A patient sat down with me last spring and told me she had been "low-carb for years" and could not understand why her HbA1c was 6.0 and her fasting insulin was 22. We walked through what she actually ate in a typical week. The food was real food — no fast food, no obvious junk. It included a daily oat-milk latte, a "high-protein" granola bar at midmorning, a wrap for lunch, sweet potato most nights, and a piece of dark chocolate after dinner. She thought she was eating maybe 60 grams of carbs a day. The honest count was somewhere between 180 and 220.
She was not lying to me, and she had not been lying to herself. She was using a vocabulary she had picked up from social media — "low-carb," "clean," "macro-friendly" — that did not match the actual carbohydrate load of the food. The metabolic numbers were responding to the food, not to the vocabulary. That is the audit this article is about.
Why this gap exists between what patients eat and what they think they eat
I have been doing this long enough — between emergency medicine and now metabolic care — to have developed a healthy skepticism of the entire low-carb conversation as it filters through general marketing. The label "low-carb" has been applied to so many foods that it has lost almost all meaning. Products with 18 grams of carbs per serving, marketed as keto-friendly, sit on the same shelf as products with 3 grams. Restaurant meals built around "lean protein" are served on a bed of rice or with a sauce sweetened with corn syrup. The patient who is genuinely trying to manage carbs is doing so against a food environment designed to make accurate counting almost impossible.
The other reason the gap exists: the body's metabolic response to carbohydrate does not care about the label. Insulin responds to the actual glucose load arriving in the bloodstream. A patient who is "doing low-carb" but consistently spiking insulin three to five times a day is producing the same metabolic signature as a patient who is not pretending. The body keeps a more accurate record than the patient does.
The categories where carbs hide most reliably
When I do a real diet audit with a patient — and this is part of every initial visit in our medical weight loss program — these are the categories that consistently turn out to carry more carbs than the patient was tracking:
Beverages. This is the single most under-counted category, by a wide margin. A 16-ounce flavored latte at a chain coffee shop runs 35-50 grams of carbohydrate. A "skinny" version still runs 15-25. Kombucha runs 8-12 grams per bottle, and most patients drink the whole bottle and call it a serving. Sweetened iced tea — especially in middle Georgia, where unsweetened by default is not the cultural norm — easily contributes 60-80 grams across a day. Sports drinks during a workout add 20-40 grams that the patient is convinced "do not count" because they are exercising.
Sauces, dressings, and condiments. Barbecue sauce runs 6-10 grams per tablespoon. Ketchup, 4. Honey mustard, 5-7. Teriyaki, 8-10. Most commercial salad dressings carry sugar. The patient who orders a salad with chicken and feels virtuous can easily layer 20 grams of carbs on top from the dressing alone.
Bars, snacks, and "high-protein" packaged foods. This category is the worst offender for patients who think of themselves as low-carb. The marketing emphasizes the protein number; the carbohydrate number is in smaller font on the back. A typical "high-protein" granola bar carries 22-30 grams of carb against 12-15 grams of protein. Protein cereal runs 25-35 grams of carb per serving. Patients who eat two or three of these per day are consuming 60-90 grams of carb they are not counting.
Fruit treated as unlimited. A medium banana is 27 grams. A cup of grapes is 27. A medium apple is 25. An "acai bowl" with toppings can run 80-100 grams. Fruit is real food and has its place, but it is not a free space metabolically, and patients who use it as their default snack frequently produce the insulin pattern of someone eating sugar.
Starchy vegetables that read as "vegetables." Sweet potato, butternut squash, corn, peas, beets — these get categorized in the patient's mental model as vegetables and treated as carb-neutral. They are not. A medium sweet potato is about 26 grams. A cup of corn is 27.
Restaurant rice and tortillas. A restaurant serving of rice is rarely the half-cup the menu describes. A burrito wrap can carry 45-55 grams of carb in the wrap alone. Sushi rice in a typical roll runs 35-45 grams.
Alcohol. Beer, wine, mixed drinks. Wine runs 3-5 grams per glass; most patients are pouring six-ounce glasses and counting them as four. Mixed drinks with juice or syrups run 15-30 grams. The metabolic effect of alcohol on hepatic glucose handling is on top of the direct carb load — the body processes the alcohol first, which means glucose disposal stalls during that window.
"Sugar-free" and "low-sugar" products that contain other carbohydrates. Sugar alcohols, maltodextrin, oligofructose. The package implies "no metabolic effect"; the bloodstream disagrees. Read the total carbohydrate line, not just the sugar line.
The mechanism: why this matters more than total caloric intake
The conventional weight-loss model treats every calorie as equivalent. Mid-life patients with insulin resistance live in a different physiology, and the carbohydrate composition of the diet matters in a way that pure caloric arithmetic does not capture.
Not sure where to start?
The Start Here pathway walks you through the most common entry points and helps you decide which consultation type is the right fit. Five minutes of self-assessment can save you a wrong-direction conversation.
Here is the mechanism. Insulin's primary job is glucose disposal — moving glucose out of the bloodstream into muscle and liver cells. When muscle and liver cells become less responsive to insulin (the definition of insulin resistance), the pancreas compensates by producing more insulin. Higher circulating insulin does several things: it strongly suppresses lipolysis (fat breakdown), it promotes lipogenesis (fat storage), and it interferes with the glucagon signaling that should mobilize stored fat between meals. The net effect: the patient is locked into fat-storage mode regardless of caloric intake, and the only relief from that signaling is meaningful periods between glucose spikes.
A diet that delivers 200 grams of carbohydrate in five or six small spikes across the day keeps insulin elevated for the majority of the patient's waking hours. The same caloric load, delivered with a different macronutrient profile and fewer eating windows, produces dramatically different insulin behavior. This is why patients who restrict calories without addressing carbohydrate composition often plateau or regain. The signaling environment is wrong, not the math.
Patients with mid-life hormonal shifts — declining estrogen and testosterone — have additional pressure on this system because both hormones contribute to insulin sensitivity. A 52-year-old woman post-menopause has a fundamentally different metabolic environment than the same woman did at 35. The carbohydrate load that was sustainable at 35 is the carbohydrate load that drives insulin resistance at 52. This is part of why hormone optimization and metabolic intervention work better in parallel than either does alone.
How I actually audit a patient's diet in clinic
When I do this with a new patient, I do not ask for a perfect log. I ask for an honest one — three normal weekdays and one weekend day, including the things they would not write down for a stranger. We sit with the list and I ask three questions.
First: where are the carbs the patient did not count? We walk through the categories above and find them. The most common surprise is the beverage category and the bar category.
Second: what is the actual carbohydrate load on a typical day? Not the goal. Not the target. The actual number. For most patients I work with, the gap between the perceived load and the actual load is somewhere between 60 and 150 grams.
Third: what is the eating window pattern? How many distinct insulin-triggering events across the day? A patient who eats six times a day at 30 grams of carb each is producing a very different insulin curve than a patient who eats three times a day at 60 grams each, even though the totals are identical.
The audit is not about restriction for its own sake. It is about producing a diet whose actual composition matches the metabolic outcome the patient is trying to achieve.
How this fits with the broader weight loss plan
The diet audit is one piece of the workup, not the entire intervention. The full medical weight loss program at our Columbus clinic and Warner Robins clinic starts with the comprehensive metabolic and hormonal panel — fasting insulin, HbA1c, fasting glucose, comprehensive metabolic panel, lipid panel including ApoB, full sex hormone panel, full thyroid panel — and the diet audit runs in parallel with the lab review.
For patients whose labs and physiology indicate that GLP-1 therapy is appropriate, the medication addresses one part of the picture. The diet audit addresses another. Both matter. A patient on semaglutide whose food choices remain heavily carbohydrate-loaded will lose weight more slowly than a patient whose composition has been adjusted, and will be more likely to regain after the medication is reduced or stopped. The structured nutritional counseling component of the program is built around producing a sustainable composition, not a temporary restriction.
For patients whose hormonal picture is also driving the metabolic resistance, hormone optimization is layered in based on what the labs show. The combination of corrected hormones, sensible carbohydrate composition, and — where appropriate — GLP-1 support produces outcomes that none of the three produce alone in this patient population.
The next step
If you have been "doing low-carb" and the metabolic numbers are not moving the way you expected, the audit is the place to start. Book a medical weight loss consultation at either the Columbus clinic or Warner Robins clinic. Before the visit, write down everything you ate and drank for three normal weekdays and one weekend day, including the small things — the splash of milk, the dressing, the after-dinner square of chocolate. Bring whatever recent lab work you have. If you do not have recent comprehensive labs, we order them at the first visit.
The first visit is the workup and the audit. The second is the plan, with both the lab data and the actual food data in front of us. From there the work is straightforward — and the carb conversation will be grounded in numbers, not vocabulary. If you want to walk through the basics before you book, the weight loss assessment is a useful first pass.
The body keeps the record. The job is to make sure the record matches what the patient actually wants it to say.
Medical disclaimer: This article is for educational purposes only and does not constitute medical advice. Individual clinical decisions should be made in consultation with a qualified healthcare provider following appropriate evaluation. References to specific treatments, dosing, or protocols are informational.
Travis spent 17+ years in high-acuity clinical medicine — emergency, cardiac ICU, and cath lab — before founding Revitalize. He is a Certified Platinum Biote hormone therapy provider, the published author of You're Not Broken — You're Unbalanced, and the founder of the Rebuild Metabolic Health Institute. His clinical writing reflects the same precision he brought to critical care: specific, honest, and built around what actually works.
Ready to talk it through with a clinician?
Book online or call either Georgia location. Every visit starts with a consultation.