The assessments were conducted on cadets from Chile's first-division professional soccer teams. These evaluations were part of the battery of tests performed by the staff, and the players signed informed consent and permission for their data to be published. The assessments were carried out following the Declaration of Helsinki standards.

Fifty-eight male cadet players (age: 16.05 + 1.46 years, height: 175.14 + 7.10 cm, body mass: 65.78 + 8.26 kg) participated in this study. Goalkeepers were discarded due to the specificity of their positions.

All athletes were evaluated in the morning after fasting for 12 h. The assessment of body mass was conducted using a Tanita precision scale, model Bc-601, with a resolution of 0.1 g. Height evaluation was performed using a SECA brand stadiometer with 1 mm precision. A Lufkin brand metal tape measure with an accuracy of 1 mm was employed to measure the body circumferences. Diameter measurements were obtained using a Rosskraft caliper, while skinfold thickness was assessed using a Harpenden caliper. The assessments were performed following the protocols established by ISAK (39) by a certified level II professional.

The 30 m acceleration evaluation was performed on a natural grass soccer field with specific footwear for this surface type. For this, Stalker ATS II model radar (Applied Concepts, Dallas, TX, USA; accuracy + 1.61 km/h, sampling 46.9 Hz) was located on a tripod 10 m from the starting line and at the height of 1 m to align with the location of the center of mass (CM) of the subject (5). The athletes were instructed to initiate a 30 m maximal sprint starting from the standing starting position. Two attempts were performed with 5 min rest between them. Sprint performance metrics (split times at 0–5, 0–15, and 0–30 m) and mechanical outputs were analyzed based on the best time trial performance.

The speed-time data obtained via radar were processed using the Excel® spreadsheet developed by Morin and Samozino (40). This tool facilitates the computation of the maximum horizontal force exerted during the sprint (F0), maximum velocity (V0), and peak horizontal power output (Pmax). Furthermore, we determined the proportion of the total lower limb force applied horizontally to the ground (RF_max) and the rate of decline in horizontal force with increasing velocity (DRF). This approach relies on the fundamental principles of motion to establish the force-velocity relationship utilizing athlete velocity and body mass (41). Radar validity for these purposes was confirmed through comparison with force platforms (absolute bias of 3%–7%) (42). In terms of reliability, the mean typical error remained small (CV ≤ 8.4%) across all kinetic and kinematic variables (43), confirming its utility as a field tool for assessing these parameters.

To model the net horizontal anteroposterior ground reaction force (FH) applied to the body's center of mass (CM) over time, we employed the following equation (43):FH(t)=m*aH(t)+Faero(t)here, m denotes the runner's body mass in kilograms, and Faero(t) represents the aerodynamic drag to be overcome during sprinting, proportional to the square of the air velocity relative to the runner:Faero(t)=k*(vH(t)−vw)2where vw is the wind velocity (if applicable), and k is the runner's aerodynamic friction coefficient. In the vertical direction, during the acceleration phase, the runner's body CM ascended from the initial to an upright running posture and remained constant between the complete strides. Thus, by utilizing fundamental dynamics laws in the vertical plane, the mean net vertical ground reaction forces (FV) applied to the body CM during each complete stride can be modeled over time as equal to the body weight:FV(t)=m*ghere, “g” denotes the gravitational acceleration (9.81 m/s²).

The mechanical effectiveness of force application during running can be quantified in each support phase or stride by the ratio (RF in%) of FH to the corresponding total resultant ground reaction forces (FRes, in N) and throughout the entire acceleration phase by the slope of the linear decrease in RF with increasing velocity (DRF, in %/s/m). Since the starting block phase (push-off and subsequent airborne time) lasts between 0.5 and 0.6 s (44, 45), occurring for an average duration of ∼0.3 s, RF and DRF values can be computed from FH and FV values modeled for t > 0.3 s.

When examining the anthropometric variables influencing 5 m sprint times, BMI and TCC were identified as the main factors. The stepwise linear regression model explained 8% of the variability in 5 m sprint times (R² = 0.08), reaching statistical significance (p = 0.0384) with an F-value of 3.46. The model's RMSE is 0.09523, indicating an acceptable fit. Analysis of anthropometric variables for the 15 m sprint highlighted MHC as a significant determinant. The stepwise linear regression model accounted for 9% of the variability in the 15 m sprint times (R² = 0.09), with statistical significance (p = 0.0128) and an F-value of 6.62, confirming the model's reliability. The model, with an R² of 0.32, showed that BMI, % MM, and % BF collectively explained 32% of the variability in the 30 m sprint times (p < 0.001, F = 10.08). Table 1 presents the results of the stepwise linear regression analyses’ results examining anthropometric variables’ influence on the 5-, 15, and 30 m sprint times. For each distance, the table displays each predictor variable's non-standardized coefficient (B), standard error, t-value, p-value, 95% confidence interval, f² and ES.

Stepwise linear regression identified the relative Pmax as the primary mechanical variable that influenced the time to run 5 m. The model explained 16.8% of the variability in 5 m sprint times (R² = 0.168). This coefficient suggests that higher values of relative Pmax are associated with shorter 5 m sprint times. For the 15 m distance, relative Pmax was also identified as the primary mechanical variable explaining sprint time (97.57%; R² = 0.9757). The analysis identified V0 as the primary mechanical variable influencing the time to run at 30 m. The model explains 55.34% of the variability (R² = 0.5534). Table 2 presents the results of stepwise linear regression.

Analysis of the mechanical and anthropometric variables affecting 5 m sprint times revealed significant results. The model explained 36.01% (R² = 0.3601) of the variance in 5 m sprint times (p < 0.0001, F = 11.69). The coefficients indicated that a higher Pmax was associated with shorter 5 m sprint times, whereas an increase in MCH was related to longer sprint times. For 15 m, no anthropometric variable was added, leaving only Pmax as the most determining variable of this distance. Our model explained 59.75% of the variability in the 30 m sprint times (R² = 0.5975, p < 0.0001, F = 29.20). The coefficients indicate that a higher V0 and a larger TCC are associated with reduced sprint times, whereas an increased MHC is linked to longer sprint times. Table 3 presents the results of stepwise linear regression for 30 m.

Regarding anthropometric variables related to sprinting, our results align with those of previous studies (48–50), where higher BMI values were negatively associated with performance in the 5 and 30-meter sprints (small and large ES, respectively). Thus, the most critical indicator is body composition, where lower-limb MM is associated with better sprint performance (51). Our results support this: TCC and MHC are anthropometric variables associated with better performance at 5 and 15 m (ES: small), respectively. Chelly et al. (52) report significant correlations between thigh muscle volume and 5 m sprint performance in young soccer players (r = 0.43, p = 0.05), and Tottori et al. (53) identified the cross-sectional area of the gluteus maximus as a predictive variable in 100-m performance in adults sprinters. Similar to other studies (54–56), our results showed a negative association between% BF and short-distance sprint times and % MM. This result contradicts our hypothesis, which can be explained by the contribution of MM to the total BM. These results provide valuable insights into MM gain, specifically in the lower limbs, reduction of %BF, and optimal BM to enhance sprint performance in football players.

Our results indicate that the relative Pmax is the variable that best explains the time to cover 5 and 15 m (medium and large ES). At the same time, V0 is the mechanical variable that has the most significant influence on the 30-meter sprint time (ES: large). These findings are consistent with those of Haugen et al. (57), who found an almost perfect inverse correlation (r = 1 ± 0.01) between the time to cover short distances (10 m) and relative Pmax. They also observed that the V0 variable correlated more significantly as distance increased. Samozino et al. (58) emphasized that the Pmax and FV profile of each athlete were the most significant variables in short-distance sprints (>30 m). Other studies have reported that F0 and V0 are the variables that contribute most to the time over short distances (59, 60). While these studies may seem contradictory, we must clarify that mechanical power is the product of applied force and velocity (61). This means that Pmax is a composite variable that reflects the capacity to generate force and achieve a high velocity. Therefore, the results were similar at first glance. Pmax integrates the ability to effectively apply force during the initial acceleration and the capacity to maintain a high maximum velocity over longer distances. The results suggest that optimizing Pmax can improve the initial acceleration and achieve higher maximum speeds.

An analysis that considered anthropometric and mechanical variables was conducted to determine their relationship with sprint performance. Our results enhanced the accuracy over distances of 5 and 30 m. Pmax, V0, and TCC negatively affected the time to cover 5 m (higher values, shorter times; ES: large). The impact of Pmax on short distances has been discussed in the previous section. For 30 m, V0 and TCC contribute to reducing time (ES: large), while MHC slightly increases it. The latter may seem contradictory, as this variable has been shown to increase V0. Studies have highlighted the inconsistency in the relationship between sprint performance and hip extensors (62), emphasizing that they relate to the sprint phase (acceleration, maximum velocity, and deceleration). According to our results, MHC is an essential variable in the initial meters of the sprint (>15 m) but not in longer sprints.