Abstract
udy compared the body composition and strength of the lower extremity parameters between novice runners (NRs) and amateur marathon runners (AMRs). A total of 50 NRs (33.84±4.32 years) and 50 AMRs (33.36±5.55 years) were analyzed cross-sectionally. Bioelectrical impedance analysis measured body composition parameters, and isokinetic testing assessed knee muscle strength. The results showed that compared to AMRs, NRs had lower fat-free mass (FFM), skeletal muscle mass (SMM), and total body water (TBW) (−15,−12, and−5%; allp< 0.01) but higher body fat percentage (PBF) and visceral fat area (VFA) (+27 and +32%; bothp< 0.01). They also had 6% lower knee extensor (KE) strength and 31% lower knee flexor (KF) strength on the dominant legs (DLs) and 14% lower KF strength on the non-dominant legs (NDLs). In addition, their hamstring–quadriceps ratio (H: Q) was 24% lower on the DLs and 9% lower on the NDLs. The NRs exhibited significant negative correlations between PBF, VFA, and knee muscle strength
6% lower knee extensor (KE) strength and 31% lower knee flexor (KF) strength on the dominant legs (DLs) and 14% lower KF strength on the non-dominant legs (NDLs). In addition, their hamstring–quadriceps ratio (H: Q) was 24% lower on the DLs and 9% lower on the NDLs. The NRs exhibited significant negative correlations between PBF, VFA, and knee muscle strength (r =−0.54 to−0.42, p< 0.01), while the AMRs had significant negative correlations only for PBF (r =−0.59 to −0.57,p< 0.001). In conclusion, the NRs exhibited lower FFM and TBW, higher PBF and VFA, and reduced muscle strength. In contrast, the AMRs exhibited the opposite pattern. These findings suggest that NRs with elevated body fat (BF) indicators should prioritize fat reduction and performance enhancement, while those with lower muscle mass require targeted programs to increase muscle capacity and joint stability. This approach may advance them toward the level of AMRs. Future studies should adopt longitudinal designs to explore how training interventions influence the physiological adaptations observed in runners at different experience levels. Keywords:amateur marathon runners; novice runners; body composition; lower limb strength 1. Introduction With the rise in public health awareness, running has become one of the most popular fitness exercises across the world [1]. Within the running community, two distinct groups emerge—novice runners (NRs) and amateur marathon runners (AMRs). NRs are commonly associated with those who run <10 miles per week, have no former experience in running, Sports2025,13, 287 https://doi.org/10.3390/sports13090287
Sports2025,13, 287 2 of 13 and never take part in a running competition [2]. On the other hand, AMRs are commonly associated with those who run consistently for more than one year and have run more than 40 km per month or have participated in at least one official marathon [3]. Although running is increasingly popular among the general population, the physiological differences between runners at different experience levels remain unclear. In particular, few studies have compared key physical characteristics—such as body composition and lower limb strength—between NRs and AMRs. Understanding these differences is important for developing individualized training strategies and for monitoring physical development as NRs progress toward higher levels of endurance performance [4,5]. Body composition and lower limb strength are core factors in running performance. Previous studies primarily focused on elite or professional athletes [6–8], leading to a limited understanding of amateur-level runners, who in fact constitute the majority of the running population. For example, Herrmann et al. [6] found that body composition was a better predictor of running performance than body mass index (BMI) in long-distance runners; they also found that adiposity was negatively correlated with running speed. In the field of exercise science, prior studies have explained the physiological characteristics of professional athletes [7,8]. For example, Chivate et al. [7] measured the lower limb strength of sprinters with different speeds and found that the faster the sprinters, the stronger their muscle strength. In addition, Stacho´n et al. [8] measured the body composition of male runners of different distances and found that, compared with sprinters, middle-distance runners had relatively low subcutaneous fat content, a difference reflecting the distinct body composition between the two types of athletes. Quantifying these traits not only provides important baseline characteristics for the majority of the running population but also enables sports scientists and coaches to develop a clearer picture of the physical attributes associated with different stages of running [9]. This understanding informs the development of appropriate training recommendations and supports individualized nutri- tional guidance for optimizing body composition [10]. Moreover, establishing benchmarks of different running levels allows for the objective assessment
majority of the running population but also enables sports scientists and coaches to develop a clearer picture of the physical attributes associated with different stages of running [9]. This understanding informs the development of appropriate training recommendations and supports individualized nutri- tional guidance for optimizing body composition [10]. Moreover, establishing benchmarks of different running levels allows for the objective assessment of physiological status and monitoring of adaptations [11]. Given the above considerations, this study aims to quantitatively analyze the body composition and lower limb strength of AMRs and NRs using body composition tests and isokinetic muscle strength tests to clarify the specific differences as well as their extent, which can help to reveal the differences in physiological characteristics of people with different running levels. Specifically, this study addresses the following research questions: (1) variations in body composition parameters between AMRs and NRs; (2) differences in knee extensor and flexor strength between the two groups when normalized to body mass; and (3) elucidating these variations may enhance the understanding of the physiological characteristics associated with different running levels. This information can inform NRs’ individualized decisions regarding training volume and intensity by providing comparative physiological profiles of more experienced AMRs, which in turn is important for developing individualized training approaches and monitoring physical development as NRs advance toward greater endurance capacity. 2. Materials and Methods 2.1. Study Design This study adopted a sex-balanced design, with male and female participants in a 1:1 ratio (25 males and 25 females in each group), in order to control for sex-related effects on physiological indicators. A stratified purposive sampling approach was used to recruit participants from local running clubs, fitness centers, and online platforms between February 2024 and December 2024. Participants were divided into two groups, NRs and
Sports2025,13, 287 3 of 13 AMRs, which compared the differences in body composition and lower limb strength parameters. All experiments were performed in the Biomechanics Laboratory, Department of Sports Medicine, Peking University Third Hospital. This study was approved by the Ethics Committee of the Third Hospital of Peking University (project identification Number: M2023268). All participants signed an informed consent form before enrollment. 2.2. Participants To minimize potential confounding variables, especially those related to sex and age, a matched enrollment strategy was implemented—each participant enrolled in one group was matched with a participant of the same sex and similar age in the other group. Participant selection was strictly based on predefined inclusion and exclusion criteria, not on test outcomes, ensuring the scientific integrity and objectivity of the sampling process. A total of 128 participants were assessed for their eligibility for inclusion in this study. Of the 128 participants, 115 completed the clinical study after excluding individuals who did not meet the inclusion/exclusion criteria. The inclusion criteria were as follows: (1) aged 18–45 years, regardless of sex; (2) not registered with the Chinese Athletic Association; (3) AMRs [3] were described as participants having been running consistently for more than one year and have run more than 40 km per month, or have participated in at least one official marathon; (4) NRs [2] were defined as running <10 miles per week, having no former experience in running, and never taking part in a running competition; and (5) voluntarysigned informed consent and commitment to cooperate with the experimental process were provided. The exclusion criteria were as follows: (1) history of lower limb musculoskeletal injuries or surgery; (2) severe cardiovascular, respiratory, or neurological diseases; and (3) inability to comply with experimental requirements or attend tests on time. Among the 115 participants who completed the study, 15 participants were excluded due to errors in the measurements of their body composition and lower limb strength. Consequently, the analytical sample comprised 50 AMRs and 50 NRs (Figure). The achieved sample size was larger than the required minimum, thereby providing adequate statistical power. All participants signed an informed consent
on time. Among the 115 participants who completed the study, 15 participants were excluded due to errors in the measurements of their body composition and lower limb strength. Consequently, the analytical sample comprised 50 AMRs and 50 NRs (Figure). The achieved sample size was larger than the required minimum, thereby providing adequate statistical power. All participants signed an informed consent form before enrollment. Figure 1.Flow chart of study sample selection. F = female; M = male. 2.3. Sample Size Calculation This study adopted a sex-balanced design, with male and female participants in a 1:1 ratio, in order to control for the sex-related effects on physiological indicators. The
Sports2025,13, 287 4 of 13 a priori sample size was calculated using G*Power 3.1 software based on the effect size (Cohen’s d = 0.65) derived from a pilot study, in which the mean difference between groups was approximately 3.69 units, with a pooled standard deviation of 5.67 units. Using an independent samplest-test, withα= 0.05 and power (1-β) = 0.80, a minimum of 34 validsamples per group (total≥68) was required to detect this effect. To account for a potential dropout rate of 20%, we planned to recruit 50 participants per group (25 malesand 25 females), totaling 100 participants. Despite excluding 15 participants due to procedural errors during testing, the final sample size of 100 exceeded the minimum requirement, ensuring sufficient statistical power. 2.4. Data Collection 2.4.1. Demographic and Body Composition Data A standardized self-reported form was filled out by all participants, including informa- tion about their age, running experience, running volume, sex, height, and dominant leg (DL) [12]; the DL is usually defined as the one that is used to kick a ball. All participants filled in a single measurement of body composition using the bioimpedance device InBody 770 (Biospace, Urbandale, IA, USA). Participants were instructed to fast for 8 h and to avoid stren- uous exercise for 24 h before testing. Upon arrival at the laboratory, they were asked to stand barefoot on the platform of the device with the soles of their feet resting on the electrodes, hold the handles with both hands, and remain stationary for 1 min while keeping the elbows fully extended and the shoulder joints abducted by approximately 30 ◦ (Figure 13]. The following data were recorded: (1) body fat percentage (PBF, %), which is defined as the proportion of fat mass relative to total body weight, reflecting the ratio of body fat to lean tissue; (2) skeletal muscle mass (SMM, kg), which represents the total amount of skeletal muscle in the body, including associated tissues; (3) fat-free mass (FFM, kg), which comprises the total non-fat tissue in the body, including muscle, bone, organs, and water; (4) total body water (TBW, kg), which is the
body weight, reflecting the ratio of body fat to lean tissue; (2) skeletal muscle mass (SMM, kg), which represents the total amount of skeletal muscle in the body, including associated tissues; (3) fat-free mass (FFM, kg), which comprises the total non-fat tissue in the body, including muscle, bone, organs, and water; (4) total body water (TBW, kg), which is the sum of all water present in the body and constitutes the primary component of fat-free body weight; and (5) visceral fat area (VFA, cm 2 ), which refers to the area of fat surrounding the abdominal visceral organs [14]. Among these indicators, previous studies have shown outstanding test–retest reliability (ICC ranging from 0.98 to 0.99 for PBF and from 0.99 to 1.00 for FFM) [15]. Earlier research has also documented a technical measurement error of 4.2% for PBF and 2.4 kg for FFM [ Figure 2.Body composition test picture.
Sports2025,13, 287 5 of 13 This device was used since a previous study reported an almost-perfect correlation using the gold-standard method (dual-energy X-ray absorptiometry, with a Pearson’s correlation coefficient >0.97) and an excellent reliability (ICC > 0.98) [13]. 2.4.2. Lower Limb Muscle Strength Features Isokinetic muscle strength testing was performed using an isokinetic test evalu- ation system ( CMV AG; Con-Trex MJ; Physiomed Elektromedizin AG, Schnaittach, Germany). Evidence confirms exceptional reliability (ICC: 0.87–0.98) for this testing methodology [17,18]. The technical error of measurement was reported to range between 6.6 and 10.1% [18]. Before data collection, participants engaged in a 5 min warm-up on a reclining bicycle (Technogym Xt Pro 600 Recline, Cesena, Italy), followed by basic stretching exercises targeting the lower body. Participants were positioned on the dynamometer with 80 ◦ hip flexion from anatomical neutral, adhering to manufacturer specifications. The lateral femoral condyle was carefully aligned with the axis of the dynamometer’s power arm. The lever arm length was individualized for each participant, and the resistance pad was placed just proximal to the medial malleolus [19]. To ensure stability, the participant’s torso, pelvis, tested thigh, and non-tested calf were securely immobilized. The calibration angle and starting position were set at a 90 ◦ knee flexion (FigureA) [ 20]. The knee’s range of motion was defined from a 10 to 90 ◦ flexion (FigureB) [ 21]. Additionally, gravity correction was applied to account for the influence of calf and power arm weights on peak torque during knee extension. Before each formal test, the participant was given a clear safety range of knee motion and was familiarized with the test procedure. After positioning, participants completed 1–2 submaximal familiarization trials. Subsequently, five maximal isokinetic contractions (60 ◦ /s) were executed for knee flexion–extension, with a 60 s rest between testing sequences. Testing commenced with the DL, followed by the non-dominant leg (NDL) after a 3 min recovery period. A single laboratory technologist conducted all isokinetic assessments to minimize inter-rater variability. Testing included knee extensors (KEs) and flexors (KFs) bilaterally at a 60 ◦ /s angular velocity. It has been shown that the
flexion–extension, with a 60 s rest between testing sequences. Testing commenced with the DL, followed by the non-dominant leg (NDL) after a 3 min recovery period. A single laboratory technologist conducted all isokinetic assessments to minimize inter-rater variability. Testing included knee extensors (KEs) and flexors (KFs) bilaterally at a 60 ◦ /s angular velocity. It has been shown that the most used angular velocity in force analysis is 60 ◦ /s [22]. Jenner et al. [23] found moderate levels of retest reliability for knee flexion and extension movements (ICC = 0.64–0.74) by performing five tests at a mean isokinetic speed of 60 ◦ /s on the knee joint while observing effect sizes (ESs) ranging from 0.01 to 0.43 and typical errors (TEs) ranging from 5.65 to 22.87, all of which were characterized by small-to-moderate levels. Figure 3.Lower extremity isokinetic strength test pictures: (A) initial phase; (B) final phase.
Sports2025,13, 287 6 of 13 Body weight-normalized peak torque (PT/BW) served as the primary outcome mea- sure for KEs and KFs. Additionally, the H:Q ratio was calculated by dividing the mean concentric KF PT by the mean concentric KE PT over the five repetitions [24,25]. The bilateral strength differences for the KFs and KEs were calculated based on the following equation [26,27]: Bilateralstrength deficit= DLpeak torque−NDLpeak torque DLpeak torque ×100(%) 2.5. Statistical Analysis Data are presented as mean±standard deviation (SD). The Kolmogorov–Smirnov test was used to check the normality of the data [28]. Pairedt-tests were conducted to evaluate the differences between the DL and NDL performance. Differences in body composition and muscular strength between novice runners and AMRs were assessed using independent- samplet-tests. Effect sizes were interpreted according to Cohen’s d thresholds [29]: Cohen’s d < 0.2 (small), 0.2≤Cohen’s d < 0.8 (moderate), and 0.8≤Cohen’s d (large). Associations between knee extensor strength and body composition metrics were examined with Pearson correlations. Correlation magnitudes were classified as small (0.1≤|r| < 0.3), medium (0.3≤|r| < 0.5), or large (0.5≤|r|≤1.0). All computations were executed in IBM SPSS 27.0 (SPSS Inc., Chicago, IL, USA), with statistical significance set atp< 0.05. 3. Results 3.1. Demographic and Anthropometric Characteristics Table terms of age, height, weight, BMI, sex distribution, or the DL proportion (allp> 0.05). In contrast, running experience and running volume differed substantially between groups (allp< 0.05). This balance was not coincidental but was intentionally achieved as part of the study design. Specifically, participants were recruited based on predefined inclusion criteria to ensure comparable distributions in age and sex across the two groups. This matching strategy was implemented to minimize potential confounding effects from non- running-specific variables, thereby enabling a clearer interpretation of group differences in running experience, volume, and physiological outcomes. Table 1.Descriptive statistics for demographic and anthropometric characteristics. AMR Group (n= 50)NR Group (n= 50) p Cohen’s d Age (years) 33.36 ±5.55 33.84 ±4.32 0.74 0.10 Running experience (years) 4.34 ±2.00 *** 0.54 ±0.21 <0.001 1.42 Running volume (km/month) 66.02 ±12.75 *** 19.48 ±6.62 <0.001 4.58 Height (cm) 171.32 ±6.14 170.48 ±8.51 0.69 0.11
running experience, volume, and physiological outcomes. Table 1.Descriptive statistics for demographic and anthropometric characteristics. AMR Group (n= 50)NR Group (n= 50) p Cohen’s d Age (years) 33.36 ±5.55 33.84 ±4.32 0.74 0.10 Running experience (years) 4.34 ±2.00 *** 0.54 ±0.21 <0.001 1.42 Running volume (km/month) 66.02 ±12.75 *** 19.48 ±6.62 <0.001 4.58 Height (cm) 171.32 ±6.14 170.48 ±8.51 0.69 0.11 Weight (kg) 64.51 ±7.06 66.49 ±11.35 0.46 0.21 BMI (kg/m 2 ) 20.95±1.77 21.75 ±2.56 0.21 0.36 Sex (male/female) 25/25 25/25 1.00 0.00 DL (right/left) 40/10 45/5 0.169 0.14 Note. BMI = body mass index; DL = dominant leg; AMR = amateur marathon runner; NR = novice runner. ***:p< 0.001. 3.2. Characterization of Body Composition in AMR and NR Groups Table of FFM, SMM, TBW, PBF, and VFA (allp< 0.01), and the effect sizes were of practical significance. The AMR group had lower PBF and VFA than the NR group (bothp< 0.001), with the effect size indicating a substantial magnitude of difference. Conversely, the AMR group had higher FFM, SMM, and TBW (allp< 0.01), and the effect sizes for these metrics
Description
The study analyzes differences in body composition and strength between novice and amateur marathon runners.