Table 1 Baseline characteristics of study participants stratified by injury status during follow-up.

From: Development and validation of a machine learning model for non-contact injury prediction based on lower limb strength asymmetry in professional football

Characteristic

Total (n = 312)

Non-injured (n = 223)

Injured (n = 89)

p-value

Demographics

    

Age (years)

24.7 ± 4.2

24.5 ± 4.1

25.2 ± 4.4

0.184

Height (cm)

181.4 ± 6.7

181.2 ± 6.6

181.9 ± 6.9

0.412

Body mass (kg)

78.2 ± 7.9

77.8 ± 7.7

79.1 ± 8.3

0.196

BMI (kg/m²)

23.7 ± 1.8

23.6 ± 1.7

23.9 ± 1.9

0.169

Playing experience (years)

6.3 ± 3.8

6.1 ± 3.7

6.8 ± 4.0

0.142

Playing Position, n (%)

   

0.328

Defender

98 (31.4)

73 (32.7)

25 (28.1)

 

Midfielder

120 (38.5)

82 (36.8)

38 (42.7)

 

Forward

75 (24.0)

55 (24.7)

20 (22.5)

 

Goalkeeper

19 (6.1)

13 (5.8)

6 (6.7)

 

Strength Asymmetry (%)

    

Knee extensors

9.6 ± 4.8

8.1 ± 3.9

12.4 ± 5.2

< 0.001

Knee flexors

10.8 ± 5.3

9.3 ± 4.5

14.7 ± 5.8

< 0.001

Hip adductors

8.4 ± 3.7

7.9 ± 3.4

9.6 ± 4.1

0.008

Hip abductors

7.2 ± 3.2

6.8 ± 3.0

8.1 ± 3.5

0.014

H: Q Ratio (%)

    

At 60°/s

61.4 ± 7.8

63.1 ± 7.2

58.2 ± 7.9

0.002

At 180°/s

68.7 ± 8.4

69.8 ± 8.1

66.3 ± 8.7

0.011

Training Load

    

Weekly hours

17.4 ± 3.1

16.9 ± 2.8

18.6 ± 3.2

0.003

ACWR (4 weeks)

1.24 ± 0.28

1.18 ± 0.24

1.42 ± 0.31

< 0.001

Load variability (CV%)

23.6 ± 6.8

21.2 ± 5.4

28.4 ± 7.2

0.006

Injury History, n (%)

    

Previous LE injury

145 (46.5)

85 (38.1)

60 (67.4)

< 0.001

Multiple injuries (≥ 2)

68 (21.8)

32 (14.3)

36 (40.4)

< 0.001

  1. Data presented as mean ± SD or n (%). BMI: body mass index; H:Q: hamstring-to-quadriceps; ACWR: acute: chronic workload ratio; CV: coefficient of variation; LE: lower extremity. Statistical comparisons performed using independent t-tests for continuous variables and chi-square tests for categorical variables.