Protein Marketing and Adolescent Nutrition

Abstract

Protein claims can describe a useful nutritional property while also inviting broader judgments about health, personal need, and physical progress. This critical narrative review distinguishes the effects of protein marketing from the effects of consuming protein. It integrates research on nutrient claims, adolescent nutrition, body image, digital media, product composition, and supplement use. Targeted searches and source verification informed an interpretive synthesis; no exhaustive search or pooled estimate is claimed. An integrated framework specifies two candidate pathways: inference from a nutrient cue to overall quality or perceived need, and appearance comparison leading to muscularity-related goals and product interest. Persuasion knowledge and differential susceptibility to media effects explain why the direction and magnitude of responses may vary. Protein-label experiments support effects on some immediate judgments; a recent TikTok experiment supports effects on supplement-use intentions in young men. Observational studies identify relevant associations but do not establish protein-specific marketing effects in adolescents. Product comparisons show that protein enrichment can improve some nutritional attributes, while composition testing identifies variable contaminant exposure. Neither product testing nor associations involving mixed muscle-building supplements establish harm from dietary protein. The principal unresolved issue is whether identifiable protein marketing produces miscalibrated expectations or unnecessary product use, and whether subsequent dietary substitution or product exposure affects health. The framework yields falsifiable predictions and separates policy measures supported by broader evidence from protein-specific proposals requiring evaluation.

Keywords: adolescent nutrition; protein claims; persuasion knowledge; body image; social media; dietary supplements; causal inference

Introduction

A protein statement can be accurate and still be insufficient to justify the decision it encourages. The amount of protein in a food does not establish its overall nutritional quality, whether a particular adolescent needs more protein, or whether purchasing it will deliver a desired physique. Yet these broader meanings may be supplied by a product name, accompanying imagery, an influencer, or the consumer’s own goals. The scientific problem is not the presence of protein. It is how information about one nutrient is translated into expectations about a product and the self.

This problem crosses research traditions that usually address different questions. Nutrition studies assess requirements and physiological responses; consumer experiments examine judgments and choices; body-image research examines appearance-related motivations; public-health studies examine commercial exposure and regulation. Combining these traditions can clarify the problem, but only if their exposures and endpoints remain distinct. Findings about nutrient intake cannot establish advertising effects, and findings about supplement-use intentions cannot establish clinical injury.

The present review offers a theory-informed account of that translation. Its contribution is an explicit separation of nutritional inference from muscularity-related persuasion, accompanied by evidence about each candidate pathway and a design agenda for testing their interaction. It does not propose a validated new theory or claim that commercial demand is uniformly manufactured. Rather, it develops an integrative framework that can accommodate useful information, inappropriate extrapolation, and consumer skepticism within the same food environment.

Adolescents are the population of interest because nutritional growth needs and appearance-related goals can coexist. We define adolescence as ages 10–19 for the organizing question, while retaining each study’s actual age range. Young adult evidence informs potential mechanisms; it does not become adolescent evidence through proximity in age. The review addresses nutrition and adolescent-health audiences and asks when protein marketing improves decisions, when it miscalibrates them, and which outcomes would demonstrate the difference.

Review approach and evidentiary boundaries

This is a critical narrative review with a documented source audit. Literature was assembled through targeted searches of publicly accessible bibliographic and publisher records, exact-title verification, and searches for theory and counterevidence, completed on October 14, 2026. Topics included protein requirements and intake, claim interpretation, online fitness content, appearance-related behavior, product composition, clinical outcomes, and regulation. Primary government guidance and legal text were consulted for policy claims. Supplementary material records executed search terms, key study characteristics, source-access limitations, and exclusions or corrections.

Selection favored sources that directly informed a framework component or challenged an emerging interpretation. Studies were interpreted by design, population, exposure, comparator, endpoint, and temporal sequence. A laboratory judgment, stated intention, purchase, habitual intake, and clinical outcome were treated as different endpoints. Conference abstracts, dissertations, institutional polls, and consumer testing were identified by source type. The synthesis follows SANRA’s emphasis on an explicit question, transparent search description, referencing, scientific reasoning, and relevant endpoint data; SANRA is an appraisal aid, not a substitute for systematic-review methods. [1]

There was no preregistered protocol, exhaustive multi-database search, duplicate screening, or formal risk-of-bias scoring. We therefore do not claim a complete evidence census, assign pooled certainty ratings, or infer the absence of a literature from an unsuccessful targeted search. Commercial affiliations are considered when interpreting independence and selective reporting, without treating affiliation as a proxy for validity. The framework and predictions are analytic proposals derived from the synthesis, rather than findings from a new empirical study.

What is being marketed and what would count as need

Separate the exposure before assessing its effects

Four exposures require separate assessment: total dietary protein; conventional foods carrying a protein statement; protein supplements; and mixed muscle-building supplements or substances. A protein-enriched yogurt is not interchangeable with a powder containing additional active ingredients. Likewise, a study combining protein powders, creatine, and other substances cannot identify a protein-specific effect. For marketing research, exposure must also specify the communication: a quantitative statement, a qualifying nutrient claim, a product name, a physique image, or an influencer endorsement.

The proposed distinction between pathways is functional rather than a rigid division by medium. Packaging can portray an idealized body, and an influencer can give a factual protein explanation. Studies should manipulate the information or imagery being conveyed, not assume that a package necessarily activates nutritional reasoning or that every social-media message activates appearance comparison.

Outcomes require equivalent precision. An unwarranted inference is clearest when a message changes a judgment about a property that did not change, such as fiber content in an otherwise identical product. Overall healthfulness is harder to classify because its meaning depends on the product, person, and comparator. Perceived need should be measured separately from liking or willingness to buy: convenience, taste, and sport-specific goals can motivate use without a belief in deficiency.

Adequacy does not establish an optimum or a safety ceiling

The US protein Recommended Dietary Allowance is an adequacy reference intended to cover nearly all healthy people in a specified group. Fixed amounts are 34 g/day at ages 9–13, 46 g/day for females aged 14–18, and 52 g/day for males aged 14–18; these use reference body weights. Neither the allowance nor an intake above it establishes a universal performance optimum or an upper safe limit. Estimating population inadequacy generally requires usual-intake distributions and the appropriate Estimated Average Requirement. [2]

Available intake reviews do not establish a widespread protein shortage in the high-income youth populations described, although mean intakes conceal individual inadequacy. The widely cited 2023 narrative review reporting high average intake was authored by researchers affiliated with Abbott Nutrition and should be interpreted alongside independent surveillance. [3] The October 2026 NCHS report estimates that US adolescents aged 12–19 obtained 14.3% of calories from protein on a given day. That energy-share estimate is not a measure of usual intake below requirements. [4]

Clinical need depends on sufficient total energy, growth, training, food access, and health status. For adolescent athletes, sports nutrition guidance favors meeting needs through a well-planned diet and assessing supplements in context. [5] A practical benefit from a product is therefore possible even when population-wide deficiency is uncommon. Conversely, satisfying a training goal does not justify an implied promise of benefit for every adolescent.

The 2025–2030 Dietary Guidelines for Americans specify a general goal of 1.2–1.6 g/kg/day adjusted to caloric requirements. Their adolescent section emphasizes nutrient-dense foods and permits fortified foods or supplements under medical guidance where access is limited; it provides no separate adolescent grams-per-kilogram target. This dietary guidance does not itself establish a need for commercial supplementation. [6] Whether consumers infer such a need is an empirical question, not a conclusion that follows from the guideline’s publication.

An integrated theoretical framework

Nutrient inference and persuasion knowledge

The first pathway concerns the interpretation of a salient nutrient cue. Consumers may use protein content as a shorthand for nutritional quality and extend its favorable meaning to unrelated attributes. The critical inference is broader than the information provided: more protein becomes healthier overall, healthier overall becomes better for me, or better for me becomes something I should buy. These are separable judgments, and evidence for one transition does not establish the next.

The persuasion knowledge model adds an important boundary condition. Consumers draw on beliefs about an advertiser’s motives and tactics when responding to a message. Recognition of persuasive intent may lead to scrutiny or skepticism rather than favorable generalization. [7] Applied here, a protein claim may increase healthfulness judgments when it seems credible and relevant, but reduce trust when it seems to disguise an indulgent product or justify an unnecessary premium. This is a theoretical prediction about protein marketing; persuasion knowledge has not been established as the mediator of all protein-label effects.

The model distinguishes nutritional understanding from recognition of persuasion. Knowing which foods contain protein does not ensure recognition of an influencer’s financial incentive. Conversely, recognizing sponsorship does not establish the ability to assess dietary need. Experiments should measure both rather than use a single nutrition-literacy score as an assumed protection against marketing.

Appearance comparison and muscularity related goals

The second pathway concerns the self rather than primarily the food. Sociocultural models propose that appearance pressures operate partly through internalization of ideals and comparison with others. A study of French adolescent boys linked sociocultural pressures, media-ideal internalization and appearance comparison with muscularity and disordered-eating constructs, but its questionnaire design did not establish causal mediation. [8]

Protein products can be situated within this pathway as perceived tools for narrowing a gap between a current and desired body. Under this account, the relevant belief may be that the product is necessary for progress rather than that the ordinary diet is nutritionally deficient. The mechanism predicts different responses to an identical nutrient statement when it is paired with physique imagery, recovery information, or a neutral meal context.

Muscularity-related aspirations are not intrinsically pathological. Their clinical significance depends on distress, rigidity, impaired functioning, and associated behaviors. The framework therefore distinguishes ordinary training goals from appearance-driven compulsion and does not classify all supplement users as disordered. It also does not assume that boys are the only affected group; effects across sex, gender, and body ideals require direct study.

Susceptibility and structural conditions

The differential susceptibility to media effects model identifies dispositional, developmental, and social influences on media selection and response. [9] Its application here predicts heterogeneous effects according to prior protein attitudes, existing use, appearance concerns, developmental stage, and household or peer practices. These moderators should be tested prospectively; chronological age alone cannot establish a universal adolescent vulnerability.

Commercial determinants of health provide the surrounding context: product formulation, pricing, availability, promotion, and governance shape the opportunities within which individual decisions occur. [10] This perspective prevents the framework from reducing outcomes to literacy or willpower. A positive purchase decision may reflect limited access to meals, and an unfavorable judgment may have little behavioral effect if the alternative is unaffordable. These conditions should be measured, rather than presumed to operate identically across households.

Intentions remain intermediate outcomes. Research on the intention–behavior gap supports separating motivation from enacted behavior. [11] Figure 1 places purchase and sustained use downstream of both pathways and identifies circumstances that can interrupt or redirect that transition. The framework is deliberately permissive about beneficial and adverse consequences: nutritional correction, convenience, unnecessary expense, dietary displacement, and exposure to a particular formulation are all possible outcomes, with different evidence requirements.

Figure 1  Proposed routes from a protein message to product use

Arrows represent proposed relationships, not an estimated causal model. Existing studies support selected proximal responses or associations; they do not validate the full sequence. Prior characteristics may cause both exposure selection and later use. Feedback is possible over time but is omitted here for clarity.

Evidence for the proposed pathways

Protein cues can change judgments but effects depend on the task

Fernan and colleagues randomized 274 participants to labeling conditions for a bar. Both a nutrient claim and protein in the product title increased perceived protein content, but only the title increased overall healthfulness; participants also inferred more fiber and iron. A traffic-light warning did not eliminate the title effect. [12] This supports a specific inference mechanism, not a universal effect of every protein statement. The nonclaimed nutrient judgments are especially informative because they identify information consumers supplied beyond the message.

Rramani and colleagues used behavioral and imaging studies of milk-mix drinks. Protein-related information increased expected healthiness and expected satiation, while perceived taste pleasantness was not changed by the claims. [13] These findings help distinguish expectations from experienced sensory response. They do not establish that protein labeling improves dietary quality or that altered neural activity predicts clinical outcomes.

An online study of 1,022 US adults comparing commercial cereals found more favorable health-related judgments and purchase intentions for the protein-labeled product. Because the products differed, it cannot isolate the label alone. [14] A 2023 dissertation experiment reported no significant effect on healthfulness or purchase intention despite changes in perceived protein and satiety; its non-peer-reviewed status limits its weight. [15] A discrete-choice experiment by Kim and colleagues found that demand for added-protein labels varied by food type and indulgent flavor context. [16]

Taken together, these studies support context-dependent signaling rather than a single directional effect. They also leave the crucial personal-need transition underexamined. Adult evaluations of bars, drinks, or cereals cannot establish whether an adolescent believes their diet is inadequate, acts on that belief, or benefits from the resulting substitution.

Digital evidence supports proximal effects and several associations

The broader food-marketing literature supplies relevant but indirect evidence. Boyland and colleagues’ synthesis included 96 studies, with 80 articles contributing to meta-analysis. Marketing was associated with greater intake (standardized mean difference 0.25, 95% CI 0.15–0.35) and food choice (odds ratio 1.77, 95% CI 1.26–2.50). Heterogeneity was substantial, evidence certainty varied, and purchasing and health endpoints were less well established. [17] These estimates should not be imported as effect sizes for protein marketing or older adolescents.

In cross-sectional research involving 2,731 Canadians aged 16–30, engagement with fitness- and weight-related content was associated with several substance-use measures, including whey protein. [18] A Czech survey of 1,526 adolescents aged 13–18 associated online information seeking and appearance-ideal internalization with purchasing muscle-building products. [19] A separate study associated muscularity-oriented online exposure and body comparison with steroid-use intentions. [20] These associations fit the framework but also fit reverse selection: an existing goal or product interest may lead people to seek relevant content.

Beos, Kemps, and Prichard randomized 282 male-identifying participants to fitness, supplement, or travel TikTok content. Planned comparisons found increased immediate intentions to use creatine and reduced nutrition satisfaction after idealized content. [21] This supports an experimental effect on a proximal outcome in young men. It does not isolate protein marketing, measure subsequent use, or establish effects in younger adolescents. Reported mediation by appearance comparison is mechanism-consistent; randomization of content does not also randomize the mediator or establish its causal effect.

The evidence therefore advances beyond a claim that online exposure is only correlated with product interest. It remains insufficient to demonstrate the full sequence from a protein advertisement to perceived nutritional shortfall, repeated consumption, and harm. Larger relative associations or statistical significance do not repair that mismatch of exposure and endpoint.

Use and beliefs provide context rather than causal attribution

Eisenberg and colleagues’ adolescent survey found muscularity-motivated protein powder or shake use in 34.7% of boys and 21.2% of girls. Its defined local sample and historical period limit national or contemporary interpretation. [22] The 2024 Mott poll reported parent-described past-year protein-product use by 41% of teens, but the category included bars, shakes, and powders; powder alone was reported for 15%. Nearly one in five parents believed their teen obtained insufficient protein. [23]

The Australian Royal Children’s Hospital report provides adolescent-reported context: 38% reported shakes, powders, or bars at least weekly, and 54% endorsed the proposition that high-protein diets are better for teens’ health. It included 1,595 children aged 12–17 and 2,285 parent respondents. [24] These results identify beliefs worth investigating. They do not establish that those beliefs originated in advertising, that users had sufficient intake, or that endorsement of a high-protein diet is equivalent to perceived deficiency.

A prevalence estimate becomes relevant to the theory only when its definition is retained. Food bars are not necessarily legally classified supplements; occasional use differs from daily use; training-related use differs from use to compensate for a perceived dietary shortfall. Combining these measures into a single prevalence would erase the distinctions the framework is intended to test.

What the products do is a separate causal question

Nutritional comparison depends on the alternative

Spanish market research classified 90.8% of protein-claim items as less healthy under the Pan American Health Organization nutrient-profile model in a database of processed foods. [25] That is a classification within a sampled product set, not a finding that 90.8% of all protein-rich foods cause harm. German research comparing high-protein and conventional processed products found lower energy density, sugar, and fat and more fiber in the high-protein group, but more salt and additives. Products were selected predominantly as ultra-processed by design. [26]

The studies answer different questions through different samples and scoring systems. Their results need not conflict. A product can be an improvement on a conventional comparator and still trigger an unfavorable nutrient profile. Protein claims should consequently be assessed against both the available alternative and an explicit measure of overall quality. Neither processing category nor the headline protein amount is sufficient alone.

Physiological evidence also requires an explicit comparison. Hägele and colleagues studied 21 healthy young adults in a randomized crossover trial of two ultra-processed diets over 54-hour periods. Relative to the normal-protein diet, the higher-protein, lower-carbohydrate diet reduced intake and increased expenditure, although both produced positive energy balance. [27] This supports a short-term difference between diet packages. Because protein replaced carbohydrate and formulations differed, it does not isolate protein as the sole cause; because no minimally processed comparator was tested, it cannot estimate that contrast.

Marketing effects and formulation effects should therefore be studied separately before they are combined. A label experiment with chemically identical foods tests communication. A feeding trial tests a dietary intervention. An actual marketplace choice can involve both, as well as price and convenience. Treating these designs as interchangeable makes an apparent marketing benefit or harm impossible to interpret.

Contamination requires exposure assessment

Consumer Reports tested 23 protein powders and shakes and reported that more than two-thirds exceeded its lead level of concern of 0.5 micrograms per day in one serving. [28] In contrast, analysis of 22 powders purchased in Hungary found no heavy-metal concentrations above the regulatory limits applied by the investigators. [29] Differences in products, sampling, analytical methods, and benchmarks prevent these studies from establishing a single market-wide prevalence.

A peer-reviewed assessment of earlier testing data generally estimated low noncarcinogenic risk under modeled assumptions. [30] Testing, risk modeling, and observed clinical injury are distinct evidence stages. Benchmark exceedance is not automatically a federal violation or a prediction of injury, while a low modeled risk is not a guarantee for other batches or patterns of use.

The relevant marketing hypothesis is therefore conditional: if a message increases sustained use of a contaminated product, it may increase exposure. The complete proposition requires measurement of the message effect, consumption, contaminant concentration, and exposure duration. Contaminants are properties of particular products and supply chains, not a toxic property of protein as a nutrient.

Mixed supplement associations cannot identify protein specific harm

Bulens and colleagues’ incident-use analysis included 3,792 participants and 18 steroid-initiation events. Mixed muscle-building supplement use predicted next-wave initiation (adjusted odds ratio 8.31, 95% CI 2.59–26.73); the underlying cohort’s baseline mean age was 20.3 years. Exposure combined protein products with creatine and other substances. Initiation was reported in 0.5% of exposed responses and 0.06% of unexposed responses, not cumulative individual risks. [31]

Temporality strengthens the association but does not establish a gateway mechanism. Adjustment for age and cohort does not eliminate shared causes such as appearance goals, sport context, or willingness to take risks. The study can justify a broader assessment when use is disclosed; it cannot establish that protein powder, a conventional food, or a protein advertisement causes steroid initiation.

Likewise, a 2025 systematic review identified little pediatric evidence on bone or kidney disease outcomes. [32] Adult syntheses did not support generalized kidney-function deterioration or bone damage within their studied populations and intake ranges. [33] [34] These findings weaken claims of established physiological toxicity in healthy adolescents without resolving long-term pediatric safety. Absence of such evidence is neither proof of harm nor proof of unrestricted safety.

Falsifiable predictions and a causal research agenda

Specify what would support or challenge the framework

The framework makes five directional predictions, summarized in Table 1. Among consumers with favorable prior protein attitudes, adding a protein cue to an identical food should increase favorable judgments of at least some nonclaimed attributes. Pairing that cue with idealized physique imagery should add to appearance comparison and product interest among those pursuing muscularity. A paid-promotion disclosure that increases recognition of commercial intent should attenuate favorable generalization where skepticism is activated. Among participants with adequate assessed intake, some messages should increase perceived product necessity without a corresponding change in nutritional need. Finally, effects on sustained use should be smaller when price or household access constrains the transition from intention to purchase. The moderators and directions require prespecification; they should not be chosen retrospectively to explain whichever result occurs.

Each prediction can fail. For example, if an identical-package experiment changes only perceived protein content, it supports effective information provision rather than broader inference. If imagery adds no effect once nutrient wording is controlled, the proposed identity route is not supported for that setting. If exposure affects intentions but not later behavior, the findings constrain the downstream account. Null or adverse effects on product interest should be reported, not classified as failed marketing exposure.

Table 1  Predictions and tests of the proposed framework

Match the design to the estimand

A factorial experiment can vary protein wording and appearance imagery independently while holding product composition, brand, price, and portion constant. The primary estimand should be the assigned message’s effect on a prespecified judgment or choice, with participant age strata defined in advance. Appropriate adolescent assent and consent, realistic stimuli, and measurement of adverse body-image responses are essential design considerations. Monetary or incentive-compatible choice can complement stated intention, but should not be mistaken for habitual intake.

Perceived nutritional need should be measured independently of muscularity goals. Candidate measures include beliefs about dietary sufficiency, expectations of incremental benefit, and perceived necessity of the advertised product. Comparison with repeated dietary assessment and an individualized requirement framework can identify miscalibration; a single recall or a value below the RDA cannot establish deficiency. The measurement model needs validation before a new scale is treated as a clinical instrument.

A longitudinal design should distinguish assigned exposure, observed commercial exposure, and self-selected information seeking. Prior product use, training, appearance concerns, food insecurity, household purchasing, and peer practices can cause both exposure and subsequent behavior. Conversely, post-exposure intentions or comparison responses may be mediators; adjusting for them when estimating a total effect would answer a different question. Temporal measurements and an explicit causal diagram should precede statistical modeling.

Clinical endpoints require follow-up after actual exposure has been established. Studies should measure what a product replaces, changes in total energy and other nutrient intake, use of additional ingredients, and expenditure. Trials of formulations should specify the replacement nutrient and comparator diet. Policy evaluations should similarly separate changes in advertising exposure, judgments, purchases, and diet; changes in one are not sufficient evidence of changes in all.

Clinical and policy implications

Clinical guidance recommends nonjudgmental assessment of muscle-building supplement use. [35] The framework suggests asking both what a young person consumes and what they expect it to accomplish. Product, dose, frequency, co-use, meal replacement, access to food, and distress about appearance help distinguish nutritional support from potentially problematic use. Ordinary interest in strength or convenient foods should not itself be pathologized. Where an actual dietary problem exists, the response should address nourishment and practical access rather than merely challenge a marketing message.

Policy also operates at distinct levels. FDA does not approve dietary supplements for safety and effectiveness before sale, but manufacturers retain safety and labeling responsibilities and the agency has enforcement powers. [36] Conventional foods and supplements must be classified before applying their respective requirements. US qualifying protein claims involve Daily Value criteria and quality-corrected protein labeling; factual amount statements and qualifying claims are not identical categories. [37] European source-of-protein and high-protein claims use thresholds of 12% and 20% of food energy. [38]

Claim compliance is necessary but does not settle consumer interpretation. A legal statement can coexist with an unsupported implication about overall quality or personal need. Potential responses include compliance surveillance, testing accompanying explanations, and assessing nutrient-profile conditions for claims. These are protein-specific proposals requiring evaluation, not demonstrated remedies. The warning result in one label experiment cannot establish that every disclosure strategy is ineffective.

WHO recommends comprehensive mandatory protection from harmful food marketing using government-led nutrient profiling. [39] This broader evidence-based policy basis can encompass unhealthy products carrying protein claims without making protein the prohibited feature. Independent product testing addresses composition, while advertising substantiation addresses promises. Literacy and cooking support may be useful complements, but this review does not establish their superiority over structural protections or justify shifting responsibility entirely to adolescents and families.

Limitations and conclusion

The review’s targeted selection and uneven full-text access permit interpretation and source checking, not a definitive estimate of the size of the problem. Evidence is concentrated in high-income settings, adult label experiments, and male or mixed-age supplement samples. These patterns constrain generalization across development, gender, resources, and culture. Institutional polls and consumer testing provide context but cannot establish advertising effects. The proposed framework has not been tested as a whole, and similarities between separate studies are not evidence of an empirically demonstrated causal chain.

Protein marketing is scientifically tractable when nutrient information, consumer inference, identity-related goals, actual use, and product effects remain distinct. Existing evidence supports some proximal effects and several relevant associations. It also demonstrates that protein enrichment can offer advantages under specified comparisons. The defensible concern is therefore conditional: a message may encourage expectations or product use that exceed what its information and a person’s nutritional circumstances justify.

The next step is to test those conditions directly. A framework that permits beneficial use, inappropriate inference, skepticism, and null effects is more informative than a universal account of manufactured demand. It directs attention to the missing transitions, identifies their competing explanations, and makes clear what evidence would warrant a stronger clinical or policy conclusion.

References

1. Baethge C, Goldbeck-Wood S, Mertens S. SANRA—a scale for the quality assessment of narrative review articles. Research Integrity and Peer Review. 2019;4:5. DOI

2. Institute of Medicine. Dietary Reference Intakes for Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids. National Academies Press; 2005. DOI

3. Garcia-Iborra M, Castanys-Munoz E, Oliveros E, Ramirez M. Optimal protein intake in healthy children and adolescents: evaluating current evidence. Nutrients. 2023;15(7):1683. DOI

4. Couch CA, Ansai N, Noiman A, Ogden CL. Protein consumption among youth and adults: United States, August 2021–August 2023. NCHS Data Brief. 2026;(573). Primary source

5. Desbrow B, McCormack J, Burke LM, et al. Sports Dietitians Australia position statement: sports nutrition for the adolescent athlete. Int J Sport Nutr Exerc Metab. 2014;24(5):570–584. DOI

6. US Department of Health and Human Services; US Department of Agriculture. Dietary Guidelines for Americans 2025–2030. January 2026. Primary guidance, pp 2 and 7. Primary source

7. Friestad M, Wright P. The persuasion knowledge model: how people cope with persuasion attempts. Journal of Consumer Research. 1994;21(1):1–31. DOI

8. Rodgers RF, Ganchou C, Franko DL, Chabrol H. Drive for muscularity and disordered eating among French adolescent boys: a sociocultural model. Body Image. 2012;9(3):318–323. DOI

9. Valkenburg PM, Peter J. The differential susceptibility to media effects model. Journal of Communication. 2013;63(2):221–243. DOI

10. Gilmore AB, Fabbri A, Baum F, et al. Defining and conceptualising the commercial determinants of health. The Lancet. 2023;401(10383):1194–1213. DOI

11. Sheeran P, Webb TL. The intention–behavior gap. Social and Personality Psychology Compass. 2016;10(9):503–518. DOI

12. Fernan C, Schuldt JP, Niederdeppe J. Health halo effects from product titles and nutrient content claims in the context of “protein” bars. Health Commun. 2018;33(12):1425–1433. DOI

13. Rramani Q, Barakat Y, Jacob G, et al. Nutrition claims influence expectations about food attributes, attenuate activity in reward-associated brain regions during tasting, but do not impact pleasantness. Brain and Behavior. 2023;13(1):e2828. DOI

14. McKeon GP, Hallman WK. Front-of-package protein labels on cereal create health halos. Foods. 2024;13(8):1139. DOI

15. Sarrouy JL. Effect of high-protein claim on consumer’s perceptions of healthful indulgences. Master’s dissertation. Iscte Instituto Universitário de Lisboa; 2023. Non-peer-reviewed dissertation. Primary source

16. Kim DE, Kalaitzandonakes M, Ellison B. Added protein, added value? Consumer demand for protein labels across product types, flavors, and consumer segments. Food Quality and Preference. 2026;139:105842. DOI

17. Boyland E, McGale L, Maden M, et al. Association of food and nonalcoholic beverage marketing with children and adolescents’ eating behaviors and health: a systematic review and meta-analysis. JAMA Pediatrics. 2022;176(7):e221037. DOI

18. Ganson KT, Nguyen L, Ali ARH, Hallward L, Jackson DB, Testa A, Nagata JM. Associations between social media use, fitness- and weight-related online content, and use of legal appearance- and performance-enhancing drugs and substances. Eat Behav. 2023;49:101736. DOI

19. Gulec H. From scroll to sale: online media engagement predicts adolescents’ online supplement purchases in a nationally representative cross-sectional study. Front Psychol. 2026;17:1722369. DOI

20. Ganson KT, Piatkowski T, Testa A, He J, Nagata JM. Social media engagement and anabolic-androgenic steroid use intentions among boys and men in Canada and the United States. Body Image. 2026;57:102057. DOI

21. Beos N, Kemps E, Prichard I. The impact of fitness and supplement TikTok content on body, nutrition and fitness satisfaction, and intentions to use muscle-building substances in young men. Body Image. 2026;57:102082. DOI

22. Eisenberg ME, Wall M, Neumark-Sztainer D. Muscle-enhancing behaviors among adolescent girls and boys. Pediatrics. 2012;130(6):1019–1026. DOI

23. Clark SJ, Gebremariam A, Schultz SL, Beel L, Woolford SJ. Teens and protein. C.S. Mott Children’s Hospital National Poll on Children’s Health. 2024;46(2). October 21, 2024. Institutional poll; parent-reported data. Primary source

24. Royal Children’s Hospital National Child Health Poll. Teens and food: do they know the facts? Poll 41. April 2026. Institutional survey report. Primary source

25. Beltrá M, Borrás F, Ropero AB. Are foods with protein claims healthy? A study of the Spanish market. Nutrients. 2024;16(24):4281. DOI

26. Koop J, Fedde S, Hägele FA, Beunink C, Müller MJ, Bosy-Westphal A. Nutritional value and environmental aspects of high-protein ultra-processed foods on the German market. Public Health Nutr. 2024;27:e211. DOI

27. Hägele FA, Herpich C, Koop J, et al. Short-term effects of high-protein, lower-carbohydrate ultra-processed foods on human energy balance. Nature Metabolism. 2025;7(4):704–713. DOI

28. Consumer Reports. Protein powders and shakes contain high levels of lead. October 14, 2025; updated January 8, 2026. Consumer testing; not a peer-reviewed population study. Primary source

29. Horváth IL, Kajner G, Galbács G, Csupor D. Analysis of heavy metal content in protein powders available on the Hungarian market: a reassuring snapshot, but not a reassuring quality guarantee. Journal of Nutritional Science. 2025;14:e49. DOI

30. Bandara SB, Towle KM, Monnot AD. A human health risk assessment of heavy metal ingestion among consumers of protein powder supplements. Toxicol Rep. 2020;7:1255–1262. DOI

31. Bulens A, Calzo JP, Eik-Nes TT, et al. Anabolic steroid initiation among boys and young men after use of muscle-building supplements. JAMA Netw Open. 2024;7(12):e2450566. DOI

32. Lamina T, Brandt S, Abdi HI, et al. The effect of protein intake on bone disease, kidney disease, and sarcopenia: a systematic review. Curr Dev Nutr. 2025;9(3):104546. DOI

33. Devries MC, Sithamparapillai A, Brimble KS, Banfield L, Morton RW, Phillips SM. Changes in kidney function do not differ between healthy adults consuming higher- compared with lower- or normal-protein diets: a systematic review and meta-analysis. J Nutr. 2018;148(11):1760–1775. DOI

34. Darling AL, Manders RJF, Sahni S, et al. Dietary protein and bone health across the life-course: an updated systematic review and meta-analysis over 40 years. Osteoporos Int. 2019;30:741–761. DOI

35. Ganson KT, Nagata JM. Adolescent and young adult use of muscle-building dietary supplements: guidance for assessment and harm reduction approaches to mitigate risks. J Adolesc Health. 2024;75(3):379–382. DOI

36. US Food and Drug Administration. FDA 101 Dietary Supplements. Primary agency explanation. Accessed October 14, 2026. Primary source

37. US Food and Drug Administration. 21 CFR §§101.13, 101.54, and 101.9(c)(7). Nutrient content statements, qualifying claims, and protein nutrition labeling. Current legal text consulted October 14, 2026. Primary source

38. European Parliament and Council. Regulation (EC) No 1924/2006 on nutrition and health claims made on foods. Annex, source of protein and high protein. Primary source

39. World Health Organization. Policies to protect children from the harmful impact of food marketing WHO guideline. Geneva: WHO; 2023. Primary source