Prolotherapy is not superior to control or placebo-based conservative treatments for rotator cuff tendinopathy: a systematic review and meta-analysis

Abstract

Background

This systematic review and meta-analysis aimed to assess the efficacy of prolotherapy compared to control or placebo-based treatments.

Methods

A comprehensive search of PubMed, Ovid, and Scopus was conducted up to April 2025. Inclusion criteria encompassed clinical studies comparing prolotherapy with control or placebo treatments and evaluating outcomes such as pain, function, and range of motion.

Results

Eight studies involving 431 participants met the inclusion criteria. Patient-reported outcomes, including pain visual analog scale and Shoulder Pain and Disability Index, showed no statistically significant differences between prolotherapy and controls. Prolotherapy demonstrated a small but statistically significant improvement in shoulder abduction (mean difference, 7.08°; 95% CI, 2.49°–11.66°). Other range of motion measures, such as forward flexion, internal rotation, and external rotation, showed no significant differences. Radiographic outcomes, including tendon thickness and elasticity, suggested potential structural benefits but did not consistently translate to clinical improvements.

Conclusions

Prolotherapy is not superior to control treatments for rotator cuff tendinopathy. While it offers minor gains in shoulder abduction, its clinical benefits are limited.

Level of evidence

III.

INTRODUCTION

Orthobiologic treatments, such as platelet-rich plasma and bone marrow aspirate concentrate, have gained significant attention for managing chronic musculoskeletal problems [1-5]. However, an alternative treatment option with lower cost and a theoretical ability to induce both inflammatory and healing processes is prolotherapy. Prolotherapy involves the injection of an irritant solution into the site of injury to stimulate tissue repair and collagen production. Originally developed to address chronic muscle and ligament pain, tendinopathy, and osteoarthritis [6,7], this method has evolved to include various concentrated solutions, such as hyperosmolar dextrose, phenol-glycerine-glucose mixtures, and morrhuate sodium. Among these, hyperosmolar dextrose is the most commonly used agent today. Typically, dextrose concentrations between 12.5% and 25% are employed in prolotherapy, particularly for treating tendons and joints, due to their efficacy in promoting tissue healing [6,8,9].

Multiple studies have shown that prolotherapy using hyperosmolar dextrose injections can significantly improve patient-reported outcomes in individuals with rotator cuff tendinopathy, enhancing pain relief and functional recovery [10,11]. This treatment involves administering a concentrated dextrose solution directly into the rotator cuff tendons, often guided by ultrasound for precision [12]. Studies indicate that hyperosmolar dextrose effectively reduces pain, particularly in the long term, outperforming some non-invasive methods and other infiltrations, such as local anesthetics [10]. While it is not consistently more effective than corticosteroids or platelet-rich plasma, hyperosmolar dextrose has a favorable safety profile, with rare minor adverse effects like transient pain or inflammation [12]. This approach supports the natural healing process and provides an alternative to conservative treatments or surgery for managing chronic rotator cuff tendinopathy. However, findings across studies remain variable, with inconsistent comparisons to corticosteroids and platelet-rich plasma. These discrepancies, along with the treatment's favorable safety profile, underscore the need for a comprehensive systematic review and meta-analysis to clarify its clinical efficacy.

The aim of this study was to compare the clinical outcomes of prolotherapy with those of control or placebo-based conservative treatments for chronic rotator cuff tendinopathy. The authors hypothesized that prolotherapy would yield superior outcomes in terms of pain reduction, functional improvement, and range of motion compared to the control treatment.

METHODS

This study did not require ethical approval or informed consent because it was a systematic review and meta-analysis based entirely on previously published studies.

Search Strategies

Two researchers (NT and TI) independently conducted detailed searches across the PubMed, Ovid, and Scopus databases to locate studies evaluating the clinical outcomes of prolotherapy versus control or placebo-based treatments for managing rotator cuff tendinopathy. Each researcher performed the search independently to ensure accuracy and comprehensive coverage, including all relevant studies published up to April 1, 2025, from the inception of the respective databases. The systematic review adhered to the 2020 PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines [13] and was registered in PROSPERO under the identifier CRD42024615400. The search terms used were: ("prolotherapy" OR "dextrose" OR "hypertonic") AND ("control" OR "placebo" OR "physiotherapy") AND ("rotator cuff" OR "supraspinatus"). The Ovid database's advanced search feature was employed with filters set to limit results to studies in the English language, full-text availability, and studies involving human subjects.

Inclusion Criteria

To be included in the review, studies had to satisfy several criteria: (1) comparative clinical research with evidence levels between 1 and 3, (2) published in English, (3) directly evaluating prolotherapy versus control or placebo-based treatments of rotator cuff tendinopathy, (4) reporting clinical outcomes following treatment, and (5) full-text availability. Studies were excluded if they (1) focused on basic science or biomechanics, (2) consisted of case series or individual case reports, or (3) were review articles.

Data Extraction

The eligibility of the studies was independently reviewed by two researchers (NT, TI), who examined the titles, abstracts, and full texts. Any discrepancies in their evaluations were addressed through consultation with a third author (DL). The extracted information encompassed (1) publication details, (2) demographic characteristics of the patients, (3) prolotherapy strategies, (4) control or placebo-based treatments, and (5) reported post-treatment clinical outcomes.

Methodological Quality Assessment

The quality of the included studies was assessed using two tools: the Methodological Index for Non-Randomized Studies (MINORS) and the Modified Coleman Methodology Score (MCMS) [14,15]. The MINORS tool includes 12 criteria for evaluating methodological rigor, with a maximum score of 24 points for comparative studies. Higher scores indicate better quality, with scores from 0 to 12 reflecting poor quality and significant methodological flaws, 13 to 17 indicating fair quality with moderate limitations, 18 to 22 signifying good quality with minimal concerns, and 23 to 24 representing excellent quality with negligible or no issues. The MCMS is a validated tool for assessing the methodological quality of studies, focusing on factors such as study design, patient selection, follow-up, and statistical analysis. It provides a score from 0 to 100, with scores below 50 indicating poor quality, 50–69 reflecting fair quality, 70–84 indicating good quality, and scores of 85 or above considered excellent. Two authors (NT, TI) independently assessed all studies using both tools, and any discrepancies were resolved with input from a third author (DL).

Statistical Analysis

The data were analyzed using RevMan for Windows (Cochrane version 5.4.1). Odds ratios (ORs) with 95% CIs were calculated for dichotomous outcomes, while mean differences (MDs) with their corresponding 95% CI were determined for continuous outcomes. Statistical heterogeneity was assessed using the chi-square test and I² statistic, with a p-value less than 0.1 or an I² greater than 50% considered indicative of significant heterogeneity among the studies. A fixed-effects model was used when no statistical or graphical evidence of heterogeneity was observed, reflecting the assumption that the true effect is consistent across studies. In contrast, a random-effects model was applied when heterogeneity was detected either statistically or visually, to account for variability in effect sizes across studies.

RESULTS

Included Studies

A total of 186 studies initially met the eligibility criteria for inclusion in the analysis. During the preliminary screening, 62 duplicate records were identified and removed using EndNote X9 for Windows. Of the remaining studies, 89 were excluded after title and abstract screening due to irrelevance, while 27 were excluded following a full-text review for not meeting the inclusion criteria (Fig. 1). Ultimately, eight articles [16-23] were included in the final analysis. Among these, one study [16] was classified as evidence level 3, three [17-19] were rated as level 2, and four [20-23] were assigned evidence level 1. The methodological quality of the studies was evaluated using two scoring systems: MINORS scores ranged from 15 to 22, indicating fair to good quality, while MCMS scores ranged from 57 to 72, also reflecting fair to good quality (Tables 1 and 2).

Fig. 1.

PRISMA (Preferred Reporting Items for Sys­tematic Reviews and Meta-Analyses) flow diagram for study selection.

Table 1.

The details of included studies

Study	LOE	Age (yr, mean±SD) (prolotherapy/control)	Clinical FU (mo)	Sample size, n (prolotherapy/control)	Loss FU, n (prolotherapy/control)	Outcome measure	MINORS	MCMS
Lee et al. (2015) [16]	3	54.1±7.8/55.8±6.6	12	57/53	NR	VAS, SPADI, abduction isometric strength, ROM (FF, ABD, IR, ER); analgesic ingestions	15	57
Bertrand et al. (2016) [20]	1	53.8±13.5/51.1±9.2	9	27/19	0/0	VAS, USPRS	21	70
Seven et al. (2017) [19]	2	50.19±12.13/46.31±10.6	Minimum 1 yr	57/44	1/5	VAS, WORC, SPADI, ROM (FF, ABD, IR, ER)	20	66
Lin et al. (2019) [23]	1	46.25±5.69/48.6±5.95	1.5	16/15	0/0	VAS, SPADI, ROM (FF, ABD, IR, ER); supraspinatus thickness (ultrasound)	21	72
Sari et al. (2020) [18]	2	52.11±10.78	6	30/30	1/1	VAS, ASES, WORC	22	72
Chang et al. (2021) [21]	1	46.40±9.59 /47.72±11.79	3	25/25	2/2	VAS, SPADI, ROM (FF, ABD), bursa thickness & elasticity of supraspinatus (ultrasound)	21	72
Kazempour Mofrad et al. (2021) [17]	2	52.5±13.9/56.9±13.6	3	32/33	1/0	Change in SPADI; percentages	21	68
Lin et al. (2022) [22]	1	49.10±8.44/52.18±9.83	3	29/28	0/0	VAS, SPADI, ROM (FF, ABD, IR, ER); supraspinatus thickness (ultrasound)	21	72

LOE: level of evidence, SD: standard deviation, FU: follow-up, MINORS: Methodology Index for Non-Randomized Studies, MCMS: Modified Coleman Methodology Score, NR: not reported, VAS: visual analog scale, SPADI: Shoulder Pain and Disability Index, ROM: range of motion, FF: forward flexion, ABD: abduction, IR: internal rotation, ER: external rotation, USPRS: Ultrasound Shoulder Pathology Rating Scale, WORC: Western Ontario Rotator Cuff Index, ASES: American Shoulder and Elbow Surgeons.

Table 2.

The detailed of included studies

Study	Inclusion	Exclusion	Prolotherapy groups’ intervention	Control groups’ intervention
Lee et al. (2015) [16]	- Adults aged >40 yr	- Other shoulder pathology	- Agent: 16.5% hypertonic dextrose	- Agent: none
	- Shoulder pain >3 mo	- Cervical radiculopathy	- Injection dose: 10 mL (8 mL of 20% dextrose and 2 mL of 1% lidocaine)	- Injection dose: none
	- Positive painful arc, impingement test, or resisted test	- Prior shoulder or cervical surgery	- Injection technique: ultrasound-guided	- Injection technique: none
	- Correlation between physical examination and ultrasonography or MRI	- Active inflammation	- Injection interval: wk 0, 2, 5, then every 4 wk until pain reduced by 50%, 8 rounds completed, or patient withdrawal	- Injection interval: none
		- Calcific tendinitis	- Physical therapy: regular exercise education by physical therapists	- Physical therapy: regular exercise education by physical therapists
		- Recent history of trauma at shoulder
Bertrand et al. (2016) [20]	- Adults aged 19–75 yr	- Allergy to local anesthetic	- Agent: 25% hypertonic dextrose + 0.1% lidocaine and normal saline	- Agent: 0.1% lidocaine and normal saline
	- Shoulder pain >3 mo	- Unwillingness to avoid anti-inflammatories before and after treatments	- Injection dose: NR	- Injection dose: NR
	- Positive Neer sign, Hawkins-Kennedy test, or painful arc	- Corticosteroid injection within the last 8 wk	- Injection technique: landmark-based injection	- Injection technique: landmark-based injection
	- Ultrasound confirmed supraspinatus tendon pathology.	- Passive shoulder abduction <100◦ or external rotation <25◦	- Injection interval: monthly, 3 times	- Injection interval: monthly, 3 times
		- Rotator cuff calcification diameter >0.8 cm on plain film or ultrasound	- Physical therapy: structured physical therapy program (7 sessions)	- Physical therapy: structured physical therapy program (7 sessions)
		- OA K-L grade II to IV
		- Type III acromion
		- Supraspinatus tear width >1.2 cm
Seven et al. (2017) [19]	- Adults aged 30–60 yr	- Rheumatic disease	- Agent: 25% hypertonic dextrose for the subacromial bursa and 15% for supraspinatus, infraspinatus, teres minor, pectoralis minor, coracobrachialis, and biceps brachii insertions.	- Agent: none
	- Shoulder pain >6 mo - Refractory to at least 3 mo of conservative treatments	- Systemic inflammatory disease	- Injection dose: 4 mL injected into the subacromial bursa and 18 mL into tendon insertions	- Injection dose: none
	- MRI showed rotator cuff lesion (tendinosis or partial tear).	- Diabetes mellitus	- Injection technique: Ultrasound-guided	- Injection technique: none
		- Osteomyelitis	- Injection interval: Injection discontinued if: pain reduced by 25%, 6 rounds completed, and patient withdrew	- Injection interval: none
		- Active infection	- Physical therapy: Home exercise program	- Physical therapy: physical therapy of 3 sessions (an average of 30 min per season) per wk for 12 wk
		- Corticosteroid injection within the last 12 wk
		- Bleeding tendency
		- Pregnancy
		- Prior shoulder surgery
Lin et al. (2019) [23]	- Adults aged >20 yr	- Adhesive capsulitis	- Agent: 20% hypertonic dextrose	- Agent: normal saline
	- Shoulder pain >6 mo - Ultrasound confirmed supraspinatus tendon pathology (tear or tendinosis).	- Prior shoulder surgery	- Injection dose: 5 mL (4 mL of 50% dextrose and 1 mL of normal saline)	- Injection dose: 5 mL
	- VAS pain >3	- Corticosteroid, hyaluronic acid, platelet-rich plasma, or other prolotherapy injection within the last 3 mo	- Injection technique: ultrasound-guided	- Injection technique: ultrasound-guided
		- Neurological disease	- Injection interval: NR	- Injection interval: NR
		- Simultaneously participating in another clinical trial	- Physical therapy: NR	- Physical therapy: NR
Sari et al. (2020) [18]	- Adults aged 18–75 yr	- MRI showed rotator cuff tear >grade I.	- Agent: 20% hypertonic dextrose	- Agent: normal saline and 1% lidocaine
	- Shoulder pain >3 mo	- Treatment with NSAIDs within 1 wk	- Injection dose: 5 mL (4 mL of 20% dextrose and 1 mL of 1% lidocaine)	- Injection dose: 5 mL (3 mL of 1% lidocaine and 2 mL of normal saline)
	- MRI showed rotator cuff lesion (bursitis, tendinosis, or partial tears grade I).	- Allergy to local anesthetic	- Injection technique: ultrasound-guided	- Injection technique: ultrasound-guided
	- Refractory to at least 2 mo of conservative treatments	- Thrombocytopenia	- Injection interval: NR	- Injection interval: NR
		- Infection	- Physical therapy: NR	- Physical therapy: NR
		- Anti-coagulation or anti-aggregation therapy
		- Glaucoma
		- Hypertension
		- Severe renal or hepatic insufficiency
		- Pregnancy
		- Uncontrolled diabetes
		- Prosthetic joint
		- Severe osteoporosis of bones adjacent to the joint
Chang et al. (2021) [21]	- Adults aged 20–65 yr	- Shoulder pain associated with trauma, adhesive capsulitis, a full-thickness rotator cuff tear, or a bicep tendon rupture	- Agent: 15% hypertonic dextrose	- Agent: normal saline and 1% lidocaine
	- Shoulder pain >3 mo	- Coagulation disorders	- Injection dose: 5 mL (4.5 mL of 15% dextrose and 0.5 mL of 1% lidocaine)	- Injection dose: 5 mL (4.5 mL of normal saline and 0.5 mL of 1% lidocaine)
	- Positive painful arc or impingement tests	- Local or systemic infections	- Injection technique: ultrasound-guided	- Injection technique: Ultrasound-guided
	- Experienced pain during daily life activities	- Corticosteroid injection or	- Injection interval: NR	- Injection interval: NR
	- Ultrasound confirmed subacromial bursa thickness >2 mm	surgical treatment for shoulder pain	- Physical therapy: regular physical therapy	- Physical therapy: regular physical therapy
		- Regular oral NSAIDs or corticosteroid treatment
Kazempour Mofrad et al. (2021) [17]	- MRI showed rotator cuff lesion (small rotator cuff tear or tendinopathy).	- Large or full-thickness rotator cuff tear	- Agent: 12.5% hypertonic dextrose	- Agent: none
	- Shoulder pain >3 mo	- History of major trauma at the shoulder	- Injection dose: 8 mL of 12.5% dextrose and 40 mg of 2% lidocaine)	- Injection dose: none
		- Allergy to local anesthetic	- Injection technique: landmark-based injection	- Injection technique: none
		- Discopathies or any other spinal pathology	- Injection interval: NR	- Injection interval: none
		- Adhesive capsulitis	- Physical therapy: exercise program	- Physical therapy: hot pack, transcutaneous electrical nerve stimulation, pulsed ultrasound, and exercise program
Lin et al. (2022) [22]	- Adults aged >20 yr	- Adhesive capsulitis	- Agent: 20% hypertonic dextrose	- Agent: normal saline
	- Shoulder pain >6 mo	- Prior shoulder surgery	- Injection dose: 5 mL of 20% hypertonic dextrose	- Injection dose: 5 mL of normal saline
	- Ultrasound confirmed chronic degenerative supraspinatus tendinosis (tendon thickening or thinning associated with abnormal echogenicity and loss of the regular parallel structure of fibers).	- Corticosteroid, hyaluronic acid, platelet-rich plasma, or other prolotherapy injection within the last 3 mo	- Injection technique: Ultrasound-guided	- Injection technique: ultrasound-guided
		- Neurological disease	- Injection interval: NR	- Injection interval: NR
		- Simultaneously participating in another clinical trial	- Physical therapy: NR	- Physical therapy: NR

MRI: magnetic resonance imaging, OA: osteoarthritis, K-L: Kellgren-Lawrence, NR: nor reported, VAS: visual analog scale, NSAID: nonsteroidal anti-inflammatory drug.

Patient-Reported Outcomes

Seven studies [16,18-23] reported post-treatment visual analog scale (VAS) scores for pain, while five studies [16,19,21-23] assessed post-treatment Shoulder Pain and Disability Index (SPADI) scores. Additionally, one study [17] provided data on the change in SPADI scores. In the pooled meta-analysis, there were no significant differences between the prolotherapy and control groups in post-treatment VAS for pain (MD, –0.57; 95% CI, –1.70 to 0.56; P=0.33) (Fig. 2A) or SPADI scores (MD, –8.57; 95% CI, –19.34 to 2.20; P=0.12) (Fig. 2B).

Fig. 2.

Forest plot comparing patient-reported outcome scores between prolotherapy and control treatment for rotator cuff tendinopathy. (A) Visual analog scale for pain. (B) Shoulder Pain and Disability Index (SPADI). SD: standard deviation, IV: inverse variance.

Range of Motion

Five studies [16,19,21-23] evaluated the range of motion of the shoulder following treatment. Among the range of motion outcomes, prolotherapy did not result in significant differences in forward flexion (MD, 4.44°; 95% CI, –0.36° to 9.23°; P=0.07) (Fig. 3A). However, it demonstrated significant benefits in abduction (MD, 7.08°; 95% CI, 2.49° to 11.66°; P=0.002) (Fig. 3B). No significant differences were observed in internal rotation (MD, 1.78°; 95% CI, –0.10° to 3.67°; P=0.06) (Fig. 3C) or external rotation (MD, 2.17°; 95% CI, –0.16° to 4.50°; P=0.07) (Fig. 3D).

Fig. 3.

Forest plot comparing range of motion (in degrees) between prolotherapy and control treatment for rotator cuff tendinopathy. (A) Forward flexion. (B) Abduction. (C) Internal rotation. (D) External rotation. SD: standard deviation, IV: inverse variance.

Radiographic Outcomes

Supraspinatus thickness by ultrasound

The outcomes of supraspinatus thickness differed between the studies by Lin et al. in 2022 [22] and 2019 [23]. In the 2022 study [22], the prolotherapy group demonstrated a significant increase in supraspinatus tendon thickness, from a pre-treatment value of 6.86±1.19 mm to 7.36±1.18 mm at 6 weeks (an increase of 0.50 mm, P<0.001) and to 7.46±1.21 mm at 12 weeks (an increase of 0.61 mm, P<0.001). In contrast, the control group (normal saline) showed no significant changes, with a pre-treatment value of 7.09±0.54 mm, 7.14±0.50 mm at 6 weeks, and 7.19±0.49 mm at 12 weeks. Conversely, the 2019 study [23] reported no significant changes in supraspinatus thickness for either group. The prolotherapy group had a pre-treatment thickness of 6.93±1.18 mm, with post-treatment values of 6.96±1.16 mm at 2 weeks and 7.43±1.18 mm at 6 weeks. The control group had a pre-treatment thickness of 7.01±0.58 mm, which measured 6.69±0.49 mm at 2 weeks and 7.16±0.51 mm at 6 weeks. These discrepancies may be due to differences in study methodologies, such as the timing of measurements and follow-up duration, or variations in patient populations, including severity of tendinopathy and baseline tendon characteristics. Further research is warranted to clarify these differences. These findings underscore significant morphological improvements in the 2022 study but not in the 2019 study [22,23].

Bursa thickness and elasticity of supraspinatus by ultrasound

One study [21] reported the pre- and post-treatment bursal thickness and elasticity of the supraspinatus tendon using ultrasound. At the 3-month follow-up, the bursal thickness decreased significantly in both groups compared to pre-treatment. In the prolotherapy group, the thickness decreased from 3.16±0.73 mm at pre-treatment to 2.19±0.51 mm at the 3-month follow-up (mean reduction, –0.96±0.76 mm; P<0.001 for time effect). In the control group, the thickness decreased from 2.82±0.56 mm at pre-treatment to 2.39±0.60 mm at the 3-month follow-up (mean reduction, –0.43±0.58 mm; P<0.001 for time effect). However, the between-group comparison showed no significant time-group interaction (P=0.059) [21].

For elasticity of the supraspinatus tendon, the prolotherapy group showed a significant increase from 60.59±17.59 kPa at pre-treatment to 89.52±24.56 kPa at the 3-month follow-up (mean increase, 28.93±23.79 kPa; P<0.001 for time effect). In the control group, elasticity increased from 56.83±27.65 kPa at pre-treatment to 69.00±22.41 kPa at the 3-month follow-up (mean increase, 12.17±15.99 kPa; P<0.001 for time effect). The between-group comparison showed a significant time-group interaction, with the prolotherapy group exhibiting a greater increase in elasticity compared to the control group (P=0.026) [21]. The results indicated that both treatments reduced bursal thickness and increased elasticity over time, but prolotherapy had a significantly greater effect on tendon stiffness compared to the control treatment.

DISCUSSION

This systematic review and meta-analysis evaluated the clinical outcomes of prolotherapy compared to control or placebo-based conservative treatments for managing chronic rotator cuff tendinopathy. The main finding of this study indicated that prolotherapy did not significantly outperform control treatments in terms of patient-reported outcomes, such as the VAS for pain and the SPADI. However, prolotherapy demonstrated a significant improvement in shoulder abduction range of motion compared to controls, with a small MD of 7.08°. No significant differences were observed in other range of motion parameters, including forward flexion, internal rotation, and external rotation. These findings suggest that, while prolotherapy may offer minimal targeted benefits in shoulder abduction, its overall clinical advantage remains limited.

Rotator cuff syndrome is one of the most common conditions affecting the shoulder, characterized by pain due to tendon degeneration, partial tears, or complete ruptures [24]. Treatment options for rotator cuff syndrome vary based on the severity and chronicity of the condition, ranging from conservative approaches such as rest, physical therapy, oral medications, and injections to more invasive interventions like surgery [25,26]. For tendinosis or partial tears, conservative treatment is typically the first line of management, involving nonsteroidal anti-inflammatory drugs, injections, and physiotherapy. In recent years, there has been growing interest in orthobiologic therapies, including platelet-rich plasma and bone marrow aspiration [1,3-5]. However, prolotherapy, a long-established treatment, remains a more cost-effective alternative to modern orthobiologic agents. Prolotherapy has been shown to induce localized inflammation in the affected area, which, rather than exacerbating the condition, may stimulate the body’s natural healing processes for damaged connective tissue [27]. Despite these promising effects, the outcomes of prolotherapy, particularly with hypertonic dextrose, remain inconsistent across existing research [28]. These inconsistencies may stem from differences in treatment protocols, including injection frequency and concentration, patient selection criteria, and lengths of follow-up periods, which can all influence observed clinical outcomes.

A previous systematic review investigating the outcomes of prolotherapy compared to a control treatment suggested that individuals with rotator cuff injuries might benefit from hypertonic dextrose injections [28]. This review included six studies [16,18-20,22,23], and both the prolotherapy and control groups showed improvements in VAS scores, SPADI scores, and range of motion, with the prolotherapy group demonstrating a significantly better outcome. However, the current study, which followed the PRISMA guidelines and included eight comparative studies [16-23], found that prolotherapy was not more effective than standard physiotherapy or placebo injections (normal saline or lidocaine).

Despite significant radiographic changes such as increased tendon thickness or elasticity after treatments like prolotherapy [22,23], these improvements often fail to translate into superior clinical outcomes [29]. This disconnect highlights the multifaceted nature of tendon pathology, where structural recovery alone may not adequately address functional impairments or pain perception. Evidence suggests that clinical improvements, such as reduced pain and enhanced function, are not consistently linked with morphological changes in tendons [29,30]. For example, studies on tendinopathy have shown that tendon thickness or neovascularization often changes independently of pain relief or functional gains, emphasizing the role of central and peripheral sensitization, altered biomechanics, and psychosocial factors in driving clinical outcomes [30]. Furthermore, mechanical properties like tensile strength and load tolerance may not correspond directly with changes observed through imaging [30].

Based on the results of the present study, the authors do not recommend prolotherapy as a first-line treatment option for chronic rotator cuff tendinopathy. While prolotherapy has shown a favorable safety profile and some potential for improving parameters such as range of motion or tendon morphology, its clinical efficacy remains inconsistent and not superior to physical therapy or placebo-based interventions. The small benefits observed, particularly in specific subdomains, do not justify its widespread adoption over control group therapies. Therefore, clinicians should prioritize evidence-based treatments that have consistently demonstrated better clinical outcomes and consider prolotherapy only in exceptional cases where alternative options are contraindicated or have failed.

To advance the understanding and application of prolotherapy in chronic rotator cuff tendinopathy, future research should prioritize assessment of the long-term efficacy and durability of clinical outcomes. Additionally, comparative effectiveness studies are needed to evaluate prolotherapy against other injection therapies, such as platelet-rich plasma, bone marrow aspirate concentrate, and corticosteroids. Identifying specific patient subgroups that may benefit most—based on tendon pathology or disease duration—is also essential. Addressing these priorities will help clarify prolotherapy’s role and optimize its use within the spectrum of conservative treatments for rotator cuff tendinopathy.

This study had several methodological strengths, including rigorous adherence to PRISMA guidelines and inclusion of high-quality evidence, with studies evaluated using validated tools such as the MINORS and MCMS scoring systems. However, the study had several limitations. First, there was significant heterogeneity among the included studies, particularly in injection protocols, follow-up durations, and outcome measures, which may have influenced the consistency of the results. Second, the reliance on subjective patient-reported outcomes, such as pain and functional scores, introduced variability that could have affected the interpretation of clinical benefits. Third, potential publication and language biases were acknowledged, as the analysis was restricted to English-language studies, possibly limiting the breadth of the evidence. Fourth, as with any systematic review, the search protocol may have overlooked relevant studies, further limiting the breadth of the findings. Last, subgroup meta-analysis was not feasible due to considerable heterogeneity in study designs, outcome measures, and surgical techniques, as well as the small number of studies available for each subgroup. This limited our ability to explore potential effect modifiers or sources of heterogeneity in greater detail.

CONCLUSIONS

Prolotherapy did not demonstrate significant improvement in patient-reported outcomes, including VAS for pain and SPADI, compared to controls. While it showed a small benefit in shoulder abduction, no significant differences were found in other range of motion parameters like forward flexion, internal rotation, or external rotation.