Efficacy of Tuina in patients with knee osteoarthritis: study protocol for a randomised controlled trial Authors: Jinlin Zhang, Bin Xue, Yan Wei, Ming Shi, Yi Luo, Weijia Chen, Qiang Wang, Xiaobing Xi, Zhijun Hu Published: 2025 ABSTRACT AbstractBackgroundTuina demonstrates clinical therapeutic efficacy in the treatment of knee osteoarthritis (KOA). However, due to the paucity of high-quality and credible evidence-based medical evidence, there is no unified consensus on the treatment of KOA with Tuina in the current medical guidelines. Therefore, we will conduct a two-arm parallel group design and a randomised controlled trial to evaluate the efficacy of Tuina for treating KOA and the effect of Tuina on various muscles of the knee joint.Methods and analysisA total of 84 participants from Ruijin Hospital Affiliated to Shanghai Jiao Tong University School of Medicine will be recruited. The participants will be randomly assigned to the Tuina group or the Infrared group in a 1:1 ratio. Participants in the Tuina group will receive manual therapy and infrared therapy, while the infrared group will receive infrared therapy exclusively. Both groups will undergo treatment twice per week during the 4-week treatment period and will be followed up at months 3 and 6 post-treatment period. All participants will receive healthcare education (encompassing knee joint exercise and protection) throughout the study duration. The primary outcome measure is the Lysholm score, which assesses the overall function of the knee joint. Secondary outcomes comprise Visual Analogue Scale, muscle tone and pressure pain threshold. All outcomes will be assessed and analysed by researchers blinded to the treatment allocation. Statistical analyses will adhere to the intention-to-treat principle.Ethics and disseminationThis trial protocol has been approved by the Ethics Committee of Ruijin Hospital (project number 2024-38). Written informed consent from study participants will be obtained prior to enrolment. The findings will be disseminated through peer-reviewed journals.Trial registration numberITMCTR2024000469. STRENGTHS AND LIMITATIONS OF THIS STUDY This study is a randomised controlled trial with a rigorous design and robust methodological quality to determine the short-term and long-term clinical effects of Tuina on knee osteoarthritis. This study is a randomised controlled trial with a rigorous design and robust methodological quality to determine the short-term and long-term clinical effects of Tuina on knee osteoarthritis. This study will incorporate a 6-month follow-up period, facilitating the observation of Tuina’s long-term effectiveness. This study will incorporate a 6-month follow-up period, facilitating the observation of Tuina’s long-term effectiveness. This study will investigate the impact of Tuina on the muscles surrounding the knee affected by osteoarthritis, particularly the popliteal muscle. This study will investigate the impact of Tuina on the muscles surrounding the knee affected by osteoarthritis, particularly the popliteal muscle. Electronic questionnaires may result in anomalous questionnaire scores due to patients’ subjective interpretation biases. Electronic questionnaires may result in anomalous questionnaire scores due to patients’ subjective interpretation biases. Treatment blinding could not be implemented. Treatment blinding could not be implemented. Introduction Knee osteoarthritis (KOA) is a degenerative joint disorder characterised by cartilage deterioration, osseous remodelling, osteophyte formation and synovial inflammation, resulting in joint pain, rigidity, oedema and impaired function.1 The incidence of osteoarthritis has increased by 132.2% in the past 30 years and is projected to increase by 60%–100% by 2050. The age of onset for new cases has been progressively decreasing.2 Osteoarthritis can cause a substantial burden of pain, limitation of activity and decreased quality of life for individuals.3 Osteoarthritis is a leading cause of years lived with disability.4 Moreover, osteoarthritis can lead to early retirement from work, as the economic burden of OA is about 0.4% of gross domestic product.5 Therefore, as the most frequent osteoarthritis,6 improved treatment and prevention of KOA warrant further investigation. The treatment strategy for KOA primarily aims to alleviate symptoms. Disease management methods include non-pharmacological intervention, pharmacological treatment and surgical treatment.7 Tuina, as a traditional Chinese non-pharmacological manual therapy, can stimulate the muscles and bones of the knee joint to achieve the effect of relaxing sinews and activating collaterals. Existing clinical evidence has demonstrated that Tuina can reduce pain, improve function, enhance motor ability and ameliorate negative emotions.8 9 Consequently, Tuina can exert a clinical therapeutic effect on KOA. However, there is no unified consensus on the treatment of KOA by Tuina in the current medical guidelines.1013 This is primarily due to the lack of high-quality and credible evidence-based medical evidence.14 Additionally, the muscles surrounding the knee joint play a significant role in the pathogenesis and progression of KOA. Restoring muscle function and quality is essential to improve the symptoms and quality of life of KOA patients.15 However, there are no definitive research results on the effect of Tuina on the muscles surrounding the knee. Therefore, it is of research value to conduct a randomised controlled trial (RCT) with rigorous design and robust methodological quality to determine the short-term and long-term clinical effect of Tuina on KOA, and to explore the impact of Tuina on the muscles surrounding the knee affected by osteoarthritis. To address this issue, we conduct this RCT of Tuina therapy in patients with KOA to observe not only the short-term treatment effect on patients, but also the long-term effect of Tuina therapy through extended follow-up. Furthermore, with the assistance of testing equipment, the effect of Tuina on different muscles of the knee joint will be objectively evaluated. This study aims to provide high-quality clinical evidence for the treatment of KOA using Tuina. Methods Study design This study employs a two-arm parallel group design and a RCT with a 1:1 allocation ratio from May 2024 to December 2025. A total of 84 participants from Ruijin Hospital Affiliated to Shanghai Jiao Tong University School of Medicine will be recruited. The participants will be randomly assigned to either the Tuina group or the Infrared group. This trial protocol was approved by the Ethics Committee of Ruijin Hospital (project number 2024-38) and registered with the International Traditional Medicine Clinical Trial Registry (ITMCTR2024000469). All participants will provide written informed consent prior to treatment. Participants in the Tuina group will receive manual therapy and infrared therapy, while those in the infrared group will receive infrared therapy. Both groups will undergo treatment twice per week during the 4-week treatment period and will be followed up at months 3 and 6 post-treatment. All participants will receive healthcare education (including knee joint exercise and protection) throughout the entire study period. The trial flow chart and detailed schedule are presented in figure 1 and table 1, respectively. Flow chart of the study procedure. Table 1Schedule of enrolment, intervention and outcome measuresTimeW-1W0W1W2W3W4W16W28PeriodScreeningBaselineInterventionFollow-upEnrolment Inclusion/exclusion criteria√ Informed consent√ Medical history√ Baseline characteristics√ Physical examination√ Allocation√Intervention Tuina group8 sessions of Tuina and infrared therapy Infrared group8 sessions of infrared therapyOutcomes Lysholm score√√√√√√ VAS√√√√√√ MT√√√√ PPT√√√√Data collection and statistical analysis Adverse event√√√√√ Cause of dropout√√√√√√√ Safety analysis√√√√ Analysis√√√√√√√√MT, muscle tone; PPT, pressure pain threshold; VAS, Visual Analogue Scale; W, week. Schedule of enrolment, intervention and outcome measures MT, muscle tone; PPT, pressure pain threshold; VAS, Visual Analogue Scale; W, week. Sample size estimation As a superiority trial, the formula for calculating sample size is n=2(Zα/2+Zβ)×σ2(μt−μc−Δ)2.16 The primary outcome data were derived from the clinical study by Gu Yiqin et al.17 The true mean difference between groups (μt-μc) was 10.81, and the expected population standard deviation (σ) was 8.57. We expected that a clinically meaningful difference (∆) was 5. This sample size ensured a statistical power of 80% (1−β=0.8), a two-sided significance level (α) set at 0.025, and accounts for a dropout rate of 20%. Based on these parameters, 84 participants are required (42 participants in each group). Recruitment Participants are recruited primarily through hospital network media platforms and offline posters from October 2024 to 30 June 2025. A designated study researcher will contact potential participants via telephone or WeChat. Concurrently, the researcher will distribute an electronic questionnaire to the potential participants, encompassing height, weight, major symptoms, VAS, Lysholm score and other relevant measures. Participants meeting the eligibility criteria based on the questionnaire responses will be invited to participate in the randomised study according to their registration time. Written informed consent will be obtained on the day of study participation. The initial patient enrolment date was 25 October 2024. Inclusion criteria Participants who meet the following criteria will be included: According to guidelines for diagnosis and treatment of osteoarthritis formulated by the Chinese Rheumatology (2010). According to guidelines for diagnosis and treatment of osteoarthritis formulated by the Chinese Rheumatology (2010). Age between 30 and 70 years old. Age between 30 and 70 years old. Kellgren-Lawrence (K-L) imaging grades between 1 and 3.18 Kellgren-Lawrence (K-L) imaging grades between 1 and 3.18 Voluntary participation in the study, signed informed consent and adherence to study procedures Voluntary participation in the study, signed informed consent and adherence to study procedures Exclusion criteria (prior to randomisation) Participants will be excluded from the trial if they meet any of the following criteria before randomisation: Patients diagnosed with malignant neoplasms. Patients diagnosed with malignant neoplasms. Diagnosed with suppurative arthritis, gout or acute knee trauma. Diagnosed with suppurative arthritis, gout or acute knee trauma. History of knee fracture surgery (particularly with internal fixation materials or artificial alterations in knee structure). History of knee fracture surgery (particularly with internal fixation materials or artificial alterations in knee structure). Concurrent participation in experimental studies of other treatments for the disease within 2 weeks. Concurrent participation in experimental studies of other treatments for the disease within 2 weeks. Dropout criteria The following events may lead to the discontinuation of the assigned treatment, but all randomised participants will still be included in the intention-to-treat (ITT) analysis for the final study assessment: Non-adherence: Persistent failure to adhere to the prescribed treatment plan. Non-adherence: Persistent failure to adhere to the prescribed treatment plan. Safety concerns: Occurrence of irreversible adverse events (eg, significant trauma, intercurrent illness) where continuing the intervention is deemed unsafe or not in the participant’s best interest. Safety concerns: Occurrence of irreversible adverse events (eg, significant trauma, intercurrent illness) where continuing the intervention is deemed unsafe or not in the participant’s best interest. Participant request: Decision by the participant to withdraw consent for further treatment. Participant request: Decision by the participant to withdraw consent for further treatment. Loss to follow-up: Inability to contact the participant for subsequent study visits. Loss to follow-up: Inability to contact the participant for subsequent study visits. Randomisation and allocation concealment Eligible patients will be randomised into one of the two groups: the Tuina group or the infrared group, using a computer-generated random number table. The resulting random numbers will be sealed in opaque paper envelopes and provided only to the therapist for patient allocation. This process will be conducted by a researcher independent of other aspects of the study. Blinding Treatment is non-blind for both the therapist and the participants. But all assessors and statisticians will remain blinded to group allocation. Intervention Participants will receive a total of eight sessions over a 4-week period, twice weekly. To mitigate the effects of variations in the application of Tuina techniques, all Tuina treatments are administered by the same therapist. The therapist possesses more than 10 years of experience in clinical Tuina therapy. Data collection will occur four times during the eight treatment periods, encompassing VAS, Lysholm score, muscle tone (MT) and pressure pain threshold (PPT) of the knee joint-related muscles of the affected limb. The MT and PPT values will be obtained by the therapist through instrument operation, with the instrument values recorded via photographic documentation. VAS and Lysholm scores will be obtained through electronic questionnaires on the day of treatment, with patients provided the opportunity to seek clarification on questionnaire contents prior to completion. During treatment, appropriate health education will be provided to patients; furthermore, based on individual patient symptoms, recommendations for various self-exercise methods and precautions will be offered. Following the 4-week treatment period, questionnaires will be administered at months 3 and 6 post-treatment. The questionnaire content primarily comprises VAS, Lysholm score and daily self-protection exercise information. Tuina group The functional activity of the knee joint requires support from tendons and bones.1921 Tendon fatigue directly results in the limitation of knee joint flexion and extension, while bone fatigue leads to joint instability. Bones serve as attachment sites for tendons. Physiologically, tendons and bones are interdependent, and pathologically, they exert mutual influence. Consequently, the theory of tendon and bone is integral to the treatment of KOA using traditional Chinese manual therapy. The Tuina therapy employed in this study is based on the experience of Wei’s Traumatology Department at Ruijin Hospital Affiliated to Shanghai Jiao Tong University School of Medicine in the treatment of KOA. The specific operational process is delineated in the subsections below: Step 1: supine position Initially, the practitioner applies manipulation to the quadriceps femoris muscle on the affected side, progressing from the proximal thigh to the superior aspect of the patella for 1 min using pressing, kneading and plucking techniques. Subsequently, the practitioner performs circular manipulation on the patella using plucking. On completion of each circle, the patella is pushed medially and laterally until further movement is impeded and the patient experiences a mild sensation of acidity and swelling, continuing for 2–3 min. Finally, the practitioner applies plucking manipulation to the surrounding muscles along the fibula of the lateral leg for 1 min. Step 2: prone position Initially, the physician administers rolling and kneading treatment on the hamstring muscle of the affected side, from the proximal thigh to the superior aspect of the popliteal fossa, for a duration of 1 min. Subsequently, the affected limb is maintained at 20–30 degrees of flexion, and the physician performs manipulation on the popliteal fossa through pressing, kneading and plucking for 2–3 min. Following this, the physician administers rolling and kneading treatment on the gastrocnemius muscle for 1 min. Lastly, the patient is instructed to grasp the bed with both hands, while the physician grasps the ankle joint of the affected limb and performs intermittent traction of the knee 3–5 times, lasting for 30 s. On completion of the Tuina manipulation, the patient proceeds to infrared treatment. The treatment protocol for this component is identical to that of the infrared group, and its treatment process will be delineated in detail in the infrared group section. Infrared group Infrared therapy is a widely used physical therapy modality for KOA.2225 It is a non-contact treatment with mild irritation. The patient population that can tolerate this treatment is relatively broad, and it is associated with a low incidence of adverse reactions. Consequently, infrared therapy was selected as the control group in this study. Infrared therapy will be administered using a therapy instrument (JLY-II Huaicheng Medical Appliance Factory, China). Two certified physical therapists with over 3 years of professional experience at Shanghai Ruijin Hospital are responsible for the infrared treatment and are blinded to the group assignments. To ensure treatment standardisation, the technical criteria for infrared therapy were established by the research team following a review of relevant literature, and both physiotherapists have undergone concurrent training. The specific operational process is outlined in the subsections below: Preparation: The patient will be seated with the skin of the unilateral knee joint exposed, centred on the patella, and extending 3 cm superiorly and inferiorly. The therapist will position the instrument vertically 5 cm above the knee joint. Treatment: On obtaining the patient’s confirmation, the therapist will initiate the treatment by activating the start button, and the therapy will proceed for 20 min. The height of the instrument can be adjusted according to the patient’s subjective sensation, with joint comfort serving as the primary criterion. Outcomes The outcome will be measured by two self-report questionnaires and the muscle condition rapid measurement system (OE-220, ITO, Japan), which can assess knee pain, dysfunction and quality of life. Seven time points will be used to evaluate outcomes, including baseline (after enrolment); weeks 2, 3 and 4 (intervention sessions); and months 3 and 6 (follow-up sessions). Primary outcome The primary outcome measurement of this study is the Lysholm score, which was initially published in 1982 in The American Journal of Sports Medicine and subsequently revised in 1985.26 The Lysholm score comprises eight questions that focus on the categories of limping, support, locking, instability, pain, swelling, stair climbing and squatting, with each question response being assigned an arbitrary score on an increasing scale. The total score is the sum of each response to the eight questions and may range from 0 to 100, where higher scores indicate a better outcome with fewer symptoms or disability. Scores are also categorised as excellent (95–100), good (84–94), fair (65–83) and poor (<65). A study conducted by Briggs et al27 demonstrated that the Lysholm questionnaire possesses acceptable test–retest reliability, floor and ceiling effects, criterion validity, construct validity and responsiveness to change. The brevity and ease of completion for patients are significant strengths of the Lysholm scale. Secondary outcome Visual Analogue Scale. Visual Analogue Scale. Visual Analogue Scale (VAS), first implemented in 1921 by Hayes and Patterson,28 is one of the most frequently used single-dimensional measurement and evaluation tools for pain intensity. The scale primarily consists of a straight line, with one end indicating ‘no pain at all’ and the other end indicating ‘the most severe pain imaginable’ or ‘pain to the extreme’. Patients will be instructed to mark the corresponding position on this line to represent the intensity of the pain they experience at that time. In this study, to facilitate the understanding and selection of some elderly patients, the facial VAS was modified by adding several cartoon expressions (happy, neutral, pain, etc) to the aforementioned linear VAS line, thereby making the score more intuitive and comprehensible. Assessment of muscle tissue status. Assessment of muscle tissue status. Tender point assessment. Tender point assessment. The assessment of muscle tissue condition mainly focuses on MT, while the tender point assessment mainly refers to the PPT of muscles.2931 The assessment will be conducted using the OE-220 muscle condition rapid measurement system. Prior to the evaluation, the instrument should be calibrated to the corresponding detection system. Subsequently, the MT and PPT of the quadriceps, biceps, popliteal and gastrocnemius muscles are measured, respectively, using the probe sensor. Four time points will be used to assess outcomes, including baseline (before treatment), weeks 2, 3 and 4 (after treatment). Quality control First, standard operating procedures will be established, and all research personnel should undergo training to ensure the continuity and progress of the study. Second, the number of participants enrolled each week will be monitored to ensure that the study is conducted in a systematic manner and to prevent overcrowding or underutilisation. Third, data monitoring protocols will be developed to ensure timely follow-up visits and access to research data. Fourth, the Ethics Committee of Ruijin Hospital Affiliated to Shanghai Jiao Tong University School of Medicine is responsible for the conduct and safety monitoring of the trial. Fifth, there are no conflicts of interest between the committee members and this study. Sixth, the implementation of this study will strictly adhere to the principles of the Declaration of Helsinki. All adverse events, whether related to the intervention or not, will be monitored and recorded by therapists through active inquiry at each visit in the case report form (CRF) form, and their severity and causal relationship with the intervention will be assessed by the principal investigator. All serious adverse events will be reported to the ethics committee within 24 hours of being discovered. Data collection and management On completion of the trial, the filled CRF and the electronic questionnaire will be transferred to two professional data managers who are blinded to the allocation and uninvolved in the other aspects of the study. Two data staff members will independently extract data from the CRF and electronic questionnaires. All electronic data will be stored on a password-protected server, while all paper data will be secured in locked filing cabinets. Statistical analysis The analysis will be performed following the ITT principle. All randomised participants will be included in the analysis according to their initially assigned groups. Participant baseline characteristics will be summarised using descriptive statistics. Continuous variables will be presented as mean±SD or median (IQR), and categorical variables as frequencies (percentages). All analyses will be conducted using R software (V.4.3.0) with the lme4 and mice packages. A two-sided p<0.05 will be considered statistically significant. Primary and secondary continuous outcomes We will employ a linear mixed-effects model to analyse changes in continuous outcomes over time and between groups. The model will include treatment group, time and their interaction as fixed effects. To account for potential confounding and improve precision, the baseline value of the outcome, age and OA grade will be included as covariates. This approach allows for a purer assessment of the treatment effect after controlling for the influence of these factors. Subject-specific intercepts will be modelled as random effects. The model assumptions will be verified by examining the normal Q-Q plot of the standardised residuals and the scatter plot of residuals vs fitted values (to test for homoscedasticity). If the residuals are not normally distributed or if there is heteroscedasticity, we will consider transforming the dependent variable or using a generalised linear mixed model. We will prespecify a model-building strategy for random effects. The model including both random intercepts and random slopes for time will be compared with the random intercepts-only model using a likelihood ratio test. The more complex model will be retained if it results in a statistically significant improvement in model fit (p<0.05). If a significant group-by-time interaction is found (p<0.05), we will perform post hoc pairwise comparisons at each time point using the estimated least-squares means, with adjustment for multiple comparisons using Tukey’s Honestly Significant Difference (HSD) method. In the absence of a significant interaction, we will interpret the main effects of group and time. Missing data handling For missing data, the linear mixed-effects model applied assumes data are missing at random by default. To assess the impact of missing data, we will perform a sensitivity analysis using multiple imputation. We will create 20 imputed datasets using chained equations, with the imputation model including the outcome variable at all time points, treatment group and key baseline covariates. The linear mixed-effects model will then be fitted to each imputed dataset, and results will be pooled using Rubin’s rules. Sensitivity analysis for continuous outcomes We will perform the following sensitivity analyses to test the robustness of our primary model: (1) Model specification: Refit the model using restricted maximum likelihood (ML) and compare with ML results. (2) Outlier influence: Refit the model after removing influential observations to ensure results are not driven by outliers. Multiple comparisons control To account for increased type I error risk due to multiple testing, we will employ the following strategies: The primary outcome analysis will maintain a two-sided α level of 0.05, with post hoc pairwise comparisons adjusted using Tukey’s HSD method. The primary outcome analysis will maintain a two-sided α level of 0.05, with post hoc pairwise comparisons adjusted using Tukey’s HSD method. For the three secondary continuous outcomes, we will apply a Bonferroni correction, dividing the α (0.05) level by 3. For the three secondary continuous outcomes, we will apply a Bonferroni correction, dividing the α (0.05) level by 3. All prespecified interaction tests will be considered significant at α=0.05 after Bonferroni correction for the number of tested interactions. All prespecified interaction tests will be considered significant at α=0.05 after Bonferroni correction for the number of tested interactions. Ethics and dissemination This study protocol has been approved by the Ethics Committee of Ruijin Hospital (project number 2024-38). The supplementary approval from the Ruijin Hospital Ethics Committee is provided as online supplemental material 1. All participants must provide written informed consent before participating in the trial. The Chinese and English versions of the consent forms are available as onlinesupplemental materials 2 3, respectively. All personal information from potential and enrolled participants out of the scope of this trial will not be collected, shared or maintained to protect confidentiality before, during and after the trial. After the study completed and the data analysed, results of this trial will be presented as articles in peer-reviewed journals or at academic conferences. Trial status The protocol was registered on 8 September 2024. Recruitment of potential participants had commenced at the time of manuscript submission. Discussion The practitioner can make a comprehensive assessment of the pain, function and muscle tension of the patient’s knee joint through palpation during Tuina, thereby achieving ‘differentiation by palpation’. There is no positive correlation between the severity of KOA and the imaging results.32 33 Even in cases where the K-L grade is high, and knee joint hyperplasia and degeneration are evident, the patient’s knee joint may still meet normal functional requirements. This phenomenon is the result of the combined action of the knee joint muscles, ligaments and bone joints. It also aligns with the relationship between tendons and bones described in traditional Chinese medicine. Among all non-pharmacological treatments, only manual therapy allows for direct and rapid assessment of knee severity. Therefore, this study will investigate the treatment of KOA with Tuina. On the one hand, the clinical efficacy of Tuina treatment for KOA is underestimated, necessitating scientific, rigorous, high-quality RCTs for further research and observation. On the other hand, it aims to explore the changes in pain threshold and muscle tension of different muscles surrounding the knee joint during Tuina treatment, in order to identify new evaluation aspects that will more conveniently and accurately reflect the severity of KOA. To comprehensively reflect the changes in muscles surrounding the knee joint during Tuina treatment, the popliteal muscle was included in the study of muscles around the knee joint. This decision was based on the fact that the popliteal muscle is a crucial stabilising muscle posterior to the knee joint, playing an indispensable role in knee joint function.34 35 However, few clinical studies have investigated it, and dysfunction of the popliteus is often underappreciated.36 Instead, research continues to focus on quadriceps femoris, biceps femoris, hamstrings, gastrocnemius and other muscles.3740 Moreover, most patients with KOA exhibit trigger points in the popliteal muscle, which elicit significant pain on palpation. This clinical sign is rarely mentioned in the literature, necessitating further clinical studies for verification. This study has several limitations. First, treatment blinding could not be implemented. Second, the follow-up period was limited to 6 months post-treatment. Although comprehensive health education was provided during the treatment, patients may lack sufficient joint protection for various reasons, potentially leading to symptom exacerbation and affecting score results during follow-up. Third, electronic questionnaires will be exclusively used for pain and function scores during the treatment and follow-up period. While convenient, this method may result in anomalous questionnaire scores due to patients’ subjective interpretation biases. Efforts will be made to ensure patients are thoroughly informed about the options prior to their initial completion of the questionnaires. Supplementary material