This study was a single-centre, single-blind randomised controlled trial with blinded outcome assessment, which was conducted at the Sport Medicine and Rehabilitation Centre, Shanghai University of Sport. It was approved by the Ethics Committee of the Shanghai University of Sport (Ref. No.: 2016-016) and registered at Chinese Clinical Trial Registry a priori as a clinical trial (ID: ChiCTR-IOR-16009234). In addition, this study adhered to CONSORT guidelines.
2.1 Participants
With settings of α = 0.05, power (1 − β) = 0.80 and effect size = 0.25, power analysis showed that three groups of 66 participants in total were the required sample size. Given the dropouts, we recruited 80 participants for the research.
The voluntary participation of KOA patients was requested via posters on community centres in Yangpu District, Shanghai, China from July 2017. The diagnosis of KOA was performed by an orthopaedic surgeon on the basis of the patients’ medical history, X-ray imaging and physical assessment. Based on the classification criteria of the American Rheumatism Association for KOA [(35)], the inclusion criteria were as follows: (1) men or women with radiographic diagnostic criteria of definite KOA (unilateral or bilateral) and who reported pain symptoms for at least 3 months; (2) mild-to-moderate KOA (Lequesne Knee Score = 1to 7); (3) aged 50–70 years old; (4) medication not expected to change during the study period; and (5) available three times a week for over 3 months. The exclusion criteria were as follows: (1) had undergone knee surgery in the past 6 months; (2) had acute symptomatic KOA; (3) had muscular, joint or neurological conditions that affect lower limb function; (4) underwent a structured exercise program specifically for KOA; (5) unable to understand the study’s procedure; and (6) had motor neuron disorders, such as Alzheimer’s and Parkinson’s diseases [(27)]. The research designer took charge of contacting the eligible participants, confirming their willingness to complete the trial and arranging the baseline assessment of outcomes.
After screening, 73 eligible participants were informed about the study procedures, and they gave written informed consent before enrolling in the study. These participants were randomly divided into WBV, strength training (ST) and health education (HE) groups by using computer-generated randomisation by the research designer.
2.2 Outcome measures
All outcome assessments were conducted by the blinded research assessor at pre-intervention and post-intervention. For the bilateral KOA, the affected side was defined as the worse side. In addition, the demographic questionnaire was collected before the intervention, including participants characteristics (sex, age, Body Mass Index, Lequesne Knee Score and affected side). After intervention, all participants were requested to return to the laboratory within 2 days and complete the post-intervention assessment. Only the participants who completed more than 80% of the intervention sessions were asked to conduct the post-intervention assessments.
Primary Outcome
2.2.1 Isokinetic muscle strength
Knee extensor and flexor strength of the affected side were measured using an isokinetic dynamometer (Physiomed, CON-TREX, TP 1000, Germany). Before the formal test, the participants were informed about the test procedure and performed a 5 min warm up. At first, the assessor secured them to the dynamometer at the upper chest, pelvis and distal femur on the tested side with straps. Then, they were instructed to familiarise themselves with the procedure through three submaximal contractions. In the formal test, the maximal concentric knee extension–flexion contractions were performed at angular velocities of 90°/s and 180°/s. The work of each trial was recorded from 80° to 10° of the knee moving angle, and 0° was considered as the full extension. During the test, the assessor verbally encouraged participants to move as forcefully as possible. Normalised by body mass, the peak torque and peak work (PT and PW; Nm/kg and W/kg, respectively) were recoded for analysis.
2.2.2 Proprioception of knee
The proprioception of knee was observed as the threshold for the detection of passive movement (TDPM) and was measured using a reliable method [(36)]. The participants were informed about the proprioception test procedure. In a sound-attenuated room, the participants were isolated to reduce any auditory or visual interference. TDPM was tested using an electrically driven movable frame. During the test, they sat on an adjustable chair, which helped them place a foot lightly on the plane. During the test, the plane moved the shank forward or backward at a velocity of 0.4º/s. The angle of knee joint was adjusted to 120º by a goniometer. During the test, when the motion of the knee was detected, the patients would press the handheld stop button immediately and tell the direction. The rotation angles of the frame were determined as the threshold for the detection of flexion and extension. The mean value of three successful trials in each direction was used for analysis.
Secondary Outcome
2.2.3 Self-reported knee pain
The 10 cm visual analogue scale (VAS) was used to assess the knee pain of the affected side over the last week. The number 0 indicated no pain, whereas 10 represented maximal pain.
2.2.4 Physical function
The Timed Up and Go test (TUG) and the 6-min Walking Distance test (6MWD) were administered to determine the function performance of participants. These methods are simple and reliable measurements used in previous studies on KOA [(27, 37)]. In the TUG test, the participants were requested to perform the following tasks as far as possible: standing up from a standard chair (40 cm height); walking 3 m; and turn around, walk back, and sit down. The total test was timed by a chronograph (in seconds). As for 6WMD, the participants were asked to walk for 6 min, and the distances were recorded. All participants must finish these tests thrice for the average values to be calculated.
2.3 Intervention
The same certified physical therapist supervised the interventions in the WBV, ST and HE groups. Each training session comprised a 5 min warm-up and 5 min cool-down in WBV and ST.
2.3.1 Whole body vibration training
Participants in the WBV group performed the training 3 days per week for 8 weeks. WBV training was conducted on a vibration device (i-vib5050; Sport Platform, China). During the training, the participants performed static squat barefoot on the platform with bent knee (30º and 60º). In addition, the distance between their feet was consistent with the shoulders. In each session, the physical therapist adjusted the angle of knee flexion, timed the duration time and prevented falling. As showed in Table 1, the duration time, sets and total time were increased progressively over the 8-week training period. The parameters of WBV were as follows: frequency of 20 Hz; and amplitude of 2 mm.
2.3.2 Strength training
The ST group undertook three training sessions per week for 8 weeks the same as the WBV group. Likewise, all sessions were under the supervision of the same physical therapist in the WBV group. However, the participants performed static squat on flat ground. The protocol of ST, including the duration time and the angle of bent knee, was parallel with WBV except for the vibration exposure (Table 1).
(Insert Table 1)
2.3.3 Health education
The participants in the HE group received 8 weeks of health education. They attended one 60-min group session per week. Each session consisted of a 30-min lecture and a 30-min discussion. The session was facilitated by the same physical therapist in the WBV and ST groups. The lectures were focused on health-related topics, such as OA, aging and nutrition. Furthermore, the participants in the HE group were required to maintain their previous lifestyle and not to attend any other regular rehabilitation programs during the study period.
2.4 Statistical analysis
The SPSS statistical software program (IBM, Chicago, IL, USA) was used for statistical analysis. The data were included for analysis of participants who completed the training. All data were expressed as means ± SD. To evaluate the normality of these data, the Shapiro–Wilk test was used. The one-way analysis of variance (ANOVA) and the chi-squared test were performed to determine the difference of demographic characteristics among the WBV, ST and HE groups. Then, the two-way repeated measurements ANOVA was used to determine the difference in outcomes among the three groups. The effect size of between-group effect was calculated by partial eta-square, and Tukey’s post hoc test was used to compare the results. Significance was set at p < 0.05. In addition, for within-group effect, the paired t-test was performed to test the difference of the outcomes within each group.