connected
Get low- or no-cost health tips, root-cause solutions, or a health audit.
Need Support?
Call us at: 202.630.3221
Email us at: frugaldr@gmail.com
Use the form on the right to share your wants/needs.
FrugalDoc
Affordable Health & Wellness
© 2025 FrugalDoc Health & Wellness. All rights reserved.
