Fasting Insulin

The earliest detectable signal of insulin resistance - moves before glucose, often before HbA1c.

What Fasting Insulin Measures

Fasting insulin is the earliest lab marker to flag developing insulin resistance - often elevated for years before fasting glucose creeps up and decades before HbA1c crosses the prediabetic threshold. A non-diabetic, "healthy" person can have fasting insulin of 15+ mIU/L and be metabolically unhealthy despite normal glucose.

The HOMA-IR calculation (fasting insulin × fasting glucose / 405) gives a quantified insulin sensitivity index. HOMA-IR >2.5 is insulin resistant; <1.5 is sensitive; <1.0 is excellent.

What Affects This Biomarker

Fasting insulin is influenced by: insulin sensitivity (the dominant variable), recent carbohydrate load (must fast 12+ hours), body fat (especially visceral), exercise (lowers reliably), sleep quality (poor sleep raises), and pharmacologic agents - metformin lowers; berberine lowers; GLP-1s lower significantly; SGLT2is lower; corticosteroids raise; growth hormone peptides and HGH can raise modestly.

In the Context of Peptide Protocols

Pull alongside fasting glucose at baseline and 3 months on any metabolic intervention (GLP-1, metformin, dietary change). On GLP-1s expect 30–60% reductions at maintenance dose. On GH protocols (especially MK-677), watch for upward drift signaling insulin sensitivity decline. Pair with fasting glucose to compute HOMA-IR for the cleanest insulin sensitivity read.

Deep Dive

Why fasting insulin moves before glucose

The pancreas is a remarkably adaptive organ. When peripheral tissues (muscle, liver, adipose) start losing sensitivity to insulin's signal - the earliest stage of insulin resistance - the pancreas responds by secreting more insulin to maintain normal blood glucose. This compensation can persist for years or even decades before fasting glucose finally creeps above 100 mg/dL and triggers a "prediabetes" diagnosis.

This is the central reason fasting insulin is the most underutilized biomarker in routine medicine. A patient with fasting glucose of 88 mg/dL and HbA1c of 5.2% looks metabolically perfect by standard screening. But if their fasting insulin is 18 mIU/L, they're already deeply insulin-resistant - the pancreas is just hiding it with effort. By the time glucose creeps up, beta-cell capacity has already started declining and reversing the trajectory is harder. Fasting insulin lets you catch the problem 5-15 years earlier.

HOMA-IR: quantifying insulin sensitivity in one number

HOMA-IR (Homeostatic Model Assessment of Insulin Resistance) was developed by Matthews et al. in 1985 as a way to use fasting insulin + fasting glucose to estimate insulin sensitivity without requiring a clamp study. The formula is simple:

**HOMA-IR = (fasting insulin in mIU/L × fasting glucose in mg/dL) / 405**

Interpretation tiers used in metabolic optimization practice: - **<1.0**: Excellent insulin sensitivity - **1.0-1.5**: Optimal - **1.5-2.5**: Borderline / early dysfunction - **>2.5**: Insulin resistant (matches clinical IR diagnosis in most populations) - **>5.0**: Severe insulin resistance, type 2 diabetes territory

HOMA-IR is one of the most validated proxies for true insulin sensitivity. The original Matthews paper showed HOMA-IR correlates ~0.88 with euglycemic clamp results (the gold standard but impractical for routine use).

The TOFI problem: thin-on-the-outside, fat-on-the-inside

A subset of normal-BMI individuals carry significant visceral fat without obvious external signs - the "TOFI" (thin outside, fat inside) phenotype. These individuals frequently have elevated fasting insulin despite normal weight, normal BMI, and normal glucose. Without a fasting insulin draw, this metabolic dysfunction goes undetected for years.

This is particularly common in lean-appearing Asian populations and in postmenopausal women with central adiposity. If you're "skinny" but have abdominal fat, sedentary habits, or a family history of type 2 diabetes, fasting insulin is the single most valuable lab to run.

What moves fasting insulin (and what doesn't)

**Reliably lowers fasting insulin:** - Sustained weight loss (especially visceral fat reduction) - Regular resistance training (improves muscle insulin sensitivity) - Aerobic exercise (acute insulin sensitivity boost lasting 24-72 hours) - Low-carb / Mediterranean dietary patterns - Adequate sleep (7-9 hours; sleep restriction can raise insulin 20-30% within days) - Metformin (typically 10-20% reduction) - Berberine (similar magnitude to metformin in trials) - GLP-1 receptor agonists - semaglutide, tirzepatide, retatrutide consistently produce 30-60% reductions at maintenance dose - SGLT2 inhibitors (modest)

**Reliably raises fasting insulin:** - Chronic high-glycemic eating patterns - Sleep deprivation - Chronic psychological stress (cortisol-driven) - Sedentary behavior - Visceral fat accumulation - Corticosteroids (oral, often dramatic) - Growth hormone secretagogues at higher doses (especially MK-677, which can raise fasting insulin via counter-regulatory effects on glucose disposal)

Peptide protocol context

Several peptides and peptide-adjacent agents directly affect fasting insulin in ways that matter for protocol design:

**GLP-1 receptor agonists (semaglutide, tirzepatide, retatrutide)**: These reliably reduce fasting insulin by 30-60% at therapeutic doses, primarily through fat loss + direct GLP-1 effects on beta cells and hepatic insulin signaling. In the SUSTAIN and STEP trials, semaglutide produced fasting insulin reductions of 25-40% at 1mg dosing. Tirzepatide in SURPASS trials showed even larger reductions (40-55% at 10-15mg). This is one of the most reliable downstream signals that GLP-1 therapy is working metabolically, not just dropping weight.

**MK-677 (ibutamoren)**: Stimulates GH and IGF-1 release, but in doing so creates a counter-regulatory environment where fasting glucose and fasting insulin both tend to rise. A 2008 study in older adults showed MK-677 raised fasting glucose by ~10 mg/dL and HOMA-IR by ~40% over 2 years (Nass et al., Annals of Internal Medicine). If you're on MK-677, fasting insulin should be tracked quarterly and a 30%+ upward drift warrants protocol re-evaluation.

**HGH (human growth hormone)**: Similar to MK-677 mechanistically - GH antagonizes insulin action peripherally. Fasting insulin rises modestly on HGH protocols, particularly at clinical doses (>2 IU/day). The MK-677 caution applies.

**MOTS-c, BPC-157, TB-500**: No known direct effect on fasting insulin in human data.

Interpretation traps

**Recent meal masquerading as fasted**: Fasting insulin requires a true 12+ hour fast. Even a small "cream in coffee" the morning of the draw can spike insulin 2-3×. If the result is unexpectedly high, the first question is always: was the fast actually 12+ hours, no caloric intake of any kind?

**Lab assay variability**: Insulin assays vary more than glucose assays between labs. If you're tracking trends, use the same lab. A "rise" of 2-3 mIU/L between two different labs may just be assay drift.

**Acute illness or inflammation**: Acute infections, surgery recovery, or significant stress can transiently raise fasting insulin. Don't make protocol decisions on a single elevated reading taken during illness.

**Pregnancy**: Fasting insulin rises significantly in pregnancy (especially 3rd trimester). Different reference ranges apply.

When to pull fasting insulin

The case for routine fasting insulin on every metabolic panel is strong: - Baseline before starting any metabolic intervention (GLP-1, metformin, dietary change) - 3 months after intervention to confirm response - Annual screening for anyone with family history of type 2 diabetes - Anyone with normal glucose + central adiposity (the TOFI screen) - Anyone on GH-axis peptides (MK-677, HGH, sermorelin) - quarterly

The cost is trivial ($15-30 at most direct-to-consumer labs). The information density is among the highest of any single biomarker. The "why isn't this on every panel?"question is largely an inertia problem in conventional primary care.

Fasting insulin trajectories that matter

Single fasting insulin readings tell you about a moment in time. Trajectories over 12-24 months tell you about underlying metabolic direction, which is the part that determines healthspan outcomes.

The favorable trajectory: starting at 12 mIU/L at age 40, gradually dropping to 6 mIU/L by age 45 through sustained lifestyle change or GLP-1 therapy. This trajectory tracks with reduced visceral fat, improved aerobic fitness, and dropping HbA1c. The 10-year risk projection for type 2 diabetes, cardiovascular events, and Alzheimer's all decline meaningfully.

The unfavorable trajectory: starting at 8 mIU/L at age 35 and gradually rising to 18 mIU/L by age 45, despite normal glucose throughout. This is the classic pre-diabetes-pre-diabetes pattern - beta cells compensating for decade-spanning insulin resistance buildup. Most people in this trajectory are completely asymptomatic and don't get the diagnosis until their glucose finally cracks 100 mg/dL, by which point they've been metabolically dysfunctional for 10+ years.

The flat-then-cliff trajectory: stable in the 4-7 mIU/L range for years, then a relatively rapid drop to 2-3 mIU/L. In the right context (no caloric restriction, no GLP-1, no metformin) this can signal beta-cell exhaustion - the pancreas losing its capacity to secrete normal insulin amounts. This is the trajectory that precedes Type 1.5 / LADA diabetes in some adults and warrants prompt clinical workup.

The fasting insulin / fasting glucose pair

Fasting insulin alone is informative. Fasting insulin paired with fasting glucose (which gives you HOMA-IR) is dramatically more informative because the pair distinguishes between four very different metabolic states.

State 1: Low insulin (3-6), normal glucose (75-90) = excellent metabolic health. Most lean fit adults under 40 fall here.

State 2: High insulin (12-25), normal glucose (75-95) = early insulin resistance, often visceral fat or sedentary lifestyle-driven. The pancreas is working overtime but succeeding. This is the highest-leverage intervention window because reversibility is high.

State 3: High insulin (15-30), elevated glucose (95-115) = established insulin resistance with beta-cell strain showing. Often coincident with elevated triglycerides, low HDL, ApoB drift. Risk of progression to clinical type 2 within 5 years is meaningful without intervention.

State 4: Low insulin (3-7), high glucose (>110) = beta-cell failure. Pancreas can no longer secrete enough insulin to compensate. Late-stage type 2, or autoimmune destruction (Type 1, LADA). Requires medical management, often insulin therapy.

The reason fasting insulin is so underutilized in conventional medicine is that primary care has been trained to chase fasting glucose and HbA1c, which only flag States 3 and 4. States 1 and 2 - which is where most metabolically suboptimal adults actually live - are invisible without the insulin draw.

C-peptide as a companion biomarker

For users tracking fasting insulin closely, C-peptide is the natural companion. C-peptide is the inactive cleavage product produced 1:1 with insulin, but it has a longer half-life (30 min vs 5-10 min for insulin) and isn't cleared by the liver. This makes C-peptide a more stable read on pancreatic insulin output, less affected by recent acute factors.

Where C-peptide is particularly useful: differentiating endogenous insulin secretion from exogenous insulin (in people who are or might be on insulin therapy), assessing beta-cell function in someone with rapidly dropping insulin levels, tracking insulin response to GLP-1 therapy over 6-12 month windows. Most direct-to-consumer labs offer C-peptide for $25-40. Pulling it alongside fasting insulin once a year provides a richer pancreatic-health picture.

*Educational reference only. Track and discuss with a qualified clinician. MyProtocolStack does not provide medical advice, diagnosis, or treatment recommendations.*

Peptides That Commonly Move Fasting Insulin

Conditions That Track Fasting Insulin

Upload any lab PDF and MyProtocolStack maps your values to Fasting Insulin and 40+ other biomarkers. StackAI interprets the trend in context of your protocol.