Exploring dextrose prolotherapy in rotator cuff disorders: A systematic review and meta-analysis

Abstract

Background and Aims:

Rotator cuff (RC) disorders have a varied range of treatment. Multiple interventional, non-surgical treatments are often opted for by patients. This study evaluates the effectiveness of dextrose prolotherapy in treating RC pathologies.

Methods:

A comprehensive review of databases (PubMed, Google Scholar, Embase, Scopus, and Cochrane databases) from 2000 to June 2025 identified 13 relevant studies involving 936 patients. The randomised clinical trials, which compared dextrose prolotherapy with interventions such as platelet-rich plasma, steroids, physiotherapy, or placebos, were included in the systematic review. A risk-of-bias analysis was conducted using the Risk of Bias Visualisation Tool. It indicated that seven studies had low bias. Studies with a high risk of bias were excluded from the meta-analysis. The protocol was registered beforehand (PROSPERO ID: CRD42024520747).

Results:

Review Manager software was used to analyse the data and generate the plots. The analysis showed that prolotherapy significantly reduced pain (MD = −0.76; 95% CI: −1.26, −0.27; P = 0.003), improved functional outcomes (shoulder pain and disability index (MD = −8.58; 95% confidence interval (CI): −13.00, −4.17; P = 0.0001)), and enhanced ultrasonography features. No major adverse effects were reported, indicating that the treatment is safe.

Conclusions:

This systematic review and meta-analysis suggests that dextrose prolotherapy may be a feasible and effective alternative for RC disorders, and it can be considered for patients with limited treatment options. However, to reach the required information size, further trials are required. Uniformity in protocols, intervention strategies, and outcome reporting is necessary for longer follow-up periods to facilitate more effective evidence synthesis. This study received no external funding.

INTRODUCTION

The rotator cuff (RC) is essential for stabilising the shoulder joint while facilitating an extensive range of motion (ROM).[1] RC tendinopathy affects approximately 0.3%–5.5% of the population, with annual prevalence rates ranging from 0.5% to 7.4%.[2] This condition results from both extrinsic and intrinsic factors, necessitating customised treatment strategies based on the severity, progression, and symptoms of the disease.[3] Available treatment options are diverse, encompassing physical therapy to surgical interventions.

Corticosteroid injections, frequently administered alongside local anaesthetics, are prevalent yet may adversely affect RC tissue, as research indicates a heightened risk of full-thickness tears.[4,5] Local anaesthetics may diminish tenocyte populations and disrupt tendon collagen architecture.[6] The application of platelet-rich plasma (PRP) for RC injuries remains a contentious issue, with inadequate evidence to substantiate its effectiveness.[7,8]

Prolotherapy is a minimally invasive procedure with regenerative capabilities. The objective is to activate the body’s intrinsic healing mechanisms by promoting the repair and regeneration of normal tissues. It is thought to facilitate healing by inducing a localised inflammatory response and fibroblast activation.[9] Dextrose prolotherapy thus offers a pharmacologically advantageous healing approach by promoting natural tissue repair without relying on synthetic pharmacologic receptor-based modulation. It is also not immunosuppressive, unlike corticosteroids, which can impact long-term healing.[5] This cost-effective, targeted intervention, with no systemic side effects, can be an alternative or adjunct to other available treatment options. This systematic review aimed to evaluate randomised controlled trials regarding the efficacy of dextrose prolotherapy in treating RC lesions, focusing on its ability to reduce pain and disability, and improve functionality, compared to placebo or control treatments, platelet-rich plasma, corticosteroids, and physiotherapy. The review focused on the following questions based on PICOS elements: Participants (P): In adults diagnosed with RC lesions; Interventions (I): does dextrose prolotherapy; Comparisons (C): In comparison to placebo or control treatments, platelet-rich plasma (PRP), corticosteroid injections, and physiotherapy; Outcomes (O): Reduces pain i.e., visual analogue scale (VAS) or numerical rating scale (NRS) and disability, shoulder pain and disability index (SPADI) and improves ROM and ultrasound-based morphological features of the RC; Study design (S): Randomised controlled trials.

METHODS

We performed a systematic review and meta-analysis in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The review protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO) database (ID: CRD42024520747).

Identification and selection of research studies

We searched the PubMed, Google Scholar, Embase, Scopus, and Cochrane databases from 2000 to 31 October 2024, and the search was updated during the review process until June 2025. The search query employed was: [(dextrose prolotherapy) OR (prolotherapy)] AND [(rotator cuff) OR (rotator cuff tendinopathy)] [Supplementary Table 1]. The first author conducted the preliminary search and removed all duplicates using Covidence systematic review software (Veritas Health Innovation, Melbourne, Australia).[10] The first and second authors assessed eligibility and reviewed the titles and abstracts, and subsequently the complete texts of qualifying articles. The disputes were settled through dialogue, and the third author rendered the final decisions. A manual review of the references from the included publications was conducted to identify any studies that may have been overlooked. The aforementioned databases were re-examined on 1 June 2025 to ascertain if any new findings had been incorporated into the literature. No novel studies were identified during this period.

Supplementary Table 1.

Search Strategy

PubMed   #1 “Dextroe Prolotherapy”[MeSH Terms]   #2 ((Dextrose Prolotherapy[MeSH Terms]) OR (prolotherapy[Title/Abstract]))   #3 #1 OR #2   #4 ”Rotator cuff tendinopathy”[Mesh]   #5 ((Rotator cuff tendinopathy[MeSH]) OR (Rotator cuff disorders[Title/Abstract]))   #6 #4 OR #5   #7 #3 AND #6 Scopus:   #1 TITLE-ABS-KEY (prolotherapy)   #2 TITLE-ABS-KEY (prolotherapy) OR KEY (dextrose AND prolotherapy)   #3 #1 OR #2   #4 TITLE-ABS-KEY (rotator AND cuff AND tendinopathy)   #5 KEY (rotator AND cuff AND tendinopathy) OR TITLE-ABS-KEY (rotator AND cuff AND disorders) OR KEY (rotator AND cuff AND disorders))   #6 #4 OR #5   #7 #3 AND # Embase   #1 ‘dextrose prolotherapy’:ti, ab, kw   #2 ‘prolotherapy: ti, ab, kw’   #3 #1 OR #2   #4 ‘rotator cuff tendinopathy’:ti, ab, kw   #5 ‘rotator cuff disorders’:ti, ab, kw   #6 #4 OR #5   #7 #3 AND #6 Cochrane Library:   #1 Dextrose Prolotherapy [Title/abstract/keyword]   #2 Prolotherapy [Keyword]   #3 #1 OR #2   #4 Rotator Cuff Tendinopathy   #5 Rotator Cuff Disorders   #6 #4 OR #5   #7 #3 AND #6 Google Scholar: (Dextrose Prolotherapy OR Prolotherapy) AND (Rotator cuff Tendinopathy OR Rotator Cuff Disorders) Search results from first five pages were only included

Inclusion criteria: Study design: randomised controlled trials; Population: adult patients having RC pathologies, including tendinopathies, lesions, disease, partial tears, and bursitis (e.g., supraspinatus tendinopathy/tendinosis, subacromial or subdeltoid bursitis); Intervention: dextrose prolotherapy; Comparison: placebo or any other interventional or non-interventional treatment modalities (e.g., corticosteroid injections, PRP, physiotherapy).

Exclusion criteria: any studies other than randomised controlled trials; full-thickness tears, post-surgical cases, non-human trials, and shoulder disorders not pertaining to the RC or dextrose prolotherapy, and languages other than English.

Outcomes

The primary outcomes were the VAS/NRS and SPADI. Secondary outcomes encompassed ROM and ultrasonic morphological changes.

Essential information from each study was recorded in an electronic spreadsheet, including the primary author, publication date, funding source, patient characteristics (pathology, duration of symptoms, method of confirmation of diagnosis), intervention (concentrations and volume of dextrose used with number of participants in the group), comparators (steroid, PRP, control/placebo, physiotherapy with number of participants), number of sessions received, outcomes assessed and assessment interval, and follow-up duration. Adverse occurrences were also seen. Comparisons were conducted between dextrose prolotherapy and alternative therapies (control/placebo, PRP, and corticosteroids) at time intervals prevalent across the studies (2–4 weeks, 1 month, and 3 months). Missing data were resolved by contacting the associated authors via email; studies without a response were excluded from the analysis.

Evaluation of bias risk and methodological quality

The primary and secondary authors independently evaluated the methodological quality and risk of bias of each trial, adhering to the guidelines of the Cochrane Handbook. The ‘robvis’ tool was employed to assess bias risk across seven domains: generation of random sequence, concealment of allocation, subject and investigator blinding, outcome blinding, incomplete data, selective reporting, and additional bias.[11] The Cohen’s kappa statistic was utilised to evaluate inter-rater agreement.[12] The evaluation findings revealed a κ value of 0.84, signifying a substantial level of agreement among the assessors. The Grading of Recommendations Assessment, Development, and Evaluation (GRADE) tool was utilised to evaluate the quality of evidence for outcomes.[13]

Statistical analysis

Data were examined using Review Manager software (RevMan version 5.4).[14] Effect sizes for continuous outcomes were assessed using mean differences (MD). An I² value exceeding 50% signifies heterogeneity, prompting the utilisation of random-effects analysis to accommodate variations in interventions and study designs. A post-hoc subgroup analysis was performed to investigate notable heterogeneity. The threshold for statistical significance was established at P < 0.05. Sensitivity analysis was conducted using the leave-one-out method to assess robustness. A post-hoc trial sequential analysis (TSA) was undertaken to assess the cumulative impact of the required information size with a type I error rate of 5% and power levels of 80% and 90% concurrently for the primary outcomes by utilising TSA software [The Copenhagen Trial Unit, Centre for Clinical Intervention Research, Copenhagen].[15]

RESULTS

Our investigation yielded 760 articles. Duplicates were eliminated, and the titles and abstracts of the resulting 657 studies were evaluated. As a result, 15 articles were meticulously examined. A manual reference check indicated the absence of further studies. Two studies provided data exclusively in visual formats or diagrams. Efforts to obtain the requisite information by contacting the corresponding authors were unsuccessful.[16,17] Consequently, we incorporated 13 papers into the systematic review. Figure 1 illustrates the PRISMA flowchart. The 13 included studies encompassed a total of 936 patients. The sample sizes ranged from 36 to 120. The characteristics of each study, including patient, intervention, comparator, and outcomes, are presented in Table 1.

Figure 1.

PRISMA flowchart. PRISMA = Preferred Reporting Items for Systematic Reviews and Meta-Analyses

Table 1.

Characteristics of included studies

Study (Author Year)	Sponsors	Participant Characteristics	Intervention	Comparator (s)	Outcomes Assessed (Interval)
Lin et al. 2022[18]	Ministry of Science and Technology, Taiwan.	Patients with chronic shoulder pain for>6 months with ultrasound evidence of supraspinatus tendinosis	20% dextrose prolotherapy: 5 mL (n=29) (1 session) Supraspinatus insertion site injection	NS/placebo: 5 ml (n=28) Supraspinatus insertion site injection	VAS SPADI ROM (2, 6, 12 weeks)
Lin et al. 2023[19]	Ministry of Science and Technology, Taiwan	Patients with chronic shoulder pain for>6 months, with ultrasound evidence of subacromial bursitis	20% dextrose prolotherapy: 3 mL (n=28) (1 session) Subacromial bursa injection	Steroid: 3 mL (n=26) Subacromial bursa injection	VAS SPADI ROM (2, 6, 12 weeks)
Sari et al. 2020[20]	Not Disclosed	Patients with chronic shoulder pain for>3 months with RC pathology: bursitis, tendinosis, partial tear	20% dextrose prolotherapy: 5 mL (n=30) (1 session) Subacromial bursa injection	Steroid: 5 mL (n=30) PRP: 5 mL (n=30) NS/Placebo: 5 ml (n=30) Subacromial bursa injection	VAS ASES WORC (3, 12, 24 weeks)
Mofrad et al. 2021[21]	Not sponsored	Patients of chronic RC tendinopathy with a small RC tear or tendinopathy on imaging scan	12.5% dextrose prolotherapy: 8 mL (n=32) (2 sessions- 1 week apart) Superficial injection	Physiotherapy (n=33)	SPADI (2, 12 weeks)
Chang et al. 2021[22]	Supported by Wan Fang Hospital, Taipei Medical University (grant no. 109-wf-eva-32)	Patients with chronic shoulder pain with ultrasound-confirmed subacromial bursitis	15% dextrose prolotherapy: 5 mL (n=25) (3 sessions- 2 weeks apart) Subacromial bursa injection	NS/placebo: 5 mL (n=25) Subacromial bursa injection	VAS SPADI ROM Ultrasound changes (1, 4, 12 weeks)
Amanollahi et al. 2020[23]	Not sponsored	Patients with small RC tears or tendinopathy documented on imaging scan	5% dextrose prolotherapy: 4 mL (n=30) (3 sessions- 1 week apart) Tendon insertion site injection	Steroid: 4 mL (n=30) Subacromial bursa injection	VAS ROM (4 weeks)
Bertrand et al. 2016[24]	Supported by WorkSafeBC (Workers’ Compensation Board of British Columbia; grant no: RS2010-OG07)	Patients with supraspinatus pathology confirmed by ultrasound	25% enth-dextrose: volume not specified (n=53) (3 sessions- at 1-month interval) Supraspinatus insertion site injection	Enth Saline (n=51) Supraspinatus insertion site injection Superficial dextrose injections (n=49)	NRS USPRS (3, 9 months)
Seven et al. 2017[25]	Not declared	Patients with RC lesions in MRI and confirmed by ultrasound	25% dextrose prolotherapy: 4–20 mL (n=60) (6 sessions interval not specified) Injections into tendon insertion sites and subacromial bursa	Physiotherapy (n=60)	VAS SPADI WORC ROM (3, 6, 12 weeks, 1 year)
Lin et al. 2019[26]	Ministry of Science and Technology, Taiwan	Patients with chronic shoulder pain with ultrasound evidence of supraspinatus tendinosis	20% dextrose prolotherapy: 5 mL (n=16) (1 session) Supraspinatus insertion site injection	NS/placebo: 5 mL (n=15) Supraspinatus insertion site injection	VAS SPADI ROM (2, 6 weeks)
Cole et al. 2018[27]	Not funded	Patients with symptomatic supraspinatus tendinopathy of at least 3 months in duration with ultrasound evidence	50% dextrose prolotherapy: 2 mL (n=17) (1 session) Supraspinatus insertion site injection	Steroid: 2 mL (n=19) Subacromial bursa injection	ROM (6, 12 weeks, 6 months)
Sabaah et al. 2020[28]	Not funded	Patients with unilateral RCT with symptoms of at least 3 months, and clinical confirmation	25% dextrose prolotherapy: 10 mL (n=20) (2 sessions- 2 weeks apart) sites of injection unspecified	Steroid: 5 mL (n=20) PRP: 5 mL (n=20) sites of injection unspecified	VAS WORC MSUS (3 months)
Taheri et al. 2018[29]	Not Disclosed	Patients of shoulder pain history at least 6 weeks, confirmed by physical inspection and imaging techniques	12.5%Dextrose prolotherapy: 8 ml (n=40) (1 session) sites of injection unspecified	Steroid: 8 mL (n=40) Physiotherapy (n=40) sites of injection unspecified	NRS SPADI (1, 4, 12 weeks)
Karim et al. 2020[30]	Funded by a grant from UMSC CARE Fund (PV062-2018), Faculty of Medicine, University of Malaya	Patients with supraspinatus tendinosis or partial tendon tear are seen on Ultrasound or MRI	16.7% dextrose prolotherapy: 2 mL (n=32) (1 session) Injections at the site of the lesion	PRP: 2 mL (n=32) Injections at the site of the lesion	NRS SPADI ROM (3, 6, 12 weeks, 6 months)

NS=Normal Saline; N=Number of Participants; RC=Rotator Cuff; PRP=Platelet-Rich Plasma; VAS=Visual Analogue Scale; SPADI=Shoulder Pain and Disability Index; ROM=Range of Motion; ASES score=American Shoulder and Elbow Surgeons score; WORC index=Western Ontario Rotator Cuff Index; NRS=Numerical Rating Scale; USPRS=Ultrasound Pathological Rating Scale; MSUS=Musculoskeletal Ultrasound

Quality assessment and bias risk: Seven studies exhibited a low risk of bias [Figure 2a and b].[18,19,20,22,24,26,30] Four studies showed a high risk of bias, while two exhibited an ambiguous risk of bias.[21,23,25,27,28,29]

Figure 2.

a: Risk of bias for individual studies (traffic light); b: Risk of bias summary plot

Dextrose concentration and regimen: The dextrose concentration varied among studies, with tendon injections ranging from 16.7% to 50%, bursal injections from 15% to 20%, and superficial injections from 5% to 12.5% [Table 1]. The number of sessions varied from one to six, with the majority of research employing a single-injection protocol. A delay of 1 week to 1 month existed between injections, although one study could not specify the exact interval.[25]

Injection sites varied among the trials. Four studies exclusively examined tendons, two investigated both tendons and muscles, and four delivered dextrose injections to the bursa, as detailed in Table 1. One study integrated tendon and superficial injections.[23] Mofrad et al.[21] concentrated exclusively on superficial dextrose prolotherapy.[21] Another study did not delineate the injection sites.[29]

Outcome

Pain

Dextrose prolotherapy compared to control/placebo injection: Control injection was utilised in five investigations.[18,20,22,24,26] Each of these studies exhibited a substantial temporal effect with enhancements from baseline [Table 2]. A notable enhancement in VAS was noted in comparison to placebo, ranging from 2 weeks to 3 months. Bertrand et al.[24] documented a greater proportion of patients exhibiting clinically meaningful enhancement in NRS within the prolotherapy cohort during 9 months. Conversely, Sari et al.[20] determined that prolotherapy was equal to the control injection.

Table 2.

Summary of outcomes

Study	Intragroup And Intergroup Results	Adverse Event	Conclusion
Lin 2022[18]	Intragroup: VAS- Significant difference from baseline at week 2 in the prolotherapy group SPADI- Significant difference from baseline at 2 and 12 weeks in prolotherapy and at 2, 6, and 12 weeks in control ROM improved in the prolotherapy group at 2, 6, 12 weeks and 2 weeks in the control group Ultrasound Morphology- improved in the prolotherapy group at 6 and 12 weeks. Intergroup: VAS- A significant improvement at week 2 in the prolotherapy group SPADI- A statistically significant improvement in the SPADI scores at week 2 in the prolotherapy group ROM-(flexion) significantly improved at week 2 in the prolotherapy group The ultrasound morphology-prolotherapy group exhibited more changes in the echogenicity ratio at 6 and 12 weeks.	Not declared	Hypertonic dextrose injection could provide short-term pain and disability relief in patients with chronic supraspinatus tendinosis. Ultrasound imaging at week 6 revealed changed tendon morphology.
Lin 2023[19]	Intragroup: VAS- Both groups showed significant improvement from baseline at 2, 6, and 12 weeks SPADI- Both groups showed significant improvement at 2 and 6 weeks, but only the steroid group showed continued improvement at 12 weeks ROM- The steroid group demonstrated significant improvements in ROM in all directions, but the prolotherapy group achieved significant improvements in ROM in only flexion and abduction at 2 and 6 weeks Ultrasound morphology- Both groups demonstrated significantly greater decreases in bursa thickness with time Intergroup: VAS- The steroid group demonstrated a significantly greater reduction in VAS scores at 2 and 6 weeks, but not at 12 weeks SPADI- Steroid group showed significantly greater decreases in SPADI scores at 2, 6, and 12 weeks ROM- Significant differences in ROM were observed between the two groups at all time points Ultrasound morphology- Significant decreases in bursa thickness at 2, 6, and 12 weeks in the steroid group, while in the prolotherapy group, significant decreases in bursa thickness at 2 and 12 weeks, but not at week 6.	Not declared	Hypertonic dextrose injection may improve morphological parameters on ultrasound imaging, but it provides only short-term pain relief. Steroid injection exerted an adverse effect on the echo-texture of the supraspinatus tendon despite exhibiting more favourable clinical effects on patients with chronic subacromial bursitis.
Sari 2020[20]	Intragroup: VAS- In each group, the decrease in VAS scores was statistically significant over time, except PRP, which was significant at 12 and 24 weeks ASES: In each group, the decrease in VAS scores was statistically significant over time WORC: Over time, the score decreased in prolotherapy, corticosteroid, and control groups at 3 and 12 weeks but improved at week 24. In the PRP group, the score improved at 3 and 24 weeks. Intergroup: VAS- At 3 weeks, it was significantly lower in the COR group than in the PRP, PRO, and lidocaine groups ASES- The difference in baseline values of the ASES score was statistically significant between groups WORC- The WORC scores of the corticosteroid group at 3 weeks were significantly lower than the PRP, prolotherapy, and control groups; however, the WORC scores at 24 weeks were significantly lower than the steroid and control groups.	Not Declared	Long-term success of PRP injection was high, but all methods used, including CS, DPT, and lignocaine, could be beneficial for treatment.
Mofrad 2021[21]	Intragroup: VAS- Pain reduction was more evident at 2 weeks than at 3 months in both groups SPADI- Both interventions were significantly beneficial in reducing mean SPADI scores Intergroup: VAS- Pain reduction in prolotherapy was significant from physiotherapy at 2 weeks, but not after 3 months SPADI- Prolotherapy seemed to be more successful in decreasing SPADI scores at 2 weeks and 3 months	No severe adverse event reported	Prolotherapy is more successful as an initial treatment, and the treatment time is much shorter compared to physiotherapy.
Chang 2021[22]	Intragroup: Pain, ROM, SPADI- Significant time effects were noted across all assessments. Ultrasonographic changes over time: A significant time effect was observed in both bursal thickness and tissue elasticity. Intergroup: Pain, ROM, SPADI- No significant differences between the two groups. Ultrasonographic changes with time: A significant increase in tissue stiffness in the supraspinatus tendon with dextrose injection	No severe adverse event reported	DPT injections may increase tissue stiffness in the supraspinatus tendon, as indicated by elastography assessment results; however, further research is needed to determine the nature of these changes in elastography findings.
Amanollahi 2020[23]	Intragroup: VAS- Both interventions were successful in decreasing pain compared to the baseline. ROM- Both interventions were advantageous in increasing the ROM Intergroup: VAS- There was no significant difference between groups ROM- Not analysed	No severe adverse event reported	The 5% dextrose treatment is at least as effective as corticosteroid for reducing pain in patients with RC tendinopathy.
Bertrand 2016[24]	Intragroup: Pain- At 9 months, the Enth-Dex group maintained improvement in pain, while the other two groups could not USPRS- no differences of significance within groups Intergroup: Pain- Enth-Dex significantly outperformed Superfic-Saline. The difference between the Enth-Dex group and the Enth-Saline group did not reach clinical significance (P=0.088). USPRS- No differences of significance between groups (P=0.734)	One subject in the Enth-Saline group developed adhesive capsulitis.	Injection of hypertonic dextrose on painful entheses resulted in superior long-term pain improvement and patient satisfaction compared with blinded saline injection.
Seven 2017[25]	Intragroup: VAS/SPADI/WORC/ROM- Both groups achieved significant improvements over baseline Intergroup: VAS- A significant difference was found at baseline, weeks 3, 6, and 12, and last follow-up SPADI/WORC- Significant differences were found at weeks 6 and 12, and the last follow-up ROM- Significant differences were found in shoulder abduction and flexion at week 12 and last follow-up, and in internal rotation at last follow-up.	No severe adverse event reported	Prolotherapy is a readily applicable and effective adjunctive method for treating chronic RC lesions.
Lin 2019[26]	Intragroup: VAS/SPADI: Prolotherapy group showed significant improvement at 2 weeks but not at 6 weeks Intergroup: VAS/ROM/SPADI: Prolotherapy group showed significant improvement compared with the control group at two weeks after the injection. However, the effect did not sustain until six weeks after the injection.	Not declared	Ultrasound-guided hypertonic dextrose injection relieved pain, disability, and improved shoulder AROM for a short period in patients with chronic supraspinatus tendinopathy, having a rehabilitation impact.
Cole 2018[27]	Intragroup: Both groups showed improved ROM at 3 months from baseline not sustained up to 6 months Intergroup: There was no significant difference in the ROM or strength between groups at either time point.	Not Declared	Both glucose prolotherapy and corticosteroid were generally well tolerated. Glucose prolotherapy may be used as an alternative to corticosteroid injection without added benefits.
Sabaah 2020[28]	Intragroup: VAS- Significant improvement in prolotherapy and steroid, no improvement in PRP. WORC- Significant improvement in all 3 groups Ultrasound morphology: Grade of tendon lesion improved in Prolotherapy and PRP; no improvement in steroid. Bursitis improved in the prolotherapy and steroid group ROM- Significant improvement in prolotherapy group, no improvement in PRP/steroid Intergroup: VAS/ROM/WORC/Ultrasound Morphology: Prolotherapy significantly better than PRP/steroid	One patient in the PRP group became worse (ROM limitations in all directions)	Prolotherapy injections improve shoulder ROM, VAS, WORC index, and RC tendon healing, while PRP injections improve WORC index and tendon healing, but steroid injection has no effect on healing.
Taheri 2018[29]	Intragroup: NRS/SPADI- Both groups achieved significant improvements over baseline over time Intergroup: NRS/SPADI- Significant difference between prolotherapy and physiotherapy at week and 12; SPADI and NRS were not remarkably different in the corticosteroid and prolotherapy groups at any time point.	Not declared	DPT is as useful as the conventional corticosteroid injection technique in pain decrease and functional ameliorating of the shoulder joint
Karim 2020[30]	Intragroup: NRS/SPADI: Significant difference within the group, between baseline and each follow-up ROM increased over time in both groups, but not significant Intergroup: NRS: More patients in the PRP group reported pain lasting more than 48 hours than in the prolotherapy group, statistically significant. SPADI: Reduced in prolotherapy but non-significant than PRP ROM: Improved in prolotherapy but non-significant than PRP	No reports of serious adverse effects.	A single intra-tendinous injection of PRP and DPT significantly improved shoulder pain and function.

PRP=Platelet-Rich Plasma; VAS=Visual Analogue Scale; SPADI=Shoulder Pain and Disability Index; ROM=Range of Motion; ASES score=American Shoulder and Elbow Surgeons score; WORC index=Western Ontario Rotator Cuff Index; NRS=Numerical Rating Scale; USPRS=Ultrasound Pathological Rating Scale

Dextrose prolotherapy compared to corticosteroid: Six studies evaluated dextrose prolotherapy against corticosteroids.[19,20,23,27,28,29] A temporal effect was observed in both groups throughout all investigations. Taheri et al.[29] determined that both groups exhibited comparability over time. Lin et al.[19] indicated that steroids surpassed dextrose at 2 weeks, but not at 12 weeks. Sari et al.[20] reported that steroids were more efficacious at 3 and 12 weeks, whereas dextrose was more efficacious at 24 weeks. Conversely, Cole et al.[27] found that dextrose prolotherapy was superior at 6 weeks and 3 months, yet comparable to steroids at 6 months. Furthermore, studies by Amanollahi et al.[23] and Sabaah et al.[28] revealed significant reductions in VAS scores with DPT compared to corticosteroids, with enhancements observed at 1 month and 3 months, respectively.

Dextrose prolotherapy compared to PRP injection: In the investigation conducted by Sabaah et al.,[28] the VAS score at 3 months was markedly elevated with PRP compared to dextrose. Karim et al.[30] noted substantial enhancement in NRS scores for both cohorts; however, the groups exhibited comparability over 6 months. Sari et al.[20] identified a significant temporal effect with dextrose prolotherapy from baseline to week 24, while PRP demonstrated notable improvement from baseline solely at 12 and 24 weeks. In contrast, dextrose proved more efficacious at week 3, whereas PRP exhibited greater effectiveness at 12 and 24 weeks.

Dextrose prolotherapy compared to physiotherapy: Seven et al.[25] demonstrated that dextrose resulted in a significant enhancement in VAS scores at all assessed time points relative to physiotherapy. Mofrad et al.’s[21] within-group analysis indicated that pain alleviation from dextrose prolotherapy was more pronounced at 2 weeks than at 3 months post-intervention. Taheri et al.[29] reported a significant and sustained reduction in NRS scores with dextrose up to 12 weeks.

Questionnaires specific to the shoulder

Dextrose prolotherapy compared to control/placebo injection: Chang et al.[22] and Sari et al.[20] determined that dextrose was equivalent to control at all measured time points. Lin et al.[18] noted a significant improvement in SPADI for 2 weeks following the ultrasound-guided dextrose injection.[26] The Western Ontario RC (WORC) index demonstrated a significant improvement from baseline in dextrose prolotherapy at both 3 and 12 weeks, according to Sari et al.[20]

Dextrose prolotherapy compared to corticosteroid: Lin et al.[19] observed that the steroid resulted in considerably higher reductions in SPADI at 2, 6, and 12 weeks. Taheri et al.[29] noted that both treatments had significant temporal effects on SPADI but were comparable at all assessed time points. Sari et al.[20] noted a progressive enhancement in the American Shoulder and Elbow Surgeons score (ASES) over time in both cohorts. The steroid demonstrated superior efficacy at 3 and 12 weeks, whereas dextrose outperformed the steroid at week 24. Additionally, both groups exhibited improvements in the WORC index at 3 and 12 weeks.

Dextrose prolotherapy compared to PRP injection: Sari et al.[20] noted a substantial enhancement in ASES scores over time with dextrose; conversely, PRP demonstrated significant improvement after 12 weeks. The WORC scores for both PRP and dextrose prolotherapy exhibited temporal enhancement, with the PRP group showing greater improvement. An intergroup analysis conducted by Karim et al.[30] revealed no significant disparity in SPADI Total scores between the PRP and dextrose groups, as both groups exhibited a notable time effect. The WORC index, assessed before and after injection in the study by Sabaah et al.,[28] showed improvement in both groups; however, the enhancement was more pronounced in the dextrose group compared to the PRP group.

Dextrose prolotherapy compared to physiotherapy: Seven et al.[25] determined that dextrose prolotherapy surpassed physiotherapy in SPADI and WORC scores, with clinically significant enhancements sustained for up to 1 year. Mofrad et al.[21] observed that although dextrose demonstrated greater efficacy in the short term, the physiotherapy cohort exhibited more consistent outcomes by the conclusion of the study. In Taheri et al.’s study,[29] SPADI scores significantly declined in comparison to physiotherapy at week 4, with improvements maintained until week 12.

Range of active movement

Dextrose prolotherapy compared to control/placebo injection: Lin et al.[18] identified a transient advantage of dextrose over control at week 2.[26] Chang et al.[22] reported a substantial temporal effect for both groups, but the two treatments were found to be equivalent.

Dextrose prolotherapy compared to corticosteroid: Cole et al.[27] noted a substantial enhancement in supraspinatus strength following dextrose prolotherapy at 3 months, which persisted at 6 months. No significant difference in ROM was seen between the groups at any time point. Lin et al.[19] indicated that corticosteroids enhanced the ROM in all directions, whereas dextrose only facilitated increases in abduction and flexion. Both Amanollahi et al.[23] and Sabaah et al.[28] observed no enhancements with the steroid 3 months post-injection; in contrast, dextrose showed considerable improvement.

Dextrose prolotherapy compared to PRP injection: Three months post-injection, Sabaah et al.[28] observed a notable enhancement in ROM with dextrose prolotherapy, whereas no improvement was noted with PRP; one patient even exhibited exacerbated ROM restrictions in all directions. Karim et al.[30] indicated that from baseline to all follow-up intervals up to 6 months, ROM increased in both cohorts, with improvements being comparable between the two groups.

Dextrose prolotherapy compared to physiotherapy: In the study by Seven et al.,[25] ROM considerably improved in both the dextrose injection and physiotherapy groups over 1 year compared to baseline, with dextrose demonstrating superior results at the 1-year follow-up.

Ultrasound morphology

The parameters for ultrasonography evaluation differed among studies. Sabaah et al.[28] evaluated pathologies before and after intervention and observed significant enhancements in tendon lesions with dextrose and PRP. Bursitis showed considerable improvement with prolotherapy and corticosteroids. Bicipital tenosynovitis improved with corticosteroids but not with dextrose or platelet-rich plasma. Chang et al.[22] identified significant temporal and group effects on bursal thickness and tissue elasticity, favouring dextrose compared to placebo. Similarly, dextrose prolotherapy demonstrated notable alterations in echogenicity ratio and supraspinatus tendon thickness at 6 and 12 weeks.[18] A marked enhancement in supraspinatus tendon appearance was observed with both dextrose and steroid treatments.[27] Bursa thickness diminished with both steroid and dextrose over time, with steroid exhibiting superior improvement in intergroup comparisons.[19]

Quantitative analysis of studies: A restricted meta-analysis of studies with minimal risk of bias was performed to enhance sensitivity. The analysis was conducted at two distinct time intervals: 1 month and 3 months.

1. Pain scores: At 1 month (MD = −0.76; 95% CI: −1.26, −0.27; P = 0.003; I² = 57%), prolotherapy resulted in a statistically significant reduction in pain levels. At 3 months (MD = 0.25; 95% CI: −0.08, 0.59; P = 0.14; I² = 16%), no significant differences in pain levels were observed between dextrose and alternative therapies [Figure 3].

Figure 3.

Forest plot of visual analogue scale (VAS) scores: A.1 at 1 month and A.2 at 3 months. SD = Standard Deviation; CI = Confidence Interval; IV = Inverse Variance

2. SPADI: At 1 month (MD = −8.58, 95% CI: −13.00, −4.17; P = 0.0001; I² =0%), the decrease in SPADI score was statistically significant with dextrose prolotherapy. The effect was non-significant at 3 months (MD = −2.68; 95% CI: −9.50, 4.14; P = 0.44; I² =20%) [Figure 4].

Figure 4.

Forest Plot of Shoulder Pain and Disability Index (SPADI) values: B.1 at 1 month and B.2 at 3 months SD = Standard Deviation; CI = Confidence Interval; IV = Inverse Variance

3. At 1 month, ROM exhibited significant enhancement in abduction (MD = 5.93; 95% CI: 1.49, 10.38; P = 0.009; I² =0%) and flexion (MD = 7.28; 95% CI: 2.87, 11.70; P = 0.001; I² =0%) with dextrose prolotherapy. Improvement in external rotation (MD = 0.78; 95% CI: −3.55, 5.11; P = 0.72; I² = 8%) was similar between dextrose and other interventions; however, internal rotation (MD = −1.94; 95% CI: −4.20, 0.32; P = 0.09; I² = 0%) showed a non-significant advantage in other interventions [Figure 5]. At 3 months, abduction (MD = 3.16; 95% CI: −2.47, 8.87; P = 0.51; I² = 0%), flexion (MD = 4.37; 95% CI: −2.20, 10.94; P = 0.19; I² =25%), and external rotation (MD = 1.78; 95% CI: −2.92, 6.49; P = 0.46; I² = 13%) demonstrated superiority in the dextrose group, albeit non-significantly. Internal rotation demonstrated superior outcomes in the alternative intervention (MD = −1.85; 95% CI: −5.67, 1.97; P = 0.34; I² = 0%); although, the difference was not statistically significant.

Figure 5.

Forest plot for shoulder range of motion (ROM): C.1 at 1 month and C.2 at 3 months. MD = Mean Deviation; CI = Confidence Interval; IV = Inverse Variance

Analysis of sensitivity

Our investigation revealed heterogeneity in the VAS score; hence, a sensitivity analysis was conducted to assess the validity of the acquired data. The leave-one-out strategy was utilised for this analysis, systematically removing specific research to assess its impact. The mean difference (MD) remained negative, the confidence interval (CI) did not encompass zero, and the P value remained statistically significant [Table 3]. This substantiates the dependability of the outcome, indicating that dextrose alleviates pain at the one-month follow-up.

Table 3.

Sensitivity analysis of pain score (leave-one-out method)

	Study excluded	Effect	P	MD [CI]
VAS at 1-month follow-up	Sari 2020[20]	2.51	0.01	−0.84 [−1.5, −0.18]
	Lin 2023[19]	3.35	0.0008	−0.81 [−1.29, −0.34]
	Lin 2022[18]	3.58	0.0003	−0.55 [−0.84, −0.25]
	Lin 2019[26]	2.12	0.03	−0.78 [−1.51, 0.06]
	Karim 2020[30]	2.55	0.01	−0.75 [−1.33, −0.17]
	Chang 2021[22]	2.80	0.005	−0.79 [−1.34, −0.24]

VAS=visual analogue scale; MD=mean difference; CI=confidence interval

Analysis of subgroups

We evaluated the VAS score at the 1-month follow-up for heterogeneity over 50%. The studies were categorised according to the pathology. The mean difference was negative in both bursitis and tendinopathy, indicating the efficacy of dextrose prolotherapy in both pathologies. The benefit of dextrose prolotherapy over other interventions was significant in chronic tendinopathy (MD = −1.01; 95% CI: −1.74, −0.28; P = 0.007; I² = 73%). However, in bursitis, it was comparable to other modalities (steroid and control) (MD = −0.41; 95% CI: −0.91, 0.09; P = 0.11; I² = 0%) [Figure 6a]. The heterogeneity in the tendinopathy group could be due to different concentrations of dextrose used. The studies were also categorised based on the interventions used. The MD was negative in both the control group and the PRP, and the CI did not include 0; thus, dextrose therapy was found to be useful in reducing pain at 1-month intervals [Figure 6b]. The heterogeneity in the control group could be due to different concentrations of dextrose used (15% or 20%) or the pathology (bursitis or chronic tendinopathy), as well as the number of sessions (3 sessions or 1 session). Further studies with a uniform protocol should be conducted to explore this topic in more detail.

Figure 6.

(a) Forest plot for subgroup analysis of visual analogue scale (VAS) at 1 month: based on pathology. MD = Mean Deviation; CI = Confidence Interval; IV = Inverse Variance. (b) Forest plot for subgroup analysis of visual analogue scale (VAS) at 1 month: based on comparators. MD = Mean Difference; CI = Confidence Interval; IV = Inverse Variance

Required information size

TSA was conducted to determine the necessary information size, utilising the traditional 95% CI. The study’s power was maintained at 80% and then increased to 90%, with an alpha error of 5%. The sample sizes for VAS at 1 month were 314 (80% power) and 420 (90% power). Our analysis encompassed solely low-risk research, resulting in a sample size of 312 [Figure 7]. The sample size achieved 80% power, with the z-score positioned within the area of benefit. The sample sizes for SPADI were 273 (80% power) and 365 (90% power). The trial sequential limit of the benefit line was surpassed, yielding a z-score indicative of benefit; nonetheless, the sample size was inadequate. Consequently, additional high-quality investigations are necessary to substantiate these findings successfully.

Figure 7.

Trial sequential analysis of included trials comparing dextrose group and with other intervention group for pain (X axis = number of patients randomised; Y axis = cumulative z score; horizontal blue lines = conventional boundaries (upper for benefit, z score = 1.96; lower for harm, z score = −1.96; two-sided); sloping red lines (80% power) and magenta lines (90% power) with black filled circles = trial sequential monitoring boundaries calculated accordingly; blue line with black filled squares = z curve; vertical red line and magenta line = required information size calculated accordingly; upper gray rectangle = area of benefit; middle blue rectangle = futility area; lower orange rectangle = area of harm)

Grading of evidence certainty

The quality of evidence was assessed utilising the GRADE approach [Table 4]. The quality of evidence regarding pain scores in studies was low due to heterogeneity, imprecision, and possible publication bias. Dextrose prolotherapy showed considerable improvement in pain scores (VAS) after one month; however, it was comparable to other modalities at the three-month follow-up. Although only high-quality randomised controlled trials were analysed, the evidence was downgraded to low quality due to heterogeneity, potential publication bias, and limited sample size. A notable enhancement in the SPADI score was recorded at the one-month follow-up. The enhancement was consistent and comparable during the three-month follow-up. The evidence was classified as low quality due to potential publication bias and imprecision. These findings should be interpreted with caution due to the limited sample size in all included studies and the multifactorial aetiology of chronic pain.

Table 4.

The GRADE level of certainty for outcomes

Certainty assessment							№ of patients		Effect		Certainty	Importance
№ of studies	Study design	Risk of bias	Inconsistency	Indirectness	Imprecision	Other considerations	Dextrose prolotherapy	Others	Relative (95% CI)	Absolute (95% CI)
PAIN SCORES (follow-up: mean 1 month; scale: 0–10)
6	RCT	Not serious	Seriousa	Not serious	Not serious	Publication bias is strongly suspectedb	158	154	-	MD 0.76 SD lower (1.26 lower to 0.27 lower)	⨁⨁◯◯ Lowa,b,	Important
PAIN SCORES (follow-up: mean 3 months; scale: 0–10)
6	RCT	Not serious	Not serious	Not serious	Seriousc	Publication bias is strongly suspectedb	169	158	-	MD 0.25 SD higher (0.08 lower to 0.59 higher)	⨁⨁◯◯ Lowb,c	Important
SPADI (follow-up: mean 1 month; assessed with SPADI scale; scale: 0–100)
5	RCT	Not serious	Not serious	Not serious	Not serious	Publication bias is strongly suspectedb	128	124	-	MD 8.58 SD lower (13 lower to 4.17 lower)	⨁⨁◯◯ Moderateb,	Important
SPADI (follow-up: mean 3 months; assessed with SPADI scale; scale: 0–100)
4	RCT	Not serious	Not serious	Not serious	Seriousc	Publication bias is strongly suspectedb	112	109	-	MD 5.29 SD lower (19.85 lower to 9.27 higher)	⨁⨁◯◯ Lowb,c	Important

CI=Confidence Interval; RCT=Randomised Control Trial; MD=Mean Difference; VAS=Visual Analogue Scale; SPADI=Shoulder Pain and Disability Index; ROM=Range of Motion; a. Quality was rated down for inconsistency because I2>30%; b. Due to a smaller sample size, the results are more likely to be; c. The 95% CI includes both a reduction and an increase in scores

Publication bias

Funnel plots were created utilising Review Manager software for the visual evaluation of publication bias [Figure 8]. The VAS and SPADI plots exhibited asymmetrical data points, suggesting a potential publication bias. This may be attributed to the restricted number of papers incorporated in the analysis (<10) and the diminutive sample size. Consequently, one must exercise caution in interpreting these data.

Figure 8.

Funnel plot (visual analogue scale at 1-month). MD = Mean Difference; SE = Standard Error

DISCUSSION

Our systematic review contributes to the current literature on dextrose prolotherapy for RC diseases, demonstrating its efficacy in alleviating pain and enhancing function in affected individuals. A recent review of six studies suggested that dextrose prolotherapy may be beneficial for individuals with RC lesions; however, the studies exhibited moderate to high heterogeneity in the outcomes evaluated.[31] Another review of five studies identified it as a potentially effective adjunct to physical therapy for tendinopathy, encompassing tendinosis, partial tears, and small full-thickness tears.[32] Further analysis of five studies corroborated the benefits in pain, function, and ROM, although some studies presented a high risk of bias and heterogeneity.[33]

This review indicated that the majority of authors favoured the injection of hypertonic dextrose into the tendon and trigger points. Dextrose concentrations fluctuated, with the majority of investigations employing amounts over 20%. The intervals for injections were irregular, varying from weekly to monthly. Standard outcome measurements were VAS, ROM, and SPADI, but ultrasound assessments differed, concentrating on supraspinatus echogenicity, tissue rigidity, and bursa thickness. Numerous investigations demonstrated improvements in pain, ROM, and SPADI ratings, accompanied by steady incremental gains in ultrasound morphology.

Dextrose prolotherapy is categorised into two primary forms, namely hypertonic and isotonic, each possessing unique processes and clinical applications. Techniques differ in concentration, volume, and agent combinations. Hypertonic concentrations of dextrose induce localised inflammatory responses, whereas isotonic amounts exert a direct analgesic effect.[34,35] Dextrose prolotherapy operates through a tri-phasic mode of action: (1) Inflammatory phase: Hypertonic dextrose (>10%) causes cellular dehydration and lysis, resulting in inflammation. This leads to the recruitment of granulocytes and macrophages, along with the release of cytokines (interleukins, prostaglandins, thromboxanes, leukotrienes), thus creating a healing environment. (2) Proliferative phase: Inflammatory mediators activate fibroblast activity, resulting in collagen deposition and extracellular matrix remodelling. This phase improves the structural integrity of ligaments and tendons.[36,37] Histological evidence indicates an elevation in fibroblast quantity and an increase in collagen fibre thickness. (3) Regeneration phase: Continuous signalling and mechanical stress direct collagen maturation and alignment. Myofibroblast-mediated fascial contraction enhances tissue rigidity and functionality. This phase facilitates long-term stabilisation and alleviation of pain, maybe via improving proprioceptive and neurosensory functions.[38] Animal studies suggest that hypertonic glucose enhances tendon regeneration by promoting tendon development and increasing strength.[39] It is a low-risk, cost-effective, and durable treatment for chronic musculoskeletal pain.

Dean et al.[4] emphasised the significant long-term harm to tendon tissue and cells associated with conventional treatments such as glucocorticoid injections. Although corticosteroid injections provide temporary pain alleviation, the restricted frequency of permissible annual injections due to their adverse effects highlights the need for superior treatment alternatives. Although it is not a major adverse effect, an escalation of pain after PRP injections may be intolerable for people. The safety observations we conducted concerning dextrose prolotherapy intervention align with previous systematic studies, which indicated that dextrose prolotherapy does not lead to any direct problems.

Limitations

The studies in our evaluation exhibit heterogeneity for the diversity of interventions compared, injection sites, employed procedures, dextrose concentrations, outcome measures, and the duration of outcome monitoring. Consequently, the meta-analysis was performed over a brief follow-up period. We used selected papers with a minimal risk of bias for the meta-analysis, which diminished the overall sample size. Distinct practice patterns, such as the administration of hypertonic dextrose injections in tendons and trigger sites following a one-to-three-injection protocol, are observable. Nevertheless, these techniques lack robust evidence and should thus be regarded as merely expert opinion.

Strength

The systematic review’s strengths encompassed well-executed randomised controlled trials from the previous decade, providing an up-to-date assessment of the efficacy and safety of dextrose prolotherapy. The seven studies included in the evidence synthesis exhibited little risk of bias. The sensitivity analysis corroborated the robustness of the findings. Our review yielded significant evidence affirming the safety of dextrose prolotherapy.

CONCLUSION

Notwithstanding possible biases in certain research, the seven meticulously executed randomised controlled trials indicated that dextrose prolotherapy alleviates pain and enhances functionality. Despite the emergence of new studies, the varied protocols and intervention strategies, coupled with inconsistent outcome reporting across different follow-up periods and significant bias, have impeded the clear definition of the restorative effects of dextrose prolotherapy. A solitary injection of 20% dextrose may provide analgesia and enhance functionality for individuals with rotator cuff injuries (evidence level: low). We propose that, given the procedure’s low-risk nature and cost-effectiveness, it could be regarded as a valuable enhancement to physiotherapy for patients who have not responded to conservative treatments, are unsuitable for surgery, or have opted against definitive surgical interventions. Dextrose prolotherapy is deemed safe, with no significant adverse events documented. We underscore the necessity of high-quality randomised controlled trials for the establishment of evidence. Future research should incorporate controls and compare dextrose prolotherapy with alternative therapies such as PRP and corticosteroids. Injection techniques must be optimised (dosage, location, and frequency) and incorporate extended follow-up durations to evaluate enduring efficacy.

Data availability

The data for this systematic review and/or meta-analysis can be obtained from the authors upon acceptable justification (contact the relevant author via email) and will be shared upon request.

Conflicts of interest

There are no conflicts of interest.

Presentation at conferences/CMEs and abstract publication

Nil.

Disclosure of use of artificial intelligence (AI)-assistive or generative tools

The authors confirm that no AI tools or language models (LLMs) were used in the writing or editing of the manuscript, and no images were manipulated using AI.

Declaration of Use of Permitted Tools

The scales, scores [VAS, NRS, SPADI] are freely available and not copyrighted.

Authors Contributions

SS was involved in conceptualization, literature search, dataextraction, quality assessment, statistical analysis, writing of original draft; MK was involved in literature search, data extraction, quality assessment, writing of original draft; ShSh was involved in supervison, validation, statistical analysis, review of manuscript; AK was involved in supervision, critical analysis; BB was involved in methodology, statistical analysis; KPM was involved in review, editing, and formatting of manuscript.

Supplementary material

This article has supplementary material and can be accessed at this link. Supplementary Material at http://links.lww.com/IJOA/A33.

Funding Statement

Nil.