Comparison of the H:Q Ratio Between the Dominant and Nondominant Legs of Soccer Players: A Meta-analysis

Abstract

Context:

Soccer players often have a dominant (D) leg, which could influence the relative strength between the quadriceps and hamstrings. The hamstring-to-quadriceps (H:Q) ratio can be assessed on a dynamometer at various velocities to provide information on injury risk.

Objective:

To assess the concentric hamstrings and concentric quadriceps strength ratio (conventional H:Q ratio) assessed in D and nondominant (ND) legs at various speeds in male soccer players.

Data Sources:

A systematic literature search was completed from inception to 2020 in PubMed, Academic Search Ultimate, CINAHL, and SPORTDiscus.

Study Selection:

Keywords associated with the H:Q ratio were connected with terms for soccer players. Titles and abstracts were screened by 2 reviewers based on inclusion and exclusion criteria related to sex, playing level, language, and measurement. A total of 81 studies were reviewed and 17 studies (21%) were used.

Study Design:

A meta-analysis with random effects modeling generated standardized mean differences with 95% CIs between legs and speeds.

Level of Evidence:

Level 3.

Data Extraction:

A total of 38 cohorts were identified, with 14, 13, and 11 cohorts assessed at low, intermediate, and high velocities, respectively. The Quality Assessment Tool for Observational Cohort and Cross-sectional Studies from the National Institutes of Health was used.

Results:

The mean H:Q ratio at low velocities was 59.8 ± 9.5% in D leg and 58.6 ± 9.9% in ND leg, 64.2 ± 10.7% (D) and 63.6 ± 11.3% (ND) at the intermediate velocity, and 71.9 ± 12.7% (D) and 72.8 ± 12.7% (ND) at the high velocity. Low, intermediate, and high velocities had small effects of 0.13, 0.10, and -0.06, respectively.

Conclusion:

Conventional H:Q ratios vary across velocities but did not differ between D and ND limbs in male soccer players. This study may provide the foundation to establish norms and clinically meaningful differences.

The sport of soccer requires different movements ranging from slow running to explosive sprinting, jumping, and kicking. Often, activities like kicking can lead to unilateral strength dominance among muscle groups and limbs. ³¹ Muscle imbalances between the quadriceps and hamstrings in the dominant (D) and nondominant (ND) limbs may enhance the likelihood of injury, particularly to the hamstrings. ⁶ Male soccer players sustain an average of 0.6 muscle injuries per season, and injuries to the hamstrings are the most frequent. ¹³ Anatomic and hormonal differences in female players may lead to a sex-specific laxity of the ligaments around the knee joint, causing diverse injuries and rates among male and female soccer players. ¹⁰ About 96% of the injuries in male players occur in noncontact situations like sprinting or kicking. ¹³ In soccer, the hamstrings are highly activated and stretched as they absorb energy from the swing limb during sprinting, which creates a possible lengthening contraction injury. Kicking requires knee extension and activation of the quadriceps, putting the hamstrings in a stretched position. ¹⁸ If the hamstrings are too weak to offset this force, the knee joint can be at a greater risk of injury.^8,18

The strength of the agonist to antagonist muscles using the hamstring-to-quadriceps (H:Q) ratio has been helpful in the identification of injury risk and mitigation of strength imbalances to reduce injuries to the lower extremity as Dauty et al, ¹¹ Sangnier and Tourny-Chollet, ²⁸ and Yildiz and Kale ³⁰ agreed that H:Q ratios below 60% increase the risk for hamstrings and anterior cruciate ligament injuries. To determine the conventional H:Q ratio, the concentric strength of the hamstrings is compared with the concentric strength of the quadriceps. ¹³ Mathematically, the maximal peak torque of the hamstrings (in newton-meter [N·m]) is divided by maximal peak torque of the quadriceps (in N·m). While the recommended values of the conventional H:Q ratio are approximately 60% or greater, it is important to consider that this may vary depending on the angular velocity at which torque was assessed since the role of the hamstring muscles in stabilizing the knee joint becomes much more pronounced at higher velocities.^5,17 The H:Q ratio can also be quantified using the peak torques of eccentric hamstrings and concentric quadriceps contraction to simulate the coactivation pattern of the muscles during the deceleration movements often performed in the sport of soccer.^5,8 However, the norms for this method are less established making it difficult to make comparisons with previous studies.

In addition to agonist to antagonist strength differences, bilateral deficits are also a concern as soccer players almost always report having D and ND legs and the movements performed by each leg are different when kicking a ball. While the hamstrings stabilize the knee joint and the hip in the support limb (typically the ND leg), the activity of the hamstrings in the kicking limb is minimized to allow a quick movement for the quadriceps, ²⁵ which could lead to imbalances due to stronger hamstrings in the ND leg and stronger quadriceps in the D leg. Therefore, it is important to evaluate the H:Q ratio between both legs. Strength imbalances greater than 15% between the legs have a greater likelihood of injury.^10,20 Studies that assessed D and ND legs in soccer players are inconsistent as some do not show significant differences in the H:Q ratio,^10,22,31 whereas others show significant differences but only at certain velocities.^9,16

Since the preferential use of the D and ND legs can play an important role in risk of injury among soccer players, the aim of this meta-analysis was to summarize the H:Q ratios found in the D and ND legs in adult male soccer players. While the H:Q ratio has been reviewed in soccer players, ⁵ no comparison of the H:Q ratio between limbs has been done. It was hypothesized that the D leg would have a lower H:Q ratio than the ND leg due to the different functions of the kicking limb and the stabilizing limb.

Methods

Experimental Approach to the Problem

A systematic review and meta-analysis was conducted to investigate the differences in the conventional H:Q ratio in the D and ND leg in male soccer players. Female players were not included in this study due to anatomic and hormonal differences likely leading to differences in injuries and risk factors. ¹⁰ Studies were examined based on the isokinetic velocities tested and categorized into 3 different velocity groups: low (30° s^-1-60° s^-1), intermediate (90° s^-1-180° s^-1), and high (240° s^-1 and higher) velocity. The categorization was based on a previous systematic review that investigated the H:Q ratio on soccer players. ⁶ The H:Q ratio of the D leg was compared with that of the ND leg.

Protocol and Registration

This meta-analysis was performed in accordance with the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines, ²³ and the study was registered in PROSPERO (CRD42020198913).

Search Strategy and Study Identification

A systematic literature search was completed from inception to July 26, 2020 in the following electronic databases: PubMed, Academic Search Ultimate, CINAHL, and SPORTDiscus. Keywords included terms associated with the H:Q ratio and was connected with terms for soccer players. The search was performed with no restrictions and the final search string was:

“H:Q ratio” OR “hamstring to quadriceps” AND soccer OR football OR futbol OR athlete AND leg OR muscle OR “knee joint” AND strength OR injur* OR asymmetr* OR imbalance*”

All studies were stored in a citation manager (Zotero) and duplicates were removed before further processing (Figure 1). Covidence was used to include and exclude studies based on the eligibility criteria.

Figure 1.

Depiction of the PRISMA diagram. PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

Eligibility Criteria

All articles were screened independently by 2 reviewers according to the following inclusion criteria: (1) healthy adult men (18+ years), (2) H:Q ratio was measured in D and ND legs, (3) isokinetic dynamometer was used, and (4) study population included soccer players (amateur, college or [semi-]professional player). Studies were excluded from this review if (1) the study population did not include soccer players or was only with female players, (2) the study was written in a language other than English, and (3) the researchers did not compare D with ND legs.

Data Extraction and Collection

After an initial screening, all retained articles were considered relevant and the following information was extracted after accessing the full texts: population characteristics (age, weight, height, playing experience), number of participants, velocities being measured, absolute torque (in N·m) from the various velocities, and the primary variable of conventional H:Q ratio expressed as a percentage. Velocities were categorized as low (30° s^-1-60° s^-1), intermediate (90° s^-1-180° s^-1) and high (240° s^-1 and higher). Asymmetry was calculated by the researchers as the percentage difference between peak torque of the D and ND legs. In cases in which multiple point assessment was performed because of an intervention, only the baseline measurement was used. If only a figure was available, the H:Q ratios were estimated based on that figure. All data were extracted and screened by one reviewer.

Subjects

A total of 227 studies were identified using the established search terms (Figure 1). After title and abstract screening by 2 independent reviewers, a total of 81 studies were deemed eligible for full-text screening. There were no conflicts between reviewers. After this, 17 studies were eligible for use in the qualitative and quantitative syntheses; 9 studies used the Biodex dynamometer, 5 used a Cybex dynamometer and the other 3 used the Kin-Com, Isoforce, or Isomed dynamometer. The participants performed various warm-up protocols before testing. Torque was always measured in N·m (Table 1). One study measured the H:Q ratio only at a low velocity (30° s^-1- 60° s^-1) ²⁷ ; 2 studies used only intermediate velocity (90° s^-1-180° s^-1)^28,29; 4 studies used both low and intermediate velocities^2,4,8,13; 3 studies measured the H:Q ratio at low and high velocities (240° s^-1 and higher)^3,8,14; one study used intermediate and high velocities ²¹ ; and 6 studies utilized low, intermediate, and high velocities.^{6,10,15,19.30,31} This resulted in data analysis of 569 participants, with 476 participants at low velocities, 401 participants at intermediate velocities, and 348 participants at high velocities. The 569 participants had an average age of 23.6 ± 1.0 years, a height of 178.7 ± 7.6 cm, and a mass of 74.3 ± 3.0 kg. Their playing experience varied between 3 years and more than 14 years, while the weekly training time varied between 4 hours per week and daily training.

Table 1.

Characteristics of the included studies (H:Q ratios correspond to stated speeds in protocol)

Study	Participants	Dynamometer Protocol	H:Q Ratio D Leg, % (± SD)	H:Q Ratio ND Leg, % (± SD)
Aginsky et al 2	28 Elite male soccer players (South Africa), age 27.6 ± 2.8 years and weight 76.3 ± 2.3 kg. Playing experience not specified	Cycle ergometer warm-up Biodex dynamometer 5 reps at 60° s-1 15 reps at 180° s-1	64.7 ± 9.3 68.8 ± 9.1	60.9 ± 11.1 63.5 ± 9.1
Aktug 3	27 Male amateur soccer players, age 23.1 ± 5 years, height 181.6 ± 4.3 cm and weight 74.3 ± 4.3 kg. Playing experience not specified	Running and stretching warm-up Biodex dynamometer 10 reps at 30° s-1 15 reps at 240° s-1	51.86 ± 6.31 66.71 ± 10.27	53.39 ± 6.45 66.76 ± 10.4
Arsenis et al 4	32 Young soccer players (Greek 1st division young championship), age 19 ± 1 years, height 178 ± 6 cm and weight 72 ± 6 kg. Playing experience not specified	5-Minute cycle ergometer warm-up and active stretching Isoforce dynamometer 3 reps at 60° s-1 2 reps at 180° s-1	62 ± 11 66 ± 15	56 ± 8 70 ± 20
Bogdanis and Kalapotharakos 6	18 Professional soccer players (Greek 1st division), age 24.2 ± 1.1 years, height 182 ± 1 cm and weight 76.9 ± 1.2 kg with playing experience of at least 12 years	10-Minute cycle ergometer warm-up Biodex dynamometer 3 reps at 60° s-1 3 reps at 180° s-1 3 reps at 300° s-1	55.54 ± 3.46 63.42 ± 4.68 70.61 ± 8.31	57.24 ± 5.23 61.62 ± 6.12 68.98 ± 8.07
Brito et al 7	18 Semiprofessional players, age 22.3 ± 4.2 years, height 176.7 ± 6.3 cm and weight 70.1 ± 5.6 kg with playing experience of 10.7 years	5-Minute cycle ergometer warm-up and submaximal contractions Biodex dynamometer 3 reps at 60° s-1 5 reps at 180° s-1	52 ± 14 62 ± 17	51 ± 10 57 ± 10
Cheung et al 8	23 Male college soccer athletes, age 22.2 ± 1.8 years, height 174 ± 5 cm and weight 65.0 ± 5.1 kg with playing experience of 7.1 years	Submaximal contractions for warm-up Cybex dynamometer 5 reps at 60° s-1 5 reps at 300° s-1	63 ± 7 75 ± 13	58 ± 8 82 ± 13
Daneshjoo et al 10	36 Professional soccer players, age 18.9 ± 1.4 years, height 181.3 ± 1.4 cm and weight 73.6 ± 6.3 kg with playing experience of at least 5 years	5-Minute cycle ergometer warm-up and dynamic stretching Biodex dynamometer 3 reps at 60° s-1 3 reps at 180° s-1 3 reps at 300° s-1	50 ± 11 51 ± 13 74 ± 22	50 ± 14 56 ± 14 75 ± 26
Delextrat et al 12	8 Male soccer players (British University), age 21.3 ± 2.3 years, height 178 ± 8 cm and weight 78 ± 9 kg with playing experience of 8 years	Submaximal contractions for warm-up Cybex dynamometer 3 reps at 60° s-1 5 reps at 180° s-1	64 ± 14 62 ± 9	65 ± 15 65 ± 13
Eniseler et al 14	14 Professional soccer players (Turkish 1st division), age 25.8 ± 3.9 years and weight 75.1 ± 3.3 kg. Playing experience not specified	15-Minute unspecified warm-up Biodex dynamometer 4 reps at 60° s-1 4 reps at 300° s-1 4 reps at 500° s-1	55.4 ± 7.8 64.1 ± 6.98 63.53 ± 18.23	55.54 ± 6.32 67.45 ± 10.84 72.97 ± 5.59
Ergün et al 15	44 Male soccer players, age 23.7 ± 5.4 years, height 178.1 ± 5.7 cm and weight 74.9 ± 9.4 kg with recreational playing experience of 5 years.	5-Minute treadmill run at 8 km h-1 Cybex dynamometer Unspecified reps at 60° s-1 Unspecified reps at 180° s-1 Unspecified reps at 300° s-1	67.2 ± 7 78.4 ± 9.8 74.2 ± 12.3	63 ± 8.4 73.5 ± 9.8 69.3 ± 9.8
Ibis et al 19	42 Amateur soccer players, age 22.7 ± 1.7 years, 179.2 ± 5.7 cm and 72.7 ± 7.3 kg with playing experience of 3+ years	5 Minutes on cycle or treadmill Biodex dynamometer 10 reps at 60° s-1 10 reps at 180° s-1 15 reps at 300° s-1	54.96 ± 6.29 62.66 ± 7.15 74.9 ± 11.13	53.88 ± 10.05 62.81 ± 9.15 78.49 ± 12.35
Magalhaes et al 21	46 Professional soccer players (Portuguese 1st division), age 25.2 ± 3.5 years, height 178.3 ± 6.0 cm and weight 75.3 ± 6.0 kg. Playing experience not specified	5 Minutes on cycle ergometer for warm-up Biodex dynamometer 3 reps at 90° s-1 5 reps at 360° s-1	57.4 ± 6.7 80.2 ± 13.3	56.1 ± 8.2 82.5 ± 13.3
Ruas et al 27	102 Male soccer players (Brazilian 1st division), age 26 ± 5 years, height 181 ± 7 cm and weight 79 ± 8 kg. Playing experience not specified	5 Minutes on cycle ergometer for warm-up and isometric contractions Cybex dynamometer 5 reps at 60° s-1	61 ± 12	59.6 ± 10
Sangnier and Tourny-Chollet 28	27 University soccer players (French), age 22.9 ± 2.6 years, height 177.4 ± 4.5 cm, and weight 73.0 ± 4.0 kg with playing experience of 14.8 years	10 Minutes on stepper and submaximal contractions for warm-up KinCom dynamometer 3 reps at 180° s-1	68 ± 11	69 ± 13
Silva et al 29	18 Professional soccer players (Portuguese 1st division), age 26.3 ± 4.5 years, height 178.4 ± 6.2 cm, and weight 76.7 ± 9 kg. Playing experience not specified	5 Minutes on cycle ergometer and submaximal contractions for warm-up Biodex dynamometer 3 reps at 90° s-1	55.64 ± 6.7	54.1 ± 5.4
Yildiz and Kale 30	57 Amateur soccer players, age 21.4 ± 1.0 years, height 175.2 ± 5.7 cm and weight 69.1 ± 6.4 kg. Playing experience not specified	10 Minutes on cycle ergometer and calisthenics and flexibility for warm-up Isomed dynamometer 5 reps at 60° s-1 5 reps at 180° s-1 5 reps at 300° s-1	67 ± 11 71 ± 16 74 ± 12	70 ± 13 71 ± 15 73 ± 13
Zakas 31	27 Professional soccer players (Greek 1st division), age 26.4 ± 3.1 years, height 178.4 ± 4.9, weight 77.4 ± 6.7 kg with playing experience of 12.52 years	10 Minutes on cycle ergometer and submaximal contractions for warm-up Cybex dynamometer 3 reps at 60° s-1 3 reps at 180° s-1 3 reps at 300° s-1	55.01 ± 7.99 55.77 ± 9.96 54.24 ± 8.64	54.3 ± 9.21 54.33 ± 9.36 52.7 ± 8.72

D, dominant; ND, nondominant; H:Q ratio, hamstring-to-quadriceps ratio. Data are shown as mean ± SD

Procedures

Quality Assessment

To assess the publication quality of the included studies, the Quality Assessment Tool for Observational Cohort and Cross-sectional Studies from the National Institutes of Health (NIH) was utilized by 2 investigators and, based on criteria, studies were rated as “good,” “fair,” or “poor.” ²⁴ The assessment tool contains 14 questions, but there are no recommendations or guidelines if the studies have to be rated as good, fair, or poor, which could affect the inter-rater reliability of this tool. ¹ Since this meta-analysis did not have interventions in all studies, some questions were not applicable and certain ratings did not apply.

Statistical Analysis

A meta-analysis model with random effects was performed for all studies to investigate the main effect and difference of the H:Q ratio in the D and ND leg. Standardized mean differences and their 95% CIs were calculated based on the H:Q ratio in both legs. All statistical analyses, including forest plots, were performed using the online version of the Cochrane Review Manager 5.4. ²⁶

Results

The characteristics of the included studies and the ranges of H:Q ratios in the D and ND limbs of the soccer players across the velocities can be found in Tables 1 and 2, respectively. Table 3 displays the absolute torque values and the asymmetry values (mean ± SD) of the D and ND limbs across the slow, intermediate, and fast velocities.

Table 2.

Range of H:Q ratio displayed as mean ± SD across velocities of included studies in D and ND legs of male soccer players

	Range of H:Q Ratio in D Leg, % ± SD		Range of H:Q Ratio in ND Leg, % ± SD
Angular Velocity	Lower	Upper	Lower	Upper
60° s -1	50 ± 11	67 ± 7	50 ± 14	70 ± 13
90° s -1	56 ± 7	57 ± 7	54 ± 6	56 ± 8
180° s -1	51 ± 13	78 ± 10	54 ± 10	74 ± 10
300° s -1	54 ± 9	75 ± 13	53 ± 9	82 ± 13

D, dominant; ND, nondominant; H:Q ratio, hamstring-to-quadriceps ratio.

Table 3.

Peak torque (N·m) and asymmetry (%) of the D and ND quadriceps and hamstrings at various velocities

Speed, ° s-1	No. of Studies Included	DPT Quadriceps, N·m Mean ± SD	NDPT Quadriceps, N·m Mean ± SD	Asymmetry of Quadriceps, % Mean ± SD	DPT Hamstrings, N·m Mean ± SD	NDPT Hamstrings, N·m Mean ± SD	Asymmetry of Hamstrings, % Mean ± SD
30	1	242.7	243.4	−0.3	125.8	128.8	−2.5
60	11	221.7 ± 33.0	218.0 ± 35.5	1.3 ± 3.6	128.8 ± 21.5	126.0 ± 19.7	2.0 ± 3.1
Average of low velocities	12	221.3 ± 32.0	221.4 ± 34.4	1.2 ± 3.5	128.5 ± 20.4	126.5 ± 18.7	1.6 ± 3.3
90	2	233.5 ± 7.7	232.7 ± 12.8	0.4 ± 2.2	129.6 ± 2.3	126.5 ± 3.5	2.4 ± 1.1
180	10	148.0 ± 12.4	146.3 ± 12.6	1.1 ± 1.9	94.3 ± 15.9	95.6 ± 16.2	−1.6 ± 6.4
Average of intermediate velocities	12	163.5 ± 36.4	162 ± 36.9	1.0 ± 1.8	100.7 ± 20.1	101.2 ± 19.2	−0.8 ± 3.5
240	1	122.3	126.5	−3.5	81.1	81.0	0.1
300	7	113.1 ± 18.7	112.6 ± 17.5	0.2 ± 2.1	74.6 ± 22.1	78.6 ± 13.5	0.8 ± 3.5
360	1	137.0	132.4	3.4	107.3	107.1	0.2
500	1	79.9	73.6	7.9	51.0	51.7	−1.5
Average of high velocities	10	113.1 ± 20.9	112.0 ± 21.3	1.0 ± 3.5	76.3 ± 22.6	79.1 ± 17.5	0.4 ± 2.9

D, dominant; ND, nondominant; N·m, newton-meter; PT, peak torque.

At the low velocity (30° s^-1-60° s^-1) analysis, there were 14 cohorts from 14 studies (Figure 2) and the average H:Q ratio was 59.8 ± 9.5% in the D leg and 58.6 ± 10% in the ND leg. The range of the H:Q ratio between the studies at 60° s^-1 was similar between the D (17%) and ND legs (20%) (Table 2). Eight of these cohorts had higher H:Q ratios in the D leg (percentage difference ranged from 1% to 5%),^{2,4,7,8,15,19,27,31} while the ND leg had higher H:Q ratios in 5 cohorts (percentage difference ranged from 0.1% to 3%),^3,6,12,14,30 and 1 study had similar ratios between legs. ¹⁰ However, only 4 studies resulted in significant differences between the D and ND legs of an average of 5% between the legs.^4,8,15,30 The overall standardized mean difference was a small effect of 0.13 (-0.03 to 0.30). Heterogeneity determined by I-squared was moderate (τ² = 0.03; I² = 34%). The torque values of the D and ND quadriceps and hamstrings were obtained from 12 of those studies and were similar between legs, with levels of asymmetry of less than 2% at slow velocities (Table 3).

Figure 2.

Standardized mean differences and Forest plot of the included cohorts for H:Q ratio of the D and ND legs at low velocities. D, dominant; ND, nondominant; H:Q ratio, hamstring-to-quadriceps ratio; Std, standardized.

The intermediate velocity (90° s^-1–180° s^-1) included 13 cohorts and had an average H:Q ratio of 64.2 ± 10.7% in the D leg and 63.6 ± 11.3% in the ND leg. The range of H:Q ratios between the D and ND legs at intermediate speeds was 27% and 20%, respectively (Table 2). Seven cohorts had greater H:Q ratios in the D leg (percentage differences ranged from 1.5% to 5.3%).^{2,6,7,15,21,29,31} The ND leg was greater in 5 cohorts (percentage difference ranged from 0.1% to 5%),^{4,10,12,19,28} and one study had similar ratios. ³⁰ Only 2 of the 13 cohorts showed significantly higher H:Q ratios in the D leg compared with the ND leg (68.8% vs 63.5% and 78.4% vs 73.5%),^2,15 with an average of 5.1%. The analysis resulted in a small effect of 0.10 (-0.05 to 0.26) (Figure 3) and the heterogeneity was low (τ² = 0.02; I² = 18%). The torque values of the D and ND quadriceps and hamstrings were obtained from 12 of those studies and were similar between legs with levels of asymmetry of 1% or less at intermediate velocities (Table 3). The torque values of the quadriceps and hamstrings at the intermediate velocities were approximately 27% and 21% lower, respectively, than the torque produced at the low velocities.

Figure 3.

Standardized mean differences and Forest plot of the included cohorts for H:Q ratio of the D and ND legs at intermediate velocities. D, dominant; ND, nondominant; H:Q ratio, hamstring-to-quadriceps ratio; Std, standardized.

The high velocity (240° s^-1 and above) analysis included 11 cohorts from 10 studies. They had an average H:Q ratio of 71.9 ± 12.7% in the D leg and 72.8 ± 12.7% in the ND leg. The range of H:Q ratios between the D and ND legs at 300° s^-1 was 21% and 29%, respectively (Table 2). In the high velocity group (240° s^-1 and higher), 4 studies resulted in higher ratios in the D leg (percentage differences ranged from 1% to 4.9%),^6,15,30,31 5 studies had higher ratios in the ND leg (percentage differences ranged from 1% to 9.5%),^{8,10,14,19,21} and one study did not find differences. ³ Of the 5 studies, 2 had significant differences between the legs (75% vs 82% and 63.5% vs 73%),^9,15 with an average of 8.25%. There was a small negative effect of -0.06 (-0.25 to -0.12) (Figure 4) and the heterogeneity was moderate (τ² = 0.03; I² = 30%). The torque values of the D and ND quadriceps and hamstrings were obtained from all 10 of those studies and were similar between legs with levels of asymmetry of less than 1% at high velocities (Table 3). The torque values of the quadriceps and hamstrings at the high velocities were approximately 49% and 40% lower, respectively, than the torque produced at the low velocities.

Figure 4.

Standardized mean differences and Forest plot of the included cohorts for H:Q ratio of the D and ND legs at high velocities. D, dominant; ND, nondominant; H:Q ratio, hamstring-to-quadriceps ratio; Std, standardized.

None of the 3 categorized velocities (low, intermediate, and high) were statistically significant in H:Q ratios as they only showed small effects and negligible differences between the D and ND legs (Figures 2-4).

The quality of the included studies was evaluated as fair, except one was good and one was of poor quality based on the quality assessment tool from NIH. ²⁷ In most of the studies, question 1 (“Was the research question or objective in this paper clearly stated”), question 2 (“Was the study population clearly specified and defined”), question 3 (“Was the participation rate of eligible persons at least 50%”), question 4 (“Were all subjects selected or recruited from the same or similar populations [. . .]”), and question 11 (“Were the outcome measures clearly defined, valid, reliable and implemented consistently across all study participants”) were answered with yes. Questions 6 and 14, which are about power description and statistical power, were answered as no in most of the studies and questions 6, 7, 8, 9, 10, 12, 13 were answered as “not applicable” or no in most of the studies. This occurred because most of the studies evaluated were observational and did not have any intervention. Further, the questions of exposure, time frame, blinding, or loss to follow-up after baseline were not applicable, which overall indicates a fair quality even though 5 questions were answered with yes and indicated good quality.

Discussion

This meta-analysis synthesized literature on the conventional H:Q ratio of the D and ND legs in male soccer players and examined whether differences between the legs exist. The test of the overall effect of differences in the conventional H:Q ratio across low, intermediate, and high contraction velocities resulted in nonsignificant small effects. Further, torque values of the quadriceps and hamstrings were rather similar in the D and ND legs as bilateral asymmetry was calculated to be less than 3%, which corresponds to being under the 5% threshold of injury risk described by Croisier et al. ⁹

The lowest H:Q ratios were observed at low velocities; however, these values were not statistically significant between legs. The average H:Q ratio at low velocities was approximately 59%, which is very close to the accepted threshold of 60% at 60° s^-1,^11,28,30 and categorized by Crosier et al ⁹ as having no strength imbalances. The H:Q ratios increased incrementally to 64% and 72% at intermediate and high velocities, respectively, and still did not differ between D and ND limbs. At low velocities of knee extension, the quadriceps are dominant in producing torque but as contraction velocity rises the involvement of the hamstrings in both concentric and eccentric movements become more prominent leading to higher H:Q ratios. Greater torque production of the hamstrings helps maintain knee joint stability at high speeds. ² Since the results of this meta-analysis found no conventional H:Q asymmetries between the D and ND legs at all velocities, our hypothesis is refuted. Soccer players may develop a muscular dominance on one side due to the preferred kicking limb and the landing after a jump. ³¹ The sport requirements of soccer are not limited to kicking a ball or landing on a single leg, as essential bilateral actions like walking, running, jumping, and tackling often occur. ²⁸ Therefore, kicking a soccer ball is only a small portion of competition, which might result in finding a lack of muscular imbalances.

The quadriceps produce a great deal of force when contracting in a concentric manner, and the response of the hamstring muscles on the D and ND limbs can vary during stabilizing, decelerating, change-of-direction movements, tackling, and kicking. ³⁰ For example, while the quadriceps of the kicking limb rapidly extend the knee concentrically, the hamstrings in the supporting limb can act eccentrically or isometrically to stabilize the knee joint and support the whole-body weight with the help of the hip extensors. ²⁵ Over time, this can lead to stiffness and strengthening of the hamstring muscles and tendons. ³⁰ It is possible that the rapid movements of the quadriceps while kicking a ball do not produce a sufficient stimulus on the quadriceps to result in enhanced muscular strength on the D limb, yet the hamstrings on the ND limb may increase strength due to the absorption of energy from the kicking limb. ¹⁸ Perhaps more information could be gained through the analysis of the functional H:Q ratio, where the eccentric torque of the hamstrings is compared with the concentric torque of the quadriceps. Crosier et al ⁹ explained that the conventional H:Q ratio may be limited in the identification of eccentric strength deficiencies in the hamstring muscle group and could fail to identify athletes at risk of injury. In such cases, the functional H:Q ratio may be a more useful tool in injury risk assessments. However, compared with the conventional H:Q ratio, there is a paucity of research evaluating the functional H:Q ratio. This could be attributed to the complexities of evaluating eccentric torque of the hamstrings on isokinetic dynamometers. These eccentric contractions on a dynamometer are not intuitive movements, pose greater risk of injury than concentric contractions when performed at high velocities, ⁹ and may require more extensive familiarization for athletes and subjects. Nevertheless, the inability to perform the meta-analysis to include the functional H:Q ratio can be considered a limitation in the overall conclusions and recommendations we can provide.

The studies in this meta-analysis displayed some variability in their conclusions as conventional H:Q ratio in some studies demonstrated significant differences,^{2,4,8,14,15,30} and others found no differences between the D and ND legs.^{3,6,7,10,12,19,21,27-39,31} The significant differences in the conventional H:Q ratio ranged from 3% to 5% at low velocities, favoring higher values in the D leg, whereas differences at intermediate velocities ranged from 5% to 6%, with sometimes the D limb having a higher H:Q ratio and vice versa. Interestingly, the significant differences in the H:Q ratio at high velocities all displayed higher values in the ND leg and ranged from 7% to 10%. This tendency to observe higher conventional H:Q ratios at higher velocities in the ND leg is intriguing and deserves more attention. All studies indicating no significant differences in H:Q ratios between D and ND legs ranged from 1% to 5% among all velocities. One should also consider that the velocities at which torque is assessed on isokinetic dynamometers are substantially lower than the velocities that players experience during training and game situations, as kicking velocities have been found to range from 730° s^-1 to 1720° s^-1. ⁷ It will be important for clinicians and researchers to firmly establish clinical meaningful differences in the H:Q ratio if it is to be used as a tool to identify injury risk in athletes.

When examining the H:Q ratio in male soccer players. playing experience may be influential. According to Fousekis et al, ¹⁶ professional soccer players displayed lower incidences of isokinetic strength asymmetries than players with less experience or at lower levels of play. We found that most of the studies that investigated soccer athletes with experience of over 10 years showed nonsignificant differences in the H:Q ratio between the legs,^6,7,10,28,31 indicating a balance in muscle strength. ¹⁶ However, Cheung et al ⁸ showed a 5% difference in the conventional H:Q ratio between the limbs of college soccer players who had only 7 years of playing experience and Arsenis et al ⁴ reported 4% to 6% differences in young players. Further, Ergün et al ¹⁵ showed a 5% difference in H:Q ratio between the limbs at all velocities in male recreational soccer players. On the other hand, Daneshjoo et al ¹⁰ is just one example of professional players that showed no differences in the H:Q ratios between the D and ND legs at all velocities. When analyzed together, we were able to conclude that overall differences in the conventional H:Q ratio between the D and ND legs of male soccer players do not exist.

These opposing results from individual studies provide evidence that perhaps differences between players from different teams, level of play, and possibly even across countries, could be considered when assessing strength imbalances and the H:Q ratio. Therefore, our inability to evaluate the data in consideration of training status can be considered a limitation to the present meta-analysis. Another limiting factor in the present study is the inclusion of only healthy soccer players with no previous injuries within in the last few months. As reinjuries are common, players with previous injuries should be considered in studies too. ¹¹ It is likely that players with previous hamstring or knee injuries are at higher risk of sustaining injuries even if their H:Q ratios appear to meet the accepted values of at 60%. Therefore, it is important for further studies to include players with previous injuries to make the studies more encompassing for the sport of soccer.

In conclusion, the assessment of the conventional H:Q ratio in male soccer players varies among the velocities at which torque is assessed, but there appear to be no differences between the D and ND legs. Our results indicate that the overall H:Q ratios of the participants in the selected studies were often lower than the recommended values of 60% to 66%,^{5,11,17,28,30} and those values differ considerably by velocity. There also appear to be no differences in the H:Q ratio between the D and ND legs of male soccer players, but individual analyses of studies suggest that playing experience could be an intriguing factor to consider. This study highlights the need to establish norms for strength asymmetries and conventional (and functional, although not measured in this study) H:Q ratios. The thorough analysis and comprehensive data provided could be used by researchers, clinicians, and coaches to not only establish norms but to determine clinically meaningful differences in strength asymmetries and the conventional H:Q ratio at various velocities in male soccer players. This, in turn, could be further used to evaluate risk of injury in soccer.