A meta-analysis of randomized controlled trials: evaluating the efficacy of isokinetic muscle strengthening training in improving knee osteoarthritis outcomes

Abstract

Background

Knee osteoarthritis (KOA) is a prevalent degenerative joint disease. The primary pathological manifestations of KOA include articular cartilage degeneration, joint space narrowing, and osteophyte formation, leading to a spectrum of symptoms, including joint pain, stiffness, reduced mobility, diminished muscle strength, and severe disability. We aimed to utilize a meta-analysis to evaluate the efficacy of isokinetic muscle strengthening training (IMST) as a rehabilitation treatment for KOA in lowland areas.

Methods

The study conducted a comprehensive search of the CNKI, WanFang Data, VIP Database, PubMed, Ovid MEDLINE (1946–), Cochrane Library, Embase, and CBM databases. The databases were conducted from establishing each database to September 31, 2024. The studies included were randomized controlled trials (RCTs) with participants from the plains who met the diagnostic criteria for KOA as outlined in the 2019 edition, with no restrictions on gender, age, or disease course, and no patients with advanced disease; studies where in the control group was either a non-intervention group or a group receiving treatment, other than IMST, and the experimental group received IMST alone or in addition to the treatment administered to the control group; and studies with at least two of the following outcome indicators: (i) knee flexors (Flex)/extensors (Ext) peak torque (PT), (ii) knee Flex/Ext total work (TW), (iii) knee Flex/Ext max rep total work (MRTW), (iv) knee Flex/Ext average power (AP), (v) visual analogue scale (VAS) for pain, (vi) Lequesne index (LI), (vii) Western Ontario and McMaster University Osteoarthritis Index (WOMAC), (viii) Lysholm Knee Scoring Scale (LKSS), (ix) range of motion (ROM) of the knee joint, and (x) 6-min walk test. We systematically reviewed the RCTs in both Chinese and English and evaluated the quality of the included literature. Data were processed and analyzed using ROB 2, RevMan 5.4, Stata17, and GRADEpro. The study protocol was registered on PROSPERO (CRD42024607528).

Results

Thirty-three (46 studies, 2,860 patients) had low-to-some concerns risk. IMST significantly improved physical therapy outcomes, including knee Flex PT and knee Ext PT at an angular velocity of 60°/second (standardized mean difference 13.19 [95% confidence interval 6.44, 19.94], P = 0.0001 and 16.34 [11.47, 21.22], P < 0.00001, respectively), and 180°/second (11.17 [2.86, 19.48], P = 0.008 and 12.62 [3.49, 21.75], P = 0.0077, respectively); knee Flex TW (79.77 [49.43, 110.10], P < 0.0001), Ext TW (86.27 [58.40, 114.15], P < 0.00001), knee Flex MRTW (9.38 [3.20, 15.56], P = 0.003), knee Ext MRTW (15.52 [8.96, 22.08], P < 0.0001), knee Flex AP (8.66 [0.70, 16.61], P = 0.03), knee Ext AP (7.27 [3.30, 11.23], P = 0.0003), knee Flex ROM (10.62 [7.94, 13.30], P < 0.00001), and LKSS scores (7.90 [5.91, 9.89], P < 0.00001). Additionally, it reduced VAS scores (− 0.70 [− 0.92, − 0.49], P < 0.00001), LI scores (− 1.24 [− 1.65, − 0.83], P < 0.00001), and WOMAC scores (− 6.05 [− 10.37, − 1.73], P = 0.006). Compared to the control group, superior clinical efficacy was noted in the experimental group. The quality of evidence the studies reported was poor, mainly due to original trials with high inter-study heterogeneity and imprecise results. The therapeutic effect of IMST on KOA remained significant after rigorous testing of subgroup and sensitivity analyses.

Conclusions

In patients with KOA, IMST improves muscle strength and relieves joint pain and stiffness. However, large-scale, high-quality, randomized controlled trials with extended observation periods are urgently needed to popularize the use of IMST in KOA patients.

Background

Knee osteoarthritis (KOA) is a prevalent degenerative joint disease [1, 2]. In western medicine, clinicians recognize KOA as a chronic condition marked by joint deformities, swelling, cartilage degeneration, defects, sclerosis of the joint edges or subchondral bone, and osteophyte formation. These symptoms significantly affect the quality of life of affected individuals. In Traditional Chinese Medicine (TCM), KOA is categorized as both “paralysis” and “bone paralysis” [3, 4], attributing its etiology to a “deficiency in origin and excess in superficiality.” “Deficiency in origin” indicates weakened functions of the liver and kidneys, whereas “excess in superficiality” is associated with wind, cold, and dampness [5, 6]. KOA is characterized by chronic and recurrent disease, often complicated by phlegm and blood stasis, and is intricately linked to the health of the liver, spleen, and kidneys. The prevalence of this condition is notably higher among the middle-aged and older populations [7]. The aging population has become a global challenge, with a generalized problem of loss of muscle strength and bone health among the older population. Loss of muscle strength leads to a loss of balance and increases the risk of illness and hospitalization, contributing to healthcare costs and the burden on health systems. Therefore, effective interventions to enhance musculoskeletal strength in older adults are critical, with significant implications for improving their quality of life and reducing the strain on socio-medical systems [8].

Three primary clinical strategies are employed to treat KOA: surgical procedures, pharmacological therapies, and non-pharmacological management [9]. Surgical treatment usually consists of arthroscopic cleanup, osteotomy, and arthroplasty in patients with advanced disease. No differential changes in blood indicators were observed before and after joint replacement, but early knee mobilization reduces the risk of joint adhesions [10]. Therefore, selecting adapted exercises that can increase the range of joint motion can contribute to postoperative recovery. Pharmacological treatments usually include articular cartilage protectors, analgesics, and nonsteroidal anti-inflammatory drugs (NSAIDs). Among them, NSAIDs, although practical, may increase the risk of venous thrombosis in the lower extremities and cardiovascular complications, which may potentially increase the rate of cardiovascular and cerebral vascular events and the overall mortality rate [11,12,13]. Therefore, people on long-term medications need to perform appropriate exercises and strengthen their lower extremities. Non-pharmacological therapies include exercise, health education, and physical therapy [7]. Among them, isokinetic muscle strengthening training (IMST), recognized as a safe and non-invasive therapeutic approach, is widely employed to rehabilitate KOA muscles and enhance muscle strength [14].

Although there are a large number of clinical studies that have explored the efficacy of IMST in the treatment of KOA, considering the improvements in physical functioning and reduced dependence on pharmacologic pain management interventions, these studies are limited in sample sizes and have a fragmented focus on specific outcome measures [15,16,17]. Notably, in China, the incidence, prevalence, and disability-adjusted life year (DALY) rates of KOA showed significant increases of 123.96%, 115.21%, and 175.24%, respectively, in 2021 compared with 1990 (Global Disease Database: https://www.healthdata.org/research-analysis/gbd). We further supported this trend with age-standardized incidence, prevalence, and DALY rates, which showed increases of 6.65%, 6.85%, and 7.18%, respectively, indicating a steady annual rise in the overall incidence of KOA in China [18]. The China Health and Retirement Longitudinal Study [19] reported a higher prevalence of KOA among women (11.2%) than among men (5.6%). Additionally, the study highlighted geographical disparities in prevalence (no data for the Tibet Autonomous Region), with the highest rates observed in the western regions (13.2%), followed by the central areas (8.1%), and the lowest in the eastern regions (5.2%). Notably, its prevalence in rural areas (10.4%) was higher than in urban areas (6.2%). These findings highlight the need for interventions focusing on KOA.

The Qinghai-Tibetan Plateau, which occupies a quarter of China’s landmass, is home to populations facing unique health challenges due to extreme environmental conditions such as low atmospheric pressure, hypoxia, and cold temperatures [20, 21]. Hypobaric and hypoxia, common in high-altitude regions, induces a range of compensatory physiological responses, including increased pulmonary ventilation, accelerated heart rate, enhanced cardiac output, and a rightward shift in the oxygen–hemoglobin dissociation curve. Acute exposure to hypoxia can lead to pulmonary vasoconstriction, whereas chronic exposure may result in sustained pulmonary vasoconstriction and vascular remodeling within the pulmonary vasculature, potentially culminating in pulmonary hypertension [22]. Accordingly, compensatory reactions can significantly increase blood viscosity and induce myocardial hypertrophy, which may ultimately lead to heart failure [23]. No data on the incidence, prevalence, and disability-adjusted life year (DALY) rates of KOA in the Tibetan Plateau are available. Notably, it is reported that the prevalence of KOA at 981 m, 2000 m, and 3200 m were 11.70% [24], 17.59% [25], and 25.24% [26], respectively, suggesting that the prevalence of KOA may increase with increasing altitude. Accordingly, the unique climatic conditions of the plateau could potentially result in slower recovery from KOA compared to plain areas [27]. Therefore, it is vital to find KOA treatments that are appropriate for local populations. Furthermore, there is a notable lack of research on effective treatments for this disease at high altitudes [28,29,30,31,32,33], and no studies have reported using IMST for KOA patients at high altitudes. Much of the literature in the plains discussed only a few indicators that do not fully reflect the therapeutic effect of IMST on KOA [15,16,17, 34]. Given these gaps, we aimed to examine the existing literature on IMST for KOA treatment to provide a more robust evidence-based foundation for clinical rehabilitation practices. Concurrently, we hope the IMST data from lowland areas will provide substantial clinical support for its application in treating KOA in the Qinghai-Tibetan Plateau.

Methods

Inclusion and exclusion criteria for literature review

The inclusion criteria were as follows: (1) study type: randomized controlled trials (RCTs); (2) study population: participants were recruited from the plains and met the diagnostic criteria for KOA as outlined in the 2019 edition [7], with no restrictions on gender, age, or disease course, and no patients with advanced disease; (3) interventions: the control group consisted of either a non-intervention group or another group receiving treatment other than IMST and the experimental group received IMST alone or in addition to the treatment administered to the control group; and (4) the included literature meeting at least two of the following outcome indicators: (i) knee flexors (Flex)/extensors (Ext) peak torque (PT), (ii) knee Flex/Ext total work (TW), (iii) knee Flex/Ext max rep total work (MRTW), (iv) knee Flex/Ext average power (AP), (v) visual analogue scale (VAS) for pain, (vi) Lequesne index (LI), (vii) Western Ontario and McMaster University Osteoarthritis Index (WOMAC), (viii) Lysholm Knee Scoring Scale (LKSS), (ix) range of motion (ROM) of the knee joint, and (x) 6-min walk test (6MWT). These indices reflect knee muscle strength (PT, TW, MRTW, AP), pain (VAS), functional scores (WOMAC, LI, LKSS), and knee mobility and physical performance (6MWT) in patients with KO. The primary indicators were PT, TW, VAS, and WOMAC.

The exclusion criteria were: (1) non-randomized controlled trials, animal experiments, and laboratory studies; (2) studies not published in Chinese or English; (3) duplicate publications, review articles, conference abstracts, and sources with incomplete information; (4) studies that conducted self-comparisons of outcomes before and after IMST, studies in which isokinetic muscle instruments were not used as training machines for cyclical muscle exercise, and studies with IMST alone and no control group; and (5) researchers exclude patients who have recently received knee injections or surgeries, those with other lower extremity pain conditions, individuals suffering from cognitive or psychiatric disorders, individuals diagnosed with severe knee osteoarthritis, and those battling critical illnesses from the study.

Search and screening process for literature selection

The search and screening process for literature selection was based on the guidelines of Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) [35]. The search period spanned from the inception of each database to September 31, 2024. The search strategy included the following keywords: “Knee Osteoarthritides,” “Knee Osteoarthritis,” “Osteoarthritis of the Knee,” “knee OA,” “KOA,” “isokinetic muscle strength,” “isokinetic exercise,” “isokinetic training,” “isokinetic muscle strength training,” “isokinetic motion,” “isokinetic technology,” “isokinetic movement,” “isokinetic,” “isokinetic muscle strength test,” and “muscle strength training.” We systematically combined the keywords with unrestricted terms to ensure a comprehensive search. Two reviewers independently searched for and selected relevant articles from eight databases: PubMed (“Osteoarthritis, Knee”[Majr]) AND (“Muscle Strength”[Majr]), Ovid MEDLINE (1946 to present, Osteoarthritis, Knee/ and isokinetic muscle strength.mp. or isokinetic exercise.mp. or isokinetic training.mp. or isokinetic muscle strength training.mp. or isokinetic technology.mp. or isokinetic movement.mp. or isokinetic.mp. or isokinetic.mp. or isokinetic muscle strength test.mp. or muscle strength training.mp.), Cochrane Library ((Knee Osteoarthritides or Knee Osteoarthritis or Osteoarthritis of Knee or Osteoarthritis of the Knee or knee OA or KOA):ti,ab,kw (Word variations have been searched) AND (isokinetic muscle strength or isokinetic exercise or isokinetic training or isokinetic muscle strength training or isokinetic motion or isokinetic technology or isokinetic movement or isokinetic or isokinetic muscle strength test or muscle strength training):ti,ab,kw (Word variations have been searched)), Embase (('knee osteoarthritides':ab,ti OR 'knee osteoarthritis':ab,ti OR 'osteoarthritis of knee':ab,ti OR 'osteoarthritis of the knee':ab,ti OR 'knee oa':ab,ti OR koa:ab,ti) AND ('isokinetic muscle strength':ab,ti OR 'isokinetic exercise':ab,ti OR 'isokinetic training':ab,ti OR 'isokinetic muscle strength training':ab,ti OR 'isokinetic motion':ab,ti OR 'isokinetic technology':ab,ti OR 'isokinetic movement':ab,ti OR isokinetic:ab,ti OR 'isokinetic muscle strength test':ab,ti OR 'muscle strength training':ab,ti)), and China Biomedical Literature Database (CBM, ((KOA) OR (knee OA) OR (Osteoarthritis of the Knee) OR (Osteoarthritis of Knee) OR (Knee Osteoarthritis) OR (Knee Osteoarthritides)) AND ((muscle strength training) OR (isokinetic muscle strength test) OR (isokinetic) OR (isokinetic movement) OR (isokinetic technology) OR (isokinetic motion) OR (isokinetic muscle strength training) OR (isokinetic training) OR (isokinetic exercise) OR (isokinetic muscle strength))), China National Knowledge Infrastructure (CNKI, Keyword = “knee osteoarthritis” OR Keyword = “Osteoarthritis of the Knee” AND Keyword = “isokinetic muscle strength” OR Keyword = “isokinetic muscle strength training”), Similar strategies were applied to WanFang Data, Wipro Chinese Scientific and Technical Journal Full Text Database (VIP Database). Two reviewers extracted information based on inclusion and exclusion criteria. If the data were unclear, they would review the full article. If they disagreed on including an article, they would try to resolve the disagreement through negotiation. A third reviewer would decide if they could not reach a consensus after discussion.

Data extraction protocol and assessment of literature quality

Two reviewers independently reviewed the relevant literature and systematically extracted the following data: name of the first author; year of publication; sample size; details of intervention measures for both treatment and control groups; duration of intervention; and outcome indicators such as PT, VAS, LKSS, and LI. According to the Cochrane Handbook [36]. We utilized the Risk of Bias 2 (RoB 2) tool (https://methods.cochrane.org/risk-bias-2, The Cochrane Collaboration, London, UK) software to assess the risk of bias in the literature, encompassing the following five items: (1) Bias in the randomization process; (2) Bias due to deviation from the intended interventions; (3) Bias due to missing outcome data; (4) Bias in outcome measurement; and (5) Bias due to selective reporting of results. Determine whether the article’s risk of bias is low, some concerns, or high risk by answering the appropriate questions: yes (Y), probably yes (PY), probably no (PN), no (N), and no information (NI).

Statistical analysis

We employed RevMan 5.4 (https://revman.cochrane.org/info, The Cochrane Collaboration, London, UK) and Stata 17 (Stata Corp, College Station, TX, USA) to analyze data. The quality of the evidence was determined using GRADEpro (https://www.gradepro.org/, Evidence Prime, Canada). The presence of heterogeneity, based on the magnitude of the P value and the extent of heterogeneity among the study results, was assessed using the I² statistic. When I² was ≤ 50% or the P value was > 0.10, the included studies exhibited good homogeneity, and the meta-analysis employed a fixed-effects model was employed in the meta-analysis. Conversely, when I² exceeded 50%, or the P value was ≤ 0.10, significant heterogeneity existed among the studies, prompting using a random-effects model in the meta-analysis. Additionally, we utilized subgroup and sensitivity analyses to explore clinical and statistical heterogeneity sources. We also assessed the presence of publication bias in the included articles by examining the distribution pattern of the “inverted funnel” plot. Finally, we assessed MCID (Minimal Clinically Important Difference, MCID = SD*0.5, 0.5 for medium effect). Data were reported as relative risk (RR) or mean difference (MD), with a 95% confidence interval (95% CI). The level of statistical significance was set at P < 0.05. The study protocol was registered on PROSPERO (CRD42024607528).

Results

Literature screening

Initially, 3388 relevant articles were retrieved, comprising 1522 Chinese and 1866 English articles. After screening the titles, 185 articles (154 in Chinese and 31 in English) were selected for further review. Subsequent screening of the abstracts narrowed the selection to 52 articles, of which 42 were in Chinese, and 10 were in English. After a thorough review of the full texts, 33 articles (26 Chinese and 7 English) were included in the study, encompassing 2860 participants (1429 in the experimental group and 1431 in the control group) with comparable baseline characteristics. Figure 1 illustrates the literature screening process.

Fig. 1

Flowchart of the literature retrieval and screening process

Characteristics and assessment of quality

Table 1 presents the key characteristics of the 33 included studies.

Table 1 Characteristics of IMST randomized trials by intervention (n = 33)

The review comprised 33 papers (46 investigations) [4, 37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68]. The risk assessment of the randomization process and bias due to deviation from the intended intervention has been conducted and is considered low risk. Regarding the bias due to missing outcome data, with only 10% of the data missing, this was regarded as low risk. However, the risk was high for bias in outcome measurement, where more than 50% of the data were affected. As for the bias due to selective reporting of results, some concerns risk, with more than 50% of the outcomes potentially subject to selective reporting. The overall risk assessment for the deviation was low and some concerns. Figures 2 and 3 illustrate the distribution of each type of bias across the included studies.

Fig. 2

Risk of bias graph. Review of author judgments on each risk of bias item presented as percentages across all included studies

Fig. 3

Risk of bias summary. Review of author judgments about each risk of bias item in each included study

Meta-analysis results

Knee Flex/Ext PT

Knee Flex/Ext PT at an angular velocity of 60°/second

Among the studies included in this review, 20 publications (28 studies) [4, 40,41,42, 44,45,46, 48,49,50,51,52,53,54, 56,57,58,59,60, 63] reported outcomes for Flex PT at an angular velocity of 60°/s (N = 1,724). The results showed that at an angular velocity of 60°/second, IMST significantly improved the flexor PT value in patients with KOA (MD = 13.19, 95% CI = 6.44 to 19.94, Z = 3.83, P = 0.0001) (Fig. 4). In total, 21 articles (29 studies) [4, 40,41,42, 44,45,46, 48,49,50,51,52,53,54, 56,57,58,59,60,61, 63] reported the Ext PT results at an angular velocity of 60°/s (N = 1,773). The meta-analysis showed that when the angular velocity was 60°/s, IMST significantly enhanced the Ext PT value in patients with KOA (MD = 16.34, 95% CI = 11.47 to 21.22, Z = 6.57, P < 0.00001) (Fig. 5).

Fig. 4

Meta-analysis of flexor peak torque (Flex PT). Analysis of patients with knee osteoarthritis (KOA) at an angular velocity of 60°/second

Fig. 5

Meta-analysis of extensor peak torque (Ext PT). Analysis of patients with knee osteoarthritis (KOA) at an angular velocity of 60°/second

Knee Flex/Ext PT at an angular velocity of 180°/second

Among the studies on flexor/extensor muscle PT, six articles (10 studies) [40, 41, 44, 53, 58, 63] reported findings for Flex PT at an angular velocity of 180°/s (N = 624). The meta-analysis showed that when the angular velocity was 180°/second, IMST significantly improved the Flex PT value of patients with KOA (MD = 11.17, 95% CI = 2.86 to 19.48, Z = 2.63, P = 0.008) (Fig. 6). A collection of seven articles (11 studies) [40, 41, 44, 53, 58, 61, 63] reported the Ext PT results at an angular velocity of 180°/s (N = 673). The meta-analysis revealed that IMST improved the Ext PT in patients with KOA at an angular velocity of 180°/second (MD = 12.62, 95% CI = 3.49 to 21.75, Z = 2.71, P = 0.007) (Fig. 7).

Fig. 6

Meta-analysis of flexor peak torque (Flex PT). Analysis of patients with knee osteoarthritis (KOA) at an angular velocity of 180°/second

Fig. 7

Meta-analysis of extensor peak torque (Ext PT). Analysis of patients with knee osteoarthritis (KOA) at an angular velocity of 180°/second

Knee Flex/Ext TW

Data from 7 articles (9 studies) [4, 42, 46, 48, 49, 60, 67] (N = 602) showed the Flex TW. The meta-analysis demonstrated that the experimental group had higher flexor TW values than the control group, indicating that IMST could improve the flexor TW in patients with KOA (MD = 79.77, 95% CI = 49.43 to 110.10, Z = 5.15, P < 0.0001) (Fig. 8). Within this review, six articles (seven studies) [4, 42, 46, 48, 49, 60] discussed Ext TW results (N = 438). The meta-analysis findings indicated that IMST had the potential to improve the Ext TW in patients with KOA (MD = 86.27, 95% CI = 58.40 to 114.15, Z = 6.07, P < 0.00001) (Fig. 9).

Fig. 8

Meta-analysis of flexor total work (Flex TW) in patients with knee osteoarthritis (KOA)

Fig. 9

Meta-analysis of extensor total work (Ext TW) in patients with knee osteoarthritis (KOA)

Knee Flex/Ext MRTW

Four articles (six studies) [42, 56, 57, 60] explored the effect of IMST on Flex MRTW in patients with KOA (N = 351). The meta-analysis showed that IMST significantly enhanced the Flex MRTW in patients with KOA (MD = 9.38, 95% CI = 3.20 to 15.56, Z = 2.97, P = 0.003) (Fig. 10). The existing literature on Ext MRTW is limited to four articles (six studies) [42, 56, 57, 60] (N = 351). According to the results of the meta-analysis, IMST could improve the Ext MRTW in patients with KOA (MD = 15.52, 95% CI = 8.96 to 22.08, Z = 4.64, P < 0.0001) (Fig. 11).

Fig. 10

Meta-analysis of flexor max rep total work (Flex MRTW) in patients with knee osteoarthritis (KOA)

Fig. 11

Meta-analysis of extensor max rep total work (Ext MRTW) in patients with knee osteoarthritis (KOA)

Knee Flex/Ext AP

Six articles (10 studies) [4, 42, 44, 56, 57, 67] showed data on Flex AP (N = 639). The meta-analysis showed that IMST could improve Flex AP in patients with KOA (MD = 8.66, 95% CI = 0.70 to 16.61, Z = 2.13, P = 0.03) (Fig. 12). Five articles (eight studies) [4, 42, 44, 56, 57] analyzed Ext AP (N = 475). The results indicated that IMST effectively improved Ext AP (MD = 7.27, 95% CI = 3.30 to 11.23, Z = 3.59, P = 0.0003) (Fig. 13).

Fig. 12

Meta-analysis of flexor average power (Flex AP) in patients with knee osteoarthritis (KOA)

Fig. 13

Meta-analysis of extensor average power (Ext AP) in patients with knee osteoarthritis (KOA)

VAS scores

In total, 24 papers (34 studies) [4, 37,38,39,40,41, 43, 45, 47, 49,50,51, 53,54,55, 58, 60,61,62,63,64, 66,67,68] evaluated the impact of the VAS scores (N = 2,117). The analysis revealed that IMST reduced the VAS scores in patients with KOA (MD = −0.70, 95% CI = − 0.92 to − 0.49, Z = 6.44, P < 0.00001) (Fig. 14).

Fig. 14

Meta-analysis of visual analog scale (VAS) scores in patients with knee osteoarthritis (KOA)

LI scores

In the assessment of the LI scores, six articles (nine studies) [37,38,39,40,41, 49] reported relevant findings (N = 481). The meta-analysis results revealed that IMST significantly decreased the LI scores in patients with KOA (MD = − 1.24, 95% CI = − 1.65 to − 0.83, Z = 5.95, P < 0.00001) (Fig. 15).

Fig. 15

Meta-analysis of Lequesne Index (LI) scores in patients with knee osteoarthritis (KOA)

WOMAC scores

Nine articles (15 studies) [4, 45, 47, 53, 56, 57, 62, 67, 68] compared the effects of IMST on WOMAC scores (N = 871). The meta-analysis showed that the IMST group exhibited a lower WOMAC score, indicating the superior efficacy of this therapy in alleviating symptoms (MD = − 6.05, 95% CI = − 10.37 to − 1.73, Z = 2.74, P = 0.006) (Fig. 16).

Fig. 16

Meta-analysis of Western Ontario and McMaster University Osteoarthritis Index (WOMAC) scores in patients with knee osteoarthritis (KOA)

LKSS scores

Sixteen publications (19 studies) [42, 43, 46, 48, 50,51,52, 54, 55, 58,59,60, 63,64,65,66] reported LKSS scores for patients with KOA after IMST (N = 1395). The meta-analysis displayed that IMST training had a more substantial improvement effect on LKSS scores in patients with KOA than other treatment methods (MD = 7.90, 95% CI = 5.91 to 9.89, Z = 7.79, P < 0.00001) (Fig. 17).

Fig. 17

Meta-analysis of Lysholm knee scoring scale (LKSS) scores in patients with knee osteoarthritis (KOA)

Knee Flex/Ext ROM

In a review of 12 articles (13 studies) [38, 39, 41, 45, 48, 51, 54, 55, 58, 62, 65, 66], the researchers associated findings with knee flexion mobility (N = 873). The meta-analysis showed that compared with the control group, IMST could improve knee Flex ROM in patients with KOA (MD = 10.62, 95% CI = 7.94 to 13.30, Z = 7.76, P < 0.00001) (Fig. 18). Four articles (five studies) [45, 58, 65, 66] involved outcome measures of knee extension ROM (N = 326). The meta-analysis indicated that IMST did not significantly improve knee extension ROM in patients with KOA (MD = 0.13, 95% CI = − 3.01 to 3.28, Z = 0.08, P = 0.93) (Fig. 19).

Fig. 18

Meta-analysis of flexor range of motion (Flex ROM) in patients with knee osteoarthritis (KOA)

Fig. 19

Meta-analysis of extensor range of motion (Ext ROM) in patients with knee osteoarthritis (KOA)

6MWT

Among the included studies, two articles (three studies) [47, 52] conducted a 6MWT analysis of patients with KOA (N = 131). The meta-analysis suggested that IMST may increase 6MWT in patients with KOA; however, the observed difference did not reach statistical significance (MD = 11.78, 95% CI = − 43.58 to 67.14, Z = 0.42, P = 0.68) (Fig. 20).

Fig. 20

Meta-analysis of 6-Minute Walk Test (6MWT) in patients with knee osteoarthritis (KOA)

In brief, a meta-analysis of the above ten outcome indicators was summarized in Table 2.

Table 2 Summary of Meta-analysis results

GRADE and MCID assessments

The GRADE quality of the study outcome indicators was poor, ranging from “very low” to “low” to “moderate” levels, mainly due to original trials with high inter-study heterogeneity and imprecision of results (Fig. 22). The MCID assessment was consistent with the results of the Meta-analysis (Table 3).

Table 3 Summary of MCID assessment of outcome indicators

Subgroup analysis

Following our subgroup analysis by total treatment frequency, the results indicated no significant difference in the Ext PT at an angular velocity of 180°/s, Flex TW, and WOMAC scores (P > 0.05) when the total frequency of IMST treatment was > 20 times. The remaining indicators maintained their consistency: the Flex/Ext PT at an angular velocity of 60°/s, the Flex PT at an angular velocity of 180°/s, Ext TW, Flex/Ext AP, VAS, WOMAC, LKSS scores, and Flex/Ext ROM of the knee joint (Table 4). The treatment effect increased with the frequency of intervention.

Table 4 Subgroup analysis: total treatment frequency

The findings revealed that concentric IMST did not significantly influence Ext PT at an angular velocity of 180°/s, the Flex MRTW, or the WOMAC scores (P > 0.05). The remaining indicators maintained their consistency: the Flex/Ext PT at an angular velocity of 60°/s, the Flex PT at an angular velocity of 180°/s, Flex TW, Flex AP, VAS, WOMAC, LKSS scores, and Flex/Ext ROM of the knee joint (Table 3). When the intervention mode was eccentric IMST, no significant differences were observed in Flex/Ext PT at an angular velocity of 60°/s or Flex PT at an angular velocity of 180°/s (P > 0.05). The remaining indicators remained consistent before and after the subgroup analysis, including MRTW, AP, VAS, LI, WOMAC, LKSS scores, and knee Flex/Ext ROM of the knee joint (P < 0.05) (Table 5). We also observed that concentric, eccentric, and combined concentric-eccentric IMST interventions had therapeutic effects on KOA. Notably, the distinct intervention modalities exhibited varying impacts on the outcome indicators.

Table 5 Subgroup analysis: intervention methods

In conclusion, the subgroup-analysis results indicated that the above factors did not contribute to the high heterogeneity observed in the studies following stratification by total treatment frequency and intervention methods.

Sensitivity analysis

Using a sensitivity analysis (Table 6), we found that only knee Ext ROM values changed by excluding the outcome metrics in the literature on a sequential article-by-article method, suggesting that IMST increased knee extension mobility. The heterogeneity tests for the LI scores, knee Ext ROM, Flex/Ext TW, and 6MWT were all I² < 50%, which was justified using a fixed-effects model for analysis. Excluding individual articles did not affect the results of the study.

Table 6 Sensitivity analysis: different outcome indicators

Publication bias

Given the extensive literature and variety of outcome indicators in this study, a meticulous approach was necessary to ascertain the precise impact of IMST on patients with KOA. Outcome indicators with ≥ 10 publications were examined for potential publication bias using a funnel plot and Egger’s test. The findings revealed that knee flexion ROM (Egger’s test, P = 0.643) and WOMAC score (Egger’s test, P = 0.069) were the only indicators without publication bias. In contrast, other indicators, such as the Flex/Ext PT at 60°/s and 180°/second angular velocities, VAS, LKSS, LI scores, and Flex AP all exhibited significant publication bias (Egger’s test, P < 0.05) (Fig. 21).

Fig. 21

Funnel plot analysis with 95% confidence intervals. Selecting outcome indicators from literature with ≥ 10 articles

Discussion

Exploring effective treatment modalities for KOA has emerged as a pivotal area of research. Exercise therapy is one of the crucial modalities of non-pharmacological treatment [69,70,71]. Research [72] has shown that well-tailored exercise programs and personalized exercise therapies offer significant benefits such as alleviating pain, enhancing joint mobility, and slowing disease progression. Moreover, these exercise therapies are safe, non-invasive, and have no side effects. In our comprehensive study, we systematically analyzed the bias condition of the included literature. Our evaluation of GRADE and MCID for the ten outcome indicators showed that the meta-analysis results might be used confidently in clinical practice. We conducted a subgroup analysis of different intervention methods for KOA, including concentric, eccentric, and combined concentric-eccentric IMST. Our research also encompassed an analysis of the frequency of these interventions to understand their impact on patient outcomes better (Fig. 22).

Fig. 22

Summary of findings table: Quality of evidence in included literature assessed by GRADE

Effects on muscle strength

In recent years, a substantial body of research [55, 57, 65] has demonstrated that the regular function of the knee joint is sustained and regulated by the strength of the surrounding muscles. The onset and progression of KOA are closely associated with muscle dysfunction. Patients frequently experience reduced joint mobility, characterized by limited knee flexion and extension, a narrowed ROM, and diminished activity, primarily due to pain. This pain-induced immobility leads to disuse muscle atrophy, which degrades muscle elasticity and strength and compromises joint stability [73]. The reduced stability results in abnormal forces on the joint surfaces, aggravating the condition of KOA and perpetuating a detrimental cycle. A recent study [53] has indicated that integrating IMST with Bushen Qiangjin Capsule (a TCM tonic for strengthening the body) significantly enhances the therapeutic efficacy for patients with KOA, outperforming the capsule alone. This combined approach helps to alleviate knee joint pain and improve joint function. Concurrently, combining IMST with acupuncture, compared with acupuncture alone, could more effectively improve clinical symptoms, restore muscle strength, and enhance joint stability in the muscles surrounding the knee joint in older patients with KOA. Furthermore, this combined therapy significantly reduces interleukin (IL)−6, IL-18, and tumor necrosis factor-α serum levels [74]. Meanwhile, the muscle strength of KOA patients should be comprehensively evaluated, including PT, TW, AP, and MRTW indexes [75].

We revealed that IMST significantly improves clinical knee flexion and extension indicators, such as PT, TW, AP, and MRTW, in patients with KOA and that increases in these indicators represented enhanced knee muscle strength and joint stability. During training, the equipment applies external resistance to ensure that the lower limb flexor and extensor muscles move uniformly at a set speed, whether in centripetal or centrifugal motion. This process increases muscle tension and contraction while maintaining dynamic balance, fostering appropriate muscle exercise.

Effects on ROM

Patients with KOA often experience pain and swelling, which can restrict joint activity, lead to reduced muscle adaptability, and increase tissue fibrosis. Furthermore, chronic absence from exercise can cause the transformation of loose connective tissue in the muscles into a dense form, creating a rigid muscle structure that restricts knee joint motion and contributes to joint dysfunction [76]. Sensitivity analysis revealed that IMST effectively improved the ROM scores of the extensor and flexor muscles. IMST restricts the knee joint’s ROM by setting an angular velocity, thereby preventing secondary injuries that might arise from training. Maintaining a consistent exercise speed, the IMST effectively strengthened the knee flexor and extensor muscles and gradually increased ROM. The quadriceps and hamstring muscles contract antagonistically throughout the training process, enhancing the adhesion of the surrounding muscles, ligaments, tissues, and joint capsule. This training also increases synovial fluid secretion, lubricates the joints, and expands ROM. This exercise extends the elasticity of the joint tissue fibers, preventing rapid retraction. Later, the doctor slowly increases the intensity of the patient’s training to improve overall muscle strength, joint mobility, and knee stability, creating a virtuous cycle.

Effects on knee dysfunction

Articular cartilage degeneration is the leading cause of joint pain and functional limitations in patients with KOA [77]. Pain is a worldwide problem that burdens national health systems. Physical activity directly affects the central nervous system (CNS), which can alter pain experience and cognitive processing. Therefore, physical activity is a potential tool for preventing and treating pain-related disorders [78, 79]. Tang et al. [58] demonstrated that IMST could break the vicious cycle commonly observed in patients with KOA, characterized by “decreased muscle strength leading to joint pain and deformation, which causes joint instability and reduced joint activity, resulting in further muscle weakness.” IMST alleviates pain associated with cartilage degeneration, reduces osteophyte formation, and enhances joint repair.

Furthermore, it strengthens the lower limb muscles, improves balance and proprioception, aids in stair climbing, maintains joint stability, and lowers the risk of KOA recurrence [80]. According to Bahşi et al. [68], IMST could improve blood circulation within the knee joint, promote the absorption of joint effusion, and reduce abnormal proliferation of cartilage, thereby reducing edema and pain. Additionally, IMST thickens the femoral and cartilaginous structures, enhances knee joint stability, and increases muscle strength. An article evaluating Orthopedic Manual Therapy (OMT) noted that OMT alone improves mechanical nociceptive hypersensitivity with moderate-quality evidence and improves temporal summation and conditioned pain modulation with low-quality evidence. However, the effects were immediate and short-term [81]. It is worth noting whether the impact of IMST on pain improvement in KOA patients is temporary or long-term and whether further investigation is needed. This meta-analysis further suggested that IMST significantly improves the LKSS scores and reduces the VAS, LI, and WOMAC scores in patients with KOA. Most studies agreed that joint pain, swelling, stiffness, and stability improved in patients after IMST-assisted therapy [37, 67, 68]. However, the results of the 6MWT analysis did not show an improvement in walking speed in KOA patients, which is inconsistent with that reported by He et al. [52], probably due to the limited number of studies included in the analysis.

Consequently, the main principles of IMST treatment can be summarized as follows: 1. Joint nourishment and cartilage health: IMST bolsters neural and biochemical regulation through regular short-arc exercises, promoting synovial fluid secretion for joint cartilage nourishment, which aids in knee function recovery and fall risk reduction [16, 64, 82]. 2. Pain relief and quality of life: IMST plays a dual role in managing KOA by providing external resistance for controlled muscle movement, maintaining a dynamic balance to prevent secondary injuries from excessive joint activity, and enhancing muscle strength and joint mobility. This approach not only expands the range of motion but also significantly reduces pain and improves the overall quality of life for KOA patients through optimal movement and posture control [52, 83, 84]. 3. Personalized treatment approach: personalized IMST techniques are crucial for scientifically managing KOA progression effectively [4]. Furthermore, previous studies [85, 86] have indicated that the escalation of blood pressure and heart rate during isokinetic exercise occurs more gradually than during isometric exercise. This characteristic makes isokinetic exercise particularly advantageous for patients with KOA and cardiovascular and pulmonary dysfunctions. Therefore, IMST may improve cardiopulmonary function in patients with KOA at high altitudes.

Study limitations

The findings of this meta-analysis are subject to potential biases and confounding factors common in the studies. Our study also had several limitations.

First, the quality of the outcome evidence analyzed in the studies was poor, with ratings varying from “very low” to “low” to “moderate.” This quality discrepancy was primarily due to significant inter-study heterogeneity and the imprecision of results. Egger’s test assessments pointed to publication bias within the compiled literature. Furthermore, the limited sample sizes in individual studies resulted in unstable precision of outcome measures, casting doubt on the conclusions’ reliability. To bolster the evidence base and enhance the validity of future findings, we must conduct large-scale clinical trials. These trials will address the current limitations and provide a more robust foundation for understanding the efficacy of interventions.

Second, most studies included in this analysis did not employ allocation concealment or double-blind methods, which are critical to reducing bias. Upon utilizing the ROB 2 tool for analysis, this study revealed that most articles carried a low risk of bias, but some were concerned about bias. This assessment was influenced by including a pain rating scale in KOA evaluation, which inherently introduces subjectivity into the outcomes. As a result, this subjectivity caused the overall bias to present as low and some worrisome bias. However, it is essential to note that some degree of bias in outcome measures is inevitable in the context of KOA assessments.

Third, the literature in this analysis lacked unified standards for intervention modalities, frequencies, and treatment durations for the IMST group. The control groups varied in design, leading to significant clinical heterogeneity among studies. Due to the lack of standardized treatment protocols and intervention models for IMST in the literature, subgroup and sensitivity analyses were conducted. The findings revealed statistically significant differences in specific outcome indicators before and after treatment. Moreover, our study indicated that higher-frequency training sessions correlate with more substantial muscle strength and function improvements. However, the optimal frequency varies based on each patient’s initial condition and their tolerance to the exercise regimen. The results underscored the necessity of individualized treatment plans to maximize the benefits of IMST while minimizing the risk of overexertion and potential injury. Grouping by different intervention modes revealed that concentric IMST may be more effective than eccentric IMST in enhancing muscle strength, which is consistent with the findings of Wu et al. [87]. Moreover, eccentric IMST significantly improved the patient’s balance and walking ability, consistent with the findings of earlier studies [38, 88, 89]. Although eccentric muscle action is physiologically part of regular activity, eccentric testing is uncommon and may lead to inhibition of maximal contraction in unaccustomed individuals [90]. Therefore, physicians must recognize the value of individualized rehabilitation programs for the varying needs of their patients. Further studies are required to validate these observations. Through this multifaceted examination, we elucidated the effectiveness of IMST in managing patients with KOA.

Fourth, during recurrent KOA, muscle function restoration is gradual; however, most of the included studies lack long-term follow-up data. Therefore, there is a shortage of long-term efficacy data for IMST and assessment of its impact on the recurrence rate of KOA. Thus, it is imperative to conduct long-term follow-ups on the rehabilitation status of patients with KOA. Lastly, a funnel plot analysis was performed on outcome indicators with ≥ 10 included articles, revealing the presence of publication bias within the literature. In the future, we should systematically explore the optimal duration or intensity of IMST to ensure optimal patient outcomes and explore innovative approaches to improve the effectiveness and accessibility of therapies.

Conclusions

Our comprehensive analysis of IMST interventions has provided valuable insights into the complex relationship between exercise type, frequency, and outcomes in KOA patients. These results contribute to the growing body of evidence supporting using IMST as a non-pharmacological therapy for KOA and can inform future clinical guidelines for managing this condition. Future research should prioritize large-scale, high-quality RCTs to substantiate the conclusions of this review, given the limited number and variable quality of existing studies. Notably, IMST has been shown to improve the onset and progression of KOA in lower-altitude regions, and its safety and substantial benefits on the quality of life of patients with KOA make it a promising therapy worthy of broader application, especially in high-altitude areas.