British Medical Bulletin, 2021, 138:96–111

doi: 10.1093/bmb/ldab004
Advance Access Publication Date: 21 April 2021

Invited Review

Prolotherapy for chronic low back pain: a review

Downloaded from https://academic.oup.com/bmb/article/138/1/96/6244772 by guest on 31 July 2023

of literature
Lorenzo Giordano1 , William D. Murrell2,3 , and Nicola Maffulli1,4,5,6
1 Department of Trauma and Orthopaedic Surgery, Azienda Ospedaliera Universitaria, San Giovanni di Dio e

Ruggi D’Aragona, Via San Leonardo 1, Salerno 84131, Italy, 2 Emirates-Integra Medical and Surgical Centre,
Dubai, United Arab Emirates, 3 Department of Orthopaedics, Podiatry, and Rehabilitation, Fort Belvoir
Community Hospital, 9300 Fort Belvoir, VA 22060, USA, 4 Queen Mary University of London, Barts and the
London School of Medicine and Dentistry, Centre for Sports and Exercise Medicine, Mile End Hospital, 275
Bancroft Road, London E1 4DG, England, and 5 Institute of Science and Technology in Medicine, Keele
University School of Medicine, Thornburrow Drive, Stoke-on-Trent ST5 5B, England
*Correspondence address. Department of Musculoskeletal Disorder, Faculty of Medicine and Surgery, University of Salerno,
Salerno, Italy. E-mail: n.maffulli@qmul.ac.uk
Received 25 March 2020; Revised 23 January 2021; Accepted 31 January 2021

Abstract
Introduction: Low back pain is common and imposes major societal burdens
for patient suffering and costs. Prolotherapy injections are used for muscu-
loskeletal conditions including tendinopathies, osteoarthritis and low back
pain to enhance soft-tissue healing. This review aims to clarify the place of
prolotherapy in chronic low back pain (CLBP).
Sources of data: Using multiple databases, a systematic search was per-
formed to identify studies detailing the use of prolotherapy to manage CLBP.
A total of 12 articles was included in the present work.
Areas of agreement: Considering the level of evidence and the quality of
the studies assessed using the modified Coleman Score, prolotherapy is
an effective management modality for CLBP patients in whom conservative
therapies failed.
Areas of controversy: The presence of co-interventions and the clinical
heterogeneity of the work contributes to confound the overall conclusions.
Growing points and areas for research: The analysis of the studies included
in the review, using appropriate tools, showed how their quality has

© The Author(s) 2021. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com
Prolotherapy for chronic low back pain, 2021, Vol. 138
decreased over the years, reflecting the need for appropriately powered well
planned and performed randomized control trials.
Key words: prolotherapy, sclerotherapy, hypertonic dextrose, CLBP, proliferative, injection
Introduction
Chronic low back pain (CLBP) is common and dis-
abling, with a marked negative impact on health care
resources.1 ,2 In most patients, conservative measures
generally resolve low back pain within 1 month, and
most of the costs occur when low back pain becomes
chronic, lasting longer than 3 months.2 ,3 CLBP can be
caused by several conditions such as lumbar radicu-
lopathy, spinal stenosis, sacroiliac joint instability,
non-specific low back pain and pain after surgery
for lower back ailments.4 Conventional therapies
can be ineffective, and some patients may experi-
ence severe unexpected adverse effects. For example,
prescription opioids for CLBP have contributed to
the ‘opioid epidemic.’5 Caudal epidural injection of
anti-inflammatory, analgesic and anesthetic drugs,6
and interventional procedures not involving injec-
tions are some treatment options that have been
assessed for CLBP, but their effectiveness is at best
suboptimal.7 ,8
An alternative procedure to manage muscu-
loskeletal conditions is prolotherapy, which has
been widely used in the past century.9 Developed
by George Hackett in the 1930s, The primary target
of the prolotherapy is the treatment of potential pain
sources within connective tissue, (ligament, tendon
or cartilage), by either the proliferation of new cells
or tissue or the improvement in the health of existing
cells or tissue. The traditional theory of the mech-
anism of proliferation induction by prolotherapy
was the initiation of a temporary inflammatory
response, followed by a healing cascade.10 However,
repeated high-quality randomized animal studies
have demonstrated the proliferation of stronger and
thicker connective tissue with normal histological
appearance using a sub-inflammatory concentration
of dextrose (10%).11–14
97
Hyperosmolar dextrose and sodium morrhuate
are irritants injected either at the attachment of
ligaments and tendons or at the intra-articularly.9,15
Downloaded from https://academic.oup.com/bmb/article/138/1/96/6244772 by guest on 31 July 2023
The inflammatory cascade is stimulated by the
commonly used concentrations of dextrose used in
clinical practice11 or by the inclusion of phenol, but
the additional therapeutic benefit of inflammation
is not clear. The mechanism of benefit of dextrose
injection may, in part, be neurogenic, as shown by
multiple recent favorable randomized controlled tri-
als (RCTs) on the treatment of compression neuropa-
thy by injection of dextrose at 5% in water16–18 and
by blinded evaluation of the effect of dextrose at 5%
in water vs normal saline and by caudal epidural
injection in low back pain.
The injected tissue respond to prolotherapy by
releasing many tissue growth factors.13,19 Recent
animal studies have shown an increased cross-
sectional area of connective tissue11,14,20,21 increased
load to rupture and increased tissue strength14 ,20,21
after 10–20% dextrose injections. Our purpose
was to review evidence regarding the efficacy of
prolotherapy, including dextrose injection alone or
dextrose in combination with a more inflammatory
component (e.g. phenol) in the treatment of CLBP.
In so doing, we performed a literature review,
considering results from studies with different
dextrose-containing injectates and modalities with
potentially different mechanisms of action.
Methods
Strategy for literature search
This systematic review of clinical investigations
followed PRISMA standardized reporting guide-
lines22 (Fig. 1). Searches of the electronic databases,
98
Fig. 1 PRISMA 2009 flow.
PubMed and Google Scholar, were conducted by
L.G. for all papers published from inception through
to January 2021. The search strategy included a wide
range of terms for prolotherapy and different terms
for CLBP, aiming for high sensitivity in order to
detect all the appropriate literature (Table 1).
Selecting studies for review
All articles were initially screened for relevance by
title and abstract, excluding articles without an
abstract and obtaining the full-text article if the
abstract did not allow the investigators to assess the
L. Giordano et al., 2021, Vol. 138
Downloaded from https://academic.oup.com/bmb/article/138/1/96/6244772 by guest on 31 July 2023
defined inclusion and exclusion criteria. The first
author reviewed the full test and extracted data. A
cross-reference search of the selected articles was
also performed to obtain other relevant articles
for the study. We included all the works in which
prolotherapy was used as the only therapy or in
association with other methods for the management
of CLBP of various etiology, without temporal,
languages or type of study limits. We excluded all
the works in which the prolotherapy was aimed at
the treatment of other diseases or conditions that
do not involve CLBP and works that do not present
clinical data on the use of prolotherapy on CLBP. The
Prolotherapy for chronic low back pain, 2021, Vol. 138
Table 1 Electronic database search terms
Terms for prolotherapy
Prolotherapy OR Dextrose OR glucose OR sugar OR regenerative OR proliferation OR injection AND
Terms for chronic
Chronic OR long Term OR recurring OR persistent AND
Terms for low back pain
Low back pain OR coccygodynia OR back pain OR sacroiliac pain
following data were extracted by the investigators:
study design, demographics, interventions, follow
up, outcome measurements (Table 2).
Evaluation of methodological quality
The authors have participated in development
of a quality assessment score, modified Coleman
Methodology Score and elected to use that tool to
evaluate the quality of each article in according to the
published criteria22 (Table 3). A modification of this
score was completed for use assessing conservative
intervention by substituting surgical procedure
with non-operative procedure for the purposes of
this study, while post-intervention rehabilitation
and outcome measurement remained the same.22
Each study was scored for each of the 10 criteria,
giving a total modified Coleman Methodology
Score ranging from 0 to 100. A score ranging
from 100 to 80 would represent a study design
that largely avoids the influence of chance, various
biases, and confounding factors; a score below
70 defines a ‘poor quality’ work while between
70 and 80 ‘fair quality’ work.22 We also used
another tool to evaluate strength of recommendation
through a patient-centered approach to grading
evidence: The Sort: Strength of recommendation
taxonomy. This grading scale addresses the quality,
quantity and consistency of evidence and allows
authors to rate individual studies. The taxonomy is
built around the information mastery framework,
which emphasizes the use of patient-oriented
outcomes that measure changes in morbidity,
mortality, symptom improvement, cost reduction
99
and quality of life (QoL).23 There are three level of
Downloaded from https://academic.oup.com/bmb/article/138/1/96/6244772 by guest on 31 July 2023
recommendation:
A-level recommendation is based on consis-
tent and good-quality patient-oriented evi-
dence;
B-level recommendation is based on incon-
sistent or limited-quality patient-oriented
evidence;
C-level recommendation is based on consen-
sus, usual practice, opinion, disease-oriented
evidence or case series for studies of diag-
nosis, treatment, prevention or screening23
(Fig. 2).
Two senior orthopedic surgeons completed the
process of scoring all the articles independently.
Disagreements were defined as a difference of
>2 points from the overall score from each indi-
vidual article, and disagreements were resolved by
consensus.
Effect measurements
In all studies, pain and disability were the primary
outcomes.
Commonly using the visual analog scale (VAS)24–28
or numerical rating scale (NRS)4 ,29–31 and McGill
Pain Questionnaire,32 efficacy studies in this field
assess change in pain intensity from baseline
with patient-reported ratings. Disability was mea-
sured using the Roland-Morris disability question-
naire24,25,33 and the Oswestry Disability Scale.31,32,34
Many investigations also reported the proportion
of patients who achieved at least 50% reduction in
pain and/or disability scores at 6 months.
 100

Table 2 Summaries of the studies reviewed

Author Study Sample Interventions Follow up Outcome Coleman Sort rec-

design Score/level ommen-
of the dation
study

Ongley RCT∗ Number of participants: (E∗ ) (40): weekly injections of 1, 3 and 6 months VAS∗ pain: 72/I B
et al.24 81 lumbopelvic ligaments with (E): 3.78 (entry)—2.13—1.77—1.50
(1987) Sex F: 43—M: 38 glucose (12.5%) glycerine (C) 3.99 (entry)—3.06—2.93—3.08.
Mean age (range) 44 (12.5%) phenol (1.25%) Roland disability questionnaire
(23–70) 0.25% lignocaine, 20 ml in (E): 11.45 (entry)—4.00—4.70—3.43;
total. 6 injection treatments. (C): 11.82 (entry) 8.37—8.49—8.29
(C∗ ) (41): the same protocol
but without glucose and with
saline 0.9%
Klein RCT Number of participants: (E) (39): 6 months after intervention VAS pain: 68/I B
et al.25 80 Weekly injections of glucose (E) 4.88 (entry)—2.29
(1993) Sex F: 36—M: 44 (12.5%) glycerine (12.5%) (C) 4.56 (entry)—2.85
Mean age: 46 phenol (1.25%) 0.25% Roland disability questionnaire
lignocaine. 6 injection (E) 9.36 (entry)—4.04
treatments. (C) 8.25 (entry)—4.38
(C) (40): the same—injections
with 0.25% lignocaine
Yelland RCT Number of participants: (E) (n = 54): Fortnightly 6–12–24 months after At 12 months, the proportions achieving > 70/I B
et al.26 110 injections of glucose (20%) commencing intervention 50% reduction in pain from baseline by
(2004) Sex: F: 57—M:41 and lignocaine (0.2%), 10–30 injection group were glucose-lignocaine: 0.46
Mean age: 50 SD (11) mls, mean number of vs saline: 0.36. By activity group these
injection treatments 7 proportions were exercise:
(C) (n = 56): the same with 0.41 vs normal activity: 0.39. Corresponding
only saline (0.9%) proportions for >50% reduction in disability
were glucose-lignocaine: 0.42 vs saline 0.36
and exercise: 0.36 vs normal activity: 0.38
Dechow RCT Number of participants: (E): Weekly injections of 1–3–6 months after VAS pain: 67/I C
et al.32 74 lumbo pelvic ligaments with intervention (E) 5.3 (entry)—5.2—5.1—5.2
(1999) Sex: F: 38—M: 36 glucose, glycerine, phenol, (C) 5.3 (entry)—4.8—5.3—4.4
Mean age: 46 lignocaine. 3 injection Oswestry disability scale:
treatments. (E) 34 (entry), 34—36—36
(C): Same but without glucose (C) 33 (entry) 33—34—35

(Continued)
L. Giordano et al., 2021, Vol. 138

Downloaded from https://academic.oup.com/bmb/article/138/1/96/6244772 by guest on 31 July 2023

Table 2 Continue

Author Study design Sample Interventions Follow up Outcome Coleman Sort rec-
Score/level ommen-
of the dation
study

Hooper Retrospective Number of participants: Solution containing 20% Ranging from 2 months to Ninety-one percent (91.0%) of patients 60/IV C
et al.29 case series 157 dextrose and 0.75% 2.5 years reported reduction in level of pain; 84.8% of
(2004) Sex: F: 111—M: 46 xylocaine patients reported improvement in activities
Mean age: 39.5 of daily living, and 84.3% reported an
improvement in ability to work
Miller Prospective Number of participants: Bi-weekly disc space injection 6 weeks—41 months NRS∗ 67/III C
et al.30 consecutive 76 of 50% dextrose and 0.25% − mean number of treatment: 3.5%
(2006) patient series Sex: F: 35—M: 41 bupivacaine − 43.4% of patients sustained improvement
Mean age (range): 55 group
(21–90) − average improvement in numeric pain
Prolotherapy for chronic low back pain, 2021, Vol. 138

scores of 71%

Lyftogt27 Prospective Number of participants: Hypertonic dextrose 20–40% 1 year VAS pain: 58/III C
(2008) clinical audit 41 mixed with lignocaine or − mean initial: 7.6
Sex: F: 17—M: 24 ropivacaine − mean last treatment: 1.4
Mean age (range): 48 Mean duration of treatment: − mean duration of treatment was 8.3 weeks
(23–73) 8.3 weeks − mean number of treatments was 6.2.
− 95% of patients improved more than 50%
and 10% less than 50%.
− 29% of patients reported no pain at the
last consultation

Kim RCT Number of participants: Intra-articular 50% dextrose Pain and disability scores The cumulative incidence of >50% pain 72/I B
et al.31 48 water prolotherapy or 40 mg were assessed at baseline, relief at 15 months was 58.7% in the
(2010) Sex: F: 14—M: 34 triamcinolone acetonide 2 weeks, and monthly after prolotherapy group and 10.2% in the steroid
Mean age: 59.5 SD injection, with a biweekly completion group
(14.5) schedule and maximum of of treatment, and then at 6,
three injections. 10 and 15 months

(Continued)
101

Downloaded from https://academic.oup.com/bmb/article/138/1/96/6244772 by guest on 31 July 2023

 102

Table 2 Continue

Author Study design Sample Interventions Follow up Outcome Coleman Sort rec-
Score/level ommen-
of the dation
study

Watson Retrospective Number of participants: Initial standard solutions used 1 year or more follow-up Both pain and QoL∗ scores were 55/IV C
et al.28 case series 140 for injections included P-25-G significantly improved at least 1 year after
(2010) Sex: (F: 89—M: 51) (phenol 2.5% plus 25% the last treatment. There were no differences
Mean age: 48 SD (12) dextrose plus 25% glycerine in outcomes as a result of age, response to
mixed 50:50 with 1% Xylocaine (lidocaine) injection, insurance
lidocaine without coverage, smoking history, or gender
epinephrine) or 15%
dextrose. The number of
injection sites varied
depending on how many sites
were tender
Cusi Prospective Number of participants: Three injection of 3–12–24 months Quebec Back Pain Disability Scale, 64/III C
et al.33 descriptive 25 hypertonic dextrose into Roland–Morris, Roland–Morris Multiform
(2014) study Sex: (F: 5—M: 20) dorsal interosseous ligament Questionnaires.
Mean age: 40 under ct control, 6 weeks Functional questionnaires significant
apart improvements for those followed-up at 3, 12
and 24 months. Clinical examination scores
showed significant improvement from start
to 3, 12 and 24 months
Maniquis RCT Number of participants: Injection of 10 ml 15 minutes; and 2, 4, and 48 h NRS∗ pain score: 66/I C
et al.4 35 of 5% dextrose or 0.9% and 2 weeks post-injection change at 15 minutes (4.4 vs 2.4 points; 2 h
(2016) Sex: (F: 11—M: 24) saline (4.6 vs 1.8_2.8 points;), 4 h (4.6_2.0 vs 1.4
Mean Age: 54 SD (10.7) points; and 48 h (3.0 vs 1 points; at 2 weeks
(2.1 vs 1.2.
85% (16/19) of dextrose recipients and 19%
(3/16) of saline recipients reported 50% pain
reduction at 4 h.
Hoffman Retrospective Number of participants: Three sacroiliac injections Average follow-up: 117 days The Oswestry disability index: 59/III C
et al.34 cohort study 103 (15% dextrose in lidocaine) 24 (23%) showed a minimum clinically
(2018) Sex: (F: 18—M: 85) at 1-month interval important improvement despite a median of
Mean age: 55 SD (14) 2 years with low back pain and a mean
pre-intervention ODI of 54 ± 15 points.
Much of the improvement was evident after
the initial prolotherapy injection
∗ RCT randomized control trial (E) Experimental group, (C) Control group, VAS Visual Analog Scale, NRS Numeric Rating Scale, QoL Quality of Life Scale, ODI Oswestry disability index.
L. Giordano et al., 2021, Vol. 138

Downloaded from https://academic.oup.com/bmb/article/138/1/96/6244772 by guest on 31 July 2023

Table 3 Description of the modified CMS

Section Number or factor Score

PART A—only one score to be given for each of the seven sections
1. Study size-number of patients (N) (if multiple follow-up, multiply >60 10
N by number of times subjects followed up) 41–60 7
20–40 4
<20 not stated 0
2. Mean follow-up (months) >24 5
12–24 2
<12, not stated, or unclear 0
3. Number of different non-operative interventions or procedures − One non-operative intervention 10
included in each reported outcome. More than one non-operative − More than one, but >90% of subjects undergoing the one procedure 7
interventions or procedures may be assessed but separated − Not stated, unclear, or < 90% of subjects undergoing the one intervention 0
outcomes should be reported
4. Type of study − Randomized control trial 15
− Prospective cohort study 10
− Retrospective cohort study 0
Prolotherapy for chronic low back pain, 2021, Vol. 138

5. Diagnostic certainty (use of ultrasound, MRI, or postoperative In all 5

histopathology to confirm diagnosis) In >80% 3
In<80%, no stated, or unclear 0
6. Description of non-operative interventions or procedures given Adequate (technique stated and necessary details given) 5
Fair (technique only stated without elaboration) 3
Inadequate, not stated, or unclear 0
7. Description of postoperative rehabilitation Well described with >80% of patients complying 10
Well described with 60–80% of patients complying 5
Protocol not reported or <60–80% of patients complying 0
Part B—scores may be given for each option in each of the three sections if applicable
1. Outcome criteria (F outcome criteria is vague and does not Outcome measures clearly defined 2
specify subjects sporting capacity, score is automatically 0 for this Timing of outcome assessment clearly stated (e.g. at best outcome after surgery or at 2
section follow-up) 3
Use of outcome criteria that has reported good reliability 3
Use of outcome with good sensitivity
2. Procedure for assessing outcomes Subjects recruited 5
Investigator independent 4
Written assessment 3
Completion of assessment by subjects themselves with minimal investigator assistance 3
3. Description of subject selection process Selection criteria reported and unbiased 5
Recruitment rate reported >80% 5
or <80% 3
Eligible subjects not included in the study satisfactorily accounted for or 100% recruitment 5
103

Downloaded from https://academic.oup.com/bmb/article/138/1/96/6244772 by guest on 31 July 2023

104
Fig. 2 Algorithm for determining the strength of a recommendation.
Secondary outcomes, including physical per-
formance testing and medication usage were not
consistently reported. The follow-up period was
usually 6 months.
Results
We identified 12 studies eligible for inclusion in the
present systematic review: 6 randomized controlled
trials, 3 prospective studies and 3 retrospective
studies.
Study population
The total number of patients considered in the stud-
ies is 960 (Table 2) of which 474 (49.3%) were
females and 486 (50.7%) were males with a mean
age of 53 years, ranging from 18 to 90. All studies
L. Giordano et al., 2021, Vol. 138
Downloaded from https://academic.oup.com/bmb/article/138/1/96/6244772 by guest on 31 July 2023
included patients who presented low back pain of
various origin for 6 months and had failed conserva-
tive management.
Evidence of efficacy
The key results for pain and disability are summa-
rized in Table 2. Over the last 30 years, only 6 studies
had a control group and were randomized, while the
majority were retrospective studies or case series. No
correlations of efficacy with age and sex were found.
In 4 works4 ,24,25,31 , the 70–80% of patients reach the
minimal clinically important difference in pain and
disability at 6 months compared to the control group
(ranged 30–45%). However, one of these works4
has a follow-up of only 2 weeks. In 2 rtc26,32 , no
differences were found in achieving the minimal clin-
ically important difference between treatment group
Prolotherapy for chronic low back pain, 2021, Vol. 138
and control group but there are some confounding
elements: The presence of cointerventions and the
method of preparation of the sclerosing solution and
injection technique.
Randomized controlled trials
Multiple-session multiple-location needling of entheses
with phenol/dextrose/glycerine vs anesthetic
Ongley et al.24 randomly assigned 82 patients
with CLBP to receive prolotherapy or control
injections, and many cointerventions. Over six
sessions, each week patients were injected with 20 ml
of prolotherapy solution into the lumbosacral liga-
ments under intravenous sedation. Cointerventions
included injection of 50 mg of triamcinolone in
the gluteus medius, spinal manipulative therapy for
the prolotherapy group, lidocaine 0.5% injection
and false spinal manipulative therapy for the
control group; also, both groups undertook standing
lumbar flexion-extension stretching exercises. The
experimental group fared significantly better for all
outcomes at 6 months. Furthermore, at 6 months 35
of 40 (88%) patients in the experimental group had
more than 50% improvement in disability score,
compared with 16 of 41 (39%) patients in the
control group. By that time, in 10 of 12 (83%)
patients of the experimental group the radiating
lower limb pain present at beginning had solved
vs 2 of 12 (17%) individuals in the control group.
Considering the cointerventions, however, it is
no easy to ascribe the positive outcomes to the
prolotherapy injections.24
Klein et al.25 studied 79 patients with CLBP
who were randomly assigned to have prolotherapy
or control injections. The experimental group
received, prolotherapy injections with a solution
containing dextrose 12.5, phenol 2.3%, glycerin
12.5%, lidocaine 0.5%. Six sessions injecting 30 ml
per session were performed every week into lum-
bosacral facets and ligaments, and sacroiliac joints,
with intravenous sedation and lidocaine. Control
injections contained a solution of saline 0.45% and
lidocaine 0.25%, simultaneously with intravenous
sedation.
105
At 6 months, results were evaluated using the
Roland Morris Disability Questionnaire, pain inten-
sity (VAS), pain grid, lumbar ROM, lumbar isometric
strength and the proportion of subjects with greater
than 50% improvement in disability and pain inten-
sity. Substantial improvement in pain intensity, pain
grid scores and disability occurred in both groups.
When success being defined as a 50% improvement
in pain intensity or disability, in 30 of 39 (77%)
Downloaded from https://academic.oup.com/bmb/article/138/1/96/6244772 by guest on 31 July 2023
patients in the experimental group vs 21 of 40 (53%)
in the control group prolotherapy was efficacious.25
In both studies24 ,25 , there were significant clin-
ically and statistically improvements in treatment
group; however, it is necessary to consider the
confounding effect of cointerventions and needling
effect, so that this would be level-B evidence of
prolotherapy benefit.
In the randomized control trial by Dechow et al.32 .
74 patients with CLBP were randomly assigned
to prolotherapy or control injections. Prolotherapy
injections for the experimental group contain 12.5%
dextrose, 12.5% glycerin, 1.2% phenol and 0.5%
lidocaine. Three sessions injecting 10 ml were inoc-
ulated every week into the L4–L5 ligaments using a
single needle insertion point, with IV sedation. The
solution administered to the control group contained
0.45% saline 0.45 and 0.5% lidocaine. At 0, 1,
3 and 6 months, outcomes were assessed with the
McGill pain questionnaire, ROM (modified Schober
test), modified somatic perception questionnaire,
modified Zung questionnaire, Oswestry Disability
Index, pain grid and pain intensity (VAS). At no
follow-up time point, the treatment and control
groups showed statistically significant differences
for any of the outcomes.32 Considering the several
limitations, one of them is that the injector was not
allowed to examine the patient prior to injection and
doubts regarding the injection technique considered,
recommendation for this study would be no more
than C.
Multiple-session multiple-location needling of entheses with
dextrose vs saline.
The randomized controlled trial by Yelland et al.26
assessed the efficacy of prolotherapy and exercise
106
in 110 patients experiencing CLBP. Patients were
randomly divided in four groups: control injection
with cointerventions or without cointerventions or
prolotherapy injections with cointervention or with-
out cointerventions. The prolotherapy solution con-
tained 20% dextrose, 0.2% lidocaine. The control
injections contained 0.9% saline. The solutions were
injected into lumbo-pelvic ligaments. At 2.5, 4, 6, 12
and 24 months, there were no significant differences
in any of the outcomes, pain intensity (VAS) and dis-
ability scores (Roland-Morris), between the groups
at each follow-up time point.26
This study can reasonably be considered level-
B evidence of efficacy due to long-term sustainable
benefits and a trend favoring dextrose, although was
underpowered due to a comparison of two active
treatments (dextrose and saline) and the significant
influence of co-interventions.
Multiple-session, single location (SI joint) injection
of dextrose vs steroid
In individuals with sacroiliac joint pain, Kim et al.31
in 2010 found a remarkable cumulative incidence of
pain decrease (≥50%) in dextrose vs steroid-injected
patients. The patients received intra-articular 50%
dextrose in water prolotherapy or 40 mg triamci-
nolone acetonide injection, with a biweekly schedule
and up to three injections. The cumulative incidence
of ≥50% pain relief at 15 months was 58.7% (38–
79.5%) in the prolotherapy group and 10.2% (6.7–
27.1%) in the steroid group (P < 0.005). These dif-
ferences among the groups were statistically signif-
icant, and effect of prolotherapy lasted longer than
that of steroid injections.31 With 15-month follow-
up and more precise pain source localization, this
successful RCT outcome is strong level-B evidence
of dextrose prolotherapy benefit.
Single-session, single-location (caudal epidural space) injection
of dextrose vs saline
In a 2016 randomized controlled trial by Maniquis
et al.4 , 35 individuals (54 ± 10.7 years old;
11 female) suffering moderate-to-severe non-surgical
CLBP received a epidural injection of 10 ml of 5%
dextrose or 0.9% saline adopting a vertical caudal
L. Giordano et al., 2021, Vol. 138
inoculation technique. Main result assessed was
the change in the NRS (0–10 points) pain score
between start and 15 min, and 2, 4 and 48 h and
2 weeks post-injection. The secondary outcome
was the percentage of patients who reached 50%
or more in pain improvement 4 h after the index
injection. 84% (16 of 19 pts) of dextrose recipients
and 19% (3 of 16 pts) of saline recipients reported
pain reduction greater than 50%.4 An important
Downloaded from https://academic.oup.com/bmb/article/138/1/96/6244772 by guest on 31 July 2023
limitations of this study its short duration of follow
up; in addition, this study cannot determine if the
analgesic effect reported by dextrose participants is
a one-time response, nor whether pain reduction can
be repeated or endure with additional injections so
the level of recommendation is C.
Case series, prospective and
retrospective studies
In a prospective case series with 76 individuals with
degenerative discogenic leg pain not responding
to other treatment, Miller et al.30 found that and
intradiscal injection of 50% dextrose in water, mixed
50:50 with bupivacaine 0.25% bi-weekly produced
sustained improvement with a reduction in numeric
rating scale pain of 71% in 43% of patients.30
In Hooper et al.29 work, 177 individuals experi-
enced a history of CLBP completed prolotherapy
treatment and were followed for 2 months to
2.5 years. One half milliliter (0.5 ml) of proliferant
containing 20% dextrose and 0.75% xylocaine
was inoculated into the facet joints of the thoracic,
cervical and lumbar spine, or combinations of these
three areas. Ninety-one percent (91.0%) (72 of 79
pts) of patients reported reduction in the level of
pain; 84.8% (68 of 79 patients) of patients reported
improvement in activities of daily living, and 84.3%
(67 of 79 patients) reported an improvement in their
ability to work.29
Lyftog et al.27 in 2008 treated recalcitrant CLBP
with a mean duration of 5.5 years in 41 consecutive
patients presenting over a 1-year period, receiving
a series of subcutaneous prolotherapy treatments.
The authors initially used hypertonic dextrose
20–40%, mixed with 0.1% lignocaine and/or
Prolotherapy for chronic low back pain, 2021, Vol. 138
ropivacaine 0.1% in normal saline. At end of the
treatment, the solution consisted of dextrose 20%,
lignocaine 0.1% and cholecalciferol 1000 IU/ml in
normal saline. The mean duration of treatment was
8 weeks. About 90% (41of 46 patients) of patients
improved by more than 50% (initial mean Visual
Analog Score 0–10 of 7.6), with 29% reaching VAS
0 after a mean treatment duration of 8.3 weeks.27
In a 2010 retrospective case series study, Watson
et al.28 assessed clinical outcomes in 140 patients
with chronic lower back pain involving ligamentous
pathology refractory to conventional medical man-
agement. Initial standard solutions used for injec-
tions included P-25-G (phenol 2.5% plus 25% dex-
trose plus 25% glycerine mixed 50:50 with 1% lido-
caine without epinephrine) or 15% dextrose (2 ml
of 1% lidocaine without epinephrine mixed with
3 ml of 50% dextrose and 5 ml of normal saline).
The number of injection sites varied depending on
how many sites were tender. The volume of solution
injected could vary from 5 to 50 ml. Injections were
spaced 2 weeks to 3 months apart. When patients
were pain free, they were seen 1 year or more after
their final treatment for follow-up, during which the
VAS for pain and QoL scale were administered. Both
pain and QoL scores were significantly improved
at least 1 year after the last treatment with no
differences in outcomes as a result of age, response to
Xylocaine (lidocaine) injection, insurance coverage,
smoking history or gender.28
Cusi et al.33 in 2014 assessed the efficacy of
prolotherapy in the management of insufficient load
transfer of the sacroiliac joint, with a positive clinical
outcome for 76% (19 of 25 patients) of patients
who attended a 3-month follow-up visit (76% at
12 months and 32% at 24 months). The interven-
tions consisted three injections of hypertonic dex-
trose solution (1.8 ml of 50% glucose solution,
2.3 ml of bupivacaine 1% and around 0.8 ml of
Isovue (iopamidol) 300 contrast medium), injected
into the dorsal interosseous ligament of the affected
sacroiliac joint, under CT control, after 6 weeks.33
In the work by Hoffman et al.34 , patients with
CLBP and diagnosed with sacroiliac joint insta-
bility received three sacroiliacs joint prolotherapy
107
injections (3 ml of 50% dextrose and 7 ml of 1%
lidocaine) at 1-month intervals. The retrospective
assessment of the outcome of the treatment also com-
pared various characteristics between those with at
least a minimum clinically important improvement
and those who experienced no improvement. In 24 of
74 patients (23%) with CLBP from symptomatic SI,
joint instability prolotherapy produced the minimum
clinically important improvement for the Oswestry
Downloaded from https://academic.oup.com/bmb/article/138/1/96/6244772 by guest on 31 July 2023
Disability Index of 15 points and so clinically
meaningful functional gains. Patients in whom
prolotherapy did not work generally presented no
improvement after the first prolotherapy injection.34
The works presented in this paragraph are struc-
turally prospective27,30,33 or retrospective28,29,34 case
series with low level of evidence and a C-level of
recommendation.
Modified Coleman Methodology Scores
The modified Coleman Methodology Scores calcu-
lated for each of the studies reviewed are shown
in Table 2. The mean mCMS is 65 (range 55.00–
72.00, standard deviation [SD] 5.4134, 95% CI
61.00–70.40), and so ‘poor quality’ range.22 There
were no studies that were scored in the excellent
or good range for study methodology, three were
scored of fair quality (25%) and the other nine of
poor quality (75%). All the studies showed positive
outcomes at various follow-up, and the Coleman
score showed evidence of a positive association with
the level of the study. In particular, the studies that
included a control group and a follow-up of more
than 6 months had a higher score than the retro-
spective studies with a follow-up greater than 1 year.
The analysis of the studies included in the review
showed how their quality has decreased over the
years, reflecting the need for appropriately powered
well planned and performed randomized control
trials on prolotherapy and CLBP.22
Discussion
Prolotherapy has been undertaken to manage CBLP
for several decades. Based on previously reported
108
results from the individual work analysis, including,
a clinically relevant reduction in pain and impair-
ment, the amount of patients who achieved and
maintained the minimal clinically difference at a
mean follow-up of 6 months and the lack of rele-
vant adverse effects other than transient irritation
of the injection site, prolotherapy with dextrose is
recommended in the management of patients with
CLBP, whereas its use as a single therapy is not
supported by high quality literature, as 70% of the
studies identified in this work were of poor quality
and only four studies24–26,31 received a B-level recom-
mendation while the other eight studies received a
C-level of recommendation.4,28–31,33–35 A large RCT
found that prolotherapy had no greater efficacy than
the normal saline.26 In 2004, the Cochrane Collabo-
ration concluded that ‘There was no evidence that
prolotherapy injections alone were more effective
than control injections alone, but in the presence of
co-interventions, prolotherapy injections were more
effective than control injections, more so when both
injections and co-interventions were controlled con-
currently.’36
A reason for this conclusion is that most stud-
ies, including randomized controlled trials which
involve the use of prolotherapy in the management
of CLBP make only a generic, not specific, clinical
diagnosis for patient selection. Indeed, patient are
selected according to the presence of pain symptoms
in the lower back: the injections are then admin-
istered at or around the painful sites. This could
explain the inconsistency of the results,26,33 as there
is no evidence that the purported proliferation of
soft tissue exerts an intrinsic analgesic effect. Fur-
thermore, the presence of co-interventions and the
clinical heterogeneity of the cohorts considered con-
tributes to confound the conclusions of the vari-
ous studies. Co-interventions included a preliminary
injection with local anesthetic followed by manip-
ulation under sedation, superficial skin injections
of local anesthetic, the injection of gluteal tender
points with triamcinolone and lignocaine, encour-
agement to perform previously painful activities,
vitamin and mineral supplements, and flexion and
extension exercises.36
L. Giordano et al., 2021, Vol. 138
The number of sites to be injected determines
the injected volumes. In some studies, better results
were obtained at higher volumes and with a higher
frequency of injections. An alleged dose response
effect which, however, cannot be confirmed by an
appropriate number of randomized controlled tri-
als.2
Recently, the understanding of the action of
hypertonic dextrose on soft tissues has increased.
Downloaded from https://academic.oup.com/bmb/article/138/1/96/6244772 by guest on 31 July 2023
Most work used a solution of 50% dextrose and
0.25% bupivacaine, with a final osmolarity of
approximately 1423 mOsm (1265 dextrose + 158.5
bupivacaine). Given this high osmolarity, Reeves
et al. sustained that the rapid onset of pain relief
may be associated to chemoneuromodulation of the
near nociceptors.32
The hyperosmolar solution may also re-modulate
the local vascular hemodynamics, and a consequent
decrease in nociceptive activity. The exact mecha-
nism leading a rapid improvement in pain experi-
enced by these patients is not known. On the other
hand, the demonstrated durability of the response is
not associated only with chemomodulatory effects:
lasting benefit results from the growth factor release
and tissue-stabilizing effects of dextrose on the
annulus fibrosus of the intervertebral disc. The
potential tissue stabilizing benefits of prolotherapy
dextrose injection may occur in ligaments, tendons
and cartilage, since in chondrocytes and fibroblasts
an anabolic response is triggered by platelet derived
growth factor, transforming growth factor beta,
insulin like growth factor, basic fibroblast growth
factor and connective tissue growth factor. The
hyperosmolar nature of the dextrose solution may
also have a certain importance. Exposure of human
cells to an osmolarity change of a little as 50 mOsm
activates enzymes such as phosphate donors, i.e.
kinases, which may exert a beneficial growth
effects.37–39
The main adverse effect associated to prolother-
apy is a momentary (12–96 h post injection) rise of
pain and/or stiffness at the site of injection: this is a
consequence of the acute inflammation produced
by the injection of prolotherapy. Other adverse
effects such as headache, increased transitory leg
Prolotherapy for chronic low back pain, 2021, Vol. 138
pain, diarrhea, nausea, minor allergic reactions and
other transient symptoms are reported.40 Adverse
events related to prolotherapy for CLBP are similar
to those reported following other common spinal
injection procedures40 : severe headache caused by
inadvertent lumbar injection, neurological leg pain,
altered sleep because of psychological trauma from
punctures, and cough. No catastrophic outcomes
have been associated to prolotherapy for low back
pain.41 None of the studies included in the work
presented a prevalence of a particular adverse effect.
A well-accepted protocol for patients with CLBP
not responding to other approaches suggests injec-
tions for six weekly of 20–30 ml dextrose with
50%/glycerin/phenol/lidocaine with spinal manipu-
lation therapy and exercise.34,35 It is unclear whether
this regimen produces different results from the other
regimens routinely employed.
The present investigation is currently the largest
collection of work for 30 years on the use of pro-
lotherapy for CLBP, and includes not only RTC but
also case series, retrospective and prospective studies.
Given the methodological imperfections and hetero-
geneity of the investigations considered, this system-
atic review could not draw definitive conclusions,
but certainly identifies some interesting clinically
relevant points on a topic which has seen exponential
growth in recent years thanks to the results obtained
for various chronic diseases, including arthritis of
the knee, rotator cuff tendinopathy, Osgood Schlat-
ter lesion, hand arthrosis, lateral epicondylosis and
sacroiliac joint dysfunction.4
Limitations
The work has several limitations especially related to
small number of studies selected for this systematic
review, their study design and the lack of high-
quality trials, as 70% of the studies identified were of
poor quality. As mentioned above, all the works dif-
fer from each other for presence of cointerventions
(e.g. physical activity, massage therapy, anesthetics),
number, techniques and site of injection, concentra-
tion of injected ingredients, time span between each
109
injection and follow-up. This extreme heterogeneity
leads to difficulty in carrying out a quantitative
analysis and negatively affects the level of evidence.
Finally, although aware of the existence of valid
bias detection systems,42 the authors are sure that
the tools used in the work22,23 are able to guaran-
tee a good quality assessment and fruibility for the
readers.
Downloaded from https://academic.oup.com/bmb/article/138/1/96/6244772 by guest on 31 July 2023
Conclusion
Dextrose prolotherapy can provide improvement in
pain, functional status and patient satisfaction in
patients with CLBP, with no long-term or perma-
nent adverse reactions reported although efficacy as
an isolated therapy is not yet supported by solid
scientific evidence.
Further appropriately powered RCT studies are
needed to set the standard of care of prolotherapy
for patients with CLBP.
Conflict of interest statement
The authors have no potential conflicts of interest.
Funding
No funds have been received for this work.
Data Availability Statement
No new data were generated or analyzed in support
of this review.
References
1. Woolf AD, Pfleger B. Burden of major musculoskeletal
conditions. Bull World Health Organ 2003;81:646–56.
2. Dagenais S, Caro J, Haldeman S. A systematic review of
low back pain cost of illness studies in the United States
and internationally. Spine J Off J North Am Spine Soc.
2008;8:8–20.
110
3. Ekman M, Jönhagen S, Hunsche E et al. Burden of
illness of chronic low back pain in Sweden: a cross-
sectional, retrospective study in primary care setting.
Spine 2005;30:1777.
4. Maniquis-Smigel L, Dean Reeves K, Jeffrey Rosen H
et al. Short term analgesic effects of 5% dextrose epidu-
ral injections for chronic low back pain: a randomized
controlled trial. Anesthesiol Pain Med 2016;7:e42550.
5. Wilkerson RG, Kim HK, Windsor TA et al. The opioid
epidemic in the United States. Emerg Med Clin North
Am 2016;34:e1–e23.
6. Rahimzadeh P, Sharma V, Imani F et al. Adjuvant
hyaluronidase to epidural steroid improves the qual-
ity of analgesia in failed back surgery syndrome: a
prospective randomized clinical trial. Pain Physician
2014;17:E75–82.
7. Patel VB, Wasserman R, Imani F. Interventional
therapies for chronic low back pain: a focused
review (efficacy and outcomes). Anesthesiol Pain Med
2015;5:e29716.
8. Veizi E, Hayek S. Interventional therapies for chronic
low back pain. Neuromodulation J Int Neuromodula-
tion Soc 2014;17:31–45.
9. Hassan F, Trebinjac S, Murrell WD et al. The effec-
tiveness of prolotherapy in treating knee osteoarthri-
tis in adults: a systematic review. Br Med Bull 01
2017;122:91–108.
10. Hackett GS. Prolotherapy in whiplash and low back
pain. Postgrad Med 1960;27:214–9.
11. Jensen KT, Rabago DP, Best TM et al. Response of knee
ligaments to prolotherapy in a rat injury model. Am J
Sports Med 2008;36:1347–57.
12. Burdakov D, Jensen LT, Alexopoulos H et al. Tandem-
pore K+ channels mediate inhibition of orexin neurons
by glucose. Neuron 2006;50:711–22.
13. Oh JH, Ha H, Yu MR et al. Sequential effects of high glu-
cose on mesangial cell transforming growth factor-beta
1 and fibronectin synthesis. Kidney Int 1998;54:1872–8.
14. Yoshii Y, Zhao C, Schmelzer JD et al. The effects of
hypertonic dextrose injection on connective tissue and
nerve conduction through the rabbit carpal tunnel. Arch
Phys Med Rehabil 2009;90:333–9.
15. Rabago D, Best TM, Zgierska AE et al. A systematic
review of four injection therapies for lateral epicondylo-
sis: prolotherapy, polidocanol, whole blood and platelet-
rich plasma. Br J Sports Med luglio 2009;43:471–81.
16. Wu Y-T, Ke M-J, Ho T-Y et al. Randomized double-
blinded clinical trial of 5% dextrose versus triamci-
nolone injection for carpal tunnel syndrome patients.
Ann Neurol 2018;84:601–10.
L. Giordano et al., 2021, Vol. 138
17. Wu Y-T, Ho T-Y, Chou Y-C et al. Six-month efficacy
of perineural dextrose for carpal tunnel syndrome: a
prospective, randomized, double-blind. Controlled Trial
Mayo Clin Proc 2017;92:1179–89.
18. Shen Y-P, Li T-Y, Chou Y-C et al. Comparison of per-
ineural platelet-rich plasma and dextrose injections for
moderate carpal tunnel syndrome: a prospective ran-
domized, single-blind, head-to-head comparative trial. J
Tissue Eng Regen Med 2019;13:2009–17.
19. Di Paolo S, Gesualdo L, Ranieri E et al. High glu-
cose concentration induces the overexpression of trans-
Downloaded from https://academic.oup.com/bmb/article/138/1/96/6244772 by guest on 31 July 2023
forming growth factor-beta through the activation of a
platelet-derived growth factor loop in human mesangial
cells. Am J Pathol 1996;149:2095–106.
20. Yoshii Y, Zhao C, Schmelzer JD et al. Effects of hyper-
tonic dextrose injections in the rabbit carpal tunnel.
J Orthop Res Off Publ Orthop Res Soc 2011;29:
1022–7.
21. Yoshii Y, Zhao C, Schmelzer JD et al. Effects of multiple
injections of hypertonic dextrose in the rabbit carpal
tunnel: a potential model of carpal tunnel syndrome
development. Hand N Y N 2014;9:52–7.
22. Moher D, Liberati A, Tetzlaff J et al. Preferred reporting
items for systematic reviews and meta-analyses: the
PRISMA statement. PLoS Med 2009;6:e1000097.
23. Coleman BD, Khan KM, Maffulli N et al. Studies of sur-
gical outcome after patellar tendinopathy: clinical signif-
icance of methodological deficiencies and guidelines for
future studies. Victorian Institute of Sport Tendon Study
Group. Scand J Med Sci Sports 2000;10:2–11.
24. Ebell MH, Siwek J, Weiss BD et al. Strength of rec-
ommendation taxonomy (SORT): a patient-centered
approach to grading evidence in the medical literature.
Am Fam Physician 2004;69:548.
25. Ongley MJ, Klein RG, Dorman TA et al. A new
approach to the treatment of chronic low back pain.
Lancet Lond Engl 1987;2:143–6.
26. Klein RG, Eek BC, DeLong WB et al. A randomized
double-blind trial of dextrose-glycerine-phenol injec-
tions for chronic, low back pain. J Spinal Disord
1993;6:23–33.
27. Yelland MJ, Glasziou PP, Bogduk N et al. Prolotherapy
injections, saline injections, and exercises for chronic
low-back pain: a randomized trial. Spine 2004;29:9–16;
discussion 16.
28. Lyftogt J. Prolotherapy for recalcitrant lumbago. Aus-
tralas Musculoskelet Med 2008;13:18.
29. Watson JD, Shay BL. Treatment of chronic low-back
pain: a 1-year or greater follow-up. J Altern Comple-
ment Med 2010;16:951–8.
Prolotherapy for chronic low back pain, 2021, Vol. 138
30. Hooper RA, Ding M. Retrospective case series on
patients with chronic spinal pain treated with dex-
trose prolotherapy. J Altern Complement Med N Y N
2004;10:670–4.
31. Miller MR, Mathews RS, Reeves KD. Treatment of
painful advanced internal lumbar disc derangement with
intradiscal injection of hypertonic dextrose. Pain Physi-
cian 2006;9:115–21.
32. Kim WM, Lee HG, Won Jeong C et al. Randomized
controlled trial of intra-articular prolotherapy versus
steroid injection for sacroiliac joint pain. J Altern Com-
plement Med 2010;16:1285–90.
33. Dechow E, Davies RK, Carr AJ et al. A random-
ized, double-blind, placebo-controlled trial of scleros-
ing injections in patients with chronic low back pain.
Rheumatol Oxf Engl 1999;38:1255–9.
34. Cusi M, Saunders J, Hungerford B et al. The use of
prolotherapy in the sacroiliac joint. Br J Sports Med
2010;44:100–4.
35. Hoffman MD, Agnish V. Functional outcome from
sacroiliac joint prolotherapy in patients with sacroil-
iac joint instability. Complement Ther Med 2018;37:
64–8.
36. Yelland MJ, Mar C, Pirozzo S et al. Prolotherapy injec-
tions for chronic low-back pain. Cochrane Database
Syst Rev 2004;CD004059.
111
37. Berl T, Siriwardana G, Ao L et al. Multiple mitogen-
activated protein kinases are regulated by hyperosmolal-
ity in mouse IMCD cells. Am J Physiol 1997;272:F305–
11.
38. Okuda Y, Adrogue HJ, Nakajima T et al. Increased
production of PDGF by angiotensin and high
glucose in human vascular endothelium. Life Sci
1996;59:1455–61.
39. Caruccio L, Bae S, Liu AY et al. The heat-shock tran-
scription factor HSF1 is rapidly activated by either
hyper- or hypo-osmotic stress in mammalian cells.
Downloaded from https://academic.oup.com/bmb/article/138/1/96/6244772 by guest on 31 July 2023
Biochem J 1997;327:341–7.
40. Dagenais S, Ogunseitan O, Haldeman S et al. Side effects
and adverse events related to intraligamentous injection
of sclerosing solutions (prolotherapy) for back and neck
pain: a survey of practitioners. Arch Phys Med Rehabil
2006;87:909–13.
41. Dagenais S, Haldeman S, Wooley JR. Intraligamen-
tous injection of sclerosing solutions (prolotherapy) for
spinal pain: a critical review of the literature. Spine J Off
J North Am Spine Soc 2005;5:310–28.
42. Jørgensen L, Paludan-Müller AS, Laursen DRT et al.
Evaluation of the Cochrane tool for assessing risk of
bias in randomized clinical trials: overview of published
comments and analysis of user practice in Cochrane and
non-Cochrane reviews. Syst Rev 2016;5:80.