Sports Nutrition

Sports Nutrition

Sports nutrition is first a problem of energy availability and of matching carbohydrate, fluid, and sodium to the work, the heat, and the sport. Protein, a short list of ergogenic compounds, and a few micronutrient deficiencies matter after that. Most of the commercial remainder — proprietary blends, the one-hour “anabolic window,” fasted-training mythology, and ketogenic promises for high-intensity performance — either fails in human performance trials, works only in a narrow stratum, or moves a surrogate that is not the race.

Compiled by South Beach Longevity · 20 August 2026 Copyright 2026 Series SBL-41 / SP-SPORTS-NUTRITION · Register A scientific article Sources peer-reviewed human trials, consensus statements, meta-analyses, and labelled animal or in-vitro work · verified NCBI records Constraint This document describes published research. It is not medical advice. No human use, dose, route or schedule is recommended anywhere in this document.

How to read this document Every finding is labelled, in the sentence that reports it, by the kind of study that produced it. A randomised time trial is not a pathway cartoon. Muscle protein synthesis over hours is not hypertrophy over months. A recreational cohort is not an elite cohort. A dietary supplement is not a drug and is not a prohibited method. Where two results conflict, both are given. Amounts and durations appear only as reported experimental parameters, always with the population attached. Nothing here is a recommendation. Findings are graded in place as established, strongly supported, emerging, plausible, or speculative. Two further labels mark careful absences rather than verdicts: not established, where the evidence is too thin to place a claim on the ladder at all — untested or insufficient, an absence of proof rather than disproof; and not supported, where the weight of evidence leans against a claim but stops short of a formal refutation.

01 What sports nutrition is, and four things it is not

Sports nutrition is the study of how dietary energy and nutrients constrain, enable, or fail to change human exercise performance, training adaptation, and athlete health. It is a branch of exercise physiology that happens to use food. The 2016 joint position of the Academy of Nutrition and Dietetics, Dietitians of Canada, and the American College of Sports Medicine is the field’s current map, not its constitution: it treats periodized energy and macronutrient intake, hydration, and a short list of evidence-graded supplements as the working core (Thomas, Erdman, and Burke, 2016). That paper is a consensus synthesis. It is strongly supported as a description of what the societies then judged the human record to say. It is not a licence to treat every number in it as a prescription, and it is not used that way here.

Four things follow immediately and will be enforced for the rest of this document.

First, sports nutrition is not a supplement catalogue. The IOC consensus on dietary supplements and the high-performance athlete is explicit that most products have little or no evidence, that a few have a performance effect in defined settings, and that contamination and inadvertent doping are part of the same record (Maughan et al., 2018). Peeling, Binnie, Goods, Sim, and Burke, writing an evidence-based supplement shortlist, kept the list short on purpose (Peeling et al., 2018).

Second, it is not pharmacology and it is not doping. Creatine monohydrate, caffeine, and a glucose–fructose drink sit in one legal and evidentiary class. Anabolic-androgenic steroids, erythropoietin, and stimulant doping sit in another. A contaminated “vitamin” tablet that contains an undeclared steroid is not sports nutrition that went slightly wrong. It is a pharmaceutical exposure sold as food (Geyer et al., 2004).

Third, it is not the same science for every sport. A 100 m, a marathon, a wrestling cut, a soccer midfield shift, and a 24-hour ultra ask different questions of glycogen, fluid, and lean mass. Holway and Spriet’s team-sport review exists because the endurance textbook does not travel unchanged into intermittent play (Holway and Spriet, 2011).

Fourth, it is not medical nutrition therapy. This document describes published research. It recommends no diet, product, dose, route, or schedule for any person.

Sibling articles in this series take macronutrients, amino-acid nutrition, micronutrients, and high-protein diets as general-nutrition problems. This title is the exercise-physiology cut of the same material. Dietary-protein evidence is not spent here as if it were isolated-amino-acid evidence. Micronutrient essentiality is not spent as if it were a reason to supplement the replete athlete.

02 Energy availability versus expenditure

Energy availability (EA) is dietary energy intake minus exercise energy expenditure, expressed relative to fat-free mass. It is not energy balance. Energy balance is intake minus total expenditure, including the resting, thermic, and non-exercise remainder. A person can be in approximate energy balance — weight stable — and still have EA low enough to disrupt endocrine function, because the brain appears to see the energy left after exercise, not the bathroom scale (Loucks, Verdun, and Heath, 1998; Loucks, 2011). That distinction is established in short-term human experiments and is the load-bearing concept of this part.

Loucks, Verdun, and Heath randomised exercising women to low EA versus a comparison condition that isolated exercise stress, and reported that luteinizing-hormone (LH) pulsatility changed with EA, not with the exercise stress itself (Loucks, Verdun, and Heath, 1998). Loucks and Thuma then set EA at 45 kcal·kg⁻¹ fat-free mass·d⁻¹ versus 10, 20, or 30 kcal·kg⁻¹ fat-free mass·d⁻¹ for five days in 29 regularly menstruating, habitually sedentary young women of normal body composition, with exercise fixed at 15 kcal·kg⁻¹ fat-free mass·d⁻¹ at 70% of aerobic capacity. LH pulsatility was unaffected at 30 and was disrupted below that threshold (Loucks and Thuma, 2003). Ihle and Loucks, in a related five-day dose–response experiment in young exercising women, reported that bone-turnover markers moved with EA in a dose-responsive way (Ihle and Loucks, 2004). These are short, tightly controlled human experiments, not season-long field studies. They are strongly supported as evidence that EA has a reproductive and bone cost on a timescale of days. They are not evidence that 30 kcal·kg⁻¹ fat-free mass·d⁻¹ is a universal clinical cut-point for every athlete, every sport, or every duration. The number is a reported experimental threshold in a defined sample. It has been asked to do more work than a five-day protocol can support, and later RED-S writing has had to live with that (Mountjoy et al., 2018; Heikura et al., 2026).

Expenditure measurement is its own failure mode. Doubly labelled water, room calorimetry, and wearable estimates do not agree closely enough that a field “calorie target” can be treated as a laboratory quantity. Self-reported intake in athletes is systematically biased. Thomas, Erdman, and Burke treat periodized fueling as a planning problem precisely because both sides of the EA equation are noisy (Thomas, Erdman, and Burke, 2016). A wearable that reports a huge session expenditure, followed by a restriction that looks like discipline, is how laboratory EA becomes a field injury.

03 RED-S, the triad, and the cost of training hungry

The Female Athlete Triad — low energy availability, with or without disordered eating; menstrual dysfunction; and low bone mineral density — is the ACSM’s older, still-valid framing (Nattiv et al., 2007). The 2014 Female Athlete Triad Coalition consensus restated treatment and return-to-play around that triad (De Souza et al., 2014). Melin and colleagues, in elite endurance athletes, found low EA and triad components at a prevalence that makes the problem ordinary rather than exotic (Melin et al., 2015).

The IOC then named Relative Energy Deficiency in Sport (RED-S): a broader syndrome in which low EA impairs physiological function beyond the triad, including metabolic rate, immunity, protein synthesis, cardiovascular health, and performance, in female and male athletes (Mountjoy et al., 2014). The 2018 IOC update kept the construct, tightened the clinical language, and recorded the argument with triad traditionalists rather than pretending it had vanished (Mountjoy et al., 2018). That argument is real. The triad is a tighter, better-measured three-component model. RED-S is a wider clinical umbrella whose extra systems are unequally evidenced. Both names point at the same first cause: not enough energy left after training. Treating the naming dispute as if it cancelled the physiology is a category error.

Established: short-term low EA disrupts LH pulsatility and moves bone-turnover markers in controlled human experiments (Loucks, Verdun, and Heath, 1998; Loucks and Thuma, 2003; Ihle and Loucks, 2004). Strongly supported: low EA clusters with menstrual dysfunction and impaired bone health in athletic cohorts (Nattiv et al., 2007; Melin et al., 2015). Emerging: the full multi-system RED-S inventory and the severity tools proposed to grade it (Mountjoy et al., 2018; Heikura et al., 2026). Male athletes are not a footnote and are not as well measured. Performance impairment from low EA can precede a diagnosis anyone will bill. Overtraining syndrome shares symptoms and is not the same disease (Stellingwerff et al., 2021, in the RED-S overlap literature).

This section does not recommend a restoration diet. It records that the human experiments restored EA, not that a reader should.

04 Body composition and weight-class sport

Body composition is a measurement problem before it is a nutritional one. Dual-energy X-ray absorptiometry, surface anthropometry, and bioimpedance do not return the same lean-mass number. Seasonal change in an elite athlete is often smaller than the error of the cheap method.

Weight-class and aesthetic sports add a rule-book constraint that the physiology did not ask for. Sundgot-Borgen and Garthe, reviewing elite athletes in aesthetic and Olympic weight-class sports, described widespread use of extreme weight-control methods and noted that the rules of some sports are themselves a risk factor (Sundgot-Borgen and Garthe, 2011). That is a narrative review of a methodologically uneven literature, and they said so. It is still the right problem statement.

Garthe, Raastad, Refsnes, Koivisto, and Sundgot-Borgen randomised elite athletes to two rates of weight loss — approximately 0.7% versus 1.4% of body weight per week — and compared body composition and strength- and power-related performance (Garthe et al., 2011). The slower rate better preserved lean mass and performance in that sample. Follow-up papers from the same group tracked longer-term composition and counselling effects (Garthe, Raastad, and Sundgot-Borgen, 2011). These are small elite randomised trials, not population nutrition. They are strongly supported as evidence that the rate of a cut changes what is lost. They are not a licence for any particular weekly percentage in any particular sport.

Making weight by dehydration is a hydration problem (Part Three) wearing a weight-class costume. Making weight by collapsing EA is a RED-S problem wearing a medal. Neither is a body-composition success.

05 Glycogen: what was measured, and what it is asked to explain

The modern glycogen story is a biopsy story. Bergström and Hultman showed, in human muscle, that exercise lowers glycogen and that diet after depletion changes how much is restored (Bergström and Hultman, 1966; Bergström et al., 1967). Bergström, Hermansen, Hultman, and Saltin then tied diet, muscle glycogen, and endurance performance in the same experimental tradition (Bergström et al., 1967). Those experiments are established. They are why carbohydrate is not a lifestyle preference in this document.

Sherman, Costill, Fink, and Miller showed that exercise–diet manipulation changes muscle glycogen and its subsequent use during performance (Sherman et al., 1981). Coyle and colleagues showed that feeding carbohydrate during prolonged strenuous exercise alters muscle glycogen use (Coyle et al., 1986). The field has been arguing about how much, when, and in which events ever since. It has not been arguing about whether glycogen is real.

What glycogen is asked to explain, and cannot, is every bad session. Central fatigue, hyperthermia, hyponatremia, inadequate protein remodeling, iron deficiency, and low EA all produce “I faded.” A low glycogen biopsy is a mechanism. It is not a diagnosis of the race.

06 Endurance feeding and carbohydrate loading

Burke, Hawley, Wong, and Jeukendrup reviewed carbohydrates for training and competition and restated the practical shape of the human evidence: daily carbohydrate needs scale with training load; pre-competition glycogen supercompensation remains useful for events that would otherwise empty the store; and during-event feeding matters once duration and intensity make endogenous carbohydrate limiting (Burke et al., 2011). Thomas, Erdman, and Burke carry the same architecture into the 2016 joint position (Thomas, Erdman, and Burke, 2016). These are syntheses. The primary biopsy and feeding trials sit underneath them. The grade is strongly supported for prolonged continuous endurance. It is weaker for brief power events, in which the store is not the limiter.

Classic loading protocols combined a depletion phase with a high-carbohydrate restoration phase. Later practice often dropped the depletion phase because the restoration, not the suffering, is what fills the tank (Sherman et al., 1981; Burke et al., 2011). Reported loading intakes in the guideline literature are experimental and consensus parameters for defined endurance events. They are not a general diet.

07 During-exercise fueling and multiple transportable carbohydrates

Intestinal glucose absorption saturates. Jeukendrup’s review of multiple transportable carbohydrates is the mechanistic and applied statement of that fact: combining glucose (or maltodextrin) with fructose uses more than one intestinal transporter, raises exogenous carbohydrate oxidation above the older ~1 g·min⁻¹ ceiling, and can improve fluid delivery (Jeukendrup, 2010). Reported exogenous oxidation in that literature reaches about 1.75 g·min⁻¹ under high-intake experimental conditions. That is a measured oxidation rate in feeding studies, not a menu.

Currell and Jeukendrup tested the performance claim in eight trained male cyclists (mean VO₂max 64.7 ml·kg⁻¹·min⁻¹). Subjects ingested water, glucose at 1.8 g·min⁻¹, or glucose plus fructose at 1.8 g·min⁻¹ during 120 min at 55% Wmax, then a time trial. The mixed-carbohydrate drink improved time-trial completion versus glucose alone and versus water (Currell and Jeukendrup, 2008). That is a small, trained, male, cycling, laboratory time-trial. It is strongly supported for that setting. It is not a soccer result, not a female result, and not a 5 km result.

Jeukendrup later argued for personalization of during-exercise carbohydrate by event duration and gut training (Jeukendrup, 2014). Gut training is a plausible practice with a thinner trial base than the oxidation work. It is emerging.

Carbohydrate products — drinks, gels, bars — are delivery vehicles for the same monomers. They are not a separate pharmacology. Their trial evidence is the feeding evidence. Their marketing evidence is not.

08 Recovery glycogen

Restoration of muscle glycogen after hard endurance work is carbohydrate-dependent in the human biopsy record (Bergström et al., 1967; Coyle et al., 1986; Burke et al., 2011). Immediate post-exercise carbohydrate accelerates early resynthesis; the 24-hour total still dominates if the next session is tomorrow rather than in two hours (Burke et al., 2011; Thomas, Erdman, and Burke, 2016). Protein co-ingestion can help when carbohydrate is suboptimal; it is not a substitute for the carbohydrate the store is made of.

This is one of the few places where “timing” is not a slogan. It is a measured delay in a biochemical restoration. It still does not make the next session’s protein shake into an emergency.

09 Train-low, fasted training, and periodized carbohydrate

Training with low carbohydrate availability can increase selected molecular signals of endurance adaptation. That is not the same as improving the race. Impey and colleagues proposed “fuel for the work required”: a theoretical and practical attempt to keep train-low sessions from becoming a chronic low-EA diet (Impey et al., 2016; Impey et al., 2018). The glycogen-threshold hypothesis in the 2018 paper is a framework, not a confirmed dose–response. Grade: emerging as a periodization idea; not established as a performance superior to consistently high carbohydrate availability in championship endurance events.

Hawley’s train-low programme and the later periodization literature sit on this tension: the session you deliberately underfuel may be a worse session, offered as payment for a later adaptation. Whether the adaptation survives into a glycogen-loaded race is the only question that matters, and it is the question with the least clean evidence.

Fasted training is a popular subset. It is not a fat-loss method with a performance guarantee. It is a low-carbohydrate-availability session whose costs include reduced high-intensity output and, if repeated carelessly, a slide toward low EA (Loucks, 2011; Impey et al., 2018). The Instagram version — “burn more fat, therefore race better” — confuses substrate selection with performance. Fat oxidation can rise while economy and high-intensity capacity fall. Part Three returns to that as the ketogenic problem.

10 Protein: requirement, MPS, resistance, endurance, older athletes, deficit

Muscle protein synthesis (MPS) is an acute tracer measurement. Hypertrophy is a tissue outcome over weeks to months. The house already made this point in the amino-acid article. It is restated here because sports-nutrition marketing spends MPS as if it were lean mass.

Moore and colleagues described an ingested-protein dose response of mixed-muscle and albumin protein synthesis after resistance exercise in young men, with a plateau in the neighbourhood of a single ~20 g whey feeding in that sample (Moore et al., 2009). Macnaughton and colleagues later reported that after whole-body resistance exercise, 40 g whey raised MPS more than 20 g (Macnaughton et al., 2016). Those two human tracer studies do not cancel each other. They say the dose that saturates MPS depends on how much muscle was just used. Areta and colleagues showed that distributing ~80 g protein as four 20 g feedings over 12 hours of recovery supported myofibrillar synthesis better than two 40 g or eight 10 g feedings in trained men (Areta et al., 2013). Moore, Churchward-Venne, Witard and colleagues, pooling laboratory data, reported that older men required a larger relative protein dose than younger men to stimulate myofibrillar synthesis (Moore et al., 2015). Witard and colleagues’ later narrative review is a map of this dose–timing–age literature, not a new trial (Witard et al., 2016).

Morton, Murphy, McKellar, Schoenfeld and colleagues’ systematic review, meta-analysis, and meta-regression asked the outcome question: does protein supplementation increase resistance-training gains in muscle mass and strength in healthy adults? They concluded that supplementation beyond habitual intake has a diminishing return, with little additional fat-free-mass benefit above a total daily intake they estimated near 1.6 g·kg⁻¹·d⁻¹ in that evidence base, and that the effect is smaller in trained people (Morton et al., 2018). That is the load-bearing hypertrophy paper in this document. It is a meta-analysis of training studies, not a tracer study. Grade: strongly supported for the diminishing-return shape; the exact 1.6 figure is a modelled estimate, not a biological constant.

Endurance athletes still remodel mitochondrial and myofibrillar protein. Their binding constraint is more often carbohydrate and EA than a failure to drink whey. Protein in an energy deficit is a preservation problem: Hector and Phillips’s line of work, and the Garthe cutting trials, say that a cut done too fast costs lean mass (Garthe et al., 2011). Older athletes inherit anabolic resistance from the ageing literature (Moore et al., 2015). None of these sentences is a dose recommendation.

11 Timing and the “anabolic window”

The one-hour post-workout emergency is a commercial object. Aragon and Schoenfeld reviewed the human evidence and concluded that the window is wider than an hour, that total daily protein dominates, and that pre-exercise protein can overlap the post-exercise period when the session is not done fasted after a long fast (Aragon and Schoenfeld, 2013). Schoenfeld later returned to the controversy and did not restore the emergency (Schoenfeld, 2018). Kerksick and colleagues’ ISSN nutrient-timing position stand is more generous to peri-workout feeding but still treats daily intake as the first term (Kerksick et al., 2017). ISSN position stands are society statements. They are cited here as statements, graded strongly supported on the narrow claim that immediate feeding is not magic, and emerging-to-speculative on any claim that timing outruns total intake.

Trommelen and van Loon reviewed pre-sleep protein as a way to feed the overnight period, when amino-acid availability otherwise limits overnight synthesis (Trommelen and van Loon, 2016). That is a mechanistic and applied review of a small human literature. It is emerging. It is not a requirement that every athlete eat cottage cheese at 23:00.

Nutrient timing for carbohydrate (sections 07–08) is a different claim with a better endurance evidence base. Collapsing both timings into one “window” is how a glycogen fact becomes a whey advertisement.

12 Fat: essential fatty acids, fat adaptation, ketogenic performance

Essential fatty acids are a deficiency topic. They are not an ergogenic topic. This document will not launder an EFA requirement into a fish-oil performance claim. McGlory and colleagues, in a randomised human study, found that fish-oil supplementation suppressed some anabolic signalling without changing myofibrillar protein synthesis after resistance exercise and feeding (McGlory et al., 2016). That is a caution, not a recommendation.

Fat adaptation is real as a substrate shift. It is not free. Phinney’s 1980 work showed that obese subjects adapted to a hypocaloric ketogenic diet could perform moderate exercise (Phinney et al., 1980). That population and that intensity are not elite race walking. Volek and colleagues described metabolic characteristics of self-selected keto-adapted ultra-endurance runners in a cross-sectional comparison (Volek et al., 2016). Cross-sectional keto-adapted cohorts are not randomised performance tests.

Burke, Sharma, Heikura and colleagues reproduced, in elite race walkers, an impairment of exercise economy and performance on a ketogenic low-carbohydrate, high-fat diet (Burke et al., 2020). Burke later stated the conclusion in the title: ketogenic diets are not beneficial for athletic performance, and then defended that reading against a published challenge (Burke, 2024; Burke, 2024 response). Shaw, Merien, Braakhuis, Maunder, and Dulson reported that a ketogenic diet reduced exercise efficiency in runners at submaximal intensities (Shaw et al., 2019). Whitfield and colleagues reported that an acute ketogenic diet and ketone-ester supplementation impaired race-walk performance (Whitfield et al., 2021). Grade: strongly supported that LCHF/ketogenic adaptation impairs economy and high-intensity endurance performance in well-trained and elite athletes in the Burke and Shaw designs; plausible that some ultra-endurance settings with low relative intensity tolerate higher fat oxidation; not established that ketosis is a performance upgrade.

13 Hydration, sweat, sodium, dehydration, overhydration, hyponatremia, heat

Sweat rate and sweat sodium vary by person, heat, clothing, and intensity. Baker’s sweat work and the ACSM fluid-replacement stands exist because a single bottle rule cannot cover that variance (American College of Sports Medicine et al., 2007; McDermott et al., 2017). The 2007 ACSM stand (Sawka, Burke, Eichner, Maughan, Montain, Stachenfeld) treats prevention of excessive dehydration and of overhydration as the same problem (American College of Sports Medicine et al., 2007). Casa’s earlier NATA statement and McDermott’s later NATA update are the athletic-training versions of the same tension (Casa et al., 2000; McDermott et al., 2017).

Dehydration impairs heat dissipation and, past a point that depends on the event, performance. Overhydration kills people. Almond, Shin, Fortescue and colleagues measured hyponatremia among Boston Marathon runners and linked it to weight gain from fluid, long race time, and small body size (Almond et al., 2005). That is a field epidemiology paper with laboratory sodium measurements. It is established as a description of race-day hyponatremia. Hew-Butler and the Third International Exercise-Associated Hyponatremia Consensus stated that the primary cause is fluid intake in excess of losses, with a contribution from non-osmotic ADH, and that treatment is not “more sport drink” (Hew-Butler et al., 2015). Later updates kept that core (Hew-Butler, Loi, Rosner, and Dugas, 2017).

Electrolyte products are not an antidote to overdrinking. They are a sodium-and-fluid delivery problem. Heat makes both dehydration and overdrinking more likely because thirst, opportunity, and fear of heat stroke travel together. Weight-class rapid-cut dehydration is a deliberate version of the first risk.

Table C in the Apparatus is the hydration matrix. The only general sentence that survives it is: drink to limit both large body-mass loss and body-mass gain, in the heat, with a plan that was tested in training. That sentence is a restatement of the consensus papers. It is not a personal fluid prescription.

14 Iron, vitamin D, calcium, and the rest of the athlete-specific list

Athlete-specific micronutrient problems are deficiency problems that training can create or reveal. They are not a licence for a multivitamin as ergogenic insurance. The micronutrients article of this series is the general frame. What belongs here is the sport cut.

Iron. Endurance training raises hepcidin after exercise and can impair iron absorption in the hours that follow (Peeling et al., 2009; Sim, Dawson, Landers, Trinder, and Peeling, 2013–2014 line). Sim and colleagues’ 2019 narrative review is the current athlete-facing map: females, endurance runners, and athletes with already low stores are the strata that matter; haemoglobin is a late marker; ferritin interpretation is inflammation-dependent (Sim et al., 2019). Grade: strongly supported that exercise-induced hepcidin and menstrual iron loss create a real deficiency risk in defined athlete groups. Not established that iron supplementation improves performance in replete athletes.

Vitamin D. Close and colleagues measured winter 25-hydroxyvitamin D in professional athletes and healthy adults and found a high prevalence of low concentrations in that UK winter sample (Close et al., 2013). Owens, Fraser, and Close reviewed the athlete literature and later reported a high-dose supplementation trial in elite athletes (Owens, Fraser, and Close, 2015; Owens et al., 2017; Owens, Allison, and Close, 2018). Bone, muscle, and infection claims beyond correction of frank deficiency remain mixed. Allison and colleagues reported no association between vitamin D deficiency and markers of bone health in one athletic sample, and a heart-size association in another — both observational (Allison et al., 2015). Grade: strongly supported that winter indoor athletes can be biochemically low; emerging to conflicting that raising 25-hydroxyvitamin D in already sufficient athletes changes performance.

Calcium. The bone limb of the triad/RED-S is a calcium-and-energy problem, not a calcium-tablet problem. Low EA moves bone turnover before a calcium supplement can be the story (Ihle and Loucks, 2004; Nattiv et al., 2007).

Other athlete-specific issues — sodium in heavy sweaters, low energy-driven micronutrient collapse, injury nutrition (Close et al., 2019) — inherit the same rule: name the deficiency or the healing constraint, or do not open the bottle.

15 Endurance

Prolonged continuous work is the textbook: glycogen, during-event carbohydrate, EA, heat, and iron. Burke et al. (2011) and Thomas, Erdman, and Burke (2016) are written for this athlete. Multiple transportable carbohydrates earn their keep here (Currell and Jeukendrup, 2008; Jeukendrup, 2010). Ketogenic substitution is a performance cost in high-intensity endurance (Burke et al., 2020; Shaw et al., 2019). Caffeine has its cleanest endurance meta-analytic signal here (Southward, Rutherfurd-Markwick, and Ali, 2018; Guest et al., 2021).

16 Resistance and physique sport

The binding constraints are progressive tension, total protein, energy surplus or a controlled deficit, and creatine. Morton et al. (2018) is the supplementation-and-hypertrophy paper. Garthe’s slower-cut trial is the physique-and-power paper (Garthe et al., 2011). The anabolic window is not the programme (Aragon and Schoenfeld, 2013). Carbohydrate still fuels the session; it is not the 1967 biopsy drama of a marathon. Beta-alanine and bicarbonate are optional and event-specific (sections 23).

17 Sprint and power

Events lasting seconds to ~60 s are phosphagen and glycolytic. Creatine has its strongest traditional claim here (Kreider et al., 2017; Branch, 2003). Bicarbonate and beta-alanine target hydrogen-ion buffering in slightly longer efforts (Hobson et al., 2012; Saunders et al., 2017; Grgic et al., 2021). Carbohydrate loading is usually idle. Protein is remodeling, not fuel. Caffeine’s strength-and-power meta-analyses show small effects that a statistician should not inflate (Grgic et al., 2018; Grgic et al., 2020).

18 Team sports

Holway and Spriet described team sport as intermittent, positional, and congested: glycogen can still limit a match or a midweek turnaround, but the feeding problem is stoppage and gut comfort, not a four-hour bike ride (Holway and Spriet, 2011). Carbohydrate, caffeine, creatine, and fluid are the recurring tools. Nitrate’s sprint-and-cognitive finding in prolonged intermittent exercise is a laboratory signal, not a league table (Thompson et al., 2015). Fixture congestion is an EA and recovery problem wearing a broadcast schedule.

19 Ultra-endurance

Hours to days invert some textbook priorities. Absolute intensity often falls; eating becomes a GI and sleep problem; hyponatremia risk rises because there is so much time to overdrink (Hoffman, Fogard, and Hew-Butler line; Almond et al., 2005; Hew-Butler et al., 2015). Volek’s keto-adapted ultra cohort is a substrate description (Volek et al., 2016). It does not cancel Burke’s economy impairment at higher intensities. Tiller’s ultra-physiology papers are a reminder that the event itself is an injury (Tiller et al., 2019; Tiller et al., 2021). Sex-difference claims of a female ultra advantage are physiological hypotheses, not nutrition protocols (Tiller et al., 2021).

20 Weight-class sport, restated as a nutritional constraint

The cut is the intervention. Sundgot-Borgen and Garthe (2011) and Garthe et al. (2011) are the evidence. Rapid dehydration plus low EA plus a same-day or next-day performance is a stacked risk: heat, hyponatremia rebound, lean-mass loss, and RED-S. “Making weight” is not a body-composition method. It is a rule compliance that nutrition science can only make less damaging, and only in trials that measured the damage.

21 Female athletes as a required stratum, not a footnote

Almost every convenient sports-nutrition trial is young, male, and trained-but-not-elite. Loucks’s EA experiments, the triad consensus, Melin’s elite endurance work, and the IOC RED-S statements exist because that convenience sample is the wrong default (Loucks and Thuma, 2003; Nattiv et al., 2007; Melin et al., 2015; Mountjoy et al., 2014, 2018). Iron (Sim et al., 2019), bone, and menstrual function are not “women’s extras.” They are part of the performance and health record. Where this document reports a male-only trial — Currell and Jeukendrup (2008), Moore et al. (2009), Areta et al. (2013) — it says so. Generalizing those numbers to female athletes without a female measurement is speculative.

22 How to read a supplement claim

Maughan and colleagues’ IOC consensus is the reading protocol (Maughan et al., 2018). Peeling and colleagues’ shortlist is the applied filter (Peeling et al., 2018). The statistical rules this title will not relax:

1. Endpoint. A time trial, a 1RM, or a match-running metric outranks an enzyme, a gene, or a “pump.”
2. Size. Championship margins are real. So is publication bias. Grgic’s caffeine umbrella review of 21 meta-analyses is the cautionary object: many small positive effects, overlapping samples, and a need to ask whether the smallest worthwhile change was predefined (Grgic et al., 2020).
3. Population. Recreational responders are not elite non-responders. Elite sample sizes are small. Both facts cut against slogans.
4. Control. Placebo, blinding, and caffeine habituation matter more in this literature than in almost any other nutrition literature.
5. Contamination. Geyer and colleagues analysed 634 non-hormonal supplements purchased in 13 countries from 215 suppliers (October 2000–November 2001) and found undeclared anabolic-androgenic steroids in a material fraction (Geyer et al., 2004). A later tablet case found high amounts of 17-methylated steroids in a product sold as a dietary supplement (Parr, Geyer, and Schänzer, 2007). Proprietary blends exist to hide dose. They also hide what is in the bottle.

23 Creatine, caffeine, beta-alanine, nitrate/beetroot, bicarbonate

Creatine. Branch’s meta-analysis reported effects on body composition and performance consistent with an increase in phosphocreatine-dependent work (Branch, 2003). The 2017 ISSN position stand (Kreider, Kalman, Antonio and colleagues) treats creatine monohydrate as the most evidenced legal ergogenic aid for high-intensity exercise and lean-mass gain, and reviews a large safety literature (Kreider et al., 2017). Grade: established for repeated high-intensity performance and training-quality support in mixed athletic samples; not a endurance-economy gift; not a steroid.

Caffeine. Guest and colleagues’ ISSN 2021 stand, built on the meta-analytic pile, treats caffeine as ergogenic across endurance, high-intensity, and some strength/power settings, with the usual caveats of sleep, anxiety, habituation, and inter-individual CYP1A2 variation (Guest et al., 2021). Southward, Rutherfurd-Markwick, and Ali’s endurance meta-analysis, and its correction, are the endurance number-stack (Southward et al., 2018). Grgic’s umbrella review is the skepticism stack (Grgic et al., 2020). Grade: established that caffeine can improve performance in many protocols; strongly supported that the typical effect is small; not established that more is better, or that a proprietary “pre-workout” blend is caffeine.

Beta-alanine. Hobson, Saunders, Ball, Harris, and Sale’s meta-analysis found a small effect, largely in efforts of 1–4 min (Hobson et al., 2012). Saunders and colleagues’ later systematic review and meta-analysis confirmed a small improvement in exercise capacity/performance and located it in the same duration band (Saunders et al., 2017). Dolan and colleagues added a risk assessment (Dolan et al., 2019). Paraesthesia is the common adverse effect in the trial record. Grade: strongly supported for a small effect in mid-duration high-intensity efforts; not established for marathon or maximal strength.

Nitrate / beetroot. Bailey, Winyard, Vanhatalo and colleagues showed, in a human laboratory study, that dietary nitrate reduced the oxygen cost of low-intensity exercise and improved high-intensity tolerance (Bailey et al., 2009). Jones’s subsequent reviews are the mechanistic and applied map (Jones, 2014; Jones, Thompson, Wylie, and Vanhatalo, 2018). Effects appear larger in recreational than in elite endurance athletes in the later narrative — a generalizability problem, not a footnote. Grade: strongly supported for a reduced oxygen cost and improved tolerance in some non-elite samples; emerging to conflicting in elite endurance; not a universal beetroot mandate.

Bicarbonate. The ISSN 2021 bicarbonate stand (Grgic et al., 2021) and the older McNaughton experiments treat sodium bicarbonate as a buffering agent for high-intensity efforts, with GI distress as the tax (McNaughton, 1992; Grgic et al., 2021). Combining bicarbonate with beta-alanine is not clearly better than either (Curran-Bowen et al., 2024). Grade: strongly supported for some high-intensity protocols; practical use is limited by the gut.

24 Carbohydrate products, protein supplements, electrolytes

These are food in a wrapper. Their evidence is the evidence in Parts Two and Three. A gel is a fructose-and-glucose delivery system (Jeukendrup, 2010). A whey isolate is a leucine-rich protein dose (Moore et al., 2009; Morton et al., 2018). An electrolyte drink is a sodium-and-fluid vehicle (American College of Sports Medicine et al., 2007). Brand, flavour, and “proprietary matrix” do not create a new compound class. When a carbohydrate product is spiked with an undeclared stimulant or steroid, it leaves this section and enters section 25.

25 Contamination, proprietary blends, and the doping boundary

Geyer et al. (2004) is the contamination type specimen: undeclared anabolic-androgenic steroids in products that were not sold as steroids. The IOC supplement consensus treats batch testing and third-party certification as risk reduction, not as a guarantee (Maughan et al., 2018). Proprietary blends prevent the only check a scientist can do without a laboratory: reading the dose.

Dietary supplements, sports foods, and prohibited pharmacological agents must stay in separate sentences. Creatine and caffeine are legal dietary compounds with trial evidence. Erythropoietin, anabolic steroids, and many stimulants are medicines or doping agents. A “prohormone” sold as a supplement is a steroid exposure. This document will not provide a pharmacology or doping protocol. It will not treat a positive test as a sports-nutrition failure. It is a supply-chain and regulatory failure.

26 What the controversies actually turn on

They turn on five substitutions.

1. EA is replaced by calories-in-calories-out, and the endocrine cost disappears.
2. MPS is replaced by hypertrophy, and a four-hour infusion becomes a twelve-week physique.
3. Fat oxidation is replaced by performance, and a ketogenic economy loss is sold as metabolic flexibility.
4. A small meta-analytic effect is replaced by a championship, without asking about bias, blinding, or the smallest worthwhile change.
5. A legal food is replaced by a bottle, and then the bottle is replaced by an undeclared drug.

The controversies are not mysterious. They are those substitutions, repeated.

Table A. Supplement efficacy (human performance or athlete-health evidence)

Quantities are study/consensus parameters, not recommendations. Effect-size language is qualitative because championship SWC and publication bias forbid a single stolen percentage.

Table B. Sport-specific intake matrix (what the problem is)

Table C. Hydration matrix

Table D. Athlete-population trial matrix

Evidence handling

Study type is named in the reporting sentence. Animal and in-vitro results, when mentioned at all, are not phrased as human outcomes. Consensus statements are secondary maps to primary trials. Meta-analyses inherit the bias of their inputs; umbrella reviews inherit that twice. Conflict is left as conflict: triad versus RED-S naming; keto economy loss versus ultra fat-oxidation anecdotes; nitrate in recreational versus elite athletes; the width of the protein-timing window.

Evidence grades used in the prose:

Adversarial resolution

The evidence and argument were tested against six critical perspectives. Each perspective's load-bearing objection, and the place in the text that answers it, are recorded below.

Unresolved: a current, large, multi-country contamination surveillance series on the post-2010 online market (Geyer 2004 remains the type specimen, not the current prevalence); a female replication of Currell and Jeukendrup’s mixed-carbohydrate time trial at the same intensity; a preregistered elite nitrate trial with a predefined smallest worthwhile change.

Limitations and sibling scope

The evidential spine of this article is peer-reviewed literature verified against NCBI records; general reference and news databases were used only to locate that literature, never as evidence in their own right. This title does not replace the macronutrients, amino-acid, micronutrients, or high-protein articles. It does not treat therapeutic peptides, doping protocols, or clinical dietetics.

Glossary

Energy availability. Intake minus exercise expenditure, relative to fat-free mass — not energy balance. RED-S. IOC name for the multi-system cost of low EA. Triad. Low EA, menstrual dysfunction, low BMD. MPS. Acute muscle protein synthesis, not hypertrophy. Multiple transportable carbohydrates. Glucose (or maltodextrin) plus fructose to use more than one intestinal transporter. EAH. Exercise-associated hyponatremia. Smallest worthwhile change. The performance difference that would matter in the event; often larger than a published p value and smaller than an advertisement.

References

Numbered, surname-sorted, generated from NCBI-verified records. In-prose citations are author–year. No identifier in this document was assigned from memory: every PMID used above was confirmed against its NCBI record.