What you eat and drink in the hours and days leading up to a glucose tolerance test can genuinely skew your results, sometimes enough to trigger a false diagnosis of diabetes or gestational diabetes. The standard instruction is to fast for eight to twelve hours before the test, but fasting alone does not cover everything. Your dietary pattern in the preceding days, your caffeine intake that morning, whether you smoked a cigarette in the parking lot, and even how much water you drank all influence how your body handles the glucose load. Getting a clean, accurate result means paying attention to more than just skipping breakfast.
Why the Fasting Window Matters
Most oral glucose tolerance tests (OGTTs) require an overnight fast, typically eight to twelve hours with nothing but water. The reason is straightforward: food raises blood sugar, and the test needs a reliable baseline before you drink the glucose solution. Research on fasting duration shows that blood glucose drops only modestly with each additional hour of not eating, roughly 0.4 mg/dL per hour in men and 0.2 mg/dL per hour in women, and the difference between someone who fasted three hours and someone who fasted overnight is small in absolute terms.1SpringerOpen / European Journal of Epidemiology. Impact of time since last caloric intake on blood glucose levels But even a small bump in your fasting baseline can push borderline numbers across a diagnostic threshold. That is why clinicians insist on the full overnight fast rather than settling for a few hours.
During the fasting window, water is fine and even encouraged. Anything with calories, sugar, or ingredients that stimulate insulin, including gum, mints, and flavored water, should be avoided. The goal is metabolic quiet: your body should be in a steady resting state when the clock starts.
The Low-Carb and Keto Trap
This is where the standard advice falls short for a growing number of people. If you have been eating a low-carbohydrate or ketogenic diet, your body adapts by becoming less efficient at clearing a big hit of glucose. When you suddenly drink the sugary solution during the test, your blood sugar spikes higher than it would if you had been eating carbs regularly. The result can look like prediabetes or diabetes when your day-to-day blood sugar is actually well controlled.
A randomized feeding trial published in Diabetes Care found that after switching from a low-carb diet to a higher-carb diet, glucose tolerance kept improving for weeks, not just a few days. Fasting glucose and the two-hour glucose reading both showed downward trends from week two all the way through week nine. The researchers specifically noted that the commonly recommended three-day carb-loading preparation, eating at least 150 grams of carbohydrates daily for three days before the test, may not be enough to avoid false positives in people who habitually eat low-carb.2Diabetes Care. Prolonged Glycemic Adaptation Following Transition From a Low- to High-Carbohydrate Diet: A Randomized Controlled Feeding Trial
A separate case report in the Journal of the Endocrine Society reinforced this concern, warning that when an OGTT is given without full understanding of the patient’s dietary preparation, it can lead to false-positive results, patient distress, and potentially a misdiagnosis of diabetes.3Journal of the Endocrine Society. Carbohydrate Intake Prior to Oral Glucose Tolerance Testing If you eat fewer than about 150 grams of carbs a day, tell your provider before the test. You may need a longer preparation period or an alternative screening method like hemoglobin A1c.
High-Fat Foods in the Days Before
A greasy dinner the night before your test is not just a fasting-window problem. Research on young healthy men showed that even five days of a high-fat diet, where about 60 percent of calories came from fat, was enough to increase fasting glucose levels. The mechanism was increased glucose output from the liver, even though the body’s ability to respond to insulin in muscle tissue remained intact.4PubMed Central. Impact of short-term high-fat feeding on glucose and insulin metabolism in young healthy men One particularly fatty meal the evening before is unlikely to have the same impact as five days of overfeeding, but if your diet in the lead-up to the test has been heavy on fried foods, fast food, or rich restaurant meals, your fasting number could be slightly inflated.
The practical takeaway is to eat normally in the days before the test, meaning something close to your usual balanced diet, rather than going out for an indulgent meal or celebrating the night before. If your usual diet is already high in fat, the test should still reflect your real metabolic state. The distortion comes from short-term dietary swings, not from your habitual pattern.
Skip the Coffee
This catches a lot of people off guard because black coffee has no calories and feels harmless during a fast. But caffeine itself interferes with how your body handles glucose. A study in obese men found that caffeine ingestion before an OGTT decreased insulin sensitivity by about 15 percent and produced a significantly greater insulin response, a pattern the researchers described as highly suggestive of insulin resistance.5The American Journal of Clinical Nutrition. Caffeine ingestion increases the insulin response to an oral-glucose-tolerance test in obese men before and after weight loss Similar effects have been documented in lean individuals. The caffeine does not spike your blood sugar directly. Instead, it makes your cells temporarily less responsive to insulin, which forces your body to produce more insulin to clear the same amount of glucose. That altered insulin response is part of what the test measures.
This applies to coffee, tea, energy drinks, and caffeine pills. Even decaf has some caffeine, though the amount is usually small enough that it is unlikely to matter. The safest approach is to stick to plain water from the time you stop eating until the test is done.
Diet Soda and Artificial Sweeteners
You might assume that a zero-calorie diet drink is safe during your fast, but artificial sweeteners have biological effects that go beyond their calorie count. A controlled study found that drinking diet soda before a glucose load significantly increased secretion of GLP-1, a gut hormone involved in insulin regulation, compared to plain carbonated water. Insulin levels also trended higher, though not to a statistically significant degree.6PubMed Central. Ingestion of diet soda before a glucose load augments glucagon-like peptide-1 secretion The artificial sweetener appears to activate sweet-taste receptors in the gut that help regulate how your body responds to incoming sugar.
Whether this would be enough to change a diagnostic outcome is not settled, but the point of fasting before a glucose test is to eliminate variables. Diet soda introduces one. Leave it out.
Protein and Large Meals
Eating a large, protein-heavy meal the night before your test is less commonly discussed than carbs or fat, but protein plays a meaningful role in blood sugar regulation. Protein ingestion stimulates both insulin and glucagon, two hormones that push glucose levels in opposite directions.7PubMed Central. Effects of protein intake on glucagon, insulin, and glucose dynamics: implications for diabetes Glucagon signals the liver to release stored glucose, while insulin tells cells to absorb it. In healthy people, these signals balance out. But the timing and size of a protein-heavy meal can influence how your liver behaves the following morning.
The standard advice to eat a “normal” dinner the night before is partly about protein. A moderate, balanced meal gives your body time to return to baseline before the fast truly begins. A very large steak dinner or a protein shake right at the start of your fasting window means your body is still processing those signals while you sleep.
Fiber Can Work Both Ways
High-fiber foods are generally considered beneficial for blood sugar control because viscous fibers, the kind found in oats, beans, and certain fruits, slow down the absorption of carbohydrates and flatten the glucose response after a meal.8PubMed Central. Dietary fiber and the glycemic response This is great for everyday health, but it becomes irrelevant once you are fasting because there is no food left to slow the absorption of. The fiber in your dinner the night before will be long past its glucose-flattening window by the time you drink the test solution the next morning.
Where fiber matters is if you eat a very high-fiber meal and then do not fast long enough. If you have a bowl of lentil soup at 10 PM and your test is at 7 AM, the slow-digesting fiber could still be affecting absorption. Eating dinner early enough to allow a genuine ten-to-twelve-hour fast sidesteps this concern.
Smoking Before the Test
If you smoke, do not light up on the morning of your glucose test. Smoking acutely impairs glucose tolerance. In a study that compared OGTTs done while smoking versus without smoking, the glucose area under the curve in smokers was roughly 17 percent higher when they smoked during the test compared to when they did not. Even in nonsmokers exposed to cigarettes during the test, glucose values rose slightly. Among habitual smokers, insulin sensitivity dropped significantly on the smoking day.9Diabetes Care. Acute Effect of Cigarette Smoking on Glucose Tolerance and Other Cardiovascular Risk Factors
Separate research in pregnant women found that those who smoked at the beginning of pregnancy had higher one-hour glucose readings on their OGTTs than nonsmokers, with the effect persisting as an independent predictor even after adjusting for other factors.10PubMed. Smoking affects the oral glucose tolerance test profile and the relationship between glucose and HbA1c in gestational diabetes mellitus Nicotine triggers the release of stress hormones that raise blood sugar and impair insulin’s effectiveness. The effect starts within minutes and can last for hours. Wait until after the test to smoke.
Stay Hydrated
Water is the one thing you should actively consume during your fast, and skipping it can actually hurt your results. A study in people with type 2 diabetes compared glucose responses during a glucose tolerance test under normal hydration versus mild dehydration. Dehydrated participants had significantly higher blood glucose readings at both baseline and the two-hour mark, with fasting glucose averaging about 10.4 mmol/L compared to 9.5 mmol/L when properly hydrated, and two-hour glucose reaching 21.0 versus 19.1 mmol/L.11PubMed Central. Reduced water intake deteriorates glucose regulation in patients with type 2 diabetes That study was in people who already had diabetes, so the absolute numbers are higher than most test-takers would see, but the direction of the effect applies broadly. Dehydration concentrates your blood and can nudge glucose readings upward. Drink water normally the evening before and have a glass or two in the morning.
Being Sick Changes Everything
If you are fighting off a cold, the flu, or any acute infection, your glucose test results will not reflect your normal metabolism. Acute infections create a state of temporary insulin resistance. A study of patients with bacterial and viral infections found that two-hour blood glucose values during OGTTs were significantly elevated during the acute illness, and the insulin resistance index increased by about 33 percent compared to when the same patients were well.12PubMed. Glucose intolerance in acute infections Research on interferon, a molecule the body produces during viral infections, confirmed the mechanism: interferon alone caused the glucose area under the curve during an OGTT to increase roughly 2.6-fold in healthy subjects.13Diabetes. Effect of Interferon on Glucose Tolerance and Insulin Sensitivity
This is not a subtle effect. If you are symptomatic with anything, from a respiratory infection to a stomach bug, ask your provider about rescheduling. Even mild illness activates inflammatory pathways that push glucose higher and make insulin work less effectively. Testing during an infection risks a diagnosis that does not reflect your actual metabolic health.
Medications and Supplements Worth Mentioning
Dozens of common medications can influence glucose metabolism, and the effects go in both directions. Corticosteroids like prednisone are well-known glucose-raisers. Thiazide diuretics, beta-blockers, and certain psychiatric medications can impair glucose tolerance. On the other side, some drugs can lower blood sugar unexpectedly. A review in Annals of Internal Medicine found that the direction of a drug’s effect on glucose depends on the patient’s pancreatic reserve, nutritional state, other medications, and even alcohol exposure.14Annals of Internal Medicine. Drug-induced disorders of glucose tolerance
You should not stop any prescribed medication without your doctor’s explicit instruction, but you should tell them everything you are taking, including over-the-counter supplements, before the test. Some providers will ask you to skip certain medications the morning of the test; others will want you to take them as usual and interpret the results accordingly. The important thing is that they know what is on board so they can account for it.
Sleep, Stress, and Test Timing
Poor sleep the night before your glucose test is not just unpleasant; it can measurably raise your blood sugar. A study in pregnant women found that each hour of reduced sleep was associated with a 4 percent increase in glucose levels on a screening glucose test, and women with sleep-disordered breathing had roughly three times the odds of developing gestational diabetes.15Europe PMC. Sleep disturbances and their relationship to glucose tolerance in pregnancy Sleep deprivation increases cortisol and other counter-regulatory hormones that oppose insulin’s action, and even one bad night can reduce insulin sensitivity the next day.
The time of day your test is scheduled also matters. Studies dating back decades have consistently shown that glucose tolerance is worse in the afternoon than in the morning. When the same people took an OGTT in the morning and then again on a separate afternoon, their blood sugar levels were significantly higher in the afternoon session.16BMJ. Diurnal Variation of Oral Glucose Tolerance: a Possible Pointer to the Evolution of Diabetes Mellitus17Br Med J. Diurnal Variation in Glucose Tolerance: Associated Changes in Plasma Insulin, Growth Hormone, and Non-esterified The afternoon impairment appears to be linked to a delayed insulin response. This is one reason glucose tolerance tests are almost always scheduled for the morning: your body handles glucose best early in the day, and the diagnostic thresholds were established using morning tests.
What a Good Pre-Test Day Actually Looks Like
Putting all of this together, the ideal preparation is less about memorizing a list of banned foods and more about avoiding disruptions to your normal metabolic state. In the three days before the test, eat your regular diet, making sure to include at least 150 grams of carbohydrates per day if you have been restricting carbs. Eat a moderate, balanced dinner the evening before, something unremarkable that includes some carbs, some protein, and is not excessively fatty. Finish eating at least ten hours before the test. Drink water normally through the evening and morning. Skip coffee, tea, diet soda, cigarettes, gum, and anything flavored. Get a reasonable night’s sleep. If you are sick, reschedule.
The glucose tolerance test is designed to measure how your body processes sugar under standardized conditions. Every variable you introduce, whether it is caffeine, dehydration, a week of keto eating, or a night of insomnia, adds noise to the measurement. Most of these variables push glucose readings higher, meaning the risk of false positives is real and worth guarding against. A clean test protects you from an inaccurate result that could lead to unnecessary treatment, extra monitoring, or anxiety about a condition you may not actually have.
Exercise on the Morning of the Test
You might think that a morning jog or brisk walk before the test would help lower your blood sugar, and in a sense it does. Exercise increases glucose uptake by your muscles and can improve insulin sensitivity for hours afterward. But that is exactly the problem: the test is meant to measure your body’s glucose handling under resting conditions, not in the afterglow of a workout. A vigorous session before the test could artificially improve your results, potentially masking a real problem with glucose tolerance. Most clinical guidelines recommend avoiding strenuous exercise for at least 24 hours before the test and arriving at the clinic in a rested state. Walking to your appointment is fine. Running five miles beforehand is not.