Research Reference

Caffeine as a Performance Enhancer — With Special Attention to Athletes 18 and Under

Jumping-off point: Robb Wolf's "Open Letter to the Energy Drink Industry"(Substack), prompted by a youth volleyball tournament where several 11–14-year-olds consumed multiple 200mg energy drinks and had a rough night's sleep and a bad next-day performance as a result.

Compiled September 2026

1. The core idea: a dose-response (inverted-U) curve

Caffeine's effects on performance and mood follow an inverted-U shape. At low-to-moderate doses, effects are almost entirely positive — reduced perceived exertion, better alertness and mood, modest performance gains. Past a certain point, additional caffeine adds side effects (jitteriness, anxiety, GI upset, elevated heart rate, poor sleep) without adding more performance benefit, and eventually those side effects actively hurt performance. This framing — from both the source article and the sports-nutrition literature below — is the organizing principle for everything that follows: the question is never "caffeine: yes or no," it's "how much, and when."

2. What the performance evidence actually says (mostly adult data)

The best-established, most heavily replicated body of evidence on caffeine and exercise performance comes from adult athletes. Key points from the International Society of Sports Nutrition (ISSN) position stand on caffeine and exercise performance:

• Effective dose range: ~3–6 mg/kg body mass, with meaningful benefit possibly as low as 2 mg/kg. Doses around 9 mg/kg produce more side effects without added benefit — the top of the inverted U.
• Timing:~60 minutes pre-exercise is the most common protocol for capsules/coffee; caffeine gum acts faster.
• Mechanism: primarily central nervous system — caffeine blocks adenosine receptors, which reduces perceived exertion, suppresses perception of muscle pain, and increases alertness and motor unit activation.
• What improves: most consistent, moderate-to-large benefit is in aerobic endurance. Also documented: muscular strength/endurance, movement velocity, sprinting/jumping/throwing power, and cognitive measures like attention and vigilance.
• Individual variability is large. The CYP1A2 gene governs how fast someone metabolizes caffeine; "fast" and "slow" metabolizers respond very differently to the same dose, and habitual intake, hormonal factors, and other genetics (e.g., ADORA2A, the adenosine receptor gene) also modulate response. Reported half-life across individuals ranges roughly 1.5–10 hours.
• Notably, the ISSN position stand makes no specific statement about adolescents — it's built on adult data.

A 2018 systematic review/meta-analysis on caffeine and cycling performance is a good illustration of "less is often enough": low doses (≤3 mg/kg) produced comparable or slightly larger improvements in time-to-completion than moderate-to-high doses (4–6 mg/kg), while adverse-effect risk rises with dose. That review's sample was adults only (mean ages 20–48) and explicitly cautioned against generalizing to adolescents.

Bottom line for the "positive" side: in adults, a modest dose (roughly 1–3 mg/kg, well under the 3–6 mg/kg "optimal" adult range) reliably produces most of the available benefit — lower perceived effort, better focus, small-to-moderate performance gains — with low side-effect risk. Going substantially higher does not proportionally add benefit.

2a. Myth-check on a few adjacent questions (relevant for a swim/endurance context)

A 2024 evidence review from the International Society of Sports Nutrition — "Common Questions and Misconceptions About Caffeine Supplementation" (Antonio et al., Journal of the ISSN ) — directly addresses several beliefs that come up around athletes and caffeine. It's an adult-focused paper (no youth data), but a few of its conclusions are directly useful for a swim-team context and worth folding in:

• Caffeine does not meaningfully dehydrate athletes at typical doses. The old "caffeine is a diuretic, don't drink it before you train/swim" belief holds only at high doses. Moderate daily intake (~3 mg/kg, roughly 250–300mg for an adult) does not increase urine output or worsen hydration markers in controlled studies — including during exercise and heat. A measurable diuretic effect only reliably appears at ≥500mg or ≥6 mg/kg in a sitting — well above what's being discussed as a youth dose in this document. This is worth knowing specifically for swimmers/aquatic athletes, where "does it dehydrate you" is a common question.
• "Does caffeine work for everyone?" — mostly yes, more than commonly claimed. Early research suggested up to a third of people are caffeine "non-responders" (perform worse than placebo). Later re-analysis, correcting for test-retest reliability of the performance tests themselves (not the caffeine), found the real non-responder rate is closer to ~5%, and even that residual group may simply need a different dose or timing rather than truly not responding at all. Practically: assume caffeine will do something for most kids, for better or worse — "my kid says it does nothing" is more likely a dose/timing issue than genuine non-response.
• Regular ("habitual") caffeine use does not appear to blunt the performance benefit of a given dose, contrary to a common assumption that a habitual user needs progressively more to get an effect. This is a relevant nuance given the consumption data in §4e — many teen athletes are already frequent energy-drink users, but the evidence doesn't support "they need more now because they're used to it" as a rationale for escalating dose.

3. What's specific to youth (≤18) — performance side

Youth-specific controlled research is much thinner than the adult literature, and this is itself worth flagging to parents/athletes: most "caffeine helps performance" claims are extrapolated from adults, not proven in adolescents.

• A dedicated systematic review and meta-analysis on acute effects and safety of caffeine ingestion in adolescent athletes exists (Frontiers in Nutrition / PMC, 2025) — it's the most direct source on this exact question, though full-text access was blocked during this research pass (see §7, Gaps). Its existence at all signals the field recognizes youth need their own evidence base rather than adult extrapolation.
• A related study (Dos Santos et al., 2022) tested this directly: 90 male adolescents were genotyped for ADORA2A (C>T) and CYP1A2 (C>A) variants, then given 6 mg/kg of caffeine vs. placebo before a battery of tests — handgrip strength, countermovement jump, spike jump, sit-ups, and the Yo-Yo Intermittent Recovery Test. Caffeine improved performance across the board regardless of genotype — neither gene predicted who benefited more. This is a useful, specific data point: in this youth sample, genetics didn't meaningfully change the performance payoff, even though (per §5) genetics is still very relevant to side effects like sleep disruption and anxiety.
• A large cardiac-screening study of 7,425 athletes aged 12–20 found that, after adjusting for demographics and BMI, caffeine consumption itself was not significantly associated with exercise-related symptoms or abnormal ECG findings — though caffeine use correlated with higher rates of ADHD diagnosis and higher BMI in that population, and soda drinkers (not caffeine per se) showed elevated exercise-symptom risk while coffee drinkers showed increased exercise-related syncope (fainting) risk. This is a reassuring data point on the specific fear of "hidden heart problem unmasked by caffeine," but it's observational, not a controlled dosing study, and it doesn't examine the high acute doses seen in energy-drink binges.

Takeaway: the performance case for caffeine in adolescents is plausible (same underlying physiology as adults) but not well quantified. Nobody has run the adult-style dose-ranging studies (2 vs. 3 vs. 6 vs. 9 mg/kg) in 14-18-year-olds. Any specific dosing guidance for youth is therefore extrapolated downward from adult data with a large safety margin, not derived from head-to-head youth performance trials.

4. The negative side: what actually goes wrong, and at what dose

4a. Acute toxicity — how much is truly dangerous

Regulatory toxicology (FSANZ/EFSA-referenced safety assessment) puts this in useful context:

• Reported deaths in adults have followed single doses in the 3,000 mg range, more typically 5,000–10,000 mg.
• ≤210 mg (~3 mg/kg for a 70kg adult) in a single dose is not generally associated with adverse effects.
• Above 210 mg: rising blood pressure, catecholamines, anxiety.
• ≥1,200 mg: more serious effects become possible — tachycardia, ventricular arrhythmia, seizures.

For a much smaller youth athlete (80–120 lb / 36–55 kg is typical for a young teen), the same absolute dose represents a much higher mg/kg exposure than it would for an adult — this is the core reason youth are more vulnerable to a given can/bottle of product, independent of any metabolism differences. A 200mg energy drink is ~2.9 mg/kg for a 70kg adult but ~5.6 mg/kg for a 36kg (80 lb) young teen — already near or above the adult "optimal ergogenic" ceiling, in one drink, before accounting for a second or third can.

4b. Metabolism/half-life in youth — a needed correction to the "kids clear caffeine slower" claim

The source article states that kids clear caffeine more slowly than adults, with citation to wide half-life ranges. This is true for infants/neonates(caffeine half-life can run up to ~100 hours in neonates, due to an immature CYP1A2 enzyme system) but the regulatory toxicology literature states that by adolescence, caffeine clearance is likely at least as fast as in adults — the CYP1A2 enzyme system matures well before the teenage years. Adult half-life is typically cited at 3–7 hours (individual range roughly 1.5–10 hours based on genetics and other factors).

This matters for the practical advice: the mechanism for "don't give a 4pm energy drink" isn't necessarily that kids metabolize caffeine unusually slowly — it's simpler than that: the dose itself is oversized for their body weight, and normal-length half-lives (several hours) mean a 4pm dose is still substantially present at bedtime and beyond regardless of age. Both explanations point to the same practical conclusion (worth being precise about which one is actually driving the effect, since parents will ask).

4c. Sleep

This is arguably the most consequential downside for youth athletes, and it compounds itself:

• A blood caffeine concentration >80 mg/L is the general threshold associated with acute toxicosis symptoms; far below that, caffeine still measurably disrupts sleep architecture (onset latency, total sleep time, sleep efficiency) in dose- and timing-dependent ways.
• Regulatory guidance flags ~1.4 mg/kg body weight/day as a threshold associated with sleep disturbance in children — a notably low bar. For a 40kg young teen, that's about 56mg — roughly the caffeine in a single 12oz cola, well under one energy drink.
• Poor sleep is independently one of the most consistently documented predictors of worse athletic performance (reaction time, decision-making, injury risk) and worse mood/emotional regulation the next day — exactly the "exhausted, emotionally wrought, played terribly" pattern described in the source article. A caffeine-driven bad night's sleep can erase, and reverse, whatever acute performance boost the caffeine itself might have provided the day before.
• Broader adolescent sleep research (independent of sports) links chronic short sleep to worse mood, higher self-harm/anxiety/depression risk, and more motor vehicle accidents — underscoring that this isn't just a "sluggish practice" issue but a real health issue when it becomes a pattern rather than a one-off.

4d. Cardiac and other acute effects

• Case reports and cardiology literature document energy-drink-associated acute coronary events, arrhythmia, and sudden cardiac arrest in young, previously healthy people, typically involving rapid consumption of multiple high-caffeine products, sometimes combined with exercise or an undiagnosed underlying cardiac condition.
• The most widely cited youth caffeine death (Davis Cripe, age 16, South Carolina, 2017) involved a large diet Mountain Dew, a café latte, and an energy drink consumed within about two hours; the coroner ruled the cause of death a caffeine-induced cardiac arrhythmia. This case is commonly used in pediatric guidance precisely because it shows death can occur from a combination of ordinary, individually "unremarkable" caffeinated beverages stacked in a short window — not just from an extreme single product.
• The Panera "Charged Lemonade" litigation (referenced in the source article) involved a young woman with a known cardiac condition; her case is a reminder that individuals with undiagnosed or known cardiac arrhythmia risk are disproportionately vulnerable, which is relevant for team settings where not every athlete's cardiac history is known to the coaching staff.
• Other acute effects reported at high youth intake: hand tremor/"the shakes" (as described directly in the source article), anxiety, GI upset, elevated heart rate — all reported at the doses common in multi-can-per-day energy drink use.
• A balancing point: the same 2024 ISSN evidence review (§2a) concludes there is "no evidence that low to moderate intake of caffeine has adverse effects on cardiac muscle" in the general population, and notes several plausible cardiovascular benefits of moderate habitual intake (better endothelial function, lower stroke/heart-failure risk associated with habitual coffee intake in adults). It frames cardiovascular risk as depending on dose, duration, mode of consumption, and individual metabolic/genetic factors — consistent with §5's point that risk concentrates specifically in high-dose and slow-metabolizer combinations, not caffeine per se. In other words: the case reports and deaths above are real, but they cluster at the high-dose, multi-product-stacking end of exposure, not at the "one small coffee" end — which supports this document's dose-and-timing framing rather than a blanket "caffeine = cardiac risk" reading.

4e. Consumption patterns among youth athletes (how common is this, really)

• A 2024 survey found 64.5% of female high school athletes (ages 14–16) reported energy drink use in the past year, with over 25% using them 3–4+ times per week — a marked rise from older estimates of 5–16% among high schoolers generally. Energy drinks in this category commonly contain 70–300 mg caffeine per 16oz serving.
• This means the scenario in the source article (a team collectively hammering energy drinks at a tournament) is not an outlier — it reflects a now-common pattern, not a fluke.

5. Genetics: why the same dose hits different kids differently

The literature separates genetic effects into two distinct axes, and it's worth keeping them separate because they explain different symptoms. Note that gut absorption of caffeine (how fast it enters the bloodstream after drinking it) is not the genetically variable step — it's fast and fairly uniform for everyone, typically peaking 30–60 minutes after ingestion. The two things genetics actually control are how long caffeine stays active in the body (clearance) and how strongly the body reacts to whatever concentration is present (receptor sensitivity).

5a. CYP1A2 — the "clearance" gene

CYP1A2 is the liver enzyme responsible for breaking down roughly 95% of ingested caffeine. The key variant is rs762551 (sometimes called -163A>C):

(These half-life bands vary somewhat by source and aren't as tightly nailed down as the genotype frequencies — treat them as directional, not exact. They're also consistent with the wider 1.5–10 hour range cited in the ISSN position stand in §2.)

A caveat worth carrying forward: classification schemes differ across the literature — some group A/C with A/A as "fast enough," others group A/C with C/C as "slow," and CYP1A2 accounts for the majority (commonly cited as >90%) but not all of caffeine metabolism. More importantly, the 2024 ISSN evidence review (§2a) notes that results on whether CYP1A2 genotype actually predicts an individual's ergogenic response to caffeine are mixed — some studies show fast metabolizers get a bigger performance boost, others show no difference, and at least one shows the opposite. There is no current consensus that genotype alone determines how much a given athlete benefits performance-wise, even though it's on firmer ground for explaining differences in clearance time and downstream effects like sleep and cardiovascular risk at high habitual intake (below). Female athletes also tend to have somewhat lower average CYP1A2 activity than males, which is a further modifier not captured by genotype alone (also affected by hormonal contraceptive use).

Practically: an A/A ("fast") teen who has a can at 4pm may have largely cleared it by dinner. A C/C ("slow") teen with the identical can may still have a meaningful amount on board at midnight. Same product, same weight, same timing — very different sleep outcome, purely from genetics. Genetics are estimated to account for roughly 75% of the variation in CYP1A2 enzyme activity between people; the remaining ~25% is environmental (smoking and cruciferous vegetables increase enzyme activity/speed clearance; some medications and acute inflammation decrease it).

A well-known cohort study (Cornelis et al., JAMA , 2006) found that heavy coffee/caffeine intake (4+ cups/day equivalent) was associated with meaningfully elevated heart attack risk in slow CYP1A2 metabolizers (odds ratio ~1.6 overall, and notably higher — OR ~2.3 — in people under 59), while fast metabolizers showed no elevated risk at any intake level studied. This was an adult cohort, not youth, and it involved sustained heavy intake rather than a single energy drink — but it's a real-world illustration that "slow metabolizer + high dose" is a specifically higher-risk combination, not just a sleep inconvenience.

5b. ADORA2A — the "sensitivity" gene

ADORA2A codes for the adenosine receptor that caffeine blocks — this is the gene that determines how strongly a given caffeine concentration is felt, independent of how quickly it's cleared. The key variant is rs5751876 (T1083C):

• C allele (~55% of people): binds caffeine more readily at the receptor, associated with greater caffeine-driven sleep disruption at a given dose.
• T allele (~45% of people): associated with a stronger caffeine-induced anxiety response — jitteriness, racing heart, that "wired" feeling — again at a given dose, independent of clearance speed.

This is the axis that explains why some kids get visibly shaky or anxious ("the shakes," as one athlete described it in the source article) on a dose that a teammate shrugs off, even if both metabolize caffeine at a similar rate. A study of adolescent athletes specifically looked at whether ADORA2A and CYP1A2 genotypes explained differences in caffeine's effect on exercise performance and did not find a clear influence on performance outcomes in that youth sample — genetics appears to more reliably predict side effects (sleep, anxiety) than it does how much of a performance boost someone gets.

5c. Putting the two together

The practical upshot is that "how does caffeine affect this athlete" really has two independent genetic answers:

1. How long does it stay active?(CYP1A2 → sleep risk window, cumulative buildup with repeat doses)
2. How hard does it hit at a given concentration?(ADORA2A → anxiety/jitters/subjective "wired" feeling)

A kid can be a fast metabolizer (clears it quickly) but still have a strong anxiety response while it's active (T-allele ADORA2A) — or a slow metabolizer (long exposure window) with a relatively mild subjective response (C/C-favorable ADORA2A) who nonetheless is quietly still caffeinated at 1am with no obvious symptoms to tip anyone off. Neither combination is visible without genetic testing, which is exactly why dose and timing rules that don't depend on knowing an individual athlete's genotype(§6) are the practical answer for a team setting — you can't eyeball who's a slow metabolizer, and the athlete who "seems fine" isn't necessarily clearing caffeine any faster than the one who's visibly jittery.

6. Existing guidance for youth (mg/day)

There's no single universal standard, but the main reference points converge in a similar range:

Two of these numbers (AAP's ~100mg/day ceiling, and the source article's 25-50mg per-dose target) are compatible: 25-50mg per serving keeps a young athlete comfortably under the 100mg/day AAP ceiling even with two servings, while staying well clear of the ~1.4 mg/kg sleep-disturbance threshold for most youth body weights. The regulatory 5.7 mg/kg "safe" ceiling is a toxicology floor, not a target — it's the line below which acute harm is unlikely, not a recommended intake.

7. Practical synthesis (for a coaching context)

1. Frame it as dose and timing, not a ban. The literature doesn't support "caffeine is dangerous, full stop" for teenagers — it supports "the doses kids are actually getting from commercial energy drinks are typically 2-8x higher than what's useful or advisable."
2. A single modern energy drink (150–300mg) is not a youth "serving" — it's several servings. Splitting a can, or choosing a lower-dose product, is a meaningfully different exposure than drinking a whole can.
3. Afternoon/evening timing is the highest-leverage lever for meets and tournaments, independent of the metabolism debate — normal half-lives alone mean anything after early-to-mid afternoon is still substantially on board at bedtime.
4. Sleep is the hidden mechanism behind most of the "caffeine hurt our performance" stories — including the one in the source article. A caffeine-driven bad night can undo the modest performance edge caffeine might otherwise provide the same day.
5. Cardiac history matters and isn't always known to coaching staff — this is a reason for blanket dose moderation at team events rather than assuming "it's fine, nothing's ever gone wrong for us."
6. The youth-specific evidence base is thin. Most quantitative guidance (including in this document) is adult data scaled down with a safety margin, not youth-specific dose-ranging research. Appropriate humility, not false precision, is warranted when giving athletes/parents numbers.

8. Gaps and caveats in this research pass

• The most directly on-point source — a 2025 systematic review/three-level meta-analysis specifically on "Acute Effects and Safety of Caffeine Ingestion in Adolescent Athletes" (Frontiers in Nutrition, indexed in PubMed/PMC) — could not be fully retrieved (blocked by an access/verification wall). It's flagged here by title/citation because it's clearly the single best-matched source to this exact question; if this document needs to go further, that's the first paper worth chasing down through a library/journal-access route.
• Most dose-response performance literature is adult-derived; direct youth dose-ranging trials are sparse.
• Case reports (cardiac events, deaths) are individually compelling but don't establish population-level incidence/risk — they illustrate mechanism and worst-case severity, not "how likely is this to happen to my team."
• "Energy drink" as a category spans a wide caffeine range (70–300mg per 16oz); specific product labels should be checked rather than assuming a flat number.
• The genotype-frequency and half-life figures in §5a are drawn from secondary genetics-education sources rather than a single primary population study, and the half-life bands in particular vary somewhat across sources — treat them as illustrative of the scale of individual difference, not precise clinical values.
• Disclosure/bias note on the 2024 ISSN evidence review used in §2a, 3, 4d, and 5a: several of its authors disclose financial relationships with caffeine- and supplement-industry funders (research grants, consulting, or speaking fees from companies that make caffeine-containing sports nutrition products), and the ISSN itself (a nonprofit that authored the paper) receives grants from such companies — disclosed transparently in the paper itself. This doesn't invalidate its findings, which are consistent with independent regulatory sources (FSANZ/EFSA) used elsewhere in this document, but it's worth reading its generally reassuring conclusions on hydration, cardiac risk, and non-responders with that funding context in mind — and it's an adult-focused paper being applied here to a youth context by extension, not direct youth evidence.

Sources

• Robb Wolf — "Open Letter to the Energy Drink Industry"(source article)
• ISSN Position Stand: Caffeine and Exercise Performance
• Effect of caffeine ingestion on cycling performance: systematic review/meta-analysis
• Acute Effects and Safety of Caffeine Ingestion in Adolescent Athletes: A Systematic Review and Three-Level Meta-Analysis (PubMed listing)
• The effect of caffeine on exercise performance is not influenced by ADORA2A genotypes, alone or pooled with CYP1A2 genotypes, in adolescent athletes
• Relationship Between Caffeine Consumption and Young Athletes' Comorbidities, Exercise-Related Symptoms, and Baseline ECG (PubMed)
• Predictors of caffeine consumption patterns in high school athletes
• Are Energy Drinks Safe for Adolescent Athletes? — Athlete Training and Health (summarizing 2024 survey data and AAP guidance)
• FSANZ Safety Assessment of Caffeine (regulatory toxicology, LD50/half-life/child-adolescent intake guidance)
• Suspected Energy Drink–Induced Acute Coronary Syndrome With Sudden Cardiac Arrest (case report)
• South Carolina Teen Died of Caffeine Overdose, Coroner Rules — NBC News (Davis Cripe case)
• Panera "Charged Lemonade" lawsuit coverage — NBC News
• CYP1A2 Gene — genotype frequencies, half-life bands, and enzyme-activity modifiers and DeepDNA: Slow Caffeine Metabolizer? CYP1A2 AA/AC/CC + Half-Life
• ADORA2A Gene — caffeine-induced anxiety and insomnia by genotype
• Coffee, CYP1A2 Genotype, and Risk of Myocardial Infarction — Cornelis et al., JAMA 2006
• Antonio, J., Newmire, D.E., Stout, J.R., et al. "Common questions and misconceptions about caffeine supplementation: what does the scientific evidence really show?" Journal of the International Society of Sports Nutrition , 2024;21(1):2323919. (Uploaded PDF; user-supplied source — hydration/dehydration myths, non-responder rates, habitual-use tolerance, cardiac-risk framing, and CYP1A2/ADORA2A genotype-performance evidence in §2a, §3, §4d, §5a)