We included 140 patients with dysfunction somewhere in the lower limbs. Table 1 describes the personal and clinical characteristics of the study patients. Most of the sample consisted of women (77.9%) aged over 50 years who were overweight, who had been experiencing pain for more than 90 months. The body site most affected with pain was the knee (71.4%) and the most common diagnoses were knee osteoarthritis (55%) and anterior knee pain (11.4%).
Table 1
Sociodemographic and clinical characteristics of patients with lower limb dysfunction (n = 140)
Variables | Mean (standard deviation) or n (%) |
Age (years) | 51.89 (19.25) |
Sex (female) | 109 (77.9%) |
Weight (kg) | 70.71 (9.86) |
Height (m) | 1.65 (0.07) |
Body mass index (kg/m2) | 25.79 (3.33) |
Chronicity (months) | 90.37 (53.59) |
Physical activity (no) | 93 (66.4%) |
Numeric Pain Rating Scale (score, 0–10) | 5.79 (1.74) |
Lower Extremity Functional Scale | |
20 items (score, 0–80) | 58.45 (11.72) |
10 items (score, 0–40) | 28.57 (6.57) |
Pain site | |
Knee | 100 (71.4%) |
Foot | 11 (7.9%) |
Leg | 9 (6.4%) |
Ankle | 9 (6.4%) |
Thigh | 8 (5.7%) |
Hip | 3 (2.1%) |
Diagnosis | |
Knee osteoarthritis | 77 (55%) |
Anterior knee pain | 16 (11.4%) |
Plantar fasciitis | 9 (6.4%) |
Muscle strain | 8 (5.7%) |
Ankle sprain | 6 (4.3%) |
Medial tibial stress syndrome | 6 (4.3%) |
Tendinitis | 5 (3.6%) |
Anterior cruciate ligament injury | 4 (2.9%) |
Meniscus tear | 3 (2.1%) |
Muscle contusion | 3 (2.1%) |
Fracture | 2 (1.4%) |
Femoroacetabular impingement | 1 (0.8%) |
The long version of the LEFS with one domain and 20 items was tested using CFA. The fit index values were inadequate (chi-square/DF > 3, TLI and CFI < 0.90, RMSEA and SRMR > 0.08), as shown in the second line of Table 2. Therefore, the LEFS reduction was performed based on the modification indices, as shown in Table 3, generating a one-dimensional structure of the LEFS with 10 items (LEFS-10).
Table 2
Confirmatory factor analysis of the five proposed Lower Extremity Functional Scale models tested (n = 140)
Models | Chi-square/DF | CFI | TLI | RMSEA (90% CI) | SRMR | AIC | BIC |
Model 1 | 1.88 | 0.975 | 0.968 | 0.079 (0.049, 0.109) | 0.058 | 3287.063 | 3345.896 |
Model 2 | 3.77 | 0.817 | 0.796 | 0.141 (0.130, 0.153) | 0.131 | 6595.364 | 6713.030 |
Model 3 | 3.77 | 0.855 | 0.830 | 0.141 (0.125, 0.157) | 0.117 | 4906.446 | 4994.695 |
Model 4 | 3.48 | 0.863 | 0.840 | 0.134 (0.118, 0.150) | 0.114 | 4914.434 | 5002.683 |
Model 5 | 3.26 | 0.870 | 0.848 | 0.128 (0.112, 0.144) | 0.111 | 4927.428 | 5015.677 |
DF: degrees of freedom; CFI: comparative fit index; TLI: Tucker-Lewis Index; RMSEA: root mean square error of approximation; CI: confidence interval; SRMR: standardized root mean square residual; AIC: Akaike information criterion; BIC: Bayesian information criterion. Model 1: Structure with one domain and 10 items proposed in this study; Model 2: Structure with one domain and 20 items published by Pereira et al. 7; Model 3: Structure with one domain and 15 items published by Bravini et al. 12; Model 4: Structure with one domain and 15 items published by Alnahdi 13; Model 5: Structure with one domain and 15 items published by Repo et al. 14. |
Table 3
Reduction of the Lower Extremity Functional Scale items based on modification indices and factor loadings from confirmatory factor analysis
Redundant items | Item description | MI | Factor loading | Item deleted |
Decision 1 | | | | |
Item 18 | Making sharp turns while running fast. | 68.962 | 0.772 | Item 19 |
Item 19 | Hopping. | 0.709 |
Decision 2 | | | | |
Item 3 | Getting into or out of the bath. | 20.693 | 0.634 | Item 3 |
Item 14 | Standing for 1 hour. | 0.650 |
Decision 3 | | | | |
Item 16 | Running on even ground. | 19.169 | 0.717 | Item 17 |
Item 17 | Running on uneven ground. | 0.649 |
Decision 4 | | | | |
Item 7 | Lifting an object, like a bag of groceries from the floor. | 17.355 | 0.652 | Item 8 |
Item 8 | Performing light activities around your home. | 0.565 |
Decision 5 | | | | |
Item 10 | Getting into or out of a car. | 16.987 | 0.690 | Item 20 |
Item 20 | Rolling over in bed. | 0.512 |
Decision 6 | | | | |
Item 10 | Getting into or out of a car. | 14.906 | 0.690 | Item 10 |
Item 18 | Making sharp turns while running fast. | 0.772 |
Decision 7 | | | | |
Item 11 | Walking 2 blocks. | 12.677 | 0.783 | Item 12 |
Item 12 | Walking a mile. | 0.686 |
Decision 8 | | | | |
Item 5 | Putting on your shoes or socks. | 11.667 | 0.694 | Item 5 |
Item 11 | Walking 2 blocks. | 0.783 |
Decision 9 | | | | |
Item 4 | Walking between rooms. | 11.552 | 0.642 | Item 4 |
Item 14 | Standing for 1 hour. | 0.650 |
Decision 10 | | | | |
Item 2 | Your usual hobbies, recreational or sporting activities. | 10.825 | 0.590 | Item 2 |
Item 18 | Making sharp turns while running fast. | 0.772 |
MI: Modification indices. |
The internal structure of LEFS-10, here called Model 1, with one domain and 10 items (1, 6, 7, 9, 11, 13, 14, 15, 16 and 18), was compared with other structures proposed in the literature (Table 2), as follows: Model 2, structure with one domain and 20 items, published by Pereira et al. 7; Model 3, structure with one domain and 15 items (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 13, 14, 19 and 20), published by Bravini et al. 12; Model 4, structure with one domain and 15 items (1, 3, 4, 5, 6, 7, 8, 9, 10, 12, 13, 14, 15, 17 and 20), published by Alnahdi 13; and Model 5, structure with one domain and 15 items (1, 2, 3, 4, 5, 6, 7, 8, 9, 12, 13, 14, 15, 19 and 20), published by Repo et al. 14.
The LEFS-10 was the only structure that showed adequate fit indices (chi-square/DF = 1.88, CFI = 0.975, TLI = 0.968, RMSEA = 0.079, and SRMR = 0.058), in addition to presenting the lowest values of the AIC and BIC (Table 2). The Brazilian Portuguese and English versions of the LEFS-10 are available in Supplements 1 and 2, respectively. Figure 1 presents the adequate factor loading (> 0.40) of the 10 items of the LEFS, explained by the functionality domain.
Criterion validity was assessed using the long version of the LEFS with 20 items as the gold standard instrument. Thus, we correlated the long version with the 10-item short version proposed in this study and observed a correlation magnitude (rho) of 0.911 (p value < 0.001) according to Spearman’s correlation coefficient.