Forty-eight subjects (34 men, 14 women) participated in this study (age 20.6 ± 1.2 years, height 177.3 ± 9.1 cm, weight 77.9 ± 12.5 kg). At the time of this study, all subjects were enrolled at the US Air Force Academy, CO. While subjects were not intercollegiate athletes, they participated in mandatory intramural athletics and military training 4–7 days/week prior to their knee injury.
Selection criteria for this study included history of unilateral ACLR performed greater than 12 months from time of participation in this study, completion of a rehabilitation program that focused on full return to military and athletic activities, and clearance from an orthopedic surgeon to return to preinjury activity levels following ACLR. Prior to participation, all subjects read and signed an informed consent document approved by the Institutional Review Board at the US Air Force Academy, CO. This study was in accordance with the ethical standards of the 1964 Declaration of Helsinki as revised in 2000.
All subjects had initially undergone ACLR using a bone–patellar tendon–bone autograft. Mean time from injury to initial ACLR for subjects in this study was 31.3 ± 28.5 days (range 10–154 days). Six subjects (13%) underwent revision ACLR, in which their ipsilateral distal semitendinosus and gracilis tendons were doubled and used for reconstruction. None of the six subjects who underwent revision ACLR had undergone any other knee surgeries. However, six other subjects had additional knee surgery at a mean of 17.2 ± 11.8 months (range 3–37 months) following ACLR, including partial meniscectomy (n = 5) and diagnostic arthroscopy (n = 1). Subjects were tested at a mean of 28.3 ± 15.1 months (range 12–69 months) following their most recent knee surgery. At time of testing, all subjects had resumed all required military and athletic activities.
This was a descriptive study that assessed the relationship between functional levels in activities of daily living and sports and fear-avoidance beliefs in patients with history of ACLR, after controlling for injury-related variables and physical impairments. The dependent variable was functional levels in activities of daily living and sports as measured by the combined scores on the Knee Outcome Survey (KOS) Activities of Daily Living Scale (ADLS) and Sports Activity Scale (SAS) [11, 12]. The independent variables were fear-avoidance beliefs, injury-related variables, and physical impairments. Fear-avoidance beliefs were assessed with a version of the physical activity subscale of the fear-avoidance beliefs questionnaire (FABQ) that was modified for the knee; the FABQ was originally used to assess fear-avoidance beliefs in patients with low-back pain. Injury-related variables in this study included whether or not the patient underwent additional surgery following initial ACLR and time from ACLR to testing. Physical impairment measures included single-leg hop for distance test performance, isokinetic quadriceps function, and anterior tibiofemoral joint laxity.
The degree of functional limitations experienced by the subjects in this study was assessed through the combined scores on the KOS ADLS and SAS [11, 12]. Items on the ADLS  relate to symptoms and functional limitations that individuals may experience during activities of daily living. Items on the SAS  assess symptoms and functional limitations that individuals may experience while playing sports. Responses regarding symptoms were graduated in terms of the amount of participation restrictions that individuals experienced during activities of daily living or sports due to the symptom. Responses regarding functional limitations were graduated from no limitation to inability to perform the specific activity.
The KOS ADLS and SAS were numerically graded on a scale of 0–100, with higher scores indicating higher functional levels [11, 12]. The combined scores of the KOS ADLS and SAS (possible range 0–200) were used for data analysis. The KOS has been shown to be a reliable and valid measure of disability in patients with knee impairments [11, 12].
Fear-avoidance beliefs questionnaire
Fear-avoidance beliefs were measured using the physical activity subscale of the FABQ. The physical activity subscale of the FABQ quantifies the level of fear about physical activity and has primarily been studied in patients with low-back pain. The FABQ also has a work subscale, which quantifies the level of fear about work-related activities. The work subscale of the FABQ was not used as a method of assessment in this study, as the subjects were full-time students who were not employed. The instrument consists of five items, which are scored from 0 to 6. Possible scores range from 0 to 30, with higher scores representing increased fear-avoidance beliefs. Previous studies have reported good reliability for the FABQ in patients with low-back pain . To apply the FABQ in patients following ACLR, we adapted the form to use in patients with knee pathology as described by Piva et al.  and van Baer et al. . We changed the descriptors of physical activities such as bending, lifting, walking, or driving to physical activities such as walking, running, kneeling, or driving, and changed the word “back” to “knee” throughout the form. Prior to the descriptive study, we determined the test–retest reliability for the FABQ to be high (intraclass correlation coefficient = 0.92).
Knee surgeries after initial anterior cruciate ligament reconstruction
For each subject, it was documented whether or not additional knee surgery was performed after the initial ACLR.
Time from anterior cruciate ligament reconstruction to testing
The number of months from the most recent ACLR to participation in this study was calculated for each subject.
Physical impairment measures
Single-leg hop test for distance
A tape measure was secured to the floor, and subjects stood on the leg being tested with the anterior aspect of their athletic shoe at the zero mark of the tape measure. While standing on one leg, subjects were instructed to hop as far as possible forward and land on the same leg. The distance hopped was then measured from the zero mark of the tape measure to the point where the subject’s heel hit the ground. No other biomechanical data were recorded. Following one practice trial, subjects performed three test trials for each lower extremity in alternating fashion, beginning with the noninvolved lower extremity. Based upon the average of the three test trials, single-leg hop test indices were calculated and used for data analysis. These indices were calculated by expressing the scores of the involved lower extremity as a percentage of the noninvolved lower extremity (i.e., involved lower extremity score/noninvolved lower extremity score × 100).
Isokinetic quadriceps test
Quadriceps peak torque was assessed with the kinetic communication isokinetic dynamometer (Chattanooga Group Inc., Hixson, TN). With subjects in a seated position, concentric isokinetic knee extension was tested at 60°/s between 90° and 0° of flexion. Subjects were requested to complete five maximal test repetitions for each lower extremity. Quadriceps peak torque indices were calculated and used for data analysis. These indices were calculated by expressing the involved lower extremity scores as a percentage of the noninvolved lower extremity (i.e., involved lower extremity score/noninvolved lower extremity score × 100).
Anterior tibiofemoral joint laxity test
Anterior tibiofemoral joint laxity was assessed with the KT-1000 knee joint arthrometer (MEDmetric Corp., San Diego, CA) while subjects were supine with the knee positioned in 20° to 30° of flexion. With the KT-1000 strapped to the leg, one hand of the examiner stabilized the thigh and patellar pad of the KT-1000 and the other hand of the examiner applied a maximal manual anteriorly directed force on the proximal calf. The amount of movement was quantified in millimeters. Three test trials were performed on each knee of the subject, with the maximum value of the three test trials used to calculate side-to-side differences (the score of the involved knee–the score of the noninvolved knee), which were used for data analysis.
Prior to the descriptive study, test–retest reliability for the measurement procedures used in this study were calculated and determined to be acceptable in subjects with a history of ACLR . Following a medical chart review, which determined the date and type of previous knee surgeries, subjects were assessed during a single testing session in the following manner: completion of the FABQ and the KOS ADLS and SAS followed by KT-1000 testing.
Each subject then completed a warm-up, which consisted of 5 min of self-paced stationary cycling, followed by quadriceps, hamstring, and calf muscle stretching of both lower extremities. Following the warm-up, isokinetic testing of the quadriceps took place followed by single-leg hop testing. For KT-1000, isokinetic, and single-leg hop testing, the noninvolved lower extremity was tested prior to the involved lower extremity. A single investigator collected all data.
A hierarchical linear regression analysis assessed the relationship between patient-reported functional levels in activities of daily living and sports and fear-avoidance beliefs after controlling for injury-related variables and physical impairment measures. Scores on the Knee Outcome Survey served as the dependent variable. Scores on the physical activity subscale of the FABQ, injury-related variables, and physical impairment measures served as the independent variables. Data pertaining to injury-related variables were entered in the first step of the regression analysis, physical impairment measures were entered in the second step of the regression analysis, and scores on the physical activity subscale of the FABQ were entered in the third step of the regression analysis. Developing a regression model in this way provides a means to analyze the unique contribution of the physical activity subscale of the FABQ, after controlling for the contribution of injury-related variables and physical impairment measures. This method of analysis is consistent with previous studies evaluating the influence of fear avoidance on function and disability [4, 17]. Statistical significance was set at P < 0.05.