Insulin Resistance Screening in Athletes: Why Fasting Glucose Under-Flags in the Trained Population, HOMA-IR Against the Training-and-Fueling State, and the C-Peptide Follow-On When HbA1c Looks Normal

A 41-year-old male masters cyclist arrives at intake three weeks after a general physical he took as reassurance and left as confusion. His body mass index is 23.4, his training volume has held at 14-16 hours per week for the past year, his resting heart rate is 51, his power-at-FTP is up 3% over 12 months, and the annual bloodwork returned a fasting glucose of 92 mg/dL, an HbA1c of 5.3%, a lipid panel with total cholesterol 198 mg/dL and triglycerides 84 mg/dL, and a basic metabolic panel that read as unremarkable. The primary-care note was "metabolically healthy, continue current lifestyle."

He is in your office because two weeks after the visit he bought a continuous glucose monitor on a friend's recommendation and started reading his glucose curves through his usual training week. What he saw does not match the "metabolically healthy" note. His fasting morning glucose sits between 88 and 96 mg/dL, which reads reassuringly against the standard reference. His 40 g oatmeal breakfast on rest days is producing peaks of 142-158 mg/dL at 45-60 minutes with slow returns to baseline at 90-120 minutes. His post-ride refuels drive spikes into the 160s. He has been experiencing what he describes as "engine drops" on longer rides — a mid-ride energy crash he cannot fuel through with his usual bar-and-drink pattern — for the past six months, and his father was diagnosed with type 2 diabetes at 58 despite a similar lean build and a similar decades-long endurance-athlete history. The athlete has been chasing training volume as a way to outrun the family risk. His question is whether the CGM pattern he is seeing means his metabolic panel is wrong, or whether he is over-reading a normal athletic response.

The standard metabolic panel that produced his "metabolically healthy" flag was the wrong instrument for the pattern he is presenting. Fasting glucose is a late-appearing marker in the natural history of insulin resistance — the sequence, well-established in the metabolic literature, runs from rising post-meal glucose excursions, to rising fasting insulin (compensating to hold fasting glucose in range), to eventual fasting-glucose elevation only after the compensation fails. HbA1c aggregates across 90 days and is dragged down by nocturnal and inter-meal periods where trained athletes hold glucose low; it under-flags exactly the postprandial-excursion pattern the CGM is surfacing. The trained endurance athlete with a family-history risk factor and rising fasting insulin can sit in the pre-diabetic compensatory window for years while the standard annual bloodwork reads as clean.

Most sports-RD intakes do not run a structured insulin-resistance workup. The panel arrives from primary care, the fasting glucose and HbA1c look normal, the case gets passed as metabolic-non-issue, and the athlete's rising insulin, deteriorating postprandial handling, and family-history risk go unaddressed until a future draw finally catches the fasting-glucose elevation and the case becomes overt.

This post is the insulin-resistance workup I run when the standard panel reads clean but the clinical picture — CGM signal, family history, symptom cluster, training pattern — does not. The six-dimension interpretive frame, the fasting-insulin and HOMA-IR follow-on the standard panel misses, the C-peptide differential when HbA1c and glucose disagree, the training-and-fueling-state overlay that recalibrates the read for the athlete, the medical-coordination triggers, the common counseling mistakes, and the SOAP pattern that documents the case defensibly.

Why the standard metabolic panel under-flags the compensating athlete

Three structural reasons.

Fasting glucose is late in the natural history of insulin resistance. The compensatory pathway — rising insulin holding glucose steady — can run for years or decades before fasting glucose crosses the pre-diabetic threshold. Screening on fasting glucose alone catches the case at the end of that runway, not the beginning. The trained athlete with elevated fasting insulin, elevated post-meal glucose excursions, and an as-yet-normal fasting glucose is exactly the case the standard panel misses. Reading fasting glucose in isolation collapses this differential.

HbA1c averages across 90 days and is dragged down by the trained state. HbA1c reflects average glycemia, and the trained endurance athlete sits at chronically depressed nocturnal and inter-meal glucose values that pull the average down even when postprandial excursions are meaningfully elevated. A trained athlete with HbA1c of 5.3% and postprandial CGM peaks routinely in the 150s reads on the standard panel as normal; the same postprandial-excursion pattern in a sedentary matched control would drive the HbA1c up and trigger the flag. The instrument was not calibrated against the trained-athlete distribution.

Fasting insulin, HOMA-IR, and C-peptide are not on the standard panel. The variables that surface the compensatory pattern early — fasting insulin, the HOMA-IR calculation, and (in the ambiguous case) C-peptide — are not on any panel most primary-care physicals order. The RD who is not requesting them in the family-history or the CGM-signal case is under-workup by convention.

The six-dimension insulin-resistance workup

Dimension 1: Read fasting glucose against fasting insulin, not in isolation. Fasting insulin is the sensitive early marker. Reference ranges vary by lab and by assay, but a fasting insulin above roughly 10 mIU/L in a lean trained athlete with a normal fasting glucose is suggestive of the compensatory pattern; above 15 mIU/L is diagnostic pending the co-marker read; under 5 mIU/L in a lean trained athlete is the expected physiology. The HOMA-IR calculation — fasting glucose (mg/dL) times fasting insulin (mIU/L) divided by 405 — collapses the two variables into a single index. Values under 1.0 sit in the healthy-lean-athlete range; 1.0 to 2.0 is a borderline zone that warrants the follow-on read against the training and fueling state; above 2.0 is consistent with meaningful insulin resistance in most athletic populations. HOMA-IR as a single number requires the same integration context as the components — the trained athlete on a hard training block reads differently than the same athlete in a taper — but the number surfaces the pattern the fasting glucose alone hides.

Dimension 2: Read HbA1c against the CGM signal or the postprandial-excursion pattern. HbA1c reflects average glycemia and misses the postprandial-excursion pattern that drives cardiovascular and metabolic risk independent of the average. Where a CGM is available, capture the 14-day mean glucose, the time-in-range at 70-140 mg/dL, the mean amplitude of glycemic excursion, and the number of daily excursions above 140 mg/dL and above 180 mg/dL. Where a CGM is not available, an oral-glucose-tolerance test with 1-hour and 2-hour draws surfaces the postprandial-excursion pattern in a single visit. A trained athlete with an HbA1c of 5.3% and time-in-range under 70% or mean-amplitude-of-glycemic-excursion above 60 mg/dL sits in a meaningfully different metabolic frame than the same HbA1c paired with tight glucose control across the day.

Dimension 3: C-peptide as the disambiguator when HbA1c and glucose disagree. C-peptide is co-secreted with insulin at a one-to-one molar ratio, is not cleared by the liver on first pass the way insulin is, and provides a stable read on endogenous insulin production. The case where HbA1c reads normal but fasting insulin sits elevated and glucose excursions run high on the CGM is well-served by the C-peptide read to confirm the endogenous-insulin-driven story. A C-peptide above 3 ng/mL (reference ranges vary) confirms hyperinsulinemia; a low C-peptide in the setting of elevated glucose points toward a different mechanism (early type 1 or LADA in the athlete population is rare but relevant to catch). C-peptide is also the correct follow-on where the fasting insulin assay is suspected of interference (exogenous insulin history is not applicable here, but the assay caveats matter).

Dimension 4: Read against the training-and-fueling-state overlay. The trained endurance athlete's insulin-sensitivity profile shifts dramatically across the training block and the fueling day. Post-training-session insulin sensitivity runs elevated for hours, and a glucose or insulin draw in the acute post-exercise window reads differently than a draw 48-72 hours into a recovery week. Carbohydrate intake in the 24 hours before the draw shifts the fasting insulin. Sleep deprivation the night before the draw shifts insulin sensitivity down. Alcohol intake in the 48 hours before the draw shifts insulin sensitivity down. Document the training day, the fueling day, the last hard session, the sleep and alcohol context, and the carb-load state on the morning of the draw. The metabolic read interpreted without this overlay misclassifies routinely — either flagging a transient training-window insulin bump as chronic, or missing a chronic insulin elevation because the draw happened at a favorable post-session moment.

Dimension 5: Read against the anthropometric and body-composition overlay. Body mass index does not carry the insulin-resistance signal it does in the sedentary population because trained athletes hold BMI within a normal range even with substantial visceral or ectopic-fat accumulation. Waist circumference and waist-to-height ratio outperform BMI in the athlete population as insulin-resistance predictors; a waist-to-height ratio above 0.5 in a lean-appearing athlete flags visceral adiposity the BMI misses. Where DEXA is available, visceral adipose tissue (VAT) mass and the android-to-gynoid ratio surface the ectopic-fat pattern directly. The metabolic read paired with an ectopic-fat signal — even in an athlete with a normal BMI — shifts the case toward the intervention side.

Dimension 6: Read against the family and personal history overlay. First-degree relative with type 2 diabetes shifts the prior probability substantially and warrants the fasting insulin and HOMA-IR follow-on at every draw regardless of the fasting-glucose read. Gestational diabetes history (in female athletes) shifts the prior probability upward similarly. Personal history of PCOS in female athletes shifts insulin-resistance risk. Personal history of sleep apnea, non-alcoholic fatty liver disease, or hypertriglyceridemia at any point shifts the read. The masters athlete with a family history of type 2 diabetes and an unremarkable panel is the exact case where the workup should expand.

The four-quadrant clinical decision matrix

The six dimensions collapse into a four-quadrant matrix that drives the action plan.

Quadrant 1: Fasting glucose normal, fasting insulin under 8 mIU/L, HbA1c under 5.4%, CGM (if available) shows time-in-range above 85% and mean-amplitude-of-glycemic-excursion under 40 mg/dL, no family or personal-history flags, waist-to-height ratio under 0.5. The metabolic picture is genuinely clean. Document the integrated read; monitor at annual cadence.

Quadrant 2: Fasting glucose normal, fasting insulin 8-15 mIU/L or HOMA-IR 1.0-2.0, HbA1c under 5.7%, CGM shows time-in-range 70-85% or intermittent post-meal excursions above 140 mg/dL, family or personal history positive, or waist-to-height ratio above 0.5. The pattern is the early-compensatory case the standard panel misses. Confirm with a repeat draw at a matched training-and-fueling context. Address through the dietetic plan (carbohydrate quality and timing, meal composition, training-adjacent fueling, sleep, alcohol), initiate CGM tracking if not already in place, recheck at 12-16 weeks with fasting insulin and HbA1c.

Quadrant 3: Fasting glucose 100-125 mg/dL, or fasting insulin above 15 mIU/L or HOMA-IR above 2.0, or HbA1c 5.7-6.4%, or CGM shows time-in-range under 70% or frequent post-meal excursions above 180 mg/dL. The pattern is consistent with pre-diabetic insulin resistance. Order C-peptide to confirm the endogenous-insulin-driven story where relevant. Address through the dietetic plan aggressively, coordinate with the primary-care team, recheck at 12 weeks, escalate to medical or endocrinology if the integrated workup does not move the markers.

Quadrant 4: Fasting glucose above 126 mg/dL on a confirmed draw, or HbA1c above 6.4%, or symptomatic hyperglycemia, or the C-peptide follow-on suggests a non-type-2 mechanism. Medical or endocrinology referral is the primary action, not the dietetic intervention. The sports-RD role is the co-management of the fueling plan within the medical framework.

The C-peptide follow-on and the LADA differential

Most insulin-resistance cases in the athlete population sit on the type 2 spectrum, and the C-peptide read confirms the hyperinsulinemic story. A subset of adult-onset diabetes cases — Latent Autoimmune Diabetes of Adults (LADA) — presents with a slower autoimmune progression, an initially preserved endogenous insulin production that declines over months to years, and a clinical picture that can be mistaken for type 2 in the early years. The athlete presenting with rising glucose excursions, a normal or low-normal BMI, a family history of type 1 rather than (or in addition to) type 2, and a C-peptide that reads lower than the hyperinsulinemic case would predict is the presentation that warrants the medical workup for LADA. The GAD-65 antibody, IA-2 antibody, and zinc-transporter-8 antibody panel is the diagnostic follow-on, ordered by the medical team. The RD role in the LADA-suspicion case is early referral, not extended dietetic trial.

When to refer to medical

Five signals warrant medical referral beyond the dietetic workup.

Fasting glucose at or above 126 mg/dL on a confirmed draw. Diabetes diagnostic threshold; medical workup and management.

HbA1c at or above 6.5% on a confirmed draw. Diabetes diagnostic threshold; medical workup and management.

C-peptide that reads inappropriately low for the glucose or clinical picture, or a family history of type 1 diabetes. LADA workup; endocrinology referral.

Symptomatic hyperglycemia (polyuria, polydipsia, weight loss, blurred vision, recurrent infection). Urgent medical evaluation regardless of the lab pattern.

The Quadrant 3 pattern that does not respond to a 12-16 week dietetic intervention. Medical re-evaluation and consideration of pharmacologic support.

Common counseling mistakes

Reading fasting glucose in isolation and calling the case metabolically clean. The compensatory pattern where insulin is doing the work to hold fasting glucose in range is the exact case the standard reading misses. The workup that stops at fasting glucose stops at the wrong variable.

Reading HbA1c in isolation in the trained athlete. The chronically depressed nocturnal and inter-meal glucose in the trained state pulls HbA1c down and hides the postprandial-excursion pattern. HbA1c alone in the athlete under-flags the compensatory case.

Failing to order fasting insulin in the family-history or the CGM-signal case. Fasting insulin is the sensitive early marker; not ordering it collapses the workup.

Reading the CGM without the training-and-fueling context. A post-ride glucose spike into the 180s in an athlete who just finished a two-hour hard effort with 60 g of carbohydrate mid-ride is not the same signal as the same spike after a resting-state 40 g oatmeal breakfast. The CGM read has to sit against the training day and the fueling context.

Prescribing a low-carbohydrate diet reflexively for the insulin-resistant endurance athlete. Carbohydrate restriction lowers postprandial glucose in the short term but undercuts training capacity, glycogen availability, and long-term performance for the endurance athlete on the high-volume block. The intervention has to prioritize carbohydrate quality, timing (training-adjacent fueling), meal composition (protein and fat pairing to blunt the excursion), and volume calibrated to the training load — not a blanket carbohydrate reduction the athlete's training does not tolerate.

Failing to capture waist-to-height ratio or visceral-adiposity context in the lean-appearing athlete. The normal BMI in the trained athlete masks visceral adiposity in a subset of cases. The waist-to-height and (where available) DEXA-VAT reads catch this.

Ignoring sleep and alcohol as insulin-sensitivity levers. Chronic sleep restriction and habitual alcohol intake meaningfully shift insulin sensitivity down. The intake that treats the metabolic pattern as a pure dietary problem and does not surface sleep and alcohol misses two of the largest levers on the case.

Delaying referral on the Quadrant 4 case. Overt diabetes in the athlete is a medical diagnosis; the dietetic plan runs in coordination with medical management, not in place of it.

Where this lands in the SOAP

Subjective section format:

```

Insulin Resistance Workup (panel reviewed YYYY-MM-DD):

• Fasting glucose: [X mg/dL]
• Fasting insulin: [X mIU/L or NOT ORDERED — flag]
• HOMA-IR: [calculated: FG x FI / 405]
• HbA1c: [X %]
• C-peptide: [X ng/mL or NOT ORDERED]
• Lipid panel — triglycerides, HDL-C, triglyceride/HDL ratio: [values]
• Liver-function panel (AST, ALT, GGT): [values or NOT ORDERED]
• Hs-CRP: [X mg/L]

CGM context (if available):

• Days of data reviewed: [X]
• 14-day mean glucose: [X mg/dL]
• Time-in-range 70-140 mg/dL: [X %]
• Mean amplitude of glycemic excursion: [X mg/dL]
• Daily excursions above 140 / 180 mg/dL: [X / X]
• Notable meal-response patterns: [narrative]

Athletic-physiology and dietary context:

• Training state: [in-season / base / build / taper / off-season]
• Training-load trajectory past 12 weeks: [description]
• Draw timing: [hours since last hard effort, fueling day context, fasted status]
• Typical fueling pattern: [carbohydrate g/day, protein g/kg, fat pattern]
• Meal composition and timing habits: [narrative]
• Training-adjacent fueling: [pre / during / post pattern]
• Sleep pattern (past 4 weeks): [avg hours, consistency]
• Alcohol pattern: [servings per week, timing]

Anthropometric context:

• Height / weight / BMI: [values]
• Waist circumference / waist-to-height ratio: [values]
• Body composition (DEXA, if available): [VAT mass, android/gynoid, lean mass]

History:

• Family history: [T2D, T1D, GDM in first-degree relatives — details]
• Personal history: [GDM, PCOS, NAFLD, sleep apnea, hypertriglyceridemia]

Quadrant: [1-4 from clinical matrix]

Clinical impression: [statement integrating glucose with fasting insulin/HOMA-IR, HbA1c with CGM pattern, training and fueling state, anthropometric overlay, family and personal history]

Action: [fasting insulin / HOMA-IR order / C-peptide follow-on / CGM initiation / dietetic intervention / medical referral / observation]

Follow-up: [recheck date, plan, escalation triggers]

```

Assessment integrates the glucose and HbA1c reads with the insulin-driven compensatory picture, the CGM signal, the training-and-fueling context, the anthropometric overlay, and the family-and-personal-history stratification. Plan documents the differential, the dietetic intervention, the medical-coordination communications, and the recheck cadence. See [SOAP notes for sports dietitians](/blog/soap-notes-for-sports-dietitians) for the broader documentation framework.

The dietetic intervention in the compensating athlete

The Quadrant 2 and Quadrant 3 athletic case does not respond to blanket carbohydrate restriction and should not be treated as a low-carb-first diagnosis. Four levers with more evidence in the endurance-trained population:

Carbohydrate quality. Shift toward whole-grain, higher-fiber, and lower-glycemic-load carbohydrate sources at the meals that sit outside the training window. Reserve the higher-glycemic sources (sports drinks, gels, refined carbohydrate) for the training-adjacent window where insulin sensitivity is elevated and rapid delivery is the point.

Meal composition. Protein and fat paired with carbohydrate blunt the postprandial excursion meaningfully. The 40 g oatmeal breakfast that peaks at 158 mg/dL as a solo carbohydrate meal reads differently when paired with 30 g protein and 15 g fat.

Training-adjacent fueling. Concentrate carbohydrate intake in the 3-hour window bracketing hard sessions when insulin sensitivity is elevated. The athlete on a 15-hour training week can maintain 5-8 g/kg carbohydrate intake with post-meal excursions well controlled if the intake is aligned to the training window rather than spread evenly across rest days.

Sleep, alcohol, and stress management. The dietetic plan that treats these as separable is missing three of the largest levers on the metabolic pattern. Screen at every visit; document the trajectory.

The intervention that moves the fasting insulin and the CGM time-in-range in the compensating endurance athlete typically shows a signal at 8-12 weeks in the fasting insulin, at 12-16 weeks in the HbA1c, and in the CGM within 2-3 weeks of the meal-composition and timing changes.

The female-athlete insulin-resistance thread

Female athletes carry an insulin-resistance risk profile that shifts across the menstrual cycle (luteal-phase insulin sensitivity runs lower than follicular in most cases), through pregnancy (gestational-diabetes history is a substantial future-risk marker), and across the perimenopausal and postmenopausal transitions (insulin sensitivity typically declines). PCOS is a distinct insulin-resistance risk factor with substantial workup and management implications. The integrated female-athlete metabolic read should sit against the [menstrual-cycle intake workup](/blog/menstrual-cycle-charting-in-female-athlete-intake), the [iron-status workup](/blog/iron-status-workup-in-female-athletes), and the [thyroid workup](/blog/thyroid-screening-sports-nutrition-intake-workup).

The masters-athlete trajectory

Insulin sensitivity declines with age independent of body composition and training. The masters athlete with a family-history flag and a decade of clean panels is exactly the case where the annual fasting-insulin and HOMA-IR add-on catches the shift before the fasting-glucose flag arrives. The trajectory over 5-10 years is the informative variable; capture the historical panels where available.

Where platform tooling helps

The bottleneck in the insulin-resistance workup at scale is the multi-system integration — the panel value read against the fasting insulin and HOMA-IR, the HbA1c read against the CGM signal, the C-peptide follow-on when the picture is ambiguous, the training-and-fueling-state overlay, the anthropometric and family-history stratification, the recheck trajectory across years. The intake that runs the integration by hand drops it on busy weeks; the compensating athlete gets passed as "metabolically clean" on a 92 mg/dL fasting glucose and surfaces months later either with a fasting insulin the workup should have caught, or with a family-history case whose annual add-on never happened.

The leverage is an insulin-resistance workup module that ingests fasting glucose, prompts for fasting insulin and calculates HOMA-IR automatically, tracks HbA1c against CGM signal where available, prompts for the C-peptide follow-on on the ambiguous case, captures the training-and-fueling context, integrates the anthropometric and family-history stratification, surfaces the four-quadrant decision matrix, pre-populates the SOAP documentation, and tracks the medical-coordination communications when referral is warranted. The RD's job is the clinical judgment and the conversation, not the spreadsheet.

The chart trail is defensible — every interpretation tied to the integration context that justified it, every referral documented with the pattern that drove it, every dietetic intervention paired with the differential reasoning behind it.

The bottom line

Insulin resistance in athletes is a multi-axis read, not a single-number screen. Fasting glucose is late in the natural history of the disease; HbA1c is dragged down by the trained-state chronic low-glycemia baseline; fasting insulin, HOMA-IR, and the CGM signal are the variables that surface the compensating case early. The trained endurance athlete with a family-history flag, a normal BMI, a chronically depressed fasting glucose, and a rising fasting insulin can sit in the pre-diabetic compensatory window for years while the standard annual bloodwork reads as clean.

The workup that catches the real cases reads fasting glucose against fasting insulin and HOMA-IR, reads HbA1c against the CGM or oral-glucose-tolerance postprandial-excursion signal, orders the C-peptide follow-on when the picture is ambiguous or the family history suggests LADA, integrates the training-and-fueling-state overlay, captures the anthropometric and family-history stratification, and coordinates with the medical team on the cases that warrant it.

The 41-year-old masters cyclist with the 92 mg/dL fasting glucose, the 5.3% HbA1c, the 14-16-hour training week, the CGM peaks routinely in the 140s and 150s, the father-diagnosed-at-58 family history, and the mid-ride energy crashes is not the "metabolically healthy" case the standard panel produced. He is a Quadrant 2 or Quadrant 3 case pending the fasting insulin and HOMA-IR follow-on, and the intake that runs the structured workup catches the difference. The intake that reads the fasting glucose and the HbA1c and stops does not.

[Calsanova's Dietitian plan](/signup?role=dietitian) ships an insulin-resistance workup module with fasting-glucose ingestion, fasting-insulin and HOMA-IR prompts, HbA1c-against-CGM integration, C-peptide follow-on triggers, training-and-fueling-context capture, anthropometric and family-history stratification, four-quadrant decision-matrix surfacing, and pre-populated SOAP documentation with medical-coordination communications. Start your 30-day free trial and turn the metabolic panel from a single-number pass-fail into the clinical instrument that catches the compensating cases the standard read collapses.