Journal Logo

APPLIED SCIENCES

Energy Deficit Required for Exercise-induced Improvements in Glycemia the Next Day

SCHLEH, MICHAEL W.1; PITCHFORD, LISA M.1; GILLEN, JENNA B.1,2; HOROWITZ, JEFFREY F.1

Author Information
Medicine & Science in Sports & Exercise: April 2020 - Volume 52 - Issue 4 - p 976-982
doi: 10.1249/MSS.0000000000002211

Abstract

Many of the exercise-associated improvements in metabolic health result from the most recent session of exercise (1,2), and enhanced insulin action and glycemic control can persist from several hours to a few days after a session of exercise. Common methods for assessing insulin sensitivity and glycemic control involve tests performed in a clinic or laboratory (e.g., oral glucose tolerance tests, hyperinsulinemic–euglycemic clamp). Although these tests may provide a valid assessments of glucose metabolism, they reflect a mere “snapshot” of glycemic regulation under very controlled conditions and do not capture the glycemic response in daily life. Recently, continuous glucose monitoring (CGM) technology has emerged as an effective tool for measuring “free-living” glycemic control over several days, and CGM devices have been used to assess free-living glycemic responses to exercise. For example, improvements in 24 h glycemic control have been observed in adults with type 2 diabetes after acute sessions of moderate-intensity continuous (3–6) and high-intensity interval exercise (7,8). Additionally, exercise-induced improvements in glycemic control have also been observed in nondiabetic individuals (9–11).

Importantly, in these previous studies, the meals consumed on both the exercise and nonexercise control days were identical. As a result, energy intake was less than energy expenditure on the days these individuals exercised, thereby inducing an energy deficit. Because being in an energy deficit can profoundly impact metabolic outcomes, including insulin-mediated glucose metabolism (12,13), an exercise-induced energy deficit precludes the ability to assess the direct versus indirect effects of exercise. Expanding our understanding of the direct effects of exercise (independent of an exercise-induced energy deficit) will advance our ability to provide evidence-based exercise and dietary recommendations for improved metabolic health. Therefore, the purpose of this study was to determine the impact of an exercise-induced energy deficit on postprandial and 24 h glucose control after a single session of moderate-intensity exercise in free-living conditions measured by CGM. We hypothesized that maintaining the exercise-induced energy deficit would enhance glycemic control the day after exercise, whereas this effect would not be evident when the energy expended during exercise was replenished after exercise.

METHODS

Participants

Fifteen healthy, young men and women were recruited to participate in the study (Table 1). All participants were between the ages of 18 to 40 yr and nonobese (body mass index = 20–30 kg·m−2). In addition, they were not taking medications known to affect glucose or fatty acid metabolism, were not at risk for cardiovascular complications, and were nonsmokers. Written informed consent was obtained before all participants began in the study, and all study procedures were approved by the University of Michigan Institutional Review Board.

T1
TABLE 1:
Subject characteristics.

Preliminary Testing

Peak oxygen consumption

Participants performed an incremental test to exhaustion on an electronically braked cycle ergometer (Examiner; Lode Holding BV, Groningen, Netherlands) to determine peak oxygen consumption (V˙O2peak). The protocol began with a 4-min warm-up at 40 W, and increased 20 W every minute until volitional fatigue. The V˙O2peak test was completed at least 1 wk before the first experimental trial to avoid confounding effects of the exercise testing session on study outcomes.

Body composition

Body composition was assessed by bioelectrical impedance (InBody 570; Inbody Co. Ltd., Seoul, South Korea). InBody proprietary equations were used to calculate fat mass and fat-free mass (FFM) from impedance measures altered by the resistance of a current affected by various tissues (i.e., fat tissue vs water/lean mass).

Experimental Procedure

All participants completed two 5-d experimental trials (Fig. 1) in a randomized, counterbalanced order, separated by at least 1 wk. By design, the two trials differed only by the caloric content of the meal consumed after exercise on day 3 of each trial, to create either an “exercise energy deficit” (ExDEF) or an “exercise energy balance” (ExBAL) condition in which the calories expended from the exercise session were “replaced” (Table 2).

F1
FIGURE 1:
Timeline of study trials. Participants wore CGM sensors during two separate trials, each lasting a total of 5 d. 24 h glycemia assessments started were from approximately 8:00 am to 8:00 am (0800h to 0800h). Day 2 of each trial represented NoEX. On day 3, participants reported to the laboratory in the afternoon and exercised to expend 350 kcals at an intensity of 65% V˙O2peak. In the meal after exercise on day 3, participants either received a standardized dinner meal, but did not receive the calories expended from exercise and remained in ExDEF, or received 350 additional kcals in addition to their standardized meal and remained in ExBAL. Glycemic responses the day after exercise (both ExDEF and ExBAL) started at 8:00 am (0800h) on day 4 and ended at 8:00 am (0800h) on day 5. CGM sensors were removed following data collection on day 5. *Standardized meals supplied.
T2
TABLE 2:
Example macronutrient content in study meals for a 70-kg subject with 20% body fat mass.*

On day 1 of each trial, participants reported to the laboratory after a minimum 2-d washout from their most recent exercise session. On arrival to the laboratory, a CGM sensor (Enlite; Medtronic Diabetes, Inc. Minneapolis, MN) was inserted into the subcutaneous tissue of the abdomen, roughly 5 cm lateral to the umbilicus. The single-use sensor, which records interstitial glucose concentrations every 5 min, was connected to a monitor (iPro2; Medtronic Diabetes, Minneapolis, MN) placed on the abdomen. Afterward, they were familiarized to perform finger stick blood glucose measurements on themselves using a portable glucometer (OneTouch; Lifescan Inc., Milpitas, CA), which is necessary for calibrating CGM data. As per manufacturer’s instructions, participants were asked to measure their blood glucose before each meal when blood glucose was likely to be stable. They then consumed a standardized dinner and evening snack in the evening of day 1.

On days 2, 3, and 4 of each trial, participants were provided standardized diets for all meals (breakfast, lunch, dinner, and an evening snack) and instructions for when to consume each meal; no other calorie-containing foods or beverages were permitted during the trials. The macronutrient content of the standardized diets was designed by the study dietician to replicate a typical western diet (~55% carbohydrate, ~30% fat, and ~15% protein) and to maintain energy balance as calculated based on FFM (14); details of the study diets are provided in Table 2. Contents of the meals mimicked a typical western diet: Breakfast—bagels, peanut butter, fruit, juice; Lunch—turkey sandwiches, potato chips, juice; Dinner—chicken, rice, juice; Evening snack—cookies and milk. On days 2 and 4 of each trial, participants were required to abstain from exercise, but they were allowed to perform their normal daily activities. Importantly, the 5-d experimental trials were conducted on Monday through Friday in order to avoid weekends, when physical activity behavior may be more variable.

On day 3 of both trials, the energy content of breakfast and lunch were designed to maintain energy balance. In the afternoon of day 3, participants arrived to the laboratory at approximately 3:00 pm to complete an exercise session on a cycle ergometer. During both trials, they exercised to expend a total of 350 kcals at an intensity of 65% of their predetermined V˙O2peak. Indirect calorimetry was performed intermittently throughout exercise to confirm that all participants were at the targeted 65% of V˙O2peak. Substrate oxidation (g·min−1) calculated during the exercise session on day 3 was determined as described by Frayn (15). The exercise sessions during the two trials for each individual were identical; however, the duration of exercise varied among participants depending on their V˙O2peak, and ranged from approximately 30 to 60 min to expend 350 kcal. Although the exercise duration differed among participants, the start times of the exercise sessions were set so that exercise would end at approximately 4:00 pm. The participants then ate dinner at 5:00 pm, and as noted above, the only difference between the two experimental trials was the caloric content consumed at dinner after the exercise sessions. During ExDEF, the caloric content consumed after exercise was identical to that consumed on days 2 and 4, and therefore did not replace the energy expended during exercise, creating an energy deficit after exercise. During ExBAL, the energy that was expended during exercise was replaced by supplementing the postexercise meal with a nutrient supplement drink (BoostPlus: Nestlé, Vevey, Switzerland; 50% carbohydrate, 35% fat, 15% protein).

As our objective was to determine the effects of exercise and energy deficit/balance on free-living 24-h glycemia the day after exercise, we focused primarily on CGM glucose measurements from 8:00 am on day 4 through 8:00 am on day 5. The postexercise 24 h glycemia measurements during ExDEF and ExBAL were also compared with 24 h glycemia measurements on day 2, when the participants did not exercise (NoEx) and consumed the exact same meals at the same times. After 8:00 am on day 5 of each trial, they then returned to the laboratory for removal of the CGM sensor.

Quantification of Free-Living Glucose Concentration from the CGM Sensors

After each trial, CGM data were uploaded and stored onto the manufacturer’s database (CareLink; Medtronic Minimed, Northridge, CA). The capillary blood glucose values each individual measured throughout each trial were entered into the software program for CGM calibration. Although the CGM sensors measured glucose in 5-min increments continuously throughout each 5-d experimental trial, the 24 h glycemic response reported in our study was from 8:00 am on day 2 to 8:00 am on day 3 (NoEx) and from 8:00 am on day 4 to 8:00 am on day 5 (ExDEF and ExBAL; Fig. 1). During these 24-h periods on days 2 and 4 of each trial, we assessed the glucose area under the curve (AUC) using the trapezoidal rule (16), mean glucose concentration, and peak glucose concentration. Twenty-four–hour mean amplitude of glycemic excursion (MAGE) was calculated as mean glycemic excursion exceeding one standard deviation of the individuals’ daily glucose variation: (17).

In addition to quantifying 24 h glycemia, we also individually analyzed 3 h postprandial glucose responses to each meal on day 2 (NoEx) and day 4 (ExDEF and ExBAL) of each study trial.

For each participant, food logs and CGM data were inspected to confirm meals were consumed at the correct time and quantity. Time 0 for each postprandial glycemia analysis was set at the time the participant reported beginning to consume the meal. One participant did not adhere to the strict dietary requirements by not completing the entire contents of one meal during their first experimental trial. To address this, during their subsequent experimental trial the contents of this same meal were modified to exactly match the contents that were consumed during their first trial.

In addition, on two occasions we believe the CGM devices may have temporarily dislodged or may have temporarily malfunctioned, based on the observation that the CGM-derived glucose concentration data was more than 1.6 mmol·L−1 lower than the capillary glucose concentration measurements made by these two participants before one meal. In both instances, the CGM-derived glucose concentration data recovered to closely match the capillary glucose concentration measurements before the next meal. Therefore, for these two participants, we only excluded the CGM data in response to the meal where the glucose measures appeared to be erroneous (the CGM data were also removed for the corresponding meals during the other measurement days for these two participants).

Statistical Analysis

A one-way analysis of variance with Fisher’s Least Significant Difference post hoc test was used to compare all measures of glycemic responses between ExDEF, ExBAL, and NoEx. Because the experimental treatments on day 2 of both ExDEF and ExBAL were similar, we used the average glycemic responses of day 2 of both trials to represent the “NoEx” response (see Test–retest reliability of CGM glucose measurements in Results and Supplemental Digital Content 1 and 2, https://links.lww.com/MSS/B838 and https://links.lww.com/MSS/B839). The absence of prior work addressing the impact of energy balance versus energy deficit on glycemic response measured by CGM precluded our ability to conduct sample size calculations beforehand; however, effect size for significant values was evaluated by Cohen’s d. Figures were generated using GraphPad Prism (Version 8, San Diego, CA), and data were analyzed using IBM’s SPSS for windows: version 24.0 (IBM Corp. Armonk, NY, USA). Data are represented as mean ± SD, and significance was set at α < 0.05.

RESULTS

Participant characteristics and exercise responses

Table 1 presents the baseline characteristics for the 15 participants who completed both experimental trials. The exercise session on day 3 of each trial was performed at 67% ± 2% V˙O2peak, which corresponded to 90 ± 24 W. The average exercise duration was 47 ± 10 min, and resulted in each participant expending 350 kcal.

Glycemic control

After breakfast, glucose AUC was lower in ExDEF compared with NoEx (P = 0.014, d = 0.38) and ExBAL (P = 0.007, d = 0.42) (Fig. 2A). In addition, 3 h average glucose concentration and peak glucose concentration were lower (P = 0.018, d = 0.46; P = 0.001, d = 0.68, respectively) after breakfast in ExDEF compared with NoEx and ExBAL (Table 3). The postprandial glycemic responses to lunch and dinner (Figs. 2B and C), as well as overnight glycemia (Table 3), were similar among trials. Despite the lower glycemic response to breakfast during ExDEF versus NoEx and ExBAL, 24-h glucose AUC (8:00 am to 8:00 am) was not different among any of these conditions (P = 0.48, d = 0.11, Fig. 3). In addition, 24-h average glucose (P = 0.49, d = 0.10), peak glucose (P = 0.11, d = 0.29), and MAGE (P = 0.86, d = 0.02) were not different among trials (Table 3).

F2
FIGURE 2:
Glycemic responses following (A) breakfast, (B) lunch, and (C) dinner, measured without exercise performed the day before (NoEx) or measured the day after a session of exercise when in energy balance (ExBAL) or deficit (ExDEF). *Significant difference from NoEx (P < 0.05). †Significant difference from ExBAL (P < 0.05). Values are expressed as mean ± SD.
T3
TABLE 3:
Analysis of postprandial, overnight, and 24 h glucose parameters.
F3
FIGURE 3:
(A) Glycemic control was measured for 24 h without exercise the day before (NoEx) and the day after a session of exercise when in energy balance (ExBAL) or deficit (ExDEF). Meal times for breakfast (8:00 am [0800h]), lunch (12:00 pm [1200h]), dinner (5:00 pm [1700h]), and an evening snack (8:00 pm [2000h]) are indicated by vertical bars (dashed). (B) AUC for glucose concentration measured throughout the 24 h CGM measurement period (8:00 am to 8:00 am [0800h to 0800h]) during the 3 experimental conditions. Values are expressed as mean ± SD.

Test–retest reliability of CGM glucose measurements

The experimental treatments on day 2 of both ExDEF and ExBAL were identical (i.e., identical diets, meal timing, and no prior exercise [NoEx]), which provided the opportunity to assess the test–retest reliability of our CGM measurements. All postprandial NoEx data significantly correlated between weeks (Pearson’s correlation coefficient) (see Figure, Supplemental Digital Content 1, Postprandial test–retest plots for control trials between weeks, https://links.lww.com/MSS/B838). Additionally, test–retest analysis demonstrated nearly all postprandial AUC values from the two NoEx control days displayed good relative reliability measured by intraclass correlation (see Table, Supplemental Digital Content 2, Test–retest reliability for postprandial glucose measurements between NoEx/control conditions during ExDEF and ExBAL, https://links.lww.com/MSS/B839). Average and peak glucose also support moderate-to-good test–retest reliability for these multiple measures of postprandial glycemia. Percent difference, coefficient of variation (CV: [SD/mean] × 100) and intraclass correlation for postprandial values are also displayed in Supplemental Digital Content 2, https://links.lww.com/MSS/B839.

DISCUSSION

Our results agree with previous work demonstrating that a session of moderate-intensity exercise in the afternoon/evening can improve glucose tolerance and/or insulin sensitivity (1,18–20) or glycemic control (6–8), into the morning of the next day. Importantly, however, because the improvement in glycemic response to breakfast the day after exercise was absent when our participants were in energy balance after exercise (i.e., when re-fed the 350 kcal they expended during the exercise session to compensate for the exercise-induced energy deficit), our main findings indicate that the improved glycemic response to breakfast the next day is primarily due to the exercise-induced energy deficit. Our findings also reveal that the improved postprandial glycemia stemming from the exercise-induced energy deficit is relatively short-lived, as no improvement was observed in response to lunch or dinner in our group of relatively healthy participants. Several previous studies using CGM as a measure for glycemic control in “free living” conditions have reported improvements in glycemic control in both individuals with obesity (9,10) and with type 2 diabetes (3,4,6,7,21) in response to exercise. To our knowledge, however, the meals consumed after exercise in these studies did not compensate for the energy expended during exercise—and therefore, it is not clear whether the improvement in glycemic control is due to the energy deficit or exercise per se.

Several factors may contribute to exercise-induced improvements in glycemic control. Muscle glycogen depletion has been clearly linked with improved insulin-mediated glucose uptake after exercise (22,23), and it has been demonstrated that enhanced insulin sensitivity after exercise is most profound when carbohydrate intake after exercise is insufficient to fully replenish muscle glycogen stores (20,24). In fact, exercise-induced improvements in insulin sensitivity may persist until muscle glycogen levels are restored above baseline levels (i.e., glycogen “supercompensation”) (25,26). Therefore, not fully restoring muscle glycogen after exercise in ExDEF may explain the improvement in postprandial glucose control observed after breakfast the next day. The mechanistic links between reduced muscle glycogen content after exercise and insulin action have not been fully elucidated. These mechanisms may be linked to increased activation of glycogen synthase (22,23); a physical association between glycogen and the insulin-responsive glucose transporter, GLUT4 (27); and potential interaction between glycogen, AMP-activated protein kinase, and distal insulin signaling pathways (28,29). Although muscle glycogen levels were not measured in our study, participants consumed only approximately 80 g of carbohydrate after exercise in ExDEF, which is likely below that necessary to fully restore muscle glycogen even with the nonexhaustive exercise session performed in this study (30). Therefore, the improved postprandial glycemia observed after breakfast during ExDEF may have been largely a consequence of a “carbohydrate deficit” rather than an “energy deficit.”

Changes in muscle glycogen stores may also help explain why the improvement in glycemic control did not persist into the afternoon the day after exercise. Studies reporting persistent elevation of insulin sensitivity for 24 to 48 h after exercise intentionally restricted dietary carbohydrate to keep muscle glycogen levels relatively low (25,31). With our relatively moderate exercise session performed the day before, the high glycemic carbohydrate in the breakfast meal the next day (~80 g CHO from white bread, packaged fruit, fruit juice, etc.) may have been sufficient to increase muscle glycogen to levels that no longer enhanced insulin sensitivity, which could explain why postprandial glycemia was no longer different between ExDEF and ExBAL at lunch or dinner. Our findings for the relatively short-lived enhancement in glycemic control align with studies that have also examined 24-h free-living glucose control after moderate-intensity exercise measured using CGM (21), reinforcing the importance for consistent exercise in order for more sustained improvement in glycemic control.

Other factors, such as exercise intensity, energy expended during exercise, macronutrient content, and glycemic content of the meals may also contribute to the magnitude and persistence of the exercise-induced improvement in glycemic control. Exercise incorporating high-intensity intervals have been reported to induce a more robust improvement in 24-h free-living glycemic control compared with more “conventional” moderate-intensity, continuous exercise (7,10). In addition to relying heavily on muscle glycogen stores for energy provision (32), high-intensity intervals also recruit a high proportion of fast-twitch muscle fibers (33), in which glycogen concentration would be depleted to a greater extent. However, it is important to acknowledge that high-intensity exercise is not always found to improve glycemic control more than conventional exercise (4,9). In addition to the type of exercise, macronutrient composition of meals ingested around the exercise session may also play a role in moderating next-day insulin sensitivity and glycemic control. While the present study provided meals composed of approximately 55% CHO, it has been suggested that consuming low-carbohydrate meals after exercise may optimize next-day glycemic control (19,20,34,35). Along these lines, our laboratory has confirmed the effect of carbohydrate as a key macronutrient deficit required to improve insulin sensitivity (24).

Although using CGM devices to assess glucose control in response to exercise documents real-time fluctuations in glucose concentrations as people live their normal lives, limitations still exist such as challenges in matching activities of daily living between measurement days, and the need for tight dietary controls (meal contents and timing). Other important factors may impact the translation of our findings, including the health status of our participants and the intensity and the type of exercise stimulus. Because our main objective was to assess the effects of exercise (with and without an energy deficit) on 24-h glucose control, we recruited a cohort of nonobese, healthy adults to avoid the confounding influence of any obesity-related metabolic complications. Importantly, differences in meals consumed after exercise are known to impact the magnitude and persistence of the insulin-sensitizing effects of exercise even in obese, insulin resistant individuals (24). However, it is not clear if our finding that an exercise-induced energy deficit was required for improvement in glycemic control the next day in our participants directly translates individuals with obesity and metabolic abnormalities (e.g., insulin resistance). Additionally, although the order of the study trials was randomized, the NoEX (control) condition always occurred before the exercise conditions. While this might introduce an ordering effect when comparing the experimental conditions to control, our primary comparisons were between the ExDEF and ExBAL trials, which were counterbalanced. In addition, we also acknowledge that differences in exercise stimuli (i.e., intensity, duration, energy expenditure) may also impact how postexercise meals impact glycemic responses the next day in energy balance (36). However, the moderate-intensity, steady-state exercise we used in our study is a very common exercise stimulus, and aligns with standard exercise recommendations (37,38).

In conclusion, the impact of energy balance after exercise must be considered when assessing the effects of exercise on glycemic control. Our findings indicate that a session of exercise lowered glycemic response to breakfast the next day, but only when meals after exercise did not replenish the energy expended during exercise. This improved postprandial glycemia after breakfast when the participants were in an energy deficit after exercise may have been a result of a lower muscle glycogen concentration; however, future studies incorporating muscle biopsies are needed to confirm this hypothesis. Importantly, even when in an energy deficit after exercise, the exercise-induced improvement in glycemic control did not persist past noon the next day (~19 h after the exercise session), and 24-h glycemic control was not different versus without exercise performed the day before. Overall, our findings support the importance of considering the effects of meals after exercise when prescribing exercise to improve glycemic control.

The authors thank Julie Jing for meal preparation and assistance with study diet, Suzette Howton, RD for designing the study diets and recruitment, and all participants for their enthusiastic support. The results of the study are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation. The results of the present study do not constitute endorsement by ACSM.

Funding Support: This study was supported by The National Institutes of Health R01DK077966, T32DK101357, P30DK089503; American Diabetes Association 1-16-ICTS-048; Canadian Institutes of Health Research (338735).

REFERENCES

1. Heath G, Gavin JI 3rd, Hinderlite JM, Hagberg J, Bloomfield S, Holloszy J. Effects of exercise and lack of exercise on glucose tolerance and insulin sensitivity. J Appl Physiol Respir Environ Exerc Physiol. 1983;55(2):512–7.
2. King D, Dalsky G, Clutter W, et al. Effects of exercise and lack of exercise on insulin sensitivity and responsiveness. J Appl Physiol. 1988;64(5):1942–6.
3. Van Dijk J, Manders R, Tummers K, et al. Both resistance- and endurance-type exercise reduce the prevalence of hyperglycaemia in individuals with impaired glucose tolerance and in insulin-treated and non-insulin-treated type 2 diabetic patients. Diabetologia. 2012;55(5):1273–82.
4. Manders R, Van Dijk J, van Loon L. Low-intensity exercise reduces the prevalence of hyperglycemia in type 2 diabetes. Med Sci Sports Exerc. 2010;42(2):219–25.
5. Mikus C, Oberlin D, Libla J, Boyle L, Thyfault J. Glycaemic control is improved by 7 days of aerobic exercise training in patients with type 2 diabetes. Diabetologia. 2012;55(5):1417–23.
6. Karstoft K, Christensen C, Pedersen B, Solomon T. The acute effects of interval- Vs continuous-walking exercise on glycemic control in subjects with type 2 diabetes: a crossover, controlled study. J Clin Endocrinol Metab. 2014;99(9):3334–42.
7. Gillen J, Little J, Punthakee Z, Tarnopolsky M, Riddell M, Gibala M. Acute high-intensity interval exercise reduces the postprandial glucose response and prevalence of hyperglycaemia in patients with type 2 diabetes. Diabetes Obes Metab. 2012;14(6):575–7.
8. Little JP, Gillen JB, Percival ME, et al. Low-volume high-intensity interval training reduces hyperglycemia and increases muscle mitochondrial capacity in patients with type 2 diabetes. J Appl Physiol. 2011;111(6):1554–60.
9. Parker L, Shaw CS, Banting L, et al. Acute low-volume high-intensity interval exercise and continuous moderate-intensity exercise elicit a similar improvement in 24-h glycemic control in overweight and obese adults. Front Physiol. 2017;7(661).
10. Little J, Jung M, Wright A, Wright W, Manders R. Effects of high-intensity interval exercise versus continuous moderate-intensity exercise on postprandial glycemic control assessed by continuous glucose monitoring in obese adults. Appl Physiol Nutr Metab. 2014;39(7):835–41.
11. Francois ME, Baldi JC, Manning PJ, et al. ‘Exercise snacks’ before meals: a novel strategy to improve glycaemic control in individuals with insulin resistance. Diabetologia. 2014;57(7):1437–45.
12. Black S, Mitchell E, Freedson P, Chipkin S, Braun B. Improved insulin action following short-term exercise training: role of energy and carbohydrate balance. J Appl Physiol. 2005;99(6):2285–93.
13. Burton F, Malkova D, Caslake M, Gill J. Energy replacement attenuates the effects of prior moderate exercise on postprandial metabolism in overweight/obese men. Int J Obes. 2008;32(3):481–9.
14. Cunningham J. Body composition as a determinant of energy expenditure: a synthetic review and a proposed general prediction equation. Am J Clin Nutr. 1991;54(6):963–9.
15. Frayn KN. Calculation of substrate oxidation rates in vivo from gaseous exchange. J Appl Physiol. 1983;55(2):628–34.
16. Tai M. A mathematical model for the determination of total area under glucose tolerance and other metabolic curves. Diabetes Care. 1994;17(2):152–4.
17. Service FJ, Molnar G, Rosevear J, Ackerman E, Gatewood L, Taylor F. Mean amplitude of glycemic excursions, a measure of diabetic instability. Diabetes. 1970;19(9):644–55.
18. Newsom S, Everett A, Hinko A, Horowitz J. A single session of low-intensity exercise is sufficient to enhance insulin sensitivity into the next day in obese adults. Diabetes Care. 2013;36(9):2516–22.
19. Taylor H, Wu C, Chen Y, Wang P, Gonzalez JT, Betts JA. Post-exercise carbohydrate-energy replacement attenuates insulin sensitivity and glucose tolerance the following morning in healthy adults. Nutrients. 2018;10(2):E123.
20. Holtz KA, Stephens BR, Sharoff CG, Chipkin SR, Braun B. The effect of carbohydrate availability following exercise on whole-body insulin action. Appl Physiol Nutr Metab. 2008;33(5):946–56.
21. Oberlin D, Mikus C, Kearney M, et al. One bout of exercise alters free-living postprandial glycemia in type 2 diabetes. Med Sci Sports Exerc. 2014;46(2):232–8.
22. Bogardus C, Thuillez P, Ravussin E, Vasquez B, Narimiga M, Azhar S. Effect of muscle glycogen depletion on in vivo insulin action in man. J Clin Invest. 1983;72(5):1605–10.
23. Devlin J, Horton E. Effects of prior high-intensity exercise on glucose metabolism in Normal and insulin-resistant men. Diabetes. 1985;34(10):973–9.
24. Newsom S, Schenk S, Thomas K, et al. Energy deficit after exercise augments lipid mobilization but does not contribute to the exercise-induced increase in insulin sensitivity. J Appl Physiol (1985). 2010;108:554–60.
25. Cartee GD, Young DA, Sleeper MD, Zierath J, Wallberg-Henriksson H, Holloszy JO. Prolonged increase in insulin-stimulated glucose transport in muscle after exercise. Am J Phys. 1989;256(4):494–69.
26. Kawanaka K, Han DH, Nolte LA, Hansen PA, Nakatani A, Holloszy JO. Decreased insulin-stimulated GLUT-4 translocation in glycogen-supercompensated muscles of exercised rats. Am J Phys. 1999;276(5):907–12.
27. Coderre L, Kandror KV, Vallega G, Pilch PF. Identification and characterization of an exercise-sensitive pool of glucose transporters in skeletal muscle. J Biol Chem. 1995;270(46):27584–8.
28. Richter EA, Derave W, Wojtaszewski JF. Glucose, exercise and insulin: emerging concepts. J Physiol. 2001;535(2):313–22.
29. Sylow L, Richter EA. Current advances in our understanding of exercise as medicine in metabolic disease. Curr Opin Physiol. 2019;12:12–9.
30. Costill DL, Sherman WM, Fink WJ, Maresh C, Witten M, Miller JM. The role of dietary carbohydrates in muscle glycogen resynthesis after strenuous running. Am J Clin Nutr. 1981;34(9):1831–6.
31. Mikines KJ, Sonne B, Farrell PA, Tronier B, Galbo H. Effect of physical exercise on sensitivity and responsiveness to insulin in humans. Am J Phys. 1988;254(3):248–59.
32. Romijn JA, Coyle EF, Sidossis LS, et al. Regulation of endogenous fat and carbohydrate metabolism in relation to exercise intensity and duration. Am J Phys. 1993;265(3 Pt 1):380–91.
33. Gibala MJ, McGee SL. Metabolic adaptations to short-term high-intensity interval training: a little pain for a lot of gain? Exerc Sport Sci Rev. 2008;36(2):58–63.
34. Gaitan JM, Weltman A, Malin SK. Enhancing exercise responsiveness across prediabetes phenotypes by targeting insulin sensitivity with nutrition. J Diabetes Res. 2017;2017:8314852.
35. Francois ME, Gillen JB, Little JP. Carbohydrate-restriction with high-intensity interval training: an optimal combination for treating metabolic diseases? Front Nutr. 2017;4:49.
36. Fisher G, Gower BA, Ovalle F, Behrens CE, Hunter GR. Acute effects of exercise intensity on insulin sensitivity under energy balance. Med Sci Sports Exerc. 2019;51(5):988–94.
37. Nelson ME, Rejeski WJ, Blair SN, et al. Physical activity and public health in older adults: recommendation from the American College of Sports Medicine and the American Heart Association. Med Sci Sports Exerc. 2007;39(8):1435–45.
38. Haskell WL, Lee I-M, Pate RR, et al. Physical activity and public health: updated recommendation for adults from the American College of Sports Medicine and the American Heart Association. Med Sci Sports Exerc. 2007;39(8):1423–34.
Keywords:

GLYCEMIC CONTROL; CONTINUOUS GLUCOSE MONITOR (CGM); EXERCISE; ENERGY BALANCE

Supplemental Digital Content

Copyright © 2019 by the American College of Sports Medicine