What Your Bloodwork Isn't Telling You About Your Health

Why "Normal" Is Not the Same as Optimal

Reference ranges are built from population statistics, not from the biology of feeling well. A value lands inside the range if it falls between the 2.5th and 97.5th percentile of a tested cohort, which means 5% of healthy people are automatically flagged as abnormal, and many genuinely unwell people sit just inside the boundary. The American Thyroid Association's 2012 guidelines acknowledge that "the appropriate upper limit of the reference range for TSH remains controversial," with some expert panels advocating a ceiling of 2.5 mIU/L for certain patient groups rather than the conventional 4.5 mIU/L [1].

That statistical reality has direct consequences. A patient with a TSH of 4.2 mIU/L, free T4 of 0.8 ng/dL (low-normal), and debilitating fatigue receives a "normal thyroid" report and no further workup. The number cleared the reference bar. The person is still struggling. Reference ranges also shift across laboratories, analyzers, and even time of day. A TSH drawn at 8 a.m. runs roughly 26% higher than one drawn at 4 p.m. because TSH secretion follows a diurnal rhythm peaking in the early morning hours [2]. Most labs do not record draw time on the printed report.

Optimal ranges, by contrast, are derived from outcomes research: the TSH level at which cardiovascular events, pregnancy complications, or symptom burden reach their lowest point. These two numbers are frequently not the same.

The Thyroid Panel Your Doctor Ordered Is Probably Incomplete

TSH alone is the most commonly ordered thyroid test. It is also the most indirect. TSH is a pituitary signal telling the thyroid to produce hormone; it does not measure how much thyroid hormone is actually circulating in your blood or reaching your cells. A normal TSH with a low Free T4 or low Free T3 still produces hypothyroid symptoms because the downstream hormones are what tissues actually use.

Free T4 measures unbound thyroxine, the main hormone secreted by the thyroid gland. Free T3 measures unbound triiodothyronine, the metabolically active form converted from T4 primarily in the liver and kidneys. In a 2013 study published in the Journal of Clinical Endocrinology and Metabolism, patients with hypothyroidism on levothyroxine monotherapy who maintained normal TSH still reported significantly lower psychological well-being and higher body mass index than controls, suggesting that TSH normalization alone does not restore full physiological function [3]. The researchers found that a combination of T4 plus T3 therapy moved quality-of-life scores closer to euthyroid norms.

Reverse T3 (rT3) is a metabolically inactive isomer produced when T4 is converted down the wrong pathway, often during physiological stress, caloric restriction, chronic illness, or high cortisol states. Elevated rT3 competes with Free T3 at the receptor level and can produce hypothyroid symptoms even when TSH and Free T4 look perfectly adequate. A 2016 retrospective analysis in Thyroid found that critically ill patients with high rT3-to-Free T3 ratios had significantly higher 28-day mortality (OR 3.4 to 95% CI 1.6, 7.2, P<0.01) compared to patients with lower ratios [4]. Most standard panels never measure it.

Thyroid peroxidase antibodies (TPO-Ab) and thyroglobulin antibodies (TgAb) identify Hashimoto's thyroiditis, the most common cause of hypothyroidism in the United States. A person can have positive antibodies, a normal TSH, and still be on a trajectory toward overt hypothyroidism years before their TSH crosses any threshold. Ordering antibodies at the first sign of thyroid symptoms gives a far earlier warning than waiting for TSH to drift out of range.

Fasting Glucose Misses Early Insulin Resistance

Fasting glucose crosses the diabetes threshold at 126 mg/dL. Pre-diabetes is flagged at 100 to 125 mg/dL. But insulin resistance, the metabolic dysfunction that precedes both conditions by a decade or more, can be present when fasting glucose sits at a comfortable 88 mg/dL. The reason: the pancreas compensates for rising insulin resistance by secreting more insulin. Blood sugar stays controlled. The compensatory hyperinsulinemia is invisible on a glucose test.

Fasting insulin is the missing marker. A fasting insulin above 7, 10 µIU/mL in a person with normal fasting glucose signals compensated insulin resistance. The HOMA-IR formula (fasting glucose in mmol/L multiplied by fasting insulin in µIU/mL, divided by 22.5) quantifies it. A HOMA-IR above 1.9 suggests early resistance; above 2.9 indicates significant resistance. A 2019 study in Diabetes Care (N=7,822) showed that HOMA-IR predicted incident type 2 diabetes over 10 years with an area under the curve of 0.80, substantially outperforming fasting glucose alone (AUC 0.72) [5].

Hemoglobin A1c catches the 90-day average blood sugar but can be falsely low in people with rapid red-cell turnover (hemolytic anemia, recent blood loss) and falsely high in iron deficiency. An A1c of 5.4% in someone with undiagnosed iron-deficiency anemia may reflect higher average glucose than the number suggests. Running ferritin alongside A1c removes that ambiguity.

Inflammation: hsCRP, Homocysteine, and What the Basic Panel Skips

A standard comprehensive metabolic panel (CMP) does not include any direct inflammation marker. The typical CBC may flag an elevated white blood cell count in acute infection but misses the low-grade chronic inflammation tied to atherosclerosis, autoimmunity, and metabolic disease.

High-sensitivity C-reactive protein (hsCRP) fills that gap. The AHA and CDC issued a joint scientific statement classifying hsCRP <1.0 mg/L as low cardiovascular risk, 1.0 to 3.0 mg/L as moderate, and above 3.0 mg/L as high risk [6]. The JUPITER trial (N=17,802) demonstrated that rosuvastatin 20 mg daily reduced major cardiovascular events by 44% in patients with normal LDL cholesterol but elevated hsCRP (>2.0 mg/L), establishing hsCRP as a clinically actionable risk marker independent of lipid levels [7]. A person with an LDL of 95 mg/dL and hsCRP of 4.8 mg/L carries a risk profile that looks nothing like a person with an LDL of 95 and hsCRP of 0.4 mg/L. Standard lipid panels do not distinguish them.

Homocysteine is an amino-acid byproduct of methionine metabolism. Levels above 10 µmol/L are associated with roughly double the cardiovascular risk, and levels above 15 µmol/L also predict cognitive decline and bone loss. A meta-analysis of 30 prospective studies (N=5,073 cases) published in BMJ found that each 5 µmol/L increase in homocysteine raised coronary heart disease risk by 32% (95% CI 19 to 47%) and stroke risk by 59% (CI 25 to 102%) [8]. Homocysteine is elevated by low folate, low B12, low B6, hypothyroidism, and certain MTHFR gene variants. Treating it costs pennies (methylfolate, methylcobalamin). Detecting it requires ordering a test that appears on almost no routine panel.

Sex Hormones: Total vs. Free, and the SHBG Problem

A testosterone test ordered without SHBG (sex hormone-binding globulin) is only half the picture. SHBG binds testosterone tightly, rendering it biologically unavailable. Two men can share a total testosterone of 450 ng/dL. If one has SHBG of 20 nmol/L and the other has SHBG of 65 nmol/L, their free testosterone values differ by more than 60%. The man with high SHBG functions hormonally as if his testosterone were much lower, regardless of what the total number says.

SHBG itself is metabolically informative. High SHBG is driven by elevated estrogen, hyperthyroidism, liver disease, and caloric restriction. Low SHBG is driven by insulin resistance, obesity, hypothyroidism, and androgen excess. A 2010 prospective cohort study in the New England Journal of Medicine (N=1,709 women followed for 12 years) found that women in the lowest SHBG quartile had a 5.8-fold higher risk of developing type 2 diabetes compared to those in the highest quartile, independent of BMI and conventional metabolic risk factors [9].

For women specifically, the free androgen index (total testosterone divided by SHBG, multiplied by 100) helps identify androgen excess in PCOS and can explain symptoms like acne, hair loss, and menstrual irregularity that persist despite "normal" total testosterone.

Micronutrients Your Panel Almost Certainly Skips

Vitamin D insufficiency affects 41.6% of U.S. adults according to NHANES survey data (N=4,495) [10]. The standard reference range for 25-hydroxyvitamin D flags deficiency below 20 ng/mL, but functional medicine guidelines and several specialty societies argue that levels below 40 ng/mL impair immune function and increase fracture risk. The Endocrine Society's 2011 clinical practice guideline states: "We define vitamin D deficiency as a 25(OH)D below 20 ng/mL and vitamin D insufficiency as a 25(OH)D of 21 to 29 ng/mL" and recommends repletion for both categories in at-risk individuals [11]. Many routine panels do not include 25-OH vitamin D at all.

Ferritin is not the same as serum iron. Serum iron fluctuates hour to hour based on recent meals, stress, and inflammation. Ferritin reflects the body's iron stores accumulated over weeks to months. A serum iron within range alongside a ferritin below 30 ng/mL produces persistent fatigue, hair loss, cold intolerance, and reduced exercise tolerance. Those symptoms overlap almost perfectly with hypothyroidism. Without ferritin, clinicians treating a "thyroid" patient may be missing iron deficiency as the primary driver or an additive factor.

Magnesium is the fourth most abundant mineral in the human body and a cofactor for over 300 enzymatic reactions, including ATP synthesis, DNA repair, and insulin signaling. Serum magnesium, the version usually reported, reflects only 1% of total body magnesium and stays normal until depletion is severe. A 2018 meta-analysis in Nutrients (34 studies, N=1,168,813 participants) found that each 100 mg/day increase in dietary magnesium was associated with a 19% lower risk of type 2 diabetes (RR 0.81 to 95% CI 0.77, 0.86) [12]. Red blood cell magnesium, though still imperfect, provides a more meaningful index of tissue stores and is rarely ordered on any standard panel.

Cortisol and the Adrenal Axis

A single 8 a.m. serum cortisol is the test most likely to be ordered when adrenal function is questioned. It screens for Addison's disease (primary adrenal insufficiency) and Cushing's syndrome. It tells you almost nothing about the dynamic patterns of cortisol output across the day that drive fatigue, sleep disruption, blood sugar dysregulation, and thyroid conversion.

A four-point salivary cortisol test (collected on waking, at noon, at 4 p.m., and at bedtime) maps the diurnal arc. A flat curve with low morning cortisol and normal afternoon values, a pattern called hypocortisol response, correlates with chronic fatigue syndrome in research by Jerjes et al. published in the Journal of Psychosomatic Research (2006, N=84) [13]. A standard serum cortisol drawn at 8 a.m. would appear within range in many of those same patients.

The DUTCH (Dried Urine Test for Comprehensive Hormones) test extends this further, measuring cortisol and cortisone metabolites, DHEA-S, and sex hormone metabolites in a single collection. It is not a replacement for serum testing in all contexts, but for patients with normal serum panels and persistent symptoms, it offers a different axis of information.

Lipids Beyond LDL: ApoB and Lipoprotein(a)

Standard lipid panels report total cholesterol, LDL-C (usually calculated, not measured), HDL-C, and triglycerides. Two markers that substantially refine cardiovascular risk prediction are almost never included: apolipoprotein B (ApoB) and lipoprotein(a) (Lp(a)).

ApoB counts the number of atherogenic particles, one ApoB per LDL, VLDL, and IDL particle. Calculated LDL-C measures the cholesterol cargo inside those particles. A person can have a normal LDL-C of 110 mg/dL but an elevated ApoB of 130 mg/dL if their LDL particles are small and dense, each carrying less cholesterol. That particle pattern is the more dangerous one. The 2021 European Society of Cardiology guidelines state: "ApoB is a better predictor of cardiovascular risk than LDL-C and should be measured when available" [14].

Lp(a) is a genetically determined lipoprotein that raises thrombotic and atherosclerotic risk. Levels above 50 mg/dL (approximately 125 nmol/L) are found in 20% of the population and represent a risk factor comparable to heterozygous familial hypercholesterolemia. Lp(a) does not respond meaningfully to statins, diet, or exercise. Knowing a patient's Lp(a) level changes risk stratification and may eventually guide the use of RNA-based therapies currently in phase 3 trials (pelacarsen, olpasiran). A single Lp(a) measurement is needed only once in a lifetime, per ESC guidance, because levels are largely genetically fixed [14].

How to Use This Information Clinically

The table below outlines a tiered framework for extending beyond the standard panel. Tier 1 represents tests that most clinicians should consider at baseline for any symptomatic adult. Tier 2 represents tests appropriate when Tier 1 results are ambiguous or symptoms persist. Tier 3 represents specialized testing best ordered in partnership with an endocrinologist or hormone-specialist provider.

Tier 1 (Expanded Baseline): TSH plus Free T4, Free T3, TPO antibodies, fasting insulin, HOMA-IR, hsCRP, ferritin, 25-OH vitamin D, ApoB, complete blood count with differential, comprehensive metabolic panel.

Tier 2 (Persistent Symptoms, Normal Tier 1): Reverse T3, SHBG, free testosterone (calculated or direct), homocysteine, Lp(a), red-cell magnesium, morning serum cortisol, Epstein-Barr virus IgG/IgM if fatigue is prominent.

Tier 3 (Specialist-Guided): Four-point salivary cortisol, DUTCH urine hormone metabolites, MTHFR genotyping, organic acids panel, thyroglobulin antibodies, sex-hormone metabolite profiling, advanced particle sizing (NMR LipoProfile).

No single tier catches everything. Symptoms guide the sequence. A patient with fatigue, weight gain, cold intolerance, and hair loss warrants a full thyroid panel plus ferritin and vitamin D as a first step, not a repeat TSH six months later.

Interpreting Results: Context Always Matters

One number in isolation is rarely enough to act on. A ferritin of 22 ng/mL in a menstruating woman with fatigue is meaningful. The same ferritin in a 55-year-old man with no symptoms may warrant monitoring but not immediate treatment. Results interpreted without the clinical story drive unnecessary intervention and false reassurance in equal measure.

Timing matters significantly. Testosterone peaks in the morning and drops by 20 to 30% by afternoon. TSH peaks around 3 a.m. and reaches its nadir by late afternoon. Cortisol is highest within 30 minutes of waking. Drawing these tests at inconsistent times and then comparing results across months adds noise to an already complex signal. When tracking hormone levels longitudinally, standardize collection to the same time of day and the same fasting state each time.

Units matter. Testosterone reported in ng/dL by U.S. labs is reported in nmol/L by European labs. The conversion factor is 28.8. A testosterone of 450 ng/dL equals 15.6 nmol/L. Patients comparing their results across labs or countries sometimes miscalculate this and generate unnecessary alarm or false reassurance.

Reference ranges are lab-specific. LabCorp and Quest Diagnostics use different assay platforms for Free T3, producing slightly different reference intervals. A Free T3 of 2.4 pg/mL may fall within LabCorp's range and below Quest's. Comparing results across labs without noting which platform generated them introduces avoidable confusion.

If your results came back "normal" but you still feel poorly, the right response is not to accept that as a complete answer. Request the actual numbers, not just "normal" or "negative." Ask which specific tests were run. Then consider whether the tests ordered match the symptoms described.