• Hip and knee surgery

Introduction: Osteoarthritis (OA) of the hip affects nearly 10% in people of 45 years of age or older. It initially manifests as abnormal joint tissue metabolism followed by anatomic alterations and symptoms of activity-related pain, impaired physical function and short-term stiffness after inactivity. Exercise is a feasible, safe and effective treatment option for reducing pain and physical disability in hip and knee OA. Clinical guidelines broadly recommend exercise, patient education and weight loss if necessary, as first-line treatment. However, the optimal exercise type and dose is unknown. Neuromuscular exercise (NEMEX) is moderately effective for improving physical function and reducing pain in hip OA and is implemented in clinical practice in several countries. Meanwhile, power-type progressive resistance training (PRT) has been shown to be feasible and has promising efficacy. Importantly, PRT is considered the most effective intervention for improving muscle strength and power, features that are markedly impaired in hip OA.
Aims and hypotheses: The aim for Paper I and II was to compare the effectiveness of twelve weeks of progressive resistance training with twelve weeks of neuromuscular exercise for improving functional performance in patients with hip osteoarthritis. The hypothesis was that PRT is superior to NEMEX for improving functional performance in patients with hip osteoarthritis measured by changes in the 30-second chair stand test (30s-CST) at twelve-week follow-up. The aims for Paper III were to investigate whether changes in leg extensor muscle power of the affected limb are associated with changes in performance-based and patient-reported measures of physical function after twelve weeks of exercise in patients with hip OA, and to explore whether these associations are dependent on the type of exercise performed. The hypotheses were that changes in leg extensor muscle power is associated with changes in performance-based and patient-reported measures of physical function, and that changes in leg extensor muscle power will explain a larger part of the differences observed in physical function after PRT compared to after NEMEX. The aim for Paper IV was to assess associations between baseline characteristics, i.e. predictive factors, and changes in hip pain and physical function after PRT or NEMEX. The hypotheses were that baseline values for the dependent variable, muscle power, 30s-CST, body mass index (BMI), and age as well as female sex, use of analgesics and no prior exercise therapy, will be positively associated with changes in hip pain and physical function.
Methods: To disseminate the investigation of the hypotheses set above, four papers will be published. First, a trial protocol paper describing a randomized controlled trial (Paper I). Second, a paper presenting the results of the comparison of PRT and NEMEX as described in the trial protocol paper (Paper II). Third, a paper containing a secondary post-hoc analysis exploring the associations between changes in leg extensor muscle power and physical function (Paper III). Fourth, a paper describing an exploratory post-hoc analysis of predictive factors for changes in pain and physical function (Paper IV).
The Hip Booster Trial, as described in Paper I and II, was a multicenter, cluster-randomized, controlled, parallel-group, assessor-blinded, superiority trial conducted at hospitals and physiotherapy clinics in Denmark. Patients with clinically diagnosed mild-to-moderate hip OA who fulfilled all inclusion criteria were invited to participate. When three-to-five participants from the same trial site had completed baseline assessments, they were cluster-randomized to perform 12 weeks of PRT or NEMEX. These exercise interventions were delivered as supervised group-based sessions for one hour, twice weekly, for 12 weeks. PRT was a power-type, high-intensity program following classic linear periodization with five generic exercises targeting muscles at the hip and knee joints. NEMEX was focused on improving control, stability, and lower extremity muscle strength through 10 individualized exercises. Outcomes were assessed at baseline and 12 weeks. The primary outcome was changes in the 30s-CST and key secondary outcomes were changes in Pain and Hip-related Quality of Life (QoL) subscales from the Hip disability and Osteoarthritis Outcome Score (HOOS). A mixed effect model was used for statistical analyses by the intention-to-treat principle.
The analyses in Paper III were performed on the whole population with the sufficient data from The Hip Booster Trial. Simple and multivariate linear regression models estimated the associations between changes in leg extensor muscle power normalized to body weight and changes in performance-based or patient-reported measures of physical function.
As in Paper III, the analyses in Paper IV included all participants for the trial with the necessary data available. Multivariate linear and binary regression models were used to estimate associations between predictive factors and changes in pain and activities of daily life function subscales from the HOOS questionnaire.
Results: The Hip Booster trial included 160 participants who were cluster-randomized to PRT (n: 82) or NEMEX (n: 78). A total of 13 participants (8%) were lost to follow-up. Of these, 7 participants were allocated to PRT (9% drop-out rate) and 6 to NEMEX (8% drop-out rate). The number of participants who attended ≥ 80% of the sessions were 52 (63%) in PRT and 53 (68%) in NEMEX. The number of participants who completed ≥ 80% of exercise sets were 50 (61%) in PRT and 38 (49%) in NEMEX.
The between-group difference in mean changes [95% confidence interval] from baseline to 12-week follow-up in number of chair stands for the 30s-CST was 0.0 [-0.8; 0.8], not amounting to a major clinically important difference (MaCID) of 2.1 or minimal important difference (MID) of 0.5 chair stands. The between-group difference in mean changes in the HOOS pain subscale was 0.7 [-4.0; 5.3] points in favor of NEMEX, not amounting to a MID of 7.7 points. The between-group difference in mean changes in the HOOS QoL subscale was 2.3 [-3.0; 7.6] points in favor of PRT, not amounting to a MID of 8.4 points.
The secondary post-hoc analyses showed that changes in leg extensor power (watt/kg) were associated with changes in all performance based (30s-CST (β: 1.97 [.99; 2.96] chair stands, R2: 0.10), 9-step timed stair climb test (β: -.94 [-1.65; -.24] seconds, R2: 0.05) and 40-meter fast-paced walk test (β: -1.81 [-2.82; -.80] seconds, R2: 0.08)) and patient-reported (HOOS Activities of daily life function (β: 8.63 [3.02; 14.25] points, R2: 0.06) and Sport and Recreation function (β: 11.34 [2.78; 19.90] points, R2: 0.05)) measures of physical function. There were no statistically significant or clearly relevant differences in these associations depending on allocation to PRT or NEMEX. Associations were also found at baseline for leg extensor power and all measures of physical function.
The exploratory analyses of predictive factors showed some indication that changes in HOOS pain and HOOS ADL function seemed positively associated with female sex (β: 4.43 [-1.84; 10.70] and 4.70 [-1.50; 10.90] points) and negatively associated with baseline levels of the dependent variable (β: -.45 [-.61; -.29] and -.44 [-.58; -.30] points), having previously engaged in exercise therapy (β: -7.27 [-13.48; -1.06] and -5.46 [-11.66; .74] points), currently using analgesics (β: -5.67 [-10.39; -.95] and -7.24 [-11.72; -2.76] points), and BMI (β: -.59 [-1.07; -.11] and -.55 [-1.03; -.08] points). There was some indication that older age and lower BMI could be associated with greater effects on pain and physical function from PRT compared to NEMEX and conversely, that lower age and higher BMI could be associated with greater effects from NEMEX compared to PRT.
Conclusion: For patients with hip OA, PRT is not superior to NEMEX for improving functional performance, hip-related QoL or reduce hip pain, as the interventions showed very similar improvements in these outcomes. In these patients engaging in PRT or NEMEX, changes in leg extensor muscle power are consistently associated with changes in physical function across performance-based and patient-reported measures. These associations do not appear to be dependent on whether PRT or NEMEX is performed. Lastly, in patients with hip OA who engage in PRT or NEMEX, there is some indication that female sex, no prior exercise therapy, no use of analgesics, and lower levels of BMI, pain, and physical function seem associated with greater improvements in pain and physical function.