Optimal vs. Normal Lab Ranges: What Your Bloodwork Should Actually Say

Normal Labs but still feel like your running at 60%

By Barbara Johnson, MD — Founder & Medical Director, The Johnson Center. Functional Health & Longevity

If your bloodwork came back “normal” but you still feel exhausted, foggy, and a decade older than your age, you are not imagining it. Here is the distinction almost no one explains: a normal lab range tells you that you do not have a diagnosable disease. An optimal range tells you where a number sits in the people who stay sharp, lean, and metabolically young for decades. Those are not the same target — and the gap between them is where most of midlife actually plays out.

On nearly every panel I read, the most important findings are not the flagged red values. They are the “normal” numbers quietly drifting in the wrong direction, years before the standard system is willing to call them a problem. Below is a marker-by-marker guide: the standard range, the optimal range, and — because this is the part that separates information from noise — the mechanism behind each one.

Key takeaways

 “Normal” ranges are built to flag disease; “optimal” ranges reflect where the healthiest, longest-lived people actually sit

 Reference ranges are statistical, not health-based — and they drift wider as the population gets sicker.

The biggest risks after 45 hide inside the “normal” range: insulin, ApoB, Lp(a), homocysteine, ferritin.

Lower is not always better — HbA1c, uric acid, and ferritin all carry U-shaped risk

 A number only means something with its mechanism. This guide gives you both.

Optimal vs. normal lab ranges at a glance

Reference ranges vary by laboratory; the values below are representative. Optimal targets reflect the longevity and functional-medicine literature and should be interpreted by a clinician in the context of your full history.

Biomarker	Standard "normal" range	Optimal range	Why optimal matters
Fasting insulin	Up to ~25 µIU/mL	Under 5 (ideally 2–5)	Earliest sign of insulin resistance
HbA1c	Below 5.7%	5.0–5.4%	U-shaped risk; lower isn't better
Fasting glucose	70–99 mg/dL	70–85 mg/dL	High-normal predicts future diabetes
Triglyceride / HDL ratio	Not flagged	Under 2	Insulin resistance + LDL particle size
hs-CRP	Under 3.0 mg/L	Under 0.5 mg/L	Chronic inflammation ('inflammaging')
GGT	Up to ~50–60 U/L	<20 men / <15 women	Oxidative stress; early fatty liver
Uric acid	Up to ~7.0 mg/dL	4.5–5.5 mg/dL	U-shaped; vascular and metabolic
Ferritin	~15–300 ng/mL	~70–80 ng/mL	Hides iron overload and deficiency
ApoB	Under ~100–130 mg/dL	Under 80 (under 60 aggressive)	Counts actual plaque-forming particles
Lp(a)	Often not tested	<75 nmol/L (30 mg/dL)	Genetic, hidden cardiovascular risk
Homocysteine	Under 15 µmol/L	5–7 µmol/L	Methylation + vascular damage
Vitamin B12	200–900 pg/mL	500–900 pg/mL	Functional deficiency hides in low half
RBC magnesium	4.0–6.4 mg/dL	High end (~5.5–6.4)	Cofactor for ATP production
Vitamin D (25-OH)	Above ~20–30 ng/mL	40–60 ng/mL	Prohormone; immune signaling
DHEA-S	Wide, age-banded	Young-adult level*	Marker of biological age
Estradiol (menopause)	"Postmenopausal" (often <30)	Individualized**	Bone and vascular protection
TSH	0.4–4.5 mIU/L	1.0–2.5 mIU/L†	Mid-range; not suppressed, not high-normal
Total testosterone (men)	264–916 ng/dL	Upper half (~500–900)‡	Muscle, bone, metabolism, mood
Free testosterone (men)	Assay-dependent	Upper third‡	The bioavailable fraction that acts
SHBG	~14–72 nmol/L	Mid-range	High locks up T; low flags insulin resistance
IGF-1	Wide, age-banded	~120–160 ng/mL§	U-shaped: too high and too low both harm

DHEA-S optimal ≈ 300–450 mcg/dL (men), 150–300 (women). * Estradiol is interpreted by menopausal status; a bone-protective threshold sits near 60 pg/mL when treatment is indicated. † TSH lowest-mortality clusters mid-range; very low/suppressed carries its own risk, and a mildly higher TSH can be normal in advanced age — read it with free T3, free T4, reverse T3, and thyroid antibodies. ‡ Testosterone “optimal” is symptom- and individual-driven, not a guideline threshold; free testosterone and SHBG matter as much as total. § IGF-1 is a Goldilocks marker — both high (cancer signaling) and low (frailty) raise mortality.

What does a “normal” lab range actually mean?

A reference range is built by measuring a marker across a large group of people and keeping the middle 95%. That is the entire method. It is a statistical description of who walked into the lab — not a definition of health. So as the population being measured becomes heavier, more insulin-resistant, and more inflamed, the “normal” range stretches to include them. Normal becomes common. Common is not the same as healthy.

Why reference ranges keep drifting

Because ranges are population-based, they move as the population's health moves — and the direction has been downhill. Average testosterone in men has fallen for decades, and the reference ranges fell with it, so a man today can be told he is “normal” against a sicker baseline than his father was measured against. The diabetes threshold itself was formally lowered in 1997 because the old cutoff was missing too much disease. The lesson is simple: chasing the reference range means aiming at a target that keeps sliding toward sicker.

The metabolic engine

Start here, because insulin resistance sits upstream of almost everything else on this list.

Fasting insulin: aim for under 5 µIU/mL

Glucose gets the headlines, but insulin moves first. For years before fasting glucose or HbA1c shifts, the pancreas compensates by releasing more and more insulin to push sugar into cells that have stopped responding — so a flawless glucose can sit on top of an insulin quietly elevated to 15 or 20. The standard range (up to roughly 25 µIU/mL) was built to catch diabetes, not to measure metabolic health. Optimal is under 5, ideally 2–5; pair it with HOMA-IR under 1.5 to catch resistance while it is still reversible.

HbA1c: aim for 5.0–5.4% (lower is not better)

HbA1c reflects your three-month average blood sugar. The system waits until 5.7% to say “prediabetes,” by which point glycation has been damaging vessels and brain tissue for years. But the biohacker reflex to drive it as low as possible is wrong: mortality is U-shaped, lowest near 5.4% and rising again once HbA1c drops below 5.0%, where a very low value often signals anemia or inflammation rather than metabolic excellence. Sit in the 5.0–5.4% window.

Fasting glucose: aim for 70–85 mg/dL

The lab will not flag fasting glucose until 100, but the risk does not wait that long. A fasting glucose sitting in the 90s — still “normal” — already predicts a meaningfully higher chance of progressing to diabetes and tracks with cardiovascular risk. Read it next to insulin and HbA1c, never alone: a normal glucose propped up by high insulin is not the same as a normal glucose that needs no propping. Optimal sits in the low 80s, roughly 70–85 mg/dL.

Triglyceride-to-HDL ratio: aim for under 2

This ratio is free — it is already on every lipid panel, and almost no one calculates it. Divide triglycerides by HDL. Under 2 reflects good insulin sensitivity and large, buoyant LDL particles; above 3 signals the small, dense LDL that drives plaque, often while fasting glucose still reads normal. One honest caveat: the ratio is less reliable in Black patients, who tend to carry lower triglycerides and higher HDL regardless of insulin resistance.

Inflammation and oxidative stress

This is where “normal” does the most quiet damage — including the inflammation behind stubborn brain fog — because none of these ranges were built to catch low-grade, long-term burn.

hs-CRP: aim for under 0.5 mg/L

hs-CRP is your systemic inflammation gauge, and the standard model ignores anything under 3.0. But chronic low-grade inflammation — “inflammaging” — is a core driver of heart disease, cancer, and brain aging. The lowest-risk groups keep hs-CRP under 1.0, ideally under 0.5. A 2.5 will not be flagged, and you are still inflamed.

GGT: aim for under 20 (men) or under 15 (women)

Most clinicians read GGT only as a liver-and-alcohol enzyme. It is also one of the cleanest oxidative-stress markers available, because GGT sits on the cell surface breaking down glutathione — your master antioxidant — a process that itself generates free radicals and oxidizes LDL. That is why GGT predicts all-cause and cardiovascular mortality even inside the “normal” range. The lab will not flag you until 50–65; lowest risk is under about 20 for men and 15 for women, and a slowly rising GGT is often the first sign of a fatty, oxidatively stressed liver.

Uric acid: aim for 4.5–5.5 mg/dL

Filed under gout and forgotten, uric acid is actually double-edged. In the bloodstream it is one of your major antioxidants; at high levels it turns pro-oxidant — depleting nitric oxide, stiffening vessels, and tracking closely with fructose intake and metabolic dysfunction. Mortality is U-shaped, so the target is a window, 4.5–5.5 mg/dL. A 7 your doctor ignores is doing vascular damage; a rock-bottom 3 has surrendered its antioxidant reserve.

Ferritin: aim for ~70–80 ng/mL, read in context

Ferritin is the most misread number on the panel because it measures two things at once: iron storage and inflammation. The standard range is enormous (~15–300), which lets it hide a problem at either end. Low (under about 40, common in menstruating women) drains cellular energy, thyroid conversion, and focus long before anemia appears. High is not robust iron — iron has no exit route, so the excess accumulates and drives oxidative damage, and a ferritin over roughly 150–200 more often flags metabolic inflammation or genetic overload than good health. Mortality bottoms out around 70–80; if yours is high, the next test is transferrin saturation, which separates true iron overload from inflammation.

The cardiovascular markers your panel skips

ApoB: aim for under 80 mg/dL

Stop staring at LDL cholesterol. ApoB counts the actual number of plaque-forming particles in your blood, and every particle that lodges in an artery wall adds to lifetime exposure. A “normal” ApoB of 110 is still grinding at your arteries every day. Optimal is under 80 — under 60 if you are serious about long-term prevention.

Lp(a): measure it once — most people never have

This is the single most important number most adults have never been told to check. Lp(a) is an LDL-like particle with a sticky, clot-promoting tail, set almost entirely by genetics — it barely moves with diet or exercise and stays fixed for life. About one in five people carry a high level that independently drives heart attack, stroke, and aortic-valve disease, and it is invisible on a standard panel. Every major lipid guideline now says to measure it at least once in every adult's lifetime. Under about 75 nmol/L (30 mg/dL) is ideal. You only need the test once, which makes it inexcusable that almost no one orders it.

Methylation and the B vitamins

Homocysteine: aim for 5–7 µmol/L

Homocysteine does triple duty. It is a direct readout of methylation — the reaction that runs DNA regulation, neurotransmitters, and biotransformation — so when it climbs, that machinery is stalling. It directly damages the lining of blood vessels and suppresses nitric oxide. And it is the earliest sign that B12, folate, or B6 are functionally inadequate, often months before anemia. The lab's ceiling of 15 was set for a rare genetic disease; in reality, all-cause mortality rises about 34% for every 5 µmol/L increase, and cardiovascular risk climbs from 6–7. Aim for 5–7.

Vitamin B12: aim for 500–900 pg/mL

The standard B12 floor of 200 is one of the worst-calibrated cutoffs in medicine — set to catch end-stage anemia, the very last stage of deficiency. Neurological symptoms appear in the 300s and 400s, and roughly a quarter of people in the “normal” 200–500 range already show elevated methylmalonic acid, the marker proving B12 is not reaching the cell. Aim for 500–900, and confirm gray-zone results with MMA and homocysteine. A very high B12 is not reassuring either — it can flag liver, kidney, or marrow problems.

Minerals and fat-soluble vitamins

RBC magnesium: aim for the high end (~5.5–6.4 mg/dL)

First, the test: serum magnesium is close to useless, because under 1% of your magnesium lives in the blood and the kidneys defend that level by pulling magnesium from bone and tissue — you can be 20–30% depleted with a “normal” serum result. RBC magnesium measures what is inside the cell, the magnesium equivalent of an HbA1c. It matters because magnesium is a cofactor for more than 300 reactions, including every step of ATP production — the literal currency of cellular energy. Target the high end (~5.5–6.4), not the barely-not-deficient bottom.

Vitamin D: aim for 40–60 ng/mL

The standard range bottoms out near 20 ng/mL — enough to prevent rickets, a 19th-century bone disease, and not much else. But vitamin D is a prohormone with receptors on nearly every immune cell, and the dose-response data on disease protection sits well above the bone floor. Target 40–60 ng/mL. A “normal” 22 means your immune signaling is running on fumes.

Thyroid: the marker your doctor over-trusts

Your thyroid sets the metabolic rate of every cell you own, which is why a thyroid running slightly slow shows up as fatigue, weight that will not move, brain fog, and cold hands long before anything earns the label of disease. The trap is that most evaluations stop at a single number.

TSH: aim for ~1.0–2.5 mIU/L

TSH is the pituitary's signal to the thyroid: when thyroid output starts slipping, the brain shouts louder, so a rising TSH is often the first sign of a struggling thyroid before the active hormones fall. The standard range runs to about 4.5, but the lowest cardiovascular and mortality risk clusters in the middle of the range — large cohort data put the sweet spot near 1.9–2.9, with the genuinely well usually sitting between 1 and 2.5. A “normal” 3.8 with symptoms deserves a real workup, not a shrug. Two honest caveats, because this is exactly where people overcorrect: lower is not automatically better — a suppressed TSH can mean an overactive or over-medicated thyroid — and in advanced age a mildly higher TSH can be normal, even protective. And TSH alone is incomplete: it tells you the signal, not whether T4 is actually converting to active T3. The full panel — free T3, free T4, reverse T3, and thyroid antibodies — is its own deep-dive, and the next article in this series.

Hormones after midlife

After 45 these shift fastest, and the reference ranges describe what is typical for your age — which, after midlife, is not the same as healthy. (More on hormone optimization.)

Testosterone (total): aim for the upper half of the range

Most panels report total testosterone, and most doctors read only whether it clears the floor. But the reference range — currently about 264–916 ng/dL for men — has drifted down with the population, so a “normal” 350 at age 50 can sit well below where a healthy young man runs. Testosterone is not just libido: it drives muscle, bone density, insulin sensitivity, mood, and drive, and low-for-age levels in middle-aged men track with metabolic syndrome and higher mortality. Optimal is the upper half of the range, roughly 500–900 ng/dL, read against symptoms rather than chased as a number. One honest limit: that mortality link is a midlife signal and does not clearly hold in the very old, so the goal is restoring a healthy midlife level, not maximizing it. (Testosterone matters for women too, at about a tenth the concentration — for libido, energy, muscle, and bone — and it is routinely ignored.)

Free testosterone and SHBG: the number total testosterone hides

Here is why total testosterone alone misleads. Only about 2–3% of your testosterone is unbound and able to act on cells; the rest is locked to a carrier protein called SHBG. And SHBG climbs with age — so an older man's total can read reassuringly normal while his free, usable testosterone is low and he has every symptom of deficiency. A man with a total of 450 and a high SHBG can carry a free testosterone in the basement. So measure free testosterone — ideally calculated from total testosterone, SHBG, and albumin, because the direct immunoassay is notoriously unreliable — and target the upper third of the range. Read SHBG in its own right, too: a high SHBG suppresses free hormone, while a low SHBG is a quiet flag for insulin resistance.

DHEA-S: aim for the level of a healthy young adult

DHEA-S is the most abundant steroid in your blood and the precursor to testosterone and estrogen, and it declines more predictably with age than almost anything else — about 1–2% per year from a mid-20s peak. That makes it a genuine marker of biological versus chronological age: people who hold higher DHEA-S for their age show better cognition, immunity, and lower mortality, most strongly in men. The honest caveat — and the line that separates a physician from a supplement seller — is that taking DHEA back up to youthful levels has not reproduced the longevity benefit in controlled trials. So read a low-for-age result as a biological-aging and adrenal-reserve signal, best interpreted alongside cortisol, not as an automatic reason to supplement. Young-adult targets: ~300–450 mcg/dL (men), 150–300 (women).

Estradiol (for women in menopause): undetectable is not the same as fine

A postmenopausal woman's estradiol can come back essentially undetectable and be labeled “normal,” because the reference range simply describes what is typical after menopause — not what is healthy. For bone, the distinction is real: below roughly 60 pg/mL, estradiol no longer restrains the cells that break bone down, and low levels predict fractures. For the cardiovascular system the benefit is real but conditional — estrogen protects arteries that are still healthy and treated early, not arteries already calcified. And the limits matter: hormone therapy is not proven to prevent dementia (the best current analysis finds no effect either way) and is not indicated for heart-disease prevention alone. The point is not a target number; it is that “undetectable” deserves a real clinical conversation about bone and vascular risk — not a shrug.

IGF-1: a window, not a maximum

IGF-1 is the growth-and-repair signal downstream of growth hormone, and it forces more nuance than any other marker in longevity medicine — because the science genuinely pulls both ways. Too low and you lose the anabolic signal that keeps muscle, bone, and brain intact: frailty, poor recovery, sarcopenia. Too high, sustained for years, and you are pushing growth pathways that raise the risk of several cancers. Large cohort data bear this out — the relationship with mortality is U-shaped, with the lowest risk clustering around 120–160 ng/mL and both tails carrying danger. This is the clearest marker on the whole panel where “more is better” is simply wrong: the goal is the middle of your age-adjusted range, supported by adequate protein and strength training, not the top of it.

Other Markers

A few more worth ordering: an omega-3 index above 8% tracks lower cardiovascular and cognitive risk, and two cheap signals already on your standard panels — RDW and albumin — quietly predict mortality even inside the normal range, a higher RDW and a lower albumin both pointing the wrong way.

Why this matters so much after 45

Two things converge in midlife. First, research from Stanford in 2024 found that human aging is not a smooth slope — it accelerates in two distinct waves, around ages 44 and 60, when the molecules governing metabolism, cardiovascular function, and tissue repair shift sharply. Second, your physiological buffer is shrinking, so the same “normal-but-drifting” number that was harmless at 35 becomes consequential at 50.

This is the window where optimal ranges stop being a luxury. It is also the logic behind the Cellular Intelligence Protocol™, the diagnostic framework we use at The Johnson Center to read these markers together — across the membranes, mitochondria, and hormones that determine how much usable energy your cells produce — rather than one flagged value at a time. Because the principle underneath all of it is simple: cellular energy controls health and longevity. Everything else is downstream.

Frequently asked questions

What is the difference between optimal and normal lab ranges?

A normal (reference) range describes the middle 95% of people tested at a lab and is designed to flag diagnosable disease. An optimal range reflects where the healthiest, longest-lived people sit. A result can be “normal” and still be far from optimal — which is where most early dysfunction hides.

Are “normal” lab results the same as healthy?

No. Reference ranges are statistical, not health-based, and they widen as the tested population becomes less healthy. A normal result means you likely do not have a diagnosable disease — not that your physiology is optimized.

What is an optimal fasting insulin level?

Optimal fasting insulin is under 5 µIU/mL (ideally 2–5), well below the standard cutoff near 25. Elevated fasting insulin is one of the earliest signs of insulin resistance, often appearing years before glucose or HbA1c change.

Is a lower HbA1c always better?

No. HbA1c mortality is U-shaped. The lowest risk sits around 5.4%, and risk rises again below 5.0%, where a very low value can reflect anemia or inflammation rather than excellent metabolic health. The optimal window is 5.0–5.4%.

What is an optimal homocysteine level?

Optimal homocysteine is 5–7 µmol/L, well under the standard ceiling of 15. All-cause mortality rises roughly 34% for every 5 µmol/L increase, and elevated homocysteine is an early sign of inadequate B12, folate, or B6.

Should I have my Lp(a) tested?

Yes — at least once. Lp(a) is genetically set, affects about one in five people, independently raises cardiovascular risk, and is invisible on a standard lipid panel. Because it stays stable for life, a single test is usually enough.

What is an optimal ferritin level?

The mortality and inflammation sweet spot is roughly 70–80 ng/mL. Ferritin reflects both iron storage and inflammation, so both low (deficiency) and high (overload or inflammation) values carry risk. A high ferritin should be followed by transferrin saturation.

Do optimal ranges differ for men and women?

For some markers, yes. DHEA-S and ferritin targets differ by sex, and estradiol is interpreted by menopausal status. The mechanisms are the same; the numbers are calibrated differently.

What is an optimal TSH level?

Standard ranges run to about 4.5 mIU/L, but the lowest cardiovascular and mortality risk sits in the middle — roughly 1.0–2.5, with large cohort data pointing near 1.9–2.9. Lower is not automatically better: a suppressed TSH can indicate an overactive or over-medicated thyroid. TSH should be read alongside free T3, free T4, and thyroid antibodies, not on its own.

What is the difference between total and free testosterone?

Total testosterone measures all the testosterone in your blood; free testosterone measures only the 2–3% that is unbound and able to act on cells. Because SHBG rises with age and binds testosterone, a man can have a normal total but a low free level — and the symptoms of low testosterone. Free testosterone, plus SHBG, usually tells the real story.

Get your numbers read the right way

---

Your labs deserve a physician, not a portal. At The Johnson Center. the Cellular Baseline™ assessment reads your full panel against optimal ranges — and against each other — to find the bottleneck driving your symptoms. Book a discovery call →

---

About the author

Dr. Barbara Johnson is a former general and trauma surgeon, specializing in cellular longevity and the founder and Medical Director of The Johnson Center. Functional Health & Longevity, with locations in Virginia Beach and Blacksburg and telemedicine across Virginia, North Carolina and West Virginia. She is the creator of the Cellular Intelligence Protocol™ and writes the No-BS Longevity newsletter. Physician, Scholar, Guide.

This article is for educational purposes and is not a substitute for individualized medical advice. Optimal ranges should be interpreted by a qualified clinician in the context of your full history.

Sources

[1] Vitamin D receptor expression in immune cells — review of extra-skeletal vitamin D physiology.

[2] Dose-response of 25(OH)D and all-cause mortality in adults aged 65+ — meta-analysis of prospective cohorts.

[3] U-shaped association between HbA1c and mortality in non-diabetic adults — NHANES-linked cohort (~16,000 adults).

[4] hs-CRP and cardiovascular risk stratification — high-sensitivity CRP thresholds.

[5] ApoB and atherosclerotic cardiovascular disease — particle-number causality.

[6] Prati, D., et al. (2002). Updated definitions of healthy ranges for serum alanine aminotransferase levels. Annals of Internal Medicine, 137(1), 1–10.

[7] National Academy of Clinical Biochemistry — laboratory medicine practice guidelines for thyroid testing (TSH reference interval).

[8] Travison, T. G., et al. (2007). A population-level decline in serum testosterone levels in American men. Journal of Clinical Endocrinology & Metabolism, 92(1), 196–202.

[9] Lokeshwar, S. D., et al. (2021). Decline in serum testosterone levels among adolescent and young adult men in the USA. European Urology Focus, 7(4), 886–889.

[10] Expert Committee on the Diagnosis and Classification of Diabetes Mellitus (1997). Report of the Expert Committee. Diabetes Care, 20(7), 1183–1197.

[11] Travison, T. G., et al. (2017). Harmonized reference ranges for circulating testosterone levels in men. Journal of Clinical Endocrinology & Metabolism, 102(4), 1161–1173.

[12] Association between homocysteine levels and all-cause mortality: a dose-response meta-analysis (2017). Scientific Reports, 7, 4769.

[13] Sumner, A. E., et al. (2005). Fasting triglyceride and the triglyceride–HDL ratio are not markers of insulin resistance in African Americans. Archives of Internal Medicine, 165(12), 1395–1400.

[14] Gamma-glutamyl transferase and the risk of all-cause and cause-specific mortality: a nationwide cohort study (2023). Scientific Reports.

[15] U-shaped association of serum uric acid with all-cause mortality: a cohort analysis of NHANES (2023).

[16] Ellervik, C., et al. (2014). Total and cause-specific mortality by moderately and markedly increased ferritin concentrations. Clinical Chemistry, 60(11), 1419–1428.

[17] Kronenberg, F., et al. (2022). Lipoprotein(a) in atherosclerotic cardiovascular disease and aortic stenosis: an EAS consensus statement. European Heart Journal, 43(39), 3925–3946.

[18] Ohlsson, C., et al. (2010). Low serum DHEA-S predicts all-cause and cardiovascular mortality in elderly Swedish men. Journal of Clinical Endocrinology & Metabolism, 95(9), 4406–4414.

[19] Nair, K. S., et al. (2006). DHEA in elderly women and DHEA or testosterone in elderly men. New England Journal of Medicine, 355(16), 1647–1659.

[20] Ettinger, B., et al. (1998). Associations between low serum estradiol, bone density, and fractures among elderly women. Journal of Clinical Endocrinology & Metabolism, 83(7), 2239–2243.

[21] Menopausal hormone therapy and risk of dementia: a systematic review and meta-analysis (2025). The Lancet Healthy Longevity.

[22] Elevated methylmalonic acid, a functional marker of B12 status, predicts all-cause mortality (2025).

[23] Spencer, C. A., et al. (2007). NHANES III: serum TSH upper reference limits may be skewed by occult thyroid dysfunction. Journal of Clinical Endocrinology & Metabolism, 92(11), 4236–4240.

[24] Xu, Y., et al. (2023). The optimal healthy ranges of thyroid function defined by the risk of cardiovascular disease and mortality. The Lancet Diabetes & Endocrinology, 11(10), 743–754.

[25] Late-life changes in the hypothalamic–pituitary–testosterone axis, health, and survival in men with longevity (2024) — Cardiovascular Health Study cohort aged 85+.

[26] Association between IGF-1 levels and all-cause mortality: a dose-response meta-analysis (2022); lowest mortality at ~120–160 ng/mL.

[27] Longo, V. D., et al. (2014). Low protein intake is associated with a major reduction in IGF-1, cancer, and overall mortality in the 65-and-younger population. Cell Metabolism, 19(3), 407–417.