A predictive model for hip abductor strength and knee extensor strength 12 months after total hip arthroplasty with an interaction term

Subjects were female patients with hip osteoarthritis who underwent THA at two hospitals (Ebina General Hospital and Zama General Hospital) between December 2015 and September 2017. Inclusion criteria were patients who (1) underwent primary unilateral cementless THA via a minimally invasive anterolateral approach; (2) were female; and (3) agreed with the purpose of the study. Subjects were limited to female patients in order to minimize the effects of gender differences in lifestyle, as roughly 90% of our patients are female. Exclusion criteria were (1) a history of lower limb or spine surgery; (2) previously diagnosed painful orthopedic disease other than hip joint disease; (3) previously diagnosed end-stage hip osteoarthritis of the non-surgical side; (4) previously diagnosed mental disease or neuromuscular disease; and (5) postoperative complications such as fracture, dislocation, infection, or nerve paralysis. For comparison, 50 healthy community-dwelling women (age range, 45-78 years) who did not have lower limb joint pain were selected as controls.

Four highly-experienced surgeons performed all THAs without a navigation system. All THAs were performed in the half lateral position (i.e., pelvis tilted 60 degrees relative to the floor) through a minimally invasive anterolateral approach independent of the implant type [19, 20]. After an incision was made to the fascia lata at 0.5 mm posteriorly to the posterior border of the tensor fasciae latae muscle, the joint was approached between the gluteus medius muscle and the tensor fasciae latae muscle without damaging the muscles. The anterior iliofemoral ligaments, the anterior part of the capsule, and the conjoined tendon were preserved. We created a hip center similar to the anatomical hip center by acetabular reaming up to the lamina interna [21], as not doing so would result in the hip center being located superior-laterally to the anatomical hip center in the dysplastic acetabulum. By creating a hip center in its accurate anatomical position, the muscles can function normally [21, 22]. All implants were cementless. For stems, the SL-Plus stem (Smith & Nephew), Short Modular Femoral Hip System (Smith & Nephew), and Global Tissue Sparing stem (Zimmer Biomet) were used in 74.4%, 12.2%, and 7.8% of cases, respectively. For acetabular components, the R3 Acetabular System (Smith & Nephew), Continuum Acetabular System (Zimmer Biomet), and G7 Acetabular System (Zimmer Biomet) were used in 57.8%, 30.0%, and 7.8% of cases, respectively. Ceramic-on-ceramic bearings (Biolox Delta, Ceramtec AG) were used in 90% of patients, and ceramic-on-highly cross-linked polyethylene bearings (Biolox Delta, Ceramtec AG or Oxiniuum, Smith & Nephew) were used for the remaining 10%. These components were selected based on patient age and X-ray images of the hip joint.

This study complied with the ethical standards of the Declaration of Helsinki (1964) and its subsequent amendments and was approved by the Ethics Review Committees of Ebina General Hospital and Zama General Hospital (No. 207). Each patient provided written informed consent to participate in this study.

This study was supported by research funds from our institution.

This study was a 12-month prospective observational study. Postoperative rehabilitation was performed according to the clinical pathway of our institution. On the day after surgery, all patients were allowed to bear full weight and underwent inpatient rehabilitation. Rehabilitation consisted of gait exercises, passive range of motion exercises, and muscle strengthening exercises. For gait exercises, patients used parallel bars in the beginning, and crutches or a walker from one week after surgery. By the time of discharge, all patients used a T cane during walking exercises. The length of hospital stay according to the clinical pathway was three weeks. In addition to inpatient rehabilitation, most patients underwent outpatient rehabilitation roughly once a week after discharge and performed activities such as getting up from the floor or ascending and descending stairs according to ability.

We collected subject background information and surgical information, including age, height, body weight, body mass index, disease duration (years), length of hospital stay (days), operative time (minutes), intraoperative blood loss (ml), leg extension (mm), and femoral offset from medical charts. Disease duration was defined as the period from pain onset to surgery, as indicated by patients on a questionnaire [15, 22]. Leg extension and femoral offset were measured by a single researcher using a previously described method [15, 23]. We measured the Japanese Orthopaedic Association hip score, strength of lower extremity muscles (hip abductor and knee extensor) on the operated side, and Timed Up and Go (TUG) test values. All measurements were performed by two physical therapists who had thoroughly practiced measurement procedures. The number of steps was used as an index of postoperative physical activity.

Lower extremity maximal isometric strength was measured using a hand-held dynamometer (μTas F-1; Anima Corp., Tokyo, Japan). We referred to the measuring method proposed by Fukumoto et al 6. For the assessment of hip abductor strength, patients were positioned on a platform in a supine position at a 0° angle, with a sensor pad attached to the distal lateral side of the thigh. For the assessment of knee extensor strength, patients were positioned on a platform in a sitting position at a 90° angle, with a sensor pad attached to the distal front side of the lower leg. After practice, isometric muscle strength during 5 seconds of isometric contraction was measured twice, and the higher value on the hand-held dynamometer was used for analysis. The length of the lever arm (m) was measured from the each of joint to the center of the sensor pad (hip abductor strength used the distance from the greater trochanter to the center of the sensor) [24]. Muscle strength was calculated as torque (Nm).

In the TUG test, the time it took to get up from the chair, go around the mark 3 m ahead, and return to the chair again was measured. The measurement was performed twice at maximum walking speed, and the minimum value was used for analysis. For the measurement of physical activity, a digital pedometer with 3-axis acceleration sensors (TH-400; YAMASA, Tokyo, Japan) was used to count the number of steps (steps/day). The validity and reliability of another device by the same manufacturer (EX-510; YAMAX, Tokyo, Japan), which uses the same algorithm, has been verified previously [25]. Patients were instructed to wear the device on their body for 24 hours a day except when bathing or sleeping (sleep time was set at 8 hours), for one to two months after surgery. Non-wearing time (in hours), if any, was recorded by patient report. If non-wearing time was 4 hours or more, the day was treated as a missing value [26]. The average value for one month was calculated, excluding missing values.

All values are presented as mean (standard deviation) or median (1st quartile to 3rd quartile). The paired t-test or Wilcoxon signed-rank sum test was used to compare Japanese Orthopaedic Association score, TUG value, hip abductor strength, and knee extensor strength 12 months after THA with those measured before and 6 months after THA. The t-test or Mann-Whitney U test was used to compare background factors and TUG values between THA patients and healthy subjects. A general linear model adjusted for body weight was used to compare muscle strength 12 months after THA with that of healthy subjects.

A generalized linear model approach was used to identify factors that determine hip abductor and knee extensor strength 12 months after THA. The generalized linear model is an extension of the linear model, and the probability distribution of any exponential family can be selected for the error structure. In this study, we selected normal distribution and gamma distribution. First, we created a model in which the objective variables were hip abductor and knee extensor strength 12 months after THA, with preoperative muscle strength as the only explanatory variable (Step 1: univariate analysis). Second, we created a model using the forced entry method, with preoperative muscle strength, age, body weight, height, disease duration, physical activity, and leg extension as explanatory variables (Step 2: multivariate analysis). Correlation matrices were calculated for these explanatory variables. Third, a model was created by centering the explanatory variables and adding interaction terms between preoperative muscle strength and age, disease duration, physical activity, and leg extension (Step 3: multivariate analysis with interaction terms). If the interaction term was significant, and residual deviations in Steps 2 and 3 significantly decreased, a simple slope analysis was performed using the method described by Cohen & Cohen [27]. The mean value ± 1SD was adopted as the condition for simple slope analysis. The model with the lowest Akaike’s Information Criterion (AIC), a selection criterion, was used as the final model. The sample size was calculated using the calculation method for the linear model (α = 0.05, detection power = 0.95, effect size = 0.3, 8 input explanatory factors), yielding a required sample size of 84. Since a previous longitudinal study, which assessed patients for up to 12 months after THA, reported a dropout rate of roughly 20-30% [5, 8], we estimated a dropout rate of about 25% and selected 115 subjects for baseline measurements. R version 3.6.2 (R Foundation for Statistical Computing, Vienna, Austria) was used for all statistical analyses.

This study has some limitations. First, the number of dropouts was high. In addition, although within the predicted range, nearly 30% of patients were excluded for not completing the long-term survey (i.e., 12 months after THA). Thus, the presence of selection bias cannot be denied. However, when preoperative hip abductor strength and knee extensor strength were compared between dropouts and analyzed patients, no significant difference was observed between the two groups, suggesting that the impact of dropouts may have been minimal. Second, since patients with orthopedic complications other than hip osteoarthritis were excluded, it is possible that only patients with a better than average course of THA were included in our analyses.