Effects of Low Versus Moderate Glycemic Index Diets on Aerobic Capacity in Endurance Runners: Three-Week Randomized Controlled Crossover Trial

glycemic index (GI) of ingested carbohydrates may in uence substrate oxidation during exercise and athletic performance. Therefore, the aim of this study was to assess the effect of low- and moderate-GI three-week diets on aerobic capacity and endurance performance in runners. We conducted a randomized crossover feeding study of matched diets differing only in GI (low vs. moderate) in 21 endurance-trained runners. Each participant consumed both, low- (LGI) and moderate-GI (MGI) high-carbohydrate (~60%) and nutrient-balanced diets for three weeks each. At the beginning and end of each diet, participants had their aerobic capacity and body composition measured and performed a 12-min running test. After LGI, time to exhaustion during incremental cycling test (ICT) and distance covered in the 12-min run were signi cantly increased. The MGI diet led to an increase in maximal oxygen uptake ( VO2max), but no performance bene ts were found after the MGI diet. The LGI and MGI diets improved time and workload at gas exchange threshold (GET) during ICT. The results indicate that a three-week high-carbohydrate LGI diet resulted in a small but signi cant improvement in athletic performance in endurance runners. Observed increase in V O2max on MGI diet did not affect performance. Keywords: glycemic index; nutrition; aerobic capacity; endurance; running performance; body composition 1. Introduction Carbohydrates and fat are the main sources of fuel oxidized in muscles during endurance exercise [1]. In

three-week high-carbohydrate LGI diet resulted in a small but signi cant improvement in athletic performance in endurance runners. Observed increase in V O2max on MGI diet did not affect performance. Keywords: glycemic index; nutrition; aerobic capacity; endurance; running performance; body composition 1. Introduction Carbohydrates and fat are the main sources of fuel oxidized in muscles during endurance exercise [1]. In contrast to fat, endogenous stores of carbohydrates are limited. Exercise-induced hypoglycemia contributes to the perception of fatigue and consequently attenuates athletic performance [2]. However, while the ingestion of adequate amounts of carbohydrates is vital to maintain glucose availability and optimize athletic performance [1], the carbohydrate quality may also play an important role because various carbohydrate-rich foods affect postprandial glycaemia differently. The glycemic index (GI) classi es carbohydrate-containing foods based on their postprandial glycemic response [3]. The GI of a food is de ned as the incremental area under the two-hour blood glucose response curve (AUC) following consumption of the tested food usually containing 50 g of CHO and divided by the AUC of a reference food containing the same amount of CHO (either glucose or white bread) and multiplied by 100. In general, foods with a high GI increase blood glucose Nutrients2018,10, 370; doi:10.3390/nu10030370

Nutrients2018,10, 370 2 of 13 concentration more rapidly than foods with a low GI. The GI of food is termed low when it is below 55, medium when it is between 55–70, and high when it is above 70 [4–6]. Initially, the GI was used to design meals and diets for patients with diabetes, but current applications include weight loss and improvement in athletic performance [7–11]. The effects of a pre-exercise meal's GI on aerobic capacity and endurance performance have been investigated with equivocal results. Some studies reported improved endurance capacity or performance after ingestion of low-GI (LGI) compared to a high-GI (HGI) meal before exercise[9,10,12,13] , while others have not [7,8,14]. These discrepancies could be caused by tested meal's carbohydrate content and GI, meal timing, study design, and type of exercise test. For example, in two studies [8,9] with similar exercise protocols, cyclists were provided with a pre-exercise meal containing 1 g kg 1 CHO and ingested 45 min before with meals, but the obtained results were different. In a study by Kern and colleagues [8], no differences in performance on the 15-min cycling trial were found, while in the study by Moore and colleagues [9], performance on the 40-km time trial signi cantly improved. This discrepancy might have been caused by differences in the GIs between the test meals (moderate vs. high and low vs. high) or by different performance cycling tests used in the Kern at al [8] and Moore et al. [9] studies (15 min vs. >90 min), respectively. Nevertheless, one meta-analysis concluded there is no clear bene t of low-GI pre-exercise meal for endurance performance regardless of carbohydrate ingestion during exercise [15], while another meta-analysis found that endurance performance following an LGI meal is superior to that following an HGI meal [16]. The available literature on the effect of diets with high versus low GI fed for 3–5 days on endurance performance or exercise capacity is limited to a few studies with inconclusive results. For example, Chen et al. [17] showed that the most important factor in improving athletic performance was ingesting a three-day high-carbohydrate

to that following an HGI meal [16]. The available literature on the effect of diets with high versus low GI fed for 3–5 days on endurance performance or exercise capacity is limited to a few studies with inconclusive results. For example, Chen et al. [17] showed that the most important factor in improving athletic performance was ingesting a three-day high-carbohydrate diet regardless of diet's GI. Hamzah et al. [18] found that consuming a ve-day high-carbohydrate diet with either high or low GI had no impact on time to exhaustion or distance covered during a treadmill test at 65% maximal oxygen uptake ( . V O2max). However, it is possible that three- or ve-day diet might be too short to elicit any metabolic changes and alter athletic performance. The intake of carbohydrates in these studies was relatively high (>70%), potentially limiting changes in substrate oxidation and restricting possible bene cial metabolic adaptations. It is possible that long-term low-GI diets might provide greater metabolic alterations than short-term diets. A plausible physiological rationale for long-term LGI diets is an increased fat oxidation caused by reduced carbohydrate availability during exercise. In addition, increased availability of non-esteri ed fatty acids could enhance the mitochondrial enzymes activity [19]. Therefore, training on a LGI diet can be a good strategy for improving endurance adaptations, but more research in this area is necessary. We hypothesized that in actively training athletes a high-carbohydrate diet with low GI compared with moderate GI consumed for three weeks would induce modest improvements in aerobic capacity (maximal oxygen uptake, gas exchange threshold). Our secondary hypotheses were that endurance performance (distance in the 12-min running test and time to exhaustion) and maximal workload in the incremental cycling test (ICT) would differ between low and moderate GI diets for three weeks. To test these hypotheses, we conducted a randomized crossover feeding trial in a group of young actively training endurance runners consuming diets with either low (LGI = 39 1) or moderate (MGI = 69 1) GI for three weeks. 2. Materials and Methods 2.1. Participants Eligibility criteria were age 17 years or older, self-reported

moderate GI diets for three weeks. To test these hypotheses, we conducted a randomized crossover feeding trial in a group of young actively training endurance runners consuming diets with either low (LGI = 39 1) or moderate (MGI = 69 1) GI for three weeks. 2. Materials and Methods 2.1. Participants Eligibility criteria were age 17 years or older, self-reported good health, current medical exam con rming person's ability to practice sports, at least three years of endurance running training, and currently training four or more times/week for 1.5–2 h/day. We included both males and females because of equal participation of both genders in endurance running.

Nutrients2018,10, 370 3 of 13 Exclusion criteria included current smoking or illicit drug use, alcohol consumption greater than 1–2 drinks/week, and dietary supplements use or being on any special diet less than three weeks before the study. For females, additional exclusion criteria were being pregnant or planning to become pregnant during the study. The primary recruitment strategy was contacting potential participants through the local runners' clubs using mailings, yers, and word-of-mouth. The recruitment goal was 20 participants completing two diet periods and all exercise tests. After initial telephone screening, 36 potential participants (13 females) were invited to the study. At the screening visit, 11 participants were found ineligible or not interested in the study. Twenty- ve runners (9 females), either professional or recreational long- (n= 14) and middle-distance (n= 11) runners, were enrolled into the study. Each running subgroup (professional and recreational, long and middle distance) followed the same training schedule for the entire study (~10 weeks). The study protocol was reviewed and approved by the local institutional review board (Bioethics Committee at Pozna´n University of Medical Sciences, reference number: 173/15 of 5 February 2015). In accordance with the Declaration of Helsinki, all participants and parents of participants younger than 18-year-old signed informed consents. The trial was conducted from February to April 2015. This trial was registered at Clinical Trials Gov (website:; Clinical Trial Identi cation Number: NCT03062527). The study was registered retrospectively since the registration was not required when the study enrolment started. The authors con rm that all ongoing and related trials for this intervention are registered. The study complies with the Consolidated Standards of Reporting Trials (CONSORT) Statement for randomized trials as shown in Figure Table S1. 2.2. Study Design and Protocol Enrolled participants were randomized to the group receiving MGI or to LGI diet (based on their lean body mass). The random allocation sequence was performed using strati ed randomization by impartial biostatistics. All participants began a 14-day run-in phase during which they ate their habitual diet and recorded all ingested food and beverages in food diaries (Table). Collected intake data were analyzed and used to

the group receiving MGI or to LGI diet (based on their lean body mass). The random allocation sequence was performed using strati ed randomization by impartial biostatistics. All participants began a 14-day run-in phase during which they ate their habitual diet and recorded all ingested food and beverages in food diaries (Table). Collected intake data were analyzed and used to determine general eating habits and assess each individual's diet compatibility with the study diets. During this period, total daily energy expenditure was estimated using a wrist-worn heart rate monitor (Polar RS-400, Polar, Vantaa, Finland) worn for four consecutive days, as described previously [20]. After the run-in phase and the rst laboratory study visit followed by an endurance test on the next day, participants started dietary intervention. In a crossover design, participants consumed in a random order MGI or LGI diets for three weeks each and separated by a two-week break (Figure). The diet length (three weeks) was chosen based on the study by Clapp and Lopez [21] in adult women showing that 20 days of low-GI diet caused signi cant metabolic alterations. The break between diets (two weeks) was chosen based on previous studies showing that 10–14 days is suf cient time to prevent carryover of potential metabolic changes [18,22]. Participants were instructed to maintain the same physical activity level throughout the study.

Nutrients2018,10, 370 4 of 13Nutrients 2018, 10, x FOR PEER REVIEW 3 of 13 weeks before the study. For females, additional exclusion criteria were being pregnant or planning to become pregnant during the study. The primary recruitment strategy was contacting potential participants through the local runners’ clubs using mailings, flyers, and word‐of‐mouth. The recruitment goal was 20 participants completing two diet periods and all exercise tests. After initial telephone screening, 36 potential participants (13 females) were invited to the study. At the screening visit, 11 participants were found ineligible or not interested in the study. Twenty‐five runners (9 females), either professional or recreational long‐ (n = 14) and middle‐distance (n = 11) runners, were enrolled into the study. Each running subgroup (professional and recreational, long and middle distance) followed the same training schedule for the entire study (~10 weeks). The study protocol was reviewed and approved by the local institutional review board (Bioethics Committee at Poznań University of Medical Sciences, reference number: 173/15 of 5 February 2015). In accordance with the Declaration of Helsinki, all participants and parents of participants younger than 18‐year‐old signed informed consents. The trial was conducted from February to April 2015. This trial was registered at Clinical Trials Gov (website: https://clinicaltrials.gov/show/NCT03062527; Clinical Trial Identification Number: NCT03062527). The study was registered retrospectively since the registration was not required when the study enrolment started. The authors confirm that all ongoing and related trials for this intervention are registered. The study complies with the Consolidated Standards of Reporting Trials (CONSORT) Statement for randomized trials as shown in Figure 1 and Table S1. Figure 1. A flow chart of the study design. Abbreviations: GI—glycemic index, MGI—moderate glycemic index, LGI—low glycemic index. Figure 1. A ow chart of the study design. Abbreviations: GI—glycemic index, MGI—moderate glycemic index, LGI—low glycemic index. Table 1. Composition of the baseline, moderate glycemic index (MGI), and low glycemic index (LGI) diets. Parameter Baseline MGI before MGI Diet Baseline LGI before LGI Diet Energy intake (kcal) 3060 632 3190 616 3062 651 3174 600 Protein (g) 141.2 33.1 109.0 20.8 * 141.6 33.8 109.7 20.6 ‡ Fat

Abbreviations: GI—glycemic index, MGI—moderate glycemic index, LGI—low glycemic index. Table 1. Composition of the baseline, moderate glycemic index (MGI), and low glycemic index (LGI) diets. Parameter Baseline MGI before MGI Diet Baseline LGI before LGI Diet Energy intake (kcal) 3060 632 3190 616 3062 651 3174 600 Protein (g) 141.2 33.1 109.0 20.8 * 141.6 33.8 109.7 20.6 ‡ Fat (g) 102.8 20.4 90.4 16.2 * 103.5 21.2 90.4 17.5 † Carbohydrates (g) 392.5 83.4 485.2 101.3 * 389.8 86.8 480.2 96.8 ‡ Dietary ber (g) 31.7 2.6 38.1 6.5 # 31.9 2.4 57.6 8.0 ‡§ Glycemic Index (GI) 60 3 63 1 # 60 3 39 1 ‡§ Values are means standard deviation (SD). The differences between diets were analyzed using Student'st-test or Wilcoxon test depending on the distribution of data (normal—not-normal). Signi cantly different from MGI before: *p< 0.0001, # p= 0.0001; Signi cantly different from LGI before: ‡ p< 0.0001, † p= 0.0001; Signi cantly different from MGI diet: § p< 0.0001. 2.2.1. Study Visits The protocol included four visits to the laboratory before and after the rst (T1and T2) and the second (T3and T4) diet. At each visit, body mass and composition were measured and followed by the exercise tests and the eld endurance test on the next day. On the following day, participants began their prescribed (either LGI or MGI) three-week diet ending with the T2visit. Three hours before T2 and T4visits, participants consumed a standardized MGI or LGI meal corresponding to a second meal

Nutrients2018,10, 370 5 of 13 on the rst day from the received menus (Tables S2 and S3). The three-hour period between the meal and exercise was chosen to prevent potential short effect of a meal on exercise performance [12,23,24]. The LGI meal (GI = 39) consisted of wholegrain rye bread, dried tomatoes, low-fat quark, cucumber, lettuce and dried apricots. The MGI (GI = 65) meal consisted of low-fat quark, wheat bread, dried tomatoes, dried raisins, and plain yoghurt. During the two-week break between the study diets, participants followed their self-selected diet. After the break, participants had their T3visit and began the second (MGI or LGI) three-week diet culminating in the T4visit. 2.2.2. Study Diets MGI and LGI diets were individualized to ensure weight maintenance throughout the dietary interventions. Diets provided 11–15% of energy from protein (1.6 g kg body mass 1 ), 25% energy from fat, and 60–64% from carbohydrate as recommended by the International Society of Sports Nutrition [25] (Table). Diets were developed considering the type and GI of foods, cooking methods, and pre-established macronutrient ratios. The GI was calculated from the published tables [4–6]. Carbohydrate-containing foods in the LGI diet included wholegrain rye bread, rolled oats, oat bran, brown rice, buckwheat, wholegrain pasta, vegetables except for corn and potatoes, and fruits such as apples, grapefruits, tangerines, dried apricots, and unripe bananas. Carbohydrate products in the LGI diet were cookedal denteto keep the GI as low as possible. The major sources of carbohydrate in the MGI diet were wheat bread and pasta, potatoes, instant oats, corn akes, white rice, vegetables, and fruits such as ripe bananas, grapes, raisins, dates, and cranberries, honey, and high-sugar jams. Carbohydrate products in MGI diet were slightly overcooked to increase the GI. Daily diets were divided into 5–6 moderate- size meals to prevent digestive problems caused by consumption of large food portions. The participants received seven–day menus (Tables S2 and S3) for the LGI or MGI three-week diet. All instructions about foods and meal preparation were provided by the study dieticians. The participants were encouraged to contact the dieticians with any questions or concerns

were divided into 5–6 moderate- size meals to prevent digestive problems caused by consumption of large food portions. The participants received seven–day menus (Tables S2 and S3) for the LGI or MGI three-week diet. All instructions about foods and meal preparation were provided by the study dieticians. The participants were encouraged to contact the dieticians with any questions or concerns about the diets. For food intake recording, the participants used food diaries and electronic kitchen scales. Study dieticians gave instructions on how to use the scales and complete the diaries to each participant individually. Participants recorded time and the amount of foods and beverages consumed at each meal, the amount of leftovers, and any deviations from the diet. The diaries were collected at T2and T4visits. A dietician reviewed and discussed the diary with each participant. The energy and nutrient intake from diaries were calculated using Dietetyk-2 software (JuMar 2006, Pozna ´n, Poland). 2.2.3. Anthropometry and Body Composition Body mass and height were measured in duplicate using a calibrated scale with a stadiometer (WPT 60/150 OW, Radwag ® , Radom, Poland) in a fasted state, to the nearest 0.1 kg and 0.1 cm. Fat mass and fat-free mass were assessed by air displacement plethysmography (Bod Pod ® , Cosmed, Rome, Italy) [26]. Total body water content was assessed using bioelectric impedance (BIA 101S, Akern-RJL, Pontassieve, Italy) [27]. 2.2.4. Daily Energy Expenditure Total daily energy expenditure was assessed using heart rate (HR) monitoring data (Polar RS-400, Vantaa, Finland), based on a previously validated method [20]. Each participant's HR was recorded minute-by-minute for four consecutive days. To eliminate any accidental errors (e.g., cell phone interference, loss of skin contact during sleep) in HR recordings, participants were asked to report time and type of habitual daily activities and training. The information was used to ll the potential gaps in HR recordings. Recorded wrist-worn HR data were downloaded to a computer equipped with the Polar ProTrainer 5 program (ver. 5.41.002, Vantaa, Finland). On a separate visit, the thresholds HR

habitual daily activities and training. The information was used to ll the potential gaps in HR recordings. Recorded wrist-worn HR data were downloaded to a computer equipped with the Polar ProTrainer 5 program (ver. 5.41.002, Vantaa, Finland). On a separate visit, the thresholds HR

This study examines the impact of low versus moderate GI diets on endurance runners' performance.