Direct answer
What roles do sugar and caffeine have?
Sugar contributes nutritional energy, while caffeine is a stimulant that may affect alertness. Neither ingredient guarantees better performance. Any measured effect depends on the consumed dose, timing, comparison drink, task and individual response.
In a sugar-containing, caffeinated drink, the two components have different roles: sugar contributes calories, measured in kilocalories or kilojoules, while caffeine acts on the central nervous system. A sugar-free product can still contain caffeine; a caffeine-free product can still contain calories.
Caffeine does not provide calories and is therefore not a nutritional energy source. Sugar provides energy but does not reproduce caffeine’s stimulant mechanism.
Sugar provides nutritional energy
Sugars are carbohydrates. The body breaks down digestible carbohydrates into glucose, which cells can use to produce energy. In a conventional sugar-containing energy drink, sugar normally provides most of the drink’s calories. One gram of available carbohydrate provides approximately four kilocalories of nutritional energy.
Calories supplied mainly by sugar.
A stimulant effect produced primarily by caffeine.
A sugar-containing drink can provide metabolic energy even if it contains no caffeine. However, sugar alone should not automatically be expected to prevent sleepiness.
“A high sugar content, low caffeine drink does not alleviate sleepiness.”
A controlled study reported that the drink did not counteract sleepiness and was associated with slower reaction times after the initial period. [5]
Caffeine and alertness
Caffeine is a central nervous system stimulant. It reduces the effects of adenosine, a signalling substance involved in sleep pressure and tiredness. Response varies with dose, timing, habitual use, sensitivity and sleep status.
EFSA’s scientific assessment identified 75 mg of caffeine per serving as the minimum dose supporting its favourable opinion on increased alertness. It reported that most studies using 75 mg or more found a significant reduction in reaction-time measures. [1] That conclusion applies to the assessed conditions and is not a guarantee of a noticeable effect in every person.
A favourable scientific opinion is not authorisation to use a health claim on an EU food label. The Commission rejected a separate application for intakes between 40 and 75 mg per serving in 2016; the European Parliament also objected to a distinct draft authorisation of caffeine alertness and concentration claims. [2] [3] EFSA lists 80 mg as an approximate example for a standard 250 ml energy drink; actual products vary. EFSA serving-size example ↗
Natural and synthetic caffeine
Caffeine may come from botanical sources—including coffee beans, tea leaves, guarana seeds, yerba mate and kola nuts—or be produced synthetically.
Regardless of origin, pure caffeine is the same chemical compound: 1,3,7-trimethylxanthine, with the molecular formula C₈H₁₀N₄O₂. When chemical identity, purity and dose are equivalent, caffeine acts through the same physiological mechanisms.
“Natural sources of caffeine may have influences on the body similar to those of synthetic caffeine.”
A randomised, double-blind crossover study comparing 200 mg of caffeine from natural botanical extracts with synthetic caffeine found broadly similar pharmacokinetic and nervous-system effects. Botanical extracts may also contain other compounds that influence the overall product response; those effects should not be attributed to the caffeine molecule without evidence. [6]
Sugar-free and caffeine-free variants
Sugar-free energy drinks
Sugar-free energy drinks replace sugar with low- or no-calorie sweeteners. They usually provide very little nutritional energy, but may contain as much caffeine as a sugar-containing version. Sugar content alone therefore does not establish stimulant dose or alertness effect.
In a randomised crossover study involving 24 healthy young adults, a sugar-containing product produced greater improvements in two combined memory measures than the tested sugar-free version and placebo. This applies to those products, participants and tests; it does not prove that every sugar-containing product is superior for every type of performance. [7]
In another randomised crossover trial, neither a non-caloric nor a sugar-containing energy drink improved maximal oxygen consumption, time to exhaustion or peak power, although the non-caloric drink improved one resistance-type fatigue measure. [8]
Caffeine-free energy drinks
A caffeine-free beverage may provide calories from sugar, vitamins, minerals, hydration and other ingredients. It does not provide an effect attributable to caffeine. A product containing neither sugar nor caffeine provides neither meaningful energy from sugar nor caffeine stimulation.
Caffeine, glucose and cognitive performance
Research examining individual ingredients supports the need to distinguish their effects. In a randomised, double-blind study, Giles and colleagues examined caffeine, glucose, taurine and their combinations.
“Caffeine enhanced executive control and working memory, and reduced simple and choice reaction time.”
Glucose produced different effects, while the effects of taurine were less consistent. Ingredients should therefore be assessed separately as well as in combination. [9]
Physical performance findings are mixed
Several controlled trials reported favourable outcomes for the exact products and protocols tested. In 18 recreational runners, a caffeinated, sugar-containing drink shortened a five-kilometre treadmill time trial versus a non-caffeinated, sugar-free placebo; because several ingredients differed, the study cannot identify the responsible component. In 12 trained cyclists, a 500 ml drink containing 160 mg caffeine and 54 g carbohydrate shortened a simulated cycling time trial versus placebo. In 19 male physical-education students, another drink improved selected reaction-time, handgrip and 30-second cycling measures. These small acute trials do not establish an effect for every formulation, athlete or task. [11] [12] [14]
In a randomised double-blind crossover trial with 11 trained male cyclists, both the tested energy drink and a matched dose of caffeine improved time-trial performance versus placebo. There was no significant difference between the energy-drink and caffeine conditions, which suggests that the extra ingredients did not add a detectable benefit in that protocol. [10]
Other controlled trials found no significant improvement in their main performance comparisons. A caffeine-taurine drink did not improve repeated sprint time or anaerobic power in 20 college football players, and neither an energy drink nor a caffeine-and-carbohydrate cola changed completion time in an 11-person simulated cycling race. [17] [18]
Taken together, these trials do not support a category-wide promise. Result direction depends on the exact drink, dose, comparator, participants and test.
What controlled studies can—and cannot—show
Controlled trials report favourable, null and mixed findings. The correct unit of interpretation is the exact drink, dose, comparator, participants, timing and outcome—not the energy-drink category as a whole.
Study-by-study comparison
Each row links to the original publication and states the tested population, exposure, comparator, result and main limitation. Result direction is not a product-quality score.
| Primary source | Population | Exposure | Comparison / design | Observed result | What it cannot prove |
|---|---|---|---|---|---|
| Quinlivan et al., 2015 · PMID 25710190 ↗ | 11 trained male cyclists | Energy drink or matched 3 mg caffeine/kg | Non-caffeinated placebo; double-blind crossover | Both active conditions improved the time trial versus placebo; no significant difference between them | Small all-male sample and one formulation; no added drink effect beyond matched caffeine was detected. |
| Gwacham & Wagner, 2012 · PMID 22349209 ↗ | 20 male college football players | Low-calorie caffeine-taurine drink | Matched non-caffeinated placebo; double-blind crossover | No significant improvement in sprint time or anaerobic power | Usual caffeine use interacted with response; small sport-specific sample. |
| Prins et al., 2016 · PMID 26937774 ↗ | 18 recreational endurance runners | 500 ml caffeinated, sugar-containing drink | Non-caffeinated sugar-free placebo; double-blind crossover | Faster five-kilometre treadmill time versus placebo | Small acute product-specific sample; multiple drink components differed, so the responsible component cannot be isolated. |
| Phillips et al., 2014 · PMID 24172719 ↗ | 11 trained male cyclists | 500 ml energy drink | Caffeine/carbohydrate cola and flavoured placebo; crossover | No significant difference in 25-mile completion time | Selected metabolic and immune measures changed; they did not establish a performance or clinical outcome. |
| Ivy et al., 2009 · PMID 19403954 ↗ | 12 trained cyclists | 500 ml drink with 160 mg caffeine and 54 g carbohydrate | Placebo; double-blind crossover | Faster simulated cycling time trial; perceived exertion did not differ | Very small acute, fasted and product-specific sample; caffeine, carbohydrate and other ingredients were not isolated. |
| Wesnes et al., 2017 · PMID 28168925 ↗ | 24 healthy young adults | Sugar-containing or sugar-free product | Matching placebo; double-blind crossover | Product-specific differences on selected combined memory measures | Small acute study; not evidence that sugar improves every cognitive task. |
| Mets et al., 2011 · PMID 21063868 ↗ | 24 enrolled; 21 analysed | Tested drink after two hours of simulated highway driving | Placebo; crossover | Less lane weaving and lower self-reported sleepiness during later simulator periods | Simulator, not a crash study; Red Bull GmbH funded the study, supplied the drinks and held the treatment code. |
| Garcia-Alvarez et al., 2020 · PMID 31990972 ↗ | 223 healthy adults aged 18–70 | One decaffeinated energy drink | Placebo; double-blind crossover | No meaningful benefit across measured mood, cognitive, behavioural or subjective-energy outcomes | One caffeine-free formulation and a five-hour observation window. |
| Chtourou et al., 2019 · PMID 31052322 ↗ | 19 male physical-education students | Caffeinated energy drink | Placebo; double-blind crossover | Improved selected reaction-time, handgrip and 30-second cycling outcomes | Small acute all-male study with multiple endpoints; the article reports no external funding and no conflict of interest. |
| Banks et al., 2024 · PMID 38197606 ↗ | 30 physically active men | Sugar-containing or non-caloric drink | Matched non-caloric placebo; crossover | No improvement in VO₂peak, time to exhaustion or peak power | A resistance-fatigue measure improved for one product; both raised resting systolic BP. |
| Scholey & Kennedy, 2004 · PMID 15549275 ↗ | 20 fasted, caffeine-deprived adults | Whole drink and ingredient fractions | Placebo; balanced five-condition crossover | Improved secondary-memory and speed-of-attention factors; other cognitive and mood outcomes were null | Small acute study with multiple outcomes; funded by GlaxoSmithKline Consumer Healthcare R&D. |
| Fletcher et al., 2017 · PMID 28446495 ↗ | 18 healthy young adults | 946 ml drink with 320 mg caffeine | 320 mg caffeine-only drink; double-blind crossover | Higher QTc at two hours and systolic BP at six hours in the drink condition | Small high-volume study, no caffeine-free placebo and no clinical-event outcome. |
| Krieger et al., 2025 · PMID 41374082 ↗ | 60 adults | 200 mg caffeine product, acutely and for 28 days | 10-kcal non-caffeinated placebo; double-blind parallel groups | Selected energy-expenditure, day-one fat-oxidation and task-specific cognitive differences; no sustained fat-oxidation difference at day 28 | Not evidence of weight loss; Accelerator Active SHOC Energy funded the study. |
| Oberhoffer et al., 2022 · PMID 35387431 ↗ | 27 healthy children and adolescents | Weight-adjusted drink with 3 mg caffeine/kg | Sugar-similar placebo; single-blind crossover | Higher mean systolic and diastolic BP during four-hour observation | Small acute study; no chronic-use or clinical-event outcome. |
Favourable product-specific findings
In a 24-person crossover simulator study, 21 participants were included in the analysis; the tested drink reduced lane weaving and self-reported sleepiness during later driving periods. This was not a road-crash study and does not make driving while dangerously tired safe. In a five-condition crossover study of 20 overnight-fasted, caffeine-deprived adults, the whole drink improved secondary-memory and speed-of-attention factors, but not the other cognitive or mood outcomes. [13] [15]
A sponsor-funded 28-day parallel-group trial in 60 adults found higher energy expenditure at selected times, day-one-only increases in fat oxidation and selected cognitive differences. It did not demonstrate weight loss. Detailed study cards, limitations and funding notes are available in the research map. [16]
Energy drink versus matched caffeine
In 11 trained male cyclists, both an energy drink and an equivalent dose of caffeine improved a time trial compared with placebo. The two active conditions did not differ significantly. The result supports an acute caffeine-related effect in that protocol, but not an added benefit from the drink’s other ingredients. [10]
Null findings in repeated sprint and endurance tests
A caffeine-taurine drink did not significantly improve repeated sprint time or anaerobic power in 20 college football players. In a separate 11-person cycling trial, completion time did not differ among an energy drink, caffeine-and-carbohydrate cola and placebo. The second study found changes in selected metabolic and immune measures, but those did not translate into faster performance. [17] [18]
Cognitive outcomes depend on formulation
A 24-person crossover study found differences among a sugar-containing drink, sugar-free drink and placebo on selected memory measures. By contrast, a 223-person trial of a decaffeinated drink found no meaningful benefit across its measured mood, cognitive, behavioural or subjective-energy outcomes. These studies tested different formulations and should not be treated as contradictory votes on one uniform product. [7] [19]
Physiological changes are not performance benefits
In the Banks trial, neither tested drink improved the main graded-exercise outcomes, although one isometric measure changed and both drinks increased systolic blood pressure at specified times. A small high-volume study also found time-specific QTc and systolic-blood-pressure differences versus caffeine alone. Such acute measures require outcome-specific interpretation and do not establish a clinical event or a performance benefit. [8] [20]
A statistically significant change in one endpoint does not make an entire trial “positive”. Primary and secondary outcomes, null findings, comparison drink, multiplicity, sample size and funding or conflicts all affect interpretation.
Individual response matters
Caffeine’s effects are not identical in every person. Body mass, habitual intake, sensitivity, genetics, sleep status, food consumption, medication use, pregnancy, health status, dose and timing can all influence response.
An energy drink may improve alertness or a specific performance measure in one situation while producing little benefit—or unwanted effects—in another. More caffeine does not necessarily produce a greater benefit.
Safety reference values are not performance targets
EFSA concluded that, for healthy adults, single caffeine doses up to 200 mg and habitual intake up to 400 mg per day do not raise safety concerns. Its exercise conclusion applies to the stated conditions and healthy-adult population. These amounts include caffeine from all dietary sources combined and are safety reference values, not recommended performance doses. [4]
Population and dose matter. In a randomised trial of 27 healthy children and adolescents, a weight-adjusted drink containing 3 mg caffeine/kg increased mean systolic and diastolic blood pressure versus a sugar-similar placebo during four hours of observation. The small acute study did not measure chronic effects or clinical events. [21]
Key conclusions
- Sugar contributes nutritional energy; caffeine provides stimulation but no calories.
- A caffeine-related alertness effect depends on dose, timing and individual response.
- Sugar-free does not mean caffeine-free, and caffeine-free does not mean calorie-free.
- Natural and synthetic caffeine are the same molecule when chemical identity and purity are equivalent; other compounds in botanical extracts may still affect a product.
- Sugar alone should not be expected to reliably counteract sleepiness.
- Some controlled trials found benefits in specific running, cycling, driving-simulation, reaction-time, memory or attention outcomes; these findings remain product-, dose-, population- and task-specific.
- Controlled trials also report null and mixed performance findings.
- A result for one product, population or task should not be generalised to the whole category.
- Acute physiological changes are not automatically performance benefits or evidence of clinical harm.
- Safety reference values are not performance targets.
Scientific sources
- EFSA — Caffeine and increased alertness (opens in a new tab)
EFSA Journal. 2014;12(2):3574. DOI: 10.2903/j.efsa.2014.3574
- European Commission — Regulation (EU) 2016/1411 (opens in a new tab)
Rejects a separate claim application with proposed conditions of use between 40 and 75 mg per serving; scientific opinion is not legal authorisation.
- European Parliament — Resolution on caffeine alertness and concentration claims (opens in a new tab)
7 July 2016: objection to a separate draft authorisation of caffeine health claims.
- EFSA — Scientific Opinion on the safety of caffeine (opens in a new tab)
EFSA Journal. 2015;13(5):4102. DOI: 10.2903/j.efsa.2015.4102
- Anderson & Horne — High sugar, low caffeine and sleepiness (opens in a new tab)
Human Psychopharmacology. 2006;21(5):299–303. PMID: 16856218
- Krieger et al. — Natural and synthetic caffeine (opens in a new tab)
Clinical and Translational Science. 2016;9(5):246–251. DOI: 10.1111/cts.12403
- Wesnes et al. — Sugar-containing and sugar-free energy drinks (opens in a new tab)
Journal of Psychopharmacology. 2017;31(2):211–221. DOI: 10.1177/0269881116681459
- Banks et al. — Commercial energy drinks and exercise performance (opens in a new tab)
JISSN. 2024;21(1):2297988. DOI: 10.1080/15502783.2023.2297988
- Giles et al. — Caffeine, taurine and glucose (opens in a new tab)
Pharmacology Biochemistry and Behavior. 2012;102(4):569–577. DOI: 10.1016/j.pbb.2012.07.004
- Quinlivan et al. — Energy drink, caffeine and cycling (opens in a new tab)
International Journal of Sports Physiology and Performance. 2015;10(7):897–901. PMID: 25710190.
- Prins et al. — Five-kilometre running performance (opens in a new tab)
Journal of Strength and Conditioning Research. 2016;30(11):2979–2990. PMID: 26937774. DOI: 10.1519/JSC.0000000000001391
- Ivy et al. — Cycling time-trial performance (opens in a new tab)
International Journal of Sport Nutrition and Exercise Metabolism. 2009;19(1):61–78. PMID: 19403954. DOI: 10.1123/ijsnem.19.1.61
- Mets et al. — Simulated prolonged driving performance (opens in a new tab)
Psychopharmacology. 2011;214(3):737–745. PMID: 21063868. DOI: 10.1007/s00213-010-2078-2
- Chtourou et al. — Reaction time and short-term performance (opens in a new tab)
Nutrients. 2019;11(5):992. PMID: 31052322. DOI: 10.3390/nu11050992
- Scholey & Kennedy — Memory and speed of attention (opens in a new tab)
Psychopharmacology. 2004;176(3–4):320–330. PMID: 15549275. DOI: 10.1007/s00213-004-1935-2
- Krieger et al. — Energy expenditure, fat oxidation and cognition (opens in a new tab)
Nutrients. 2025;17(23):3793. PMID: 41374082. DOI: 10.3390/nu17233793
- Gwacham & Wagner — Repeated sprint performance (opens in a new tab)
International Journal of Sport Nutrition and Exercise Metabolism. 2012;22(2):109–116. PMID: 22349209.
- Phillips et al. — Simulated cycling race (opens in a new tab)
Journal of Strength and Conditioning Research. 2014;28(5):1443–1453. PMID: 24172719.
- Garcia-Alvarez et al. — Decaffeinated drink and mental energy (opens in a new tab)
American Journal of Clinical Nutrition. 2020;111(3):719–727. PMID: 31990972.
- Fletcher et al. — High-volume drink, ECG and blood pressure (opens in a new tab)
Journal of the American Heart Association. 2017;6:e004448. PMID: 28446495.
- Oberhoffer et al. — Acute paediatric blood-pressure response (opens in a new tab)
Frontiers in Cardiovascular Medicine. 2022;9:862041. PMID: 35387431.