A Sports Drink’s Only Edge in a Normal Gym Session Is Its Calories

What this page rests on. Fluid: one systematic meta-analysis of drinking during continuous exercise (Rowlands et al. 2022). Fuel: the same research group's corrected meta-analysis of carbohydrate feeding in resistance exercise (King, Helms and Jukic 2026, a preprint) and a 49-study systematic review that separates carbohydrate from calories (Henselmans et al. 2022). Price: USDA FoodData Central. Exceptions: the International Society of Sports Nutrition's nutrient-timing position stand (Kerksick et al. 2017) and Henselmans et al. 2022. Declared funding and interests are listed with each source; the USDA record is the only source on this page described as independent.

Fluid: Rowlands, Kopetschny and Badenhorst 2022. The Hydrating Effects of Hypertonic, Isotonic and Hypotonic Sports Drinks and Waters on Central Hydration During Continuous Exercise: A Systematic Meta-Analysis and Perspective. Sports Medicine 2022;52(2):349–375 (online 30 October 2021); DOI 10.1007/s40279-021-01558-y; PMID 34716905. 28 qualifying studies and 68 drink treatment effects; random-effects meta-analysis of delta plasma volume (dPV) over 0–180 min of continuous exercise, adjusted for drink osmolality, ingestion rate and metabolic rate, with a weakly informative Bayesian prior. 210 men and 16 women (average study size eight); cycling ergometry except four treadmill studies; all sub-maximal at 48–85% VO2max; 11 trials in hot conditions. Mean dPV: hypertonic −7.4% (90% CL −8.5, −6.3), isotonic −8.7% (−10.1, −7.4), hypotonic −6.3% (−7.4, −5.3), water −7.5% (−8.5, −6.4). Contrasts against a smallest important effect of 0.75% dPV: water–isotonic 1.3% (0.0, 2.5; p+ = 0.76), hypertonic–water 0.1% (−0.8, 1.0; p+ = 0.12), hypotonic–water 1.1% (0.1, 2.0; p+ = 0.72). The hypotonic–water contrast was more likely trivial after between-study random effects; the water–isotonic contrast became unclear in the individual-study setting but stayed compatible with a likely substantial benefit in the sensitivity analysis. Each 100 mOsm/kg rise in ingested osmolality lowered dPV by 1.1% (90% CL −1.4, −0.8); volume and osmolality matter most and sodium less during exercise, though electrolytes raised dPV later in exercise in the carbohydrate drinks and a role for sodium is described as likely in prolonged exercise and in the heat.

Stated limits: dPV stands in for central body water and is not performance; most studies had 12 or fewer participants and dPV was not their primary outcome; unaccounted study effects moved the water and hypertonic estimates (between-study heterogeneity equivalent to an SMD of 1.03); drinks above 50 mEq/L sodium were excluded; sweat rates, acclimation, prior diet and urine output were not modelled; intermittent exercise was excluded, and in it drink differences appear to be neutralised. In the lead author's own included 2011 trial (reference 34), the Powerade tested was isotonic (7.6% carbohydrate, 281 mOsmol/kg) and the Gatorade mildly hypertonic (6.0%, 327 mOsmol/kg), both with substantially lower plasma D2O accumulation than a hypotonic drink; those formulations describe that trial, not any current bottle. Funding and interests: no external funding; the meta-analysis was completed by the lead author, David Rowlands, who had previously received funding from Frucor Ltd (NZ) and Lucozade-Ribena-Suntory (UK) to report upon hypotonic drink effects on hydration; the review's headline favours hypotonic drinks.

Carbohydrate versus nothing: King, Helms and Jukic 2026 (preprint). An Updated Meta-analysis on The Ergogenic Effects of Acute Carbohydrate Feeding on Resistance Exercise Performance. SportRxiv preprint 696; DOI 10.51224/SRXIV.696; not peer reviewed; searches to 15 November 2025. Crossover trials of acute carbohydrate before and/or during resistance exercise against a ≤ 25 kcal placebo or water-only control (the inclusion rule carried over from 2022). Results-section values: total session volume SMD 0.28 (95% CI 0.13, 0.44; 95% PI 0.05, 0.51; p = 0.002; I² = 7.2%; k = 16; high certainty; 174 participants); without Bird et al. 2013, whose carbohydrate arm also contained caffeine, 0.25 (0.09, 0.40; p = 0.004; high certainty). Sessions over 45 min 0.38 (0.20, 0.56; high) versus under 45 min 0.23 (−0.04, 0.50; p = 0.09; moderate), interaction p = 0.27; after a fast of 8 h or more 0.27 (0.06, 0.48; high) versus within 8 h of eating 0.28 (−0.02, 0.58; p = 0.06; moderate), interaction p = 0.97; the over-45-min sensitivity estimate was 0.30 (0.19, 0.42) with a prediction interval of −0.02 to 0.62. Maximal-effort sets (b = 0.02, p = 0.07), dose (b = −0.03, p = 0.79) and load (b = −0.01, p = 0.47) were not significant moderators; if the sets slope is real, 10 more maximal-effort sets would add about 0.2 to the predicted SMD, which the authors place in the trivial-to-small range. They call the sub-group differences inconclusive and the moderators incompletely characterised, and note that small strength-and-conditioning effects generally run SMD 0.15–0.25. Standardised effect sizes only; FitChef does not convert them into repetitions, kilograms or percentages.

Post-exercise blood glucose rose with carbohydrate (SMD 0.93; very low certainty, with evidence of publication bias); feeding only before exercise did not significantly raise it (−0.10; k = 3) while feeding before and/or during did (1.18; k = 11; high certainty), and the size of the rise did not significantly predict the volume benefit (b = 0.17, p = 0.14; k = 10). Lactate was modestly higher (SMD 0.24; high certainty), read by the authors as more work performed; its publication-bias check had fewer than the 10 studies it needs. Candidate mechanisms named: oral carbohydrate sensing and improved subjective appetite, the latter from a handful of trials by one group. Funding and interests: no external funding; Andrew King and Ivan Jukic declare no relevant conflicts; Eric Helms owns or co-owns online coaching, course, book and research review businesses relevant to sports nutrition and is an educational ambassador for Optimum Nutrition, a subsidiary of Glanbia plc. One of the six new studies is the authors' own trial, bias-rated by a researcher from outside the team alongside the two original assessors.

The corrected history: King, Helms, Zinn and Jukic 2022. The Ergogenic Effects of Acute Carbohydrate Feeding on Resistance Exercise Performance: A Systematic Review and Meta-analysis. Sports Medicine 2022;52(11):2691–2712; DOI 10.1007/s40279-022-01716-w; PMID 35809162; 21 studies, 226 participants (214 men, 12 women). Superseded headline estimates: pooled SMD 0.61 (95% CI 0.11, 1.11; I² = 79%; k = 12; low certainty); over 45 min 1.02 (0.07, 1.97; low), under 45 min 0.23 (−0.21, 0.67; not significant); after a fast of 8 h or more 0.39 (0.06, 0.72; moderate), within 8 h 0.76 (−0.19, 1.71; not significant); sets a significant moderator (b = 0.11, p = 0.005). All five large-effect studies were lower-body-only, and two to four sets of lower-body lifting to failure showed no improvement. Its practical lines (over 45 min with at least 8–10 sets; after an 8-h fast) came from those sub-groups. The 2026 authors found standard errors extracted as standard deviations for four studies (Bird et al. 2013 and three others), one study left out (dos Santos et al. 2019) and caffeine in Bird et al.'s carbohydrate arm, and attribute the fall from 0.61 to 0.28 to these corrections, the new trials and, in the sensitivity analysis, removal of the confounded data. Effect sizes without raw data still use correlations imputed from the authors' laboratory (0.78 for volume). The six new studies: one low risk of bias, four some concerns, one high; three male-only, three mixed-sex. The 2022 paper declared no conflicts and only open-access publication funding; whether the later-declared industry roles existed in 2022 is not stated.

Carbohydrate versus the same calories: Henselmans et al. 2022. The Effect of Carbohydrate Intake on Strength and Resistance Training Performance: A Systematic Review. Nutrients 2022;14(4):856; DOI 10.3390/nu14040856. Of 49 studies, 39 found no significant benefit of carbohydrate manipulation, including all 16 isocaloric comparisons; 10 suggested benefits in specific contexts (otherwise fasted training, over 10 sets per muscle group, bi-daily workouts), and in tests of 11–17 sets per muscle group three studies found a benefit or trend and one did not. Every study finding a benefit also supplied more energy, which may act through hunger suppression; there was no dose-response effect. Conclusion: carbohydrate per se, independent of energy intake, is mechanistically and statistically unlikely to acutely affect resistance training performance in a fed state for workouts up to 10 sets per muscle group, and more isocaloric research with realistic placebos is needed. For 11 or more sets per muscle group, or a second hard session for the same muscles that day, up to 1.2 g/kg/h between workouts may be warranted to restore glycogen. King et al. included only crossover trials against a ≤ 25 kcal placebo or water; Henselmans et al. also included parallel trials and isocaloric comparators, which is why the two reviews read differently.

The exceptions and the ISSN's two voices: Kerksick et al. 2017. International society of sports nutrition position stand: nutrient timing. Journal of the International Society of Sports Nutrition 2017; DOI 10.1186/s12970-017-0189-4. Body text on resistance exercise: one study found carbohydrate before and during lifting helped, only in a second session on the same day, while multiple studies found no improvement, and performance results in intermittent high-intensity activity remain mixed. Its position-statement bullet is more favourable (euglycemia, higher glycogen stores, less muscle damage, greater adaptations), and it adds that with sufficient protein carbohydrate may add no further adaptive benefit. Its duration tiers come from endurance research: carbohydrate in bouts under 70 min works in a largely mixed way; a consistent benefit from a ~6–8% solution appears at 90 min or more, particularly when starting fed; 30–60 g/h in a 6–8% carbohydrate-electrolyte solution is recommended for bouts over 60 min above 70% VO2max; high-intensity exercise in hot and humid conditions demands aggressive carbohydrate and fluid replacement. These tiers are separate from the lifting meta-analyses' 45-min sub-group line and from FitChef's electrolyte claim, whose normal gym session is under 75 minutes, indoor and temperate. Funding and interests: the Society is supported in part by grants from raw-goods suppliers and branded companies, and its CEO and co-founder is a co-author.

The price: USDA FoodData Central, Sports drink, NFS. Survey (FNDDS) 2021–2023, FDC ID 2710771, food code 95321000, published 31 October 2024. Per 100 g: 26 kcal, carbohydrate (by difference) 6.43 g, total sugars 5.24 g, protein 0 g, fat 0 g, sodium 39 mg, potassium 15 mg, calcium 1 mg, magnesium 0 mg, water 93.38 g, caffeine 0 mg. USDA weighs a 20-fl-oz bottle at 620 g: 161.2 kcal, 39.87 g carbohydrate, 32.49 g sugars, 241.8 mg sodium and 93.0 mg potassium per bottle (FitChef's arithmetic). The record's single ingredient is 'Beverages, PEPSICO QUAKER, Gatorade, G performance O 2, ready-to-drink.', so these values are one brand's formulation, not an average across sports drinks. USDA FoodData Central is a U.S. government reference database, independent of the drink makers.

Fluid still matters. The numbers belong to FitChef's claim 'Does Dehydration Make You Weaker in the Gym?': in Savoie et al. 2015 (Sports Medicine; DOI 10.1007/s40279-015-0349-0) hypohydration decreased muscle strength by 5.5 ± 1.0% (upper body −3.7%, lower body −6.2%) and muscle endurance by 8.3 ± 2.3% at a mean body-mass loss of 3.0% (range 1.0–5.0%), mostly after dehydration imposed before exercise (a limitation from FitChef's extraction, not stated by the authors). The claim's synthesis notes that the cost comes from starting in a deficit, that water is sufficient for the gym, and that lab protocols may overstate real-world impact. In endurance exercise of up to 2 h in cyclists and runners aged 20 to 40, drinking to thirst beat a programmed plan by 0.98 ± 0.44% (95% CI 0.11–1.84%; p = 0.03), a difference the authors judged trivial, on 804 mL against 1725 mL (Goulet and Hoffman 2019, Sports Medicine; DOI 10.1007/s40279-018-01051-z). Electrolyte loading before a normal gym session has never been tested (38 studies, none on a gym session or on resistance training).

What this page does not claim. That sports drinks hurt performance; that electrolytes were shown not to help; that carbohydrate works only when fasted or only past 45 minutes; that fluid does not matter; any conversion of SMD or plasma-volume points into repetitions, kilograms, litres or percentages of performance; or any finding specific to women, since the evidence is mostly men (Rowlands 210 men and 16 women; King 2022, 214 men and 12 women). Brand-label figures and a widely quoted post-exercise fluid-retention trial are not used, because neither is in FitChef's locked evidence.