On the first day of the protocol, participants underwent anthropometric and body composition assessments to estimate BM and quantify FFM levels. At this time, the first 24h-R was also administered to characterize the participants' dietary pattern prior to the intervention. After a 24-h interval, resting blood samples were collected. Subsequently, participants were taken to an official athletics track, where they performed a brief five-minute warm-up consisting of jumps and short walks. This was followed by the RSE protocol, which consisted of three sets of six sprints, with 10 s of passive rest between sprints and five minutes of passive rest between sets. Additional blood samples were collected immediately after the RSE, as well as 2 h and 24 h after exercise, to analyze serum cytokine levels at different time points (pre, immediately after, 2 h, and 24 h after RSE). It is worth noting that, at the 24-h after protocol time point, participants completed the second 24h-R, allowing for the monitoring of dietary intake during the after-intervention period as well (Figure 1).

To ensure blinding regarding the implementation of the RSE protocol, the intervention was conducted by an external collaborator not affiliated with the research team, thereby guaranteeing impartiality at this stage of the study. The 24h-R were administered by a registered dietitian who was also external to the research team. Blood sample collection was carried out by an independent professional, ensuring the neutrality of the procedure. During the laboratory analyses of cytokines, all data were coded to conceal participant identities, preventing evaluators from knowing the origin of each sample. Similarly, during statistical analyses, the data remained masked, making it impossible to identify the groups or individual participants. These measures reinforced the objectivity and integrity of the results obtained.

Body mass was measured using a Filizola® digital scale with a 200 kg capacity and 0.10 kg precision (São Paulo, Brazil). For this assessment, participants were barefoot and wore light clothing. Height was measured using a Sanny® stadiometer with a precision of 0.1 mm (São Paulo, Brazil). All measurements were taken by a single evaluator following the protocols of the International Society for the Advancement of Kinanthropometry (ISAK) (24). The intra-observer technical error for anthropometric measurements was ≤1.0% (25). FFM levels were assessed using dual-energy x-ray absorptiometry (DXA) with a LUNAR®/GE PRODIGY - LNR 41.990 device (Washington, DC, USA), employing the enCORE software, GE Healthcare®, version 15.0 (Madison, WI, USA). For adolescent participants, population-specific algorithms were applied (26). During the assessments, the DXA device operated under the following configuration: whole-body scan; voltage (kV): 76.0; current (mA): 0.150; radiation dose (µGγ): 0.4 (very low, with no health risk).

The BM profile was assessed through estimates of peak height velocity (PHV), puberty scores, and skeletal maturity. PHV stage was determined using the mathematical model proposed by Moore et al. (27), for males aged between 8 and 18 years: PHV = −8.128741 + [0.0070346 × Age[years] × Sitting height(cm)], where: (cm) = centimeters. Based on the outcome of this model, PHV classification was defined as: Pre PHV (<−1), Circum PHV (entre −1 e 1) e Post PHV (>1). The puberty score was calculated using the model proposed by Almeida-Neto et al. (28), for males aged between 6 and 18 years: Puberty (score) = −17.357 + (0.603 × Age(years)] + [0.127 × Sitting height(cm)]. From these scores, Tanner stages were classified as follows: I ≤ −1.815; II = −1.816 to −0.605; III = −0.606–0.695; IV = 0.696–3.410 & V > 3.410. Stage I corresponds to the prepubertal phase, stages II to IV correspond to the pubertal phase, and stage V represents the postpubertal phase (28). Skeletal age was determined based on the mathematical model proposed by Cabral et al. (29): Skeletal age = −11.620 + 7.004 × (Height_(m)) + 1.226 × (Dsex) + 0.749 × (Age_(years)) - 0.068 × (Triceps skinfold_(mm)) + 0.214 × (Corrected arm circumference_(cm)) - 0.588 × (Humerus diameter_(cm)) + 0.388 × (Femoral diameter_(cm)), where, Dsex: For male sex = 0; for female sex = 1. (m): meters. (mm): millimeters. (cm): centimeters. The corrected arm circumference was calculated using the formula: Corrected Arm Circumference_(cm) = Contracted biceps circumference_(cm) - (Triceps skinfold_(mm)/10), where, (mm): millimeters. (cm) centimeters. After determining skeletal age, skeletal maturity was classified using the equation proposed by (30): Skeletal maturity = Skeletal Age – Chronological age. Based on this difference, skeletal maturation was categorized as follows: Delayed (<−1.0), Synchronous (between −1.0 and 1.0), and Accelerated (>1.0).

On the day of the RSE protocol, prior to the first blood collection, upper respiratory tract infection symptoms (URTIS) were assessed using the Wisconsin Upper Respiratory Symptom Survey-21 (WURSS-21) questionnaire (31, 32). Only volunteers who showed no signs of URTIS were retained as participants in the study. It is worth noting that this tool has been previously used in a study involving young athletes and validated for pediatric populations (33, 34). In the present study, none of the participants needed to be excluded due to URTIS symptoms on the day of RSE.

The physical activity level was assessed using the web-based version of the International Physical Activity Questionnaire (IPAQ) (33), validated for estimating metabolic equivalents (METs) related to energy expenditure from physical activities (35). The questionnaire covers activities at work, during transportation, household tasks, and leisure/sports. The web-IPAQ automatically calculates METs for habitual energy expenditure, providing scores for each domain and a total score. Based on total METs, the web-IPAQ classifies physical activity level as low, moderate, or high. Participants completed the web-IPAQ in the presence of at least one researcher and, in the case of adolescents, with a guardian to assist with interpreting the questions. All participants in the present study reported a high level of physical activity.

Sleep patterns were assessed using a sleep diary, in which participants recorded the time they went to bed, perceived time to fall asleep, wake-up time, total time in bed, and whether there were interruptions during the night (36). Based on these entries, sleep efficiency was calculated (considered good when >85%) (36). Records were made on the day of the RSE (referring to the previous night) and on the day following the RSE (referring to the intervention night). All participants in this study presented sleep efficiency between 85% and 92%.

The dietary pattern was assessed using two 24h-R, applied on non-consecutive days, including one typical day and one atypical day (37). To estimate habitual intake, the Brazilian Food Composition Table (38) and the virtualized Multiple Source Method (version 1.0.1.87) were used to adjust for intra- and interpersonal variability (39). The inflammatory potential of the diet was measured using the DII, which quantifies the pro- or anti-inflammatory effect of diet based on scores associated with inflammatory biomarkers and health outcomes (40).

The DII calculation was based on the regional dietary pattern of the city of Natal (Northeast Brazil), where the participants of the present study live and train. The specific version used was characterized and validated by the BRAZUCA study (22), which includes 35 of the 45 items from the original DII model described by Shivappa et al. (41). It is noteworthy that the 10 additional items from the original model were not assessed in the 24h-R analyses of the participants in this study, aligning with the specific parameters adapted from the BRAZUCA study (22).

Each parameter was assigned a score of +1 (pro-inflammatory), −1 (anti-inflammatory), or 0 (neutral). The DII score was calculated based on the Z-score of individual intake, converted to a centered percentile and weighted by the nutrient's inflammatory effect score. The global DII corresponds to the sum of the individual parameter values and is interpreted as follows: anti-inflammatory diet (<0) or pro-inflammatory diet (≥0) (42). Details of the DII analysis are provided in Supplementary Material 1 (Data spreadsheet).

In the 24 h prior to the RSE protocol, participants were instructed to refrain from engaging in high-intensity physical activities. The RSE involved 18 high-intensity sprints performed on an official athletics track, under an average temperature of 26.8 ± 0.8°C. The protocol was divided into three sets, each consisting of six 35-m sprints with 10 s of passive rest between repetitions. Between sets, participants were given a 5-min passive rest period, during which they remained seated on chairs with back support and knees flexed at 90°. During the RSE, participants received verbal encouragement, and the rest intervals between each sprint were monitored by two timekeepers. A third timekeeper monitored the rest periods between sets. All timekeepers used digital stopwatches (TecTime®, São Paulo, Brazil). Participants were instructed to sprint at maximal effort to ensure the high intensity of the exercise.

The Rating of Perceived Exertion (RPE) scale, proposed by Borg (43), was used to assess the subjective intensity of the RSE. The scale was applied immediately after the final sprint. This visual, monochromatic scale ranges from 6 to 20, where 6 corresponds to complete rest and 20 to maximal effort. Participants underwent prior familiarization with the RPE scale 72 h before the RSE and were re-familiarized on the day of the protocol. All participants reported RPE scores between 17 and 20, indicating a high-intensity effort during RSE.

At each time point in the present study (pre, immediately after, 2 h, and 24 h after RSE), 10 ml of peripheral blood was collected from the antecubital vein using the vacuum method. The blood was stored in dry tubes (BD-Vacutainer, 5.4 mg Plus Plastic) and processed for laboratory analysis. We analyzed the serum concentrations of the cytokines [interleukin (IL)] IL-1β, IL-6, IL-8 and IL-10 (44). For this purpose, peripheral blood samples were assessed by flow cytometry using a BD Biosciences® device (Model: BD FACSCanto™ II Flow Cytometer, Serial Number: V96100619, New York, USA), with the BD™ Cytometric Bead Array (CBA) Human Inflammatory Cytokines kit (São Paulo, Brazil). All procedures followed the manufacturer's instructions provided in the kit manual (45). After flow cytometry, the data were processed using BD Biosciences® FCAP Array™ software (Version 3.0; New York, USA).

All statistical analyses were performed using the open-source software JAMOVI® (version 2.3, Sidney, Australia), and figures were produced using GraphPad Prism® (version 8.01, California, USA). The significance level was set at p < 0.05.

The results of this study demonstrate that FFM plays a crucial role in immunoregulation, as reflected in the modulation of cytokine levels in response to exercise (Table 3). Both group comparisons (Figures 2, 3 and Table 3) and correlation analyses (Table 1) indicate that individuals with higher FFM exhibit elevated IL-6 levels at rest and immediately after RSE. These findings may be attributed to the endocrine function of muscle tissue, which releases IL-6 in response to mechanical and metabolic stress (2). However, the negative correlation between FFM and IL-10 observed in adults (pre, 2 h, and 24 h after RSE) deserves special attention. Although IL-10 is classically considered an anti-inflammatory cytokine, its reduction in individuals with higher FFM may suggest a distinct immunological adaptation in adult athletes, in which there is a greater demand for acute inflammatory responses at the expense of chronic anti-inflammatory modulation. This hypothesis warrants further investigation (5, 44).

During adolescence, the growth and maturation of muscle tissue may influence inflammatory responses (14). This may serve as a metabolic adaptation and repair mechanism, which could explain why adolescents with higher FFM showed negative correlations with IL-6 and IL-8 at 24 h after RSE (Tables 1, 3), suggesting a more efficient inflammatory recovery (4). IL-8, a cytokine associated with neutrophil recruitment and maintenance of the inflammatory response (44, 50), may have been downregulated to prevent a prolonged inflammatory state in better-conditioned athletes (51, 52).

In contrast, adolescents with lower FFM showed a greater percent increase (Δ%) in IL-6 at 24 h after RSE compared to their peers (Figure 2B), suggesting that lower muscle reserves may compromise repair mechanisms and prolong the inflammatory response (2). Among adults, those with higher FFM consistently exhibited lower IL-10 levels (pre, 2 h, and 24 h after RSE) and a positive association between FFM and IL-6 at 2 h after RSE (Table 1), reinforcing the importance of additional regulatory mechanisms in the modulation of the inflammatory response (1, 4, 5). It is worth noting that the absence of significant correlations between FFM and IL-1β in adolescents may indicate that this cytokine is less sensitive to variations in body composition at this age or that other factors (e.g., training load or nutritional status) exert a greater influence (Tables 1, 3) (53).

When controlling for BM (Table 1), a positive association was observed between FFM and IL-10 levels, suggesting that adolescents with higher FFM may present an anti-inflammatory predisposition, regardless of maturation stage. This adjustment also revealed a similar correlation for IL-6 levels, indicating that individuals with more muscle mass exhibit an acute IL-6 response to exercise, reinforcing the role of muscle tissue as a potential source of this cytokine (Table 3) (1). In this context, the inverse correlation with IL-1β at 24 h after RSE in adolescents suggests that FFM may attenuate late inflammatory responses, possibly due to greater tissue repair capacity (44, 53).

Consistently, less mature adolescents showed a greater percent increase (Δ%) in IL-6 at 24 h after RSE compared to their more mature peers (Figure 3B), suggesting that inflammatory markers may be influenced by BM (10), possibly due to a still developing physiological system (14). Less mature individuals also tend to have lower FFM levels compared to their more biologically advanced peers (54), which may explain the similarity of findings when adolescents were grouped by FFM levels and BM status (Figures 2B, 3B). These findings may not be exclusively linked to lower FFM, but also to hormonal variations (e.g., testosterone and GH) that influence both muscle development and immune activity (53, 54).

Furthermore, adolescents with higher FFM appeared to exhibit lower IL-1β levels at 24 h after RSE (Table 1). IL-1β is a pro-inflammatory cytokine involved in initiating the inflammatory response and muscle soreness (44, 53). Lower concentrations at 24 h after RSE in individuals with higher FFM may indicate a more controlled inflammatory response and more efficient recovery, potentially associated with training-induced adaptations and greater anti-inflammatory capacity (1, 2).

The results of this study demonstrate that DII significantly influenced cytokine responses following RSE (Figures 4C,H). Specifically, lower DII scores (indicating a more anti-inflammatory dietary pattern) were associated with higher IL-8 levels in adolescents and higher IL-10 levels in adults. These findings align with literature showing an inverse relationship between anti-inflammatory diets and pro-inflammatory markers such as IL-1β, IL-6, IL-8, TNF-α, and C-reactive protein (CRP) (40, 41). The production of pro-inflammatory interleukins by muscle tissues during exercise is well documented and is considered an adaptive response that triggers anti-inflammatory pathways, mobilizes glycogen, and promotes metabolic adaptations (1, 55).

Among adults with higher FFM and anti-inflammatory dietary patterns, the observed increase in IL-10 after exercise may reflect enhanced anti-inflammatory responses. This aligns with previous evidence that IL-6 induces IL-10 and IL-1ra, modulating the inflammatory cascade (56, 57). Individuals with less inflammatory dietary patterns tend to present a more favorable redox and immune basal state, facilitating effective immune responses to physical stress (52, 58–61).

Individuals with less inflammatory dietary patterns tend to present a more favorable redox and immune basal state, facilitating effective immune responses to physical stress (14, 61). The DII reflects the inflammatory potential of various macronutrients and micronutrients, which directly affect cytokine levels (41). Diets rich in saturated fats, refined sugars, and low in fiber are associated with higher IL-6, IL-8, CRP, and TNF-α, whereas diets rich in fruits, vegetables, and omega-3 fatty acids promote anti-inflammatory responses, including IL-10 (40, 60).

Although the exact mechanisms remain to be clarified, it is presumed that an anti-inflammatory dietary pattern contributes to redox balance, reduced baseline expression of nuclear factor kappa B (NF-κB), and lower secretion of pro-inflammatory adipokines (e.g., TNF-α, IL-1β), thereby enabling a rapid and transient IL-6 response to exercise without prolonged inflammation (62–64). During the recovery phase, the modulatory effect of DII is manifested through more efficient IL-10 expression, aiding inflammation resolution.

The findings of this study provide relevant insights for training load management, particularly in adolescent and adult athletes involved in intermittent sports. The observation that individuals with lower FFM exhibit greater inflammatory variation (especially IL-6 at 24 h after RSE) suggests a slower inflammatory recovery and a potentially higher risk of immune overload. This implies that athletes with lower muscle development may require longer recovery intervals between high-intensity sessions or complementary interventions (e.g., nutritional and sleep strategies) that support inflammation resolution.

In adolescents, the influence of BM on inflammatory responses reinforces the need to individualize training loads according to maturation status. Biologically less mature athletes showed more prolonged inflammatory responses (Figure 3B), which may hinder training adaptations. Monitoring PHV status, skeletal maturation, and puberty stages can help coaches and strength professionals avoid overloading immature athletes, promoting safer and more sustainable development.

In adults, greater FFM was associated with more effective IL-10 modulation, indicating a more favorable anti-inflammatory profile and enhanced recovery potential. This information may assist in adjusting training load based on athletes' morphological status, as individuals with higher FFM may tolerate greater volumes and intensities with lower risk of chronic inflammation.

Finally, the DII emerges as a practical and promising tool for sports science. Our findings indicate that dietary patterns with lower inflammatory potential (as reflected in lower DII values) are associated with more efficient immune responses to exercise, particularly in adults with higher FFM (greater IL-10) and adolescents (lower IL-8) (Figures 4C,H). This reinforces the need to incorporate regular nutritional assessments into training load monitoring. Nutritional intervention programs focused on healthy and anti-inflammatory diets, aimed at reducing DII, may serve as important complementary strategies. This nutritional approach seeks to reduce exercise-induced immune overload, optimizing recovery and sports performance across age groups and maturation stages.

Despite its relevant findings, the present study presents the following limitations: (i) Only athletes from intermittent sports were included, requiring caution when generalizing the findings to endurance sports; (ii) Only male athletes were analyzed, and extrapolation to female athletes should be made with caution; (iii) Although intragroup comparisons, statistical corrections, post hoc power analyses, and control for interaction variables (FFM, BM, and DII) were performed, caution is warranted when interpreting the results of the adult group due to the small sample size. We note that while the sample size and group imbalance are considerable limitations, the statistical methodology employed (GLMM) was chosen to minimize the impact of these factors, thereby increasing the internal validity of the results with the observation of 120 events when considering the time points (pre, immediately after, 2 h, and 24 h after RSE); (iv) Finally, a limitation to be considered is the absence of a control group for the RSE protocol; however, the repeated-measures design, combined with the GLMM analysis, allowed each participant to serve as their own control, strengthening the study's internal validity.

Future studies should seek to address the limitations outlined in this section.