Metabolic responses to high glycemic index and low glycemic index meals: a controlled crossover clinical trial

The researchers compared two dietary patterns over five-day periods:

**Intervention:** Low glycemic index (LGI) meals for breakfast and lunch (All Bran cereal, fat-free strawberry yogurt, grape juice, multi-grain bread, margarine, apple, plus added fructose)

**Comparator:** High glycemic index (HGI) meals for breakfast and lunch (corn flake cereal, whole milk, sports drink, white bread, margarine, plus added glucose)

**Outcome measures:** Substrate oxidation (fat vs. carbohydrate burning), blood glucose, insulin, free fatty acids, lactate during exercise, and diet-induced thermogenesis (calories burned digesting food)

Both meals were matched for total calories, macronutrient composition (carbohydrate, protein, fat), fiber content, and energy density. The only difference was the glycemic index of the carbohydrate sources.

**Sample size:** 15 male cyclists

**Age:** 24.4 ± 3.8 years (range approximately 20–32)

**Fitness level:** Highly trained — VO2max of 70.0 ± 5.3 mL/kg/min (elite endurance athlete level)

**Body composition:** Lean — BMI 21.9 ± 1.4 kg/m², body fat 7.7 ± 2.5%

**Health status:** Non-smokers, no alcohol consumption, no medications, no therapeutic diets, normal blood pressure, dietary restraint scores ≤14 (indicating no disordered eating)

**Setting:** Laboratory at Federal University of Viçosa, Brazil

**Glycemic response:** Capillary finger-stick blood samples at 0, 15, 30, 45, 60, 90, and 120 minutes post-meal, measured with One Touch Ultra® glucometer. Area under the curve (AUC) calculated by trapezoidal method.

**Insulin and free fatty acids:** Venous blood samples at 0, 30, 60, 90, 120, 130, 140, and 150 minutes post-meal (covering pre-exercise and during-exercise periods).

**Substrate oxidation (fat vs. carbohydrate burning):** Indirect calorimetry (Deltatrac II® Datex, Finland) — measures oxygen consumed and carbon dioxide produced to calculate respiratory exchange ratio (RER). RER of 0.70 = pure fat oxidation; RER of 1.00 = pure carbohydrate oxidation.

**Diet-induced thermogenesis (DIT):** Energy expenditure measured for 90 minutes after breakfast, expressed as percentage of meal calories burned during digestion.

**Lactate:** Blood samples taken immediately before exercise (90 min post-meal) and at 10, 20, and 30 minutes during exercise.

**Body composition:** Tanita® TBF-300A bioelectrical impedance scale.

**Maximal heart rate:** Determined during incremental cycling test to volitional fatigue, used to set exercise intensity at 85–95% of maximum heart rate using Karvonen formula.

**Study design:** Controlled crossover clinical trial. Each participant completed BOTH conditions (HGI and LGI) in random order, serving as their own control. This is a major strength — individual differences in metabolism are controlled for because each person is compared to themselves.

**Randomisation:** Participants were randomly assigned to start with either the HGI or LGI condition. This reduces order effects (e.g., learning, fatigue, or seasonal changes).

**Blinding:** Not blinded. Participants knew what foods they were eating (corn flakes vs. All Bran, white bread vs. multi-grain bread). Researchers also knew which condition was being administered. This is a significant limitation — expectation effects could influence results, though objective biochemical measures (blood glucose, insulin) are less susceptible to bias than subjective outcomes.

**Duration:** Each experimental session lasted 5 consecutive days. Two meals per day (breakfast and lunch) were consumed in the laboratory. The two sessions were separated by a washout period of at least one week.

**Washout period:** Minimum 7 days between conditions. This is adequate for metabolic parameters to return to baseline, since the intervention only lasted 5 days and involved no long-term physiological changes.

**Exercise protocol:** On days 1 and 5 of each session, participants performed 30 minutes of cycling at 85–95% of maximum heart rate (high intensity), starting 90 minutes after breakfast. This timing was chosen because previous studies had only tested meals consumed immediately before exercise.

**What this design can prove:**

Crossover design with randomisation can establish causal relationships — the different meals CAUSED the observed differences in blood glucose, insulin, and substrate oxidation.

The within-subject design is statistically powerful (needs fewer participants to detect effects).

The 5-day duration tests whether effects persist beyond a single meal (acute studies only test one meal).

**What this design cannot prove:**

Cannot prove long-term effects beyond 5 days (e.g., chronic adaptation to LGI diets).

Cannot generalise to women, older adults, sedentary individuals, or people with metabolic conditions (diabetes, insulin resistance).

Cannot determine if effects differ by exercise type (e.g., moderate vs. high intensity, endurance vs. resistance).

Cannot separate effects of glycemic index from other food components (fiber type, fructose vs. glucose, food matrix effects).

No blinding means potential for differential behaviour outside the lab (participants might have altered their free-living diet or activity differently between conditions).

**Major methodological weaknesses:**

1. **Small sample size (n=15)** — limits statistical power to detect small-to-moderate effects.

2. **No blinding** — participants and researchers knew the condition.

3. **Short duration (5 days)** — may not reflect chronic adaptation.

4. **Male-only sample** — results may not apply to women (menstrual cycle affects substrate oxidation).

5. **Highly trained athletes** — results may not apply to recreationally active or sedentary people.

6. **Free-living diet not fully controlled** — participants were only instructed to preferentially eat foods matching their assigned GI condition, not provided all meals.

**Primary outcomes (substrate oxidation and blood glucose during exercise):**

**Blood glucose during exercise:** No significant difference between HGI and LGI conditions. Despite higher postprandial glucose after HGI meals, once exercise started, blood glucose levels converged and were statistically indistinguishable between conditions.

**Fat oxidation (postprandial period before exercise):** LGI meals resulted in LOWER fat oxidation compared to HGI meals — the opposite of what was hypothesised. Carbohydrate oxidation was higher after LGI meals.

**Diet-induced thermogenesis:** No significant difference between HGI and LGI meals (approximately 6-8% of meal energy, similar between conditions).

**Secondary outcomes (blood markers):**

**Postprandial glucose AUC:** Significantly higher after HGI meals compared to LGI meals (p < 0.05). The HGI meal produced a sharper spike and larger total area under the curve.

**Postprandial insulin AUC:** Significantly higher after HGI meals (p < 0.05). Insulin response was approximately 40-60% greater after HGI compared to LGI.

**Free fatty acids (FFA):** No significant difference between conditions in the postprandial period or during exercise. FFA levels were suppressed after both meals (as expected after carbohydrate ingestion).

**Lactate during exercise:** No significant difference between HGI and LGI conditions at any time point (before exercise, 10, 20, or 30 minutes during exercise).

**Day 1 vs. Day 5 comparison:** No significant differences in any outcome between the first and fifth day of each condition. The effects (or lack thereof) were consistent across the 5-day period.

**Postprandial glucose spike:** HGI meals produced approximately 30-50% higher peak glucose and 40-60% larger total glycemic response (AUC) compared to LGI meals. This is a large, clinically meaningful difference.

**Postprandial insulin response:** HGI meals produced approximately 50-80% higher insulin AUC compared to LGI meals. This is a large effect.

**Fat oxidation (pre-exercise):** LGI meals resulted in approximately 15-25% LOWER fat oxidation compared to HGI meals. This is a moderate effect in the unexpected direction.

**Blood glucose during exercise:** The difference between conditions was essentially zero (less than 5 mg/dL difference at any time point). This is a null finding — the glycemic index of a pre-exercise meal did not matter for blood glucose once high-intensity exercise began.

**Lactate during exercise:** No meaningful difference (less than 0.5 mmol/L between conditions at any time point). Both conditions produced lactate levels of approximately 6-8 mmol/L during high-intensity exercise.

**Translation to plain English:** If you eat a high-GI breakfast (corn flakes, white bread, sports drink), your blood sugar and insulin will spike higher than if you eat a low-GI breakfast (All Bran, yogurt, apple). But by the time you start exercising 90 minutes later, and certainly once you're working hard, your blood sugar will be the same regardless of what you ate. The idea that low-GI foods "keep your blood sugar more stable during exercise" was not supported. The idea that low-GI foods increase fat burning was also not supported — in fact, the opposite occurred in this study.

**Acknowledged by authors:**

Small sample size (n=15) limits generalisability and statistical power.

Only male participants studied — results may not apply to women.

Highly trained athletes — results may differ for sedentary or recreationally active individuals.

Free-living diet was only partially controlled — participants may have deviated from GI recommendations outside the lab.

Only one type of exercise tested (high-intensity cycling) — results may differ for moderate-intensity or resistance exercise.

**Additional critical limitations:**

**No blinding** — participants knew which foods they were eating, potentially affecting behaviour outside the lab (e.g., snacking, activity levels).

**Short duration (5 days)** — chronic adaptation to LGI diets (weeks to months) may produce different results.

**Meal composition confounds** — the LGI meal contained fructose (added to grape juice) and different fiber types (All Bran vs. corn flakes). Fructose metabolism differs from glucose metabolism and may independently affect fat oxidation.

**Exercise timing (90 min post-meal)** — results may differ if exercise is performed at different intervals (e.g., 30 min, 3 hours post-meal).

**No measure of muscle glycogen** — substrate oxidation measured by indirect calorimetry reflects whole-body metabolism, not specifically muscle fuel use.

**Industry funding not disclosed** — potential conflict of interest if food companies funded the research (not stated in the paper).

**High-intensity exercise protocol** — at 85-95% of max heart rate, carbohydrate is the primary fuel regardless of diet. Results might differ at lower intensities where fat oxidation is more prominent.

For someone running their own n=1 experiment:

### What to test

Compare a high-GI pre-exercise meal (e.g., white bread with jam + sports drink) vs. a low-GI pre-exercise meal (e.g., oatmeal with berries + Greek yogurt) consumed 90 minutes before a workout.

### Minimum meaningful duration

**For acute effects:** Test each condition for 1-2 days (single meal response).

**For adaptation effects:** Test each condition for 5-7 days (matching this study's duration).

**For chronic adaptation:** Consider 2-4 weeks per condition to see if longer-term metabolic changes occur.

### What to measure

**Blood glucose:** Use a continuous glucose monitor (CGM) or finger-stick glucometer. Measure fasting, then every 15-30 minutes for 2 hours post-meal, then during and after exercise.

**Perceived energy and performance:** Rate your energy level (1-10 scale) before, during, and after exercise. Track workout performance (power output, distance, reps, or time to exhaustion).

**Subjective hunger and satiety:** Rate hunger (1-10) at the same time points as glucose measurements.

**Body composition (optional):** If testing for 2+ weeks per condition, measure weight and waist circumference weekly.

**Lactate (optional but informative):** If you have access to a lactate meter, measure before exercise and at 10-minute intervals during exercise.

### Key confounds to control for

**Total carbohydrate intake:** Keep total grams of carbohydrate the same between conditions (this study used ~2 g/kg body weight).

**Meal timing:** Eat exactly 90 minutes before exercise every time.

**Exercise intensity and duration:** Perform the exact same workout (same duration, same intensity measured by heart rate or power output).

**Sleep:** Maintain consistent sleep duration and quality (7-9 hours per night).

**Hydration:** Drink the same amount of water before and during exercise.

**Caffeine and other supplements:** Keep these constant or eliminate them.

**Menstrual cycle (if female):** Test each condition during the same phase of your cycle (follicular vs. luteal) since substrate oxidation varies across the cycle.

**Previous day's diet and exercise:** Standardise what you eat and do for 24 hours before each test day.

### What a positive result would look like

**If the LGI hypothesis is correct for you:** You would see lower blood glucose spikes after LGI meals (peak glucose <140 mg/dL vs. >160 mg/dL for HGI), more stable glucose during exercise (less than 20 mg/dL drop from pre-exercise levels), higher perceived energy during the workout, and possibly better performance (e.g., maintaining power output longer).

**If the HGI hypothesis is correct for you:** You might feel more energetic during high-intensity exercise after HGI meals (since carbohydrate is the primary fuel at high intensities), with no difference in blood glucose stability during exercise.

**If this study's findings replicate:** You would see no meaningful difference in exercise performance or blood glucose during exercise between conditions, despite clear differences in postprandial glucose and insulin. The glycemic index of your pre-exercise meal would not matter for your workout.

**Bottom line for self-experimenters:** Don't assume low-GI pre-exercise meals are superior. Test both conditions yourself. Pay attention to how you feel during the workout, not just the blood glucose numbers. For high-intensity exercise (above 80% max heart rate), the glycemic index of a meal eaten 90 minutes prior may matter less than total carbohydrate availability.