Chester Clinic

 

Efficacy of low-level laser therapy in the management of neck pain: a systematic review and meta-analysis of randomised placebo or active-treatment controlled trials

Effi cacy of low-level laser therapy in the management of
neck pain: a systematic review and meta-analysis of
randomised placebo or active-treatment control ed trials
Roberta T Chow, Mark I Johnson, Rodrigo A B Lopes-Martins, Jan M Bjordal
Background Neck pain is a common and costly condition for which pharmacological management has limited Lancet 2009; 374: 1897–908
evidence of effi cacy and side-eff ects. Low-level laser therapy (LLLT) is a relatively uncommon, non-invasive treatment Published Online
for neck pain, in which non-thermal laser irradiation is applied to sites of pain. We did a systematic review and meta- November 13, 2009
analysis of randomised controlled trials to assess the effi cacy of LLLT in neck pain.
See Comment page 1875
We searched computerised databases comparing effi cacy of LLLT using any wavelength with placebo or with
active control in acute or chronic neck pain. Eff ect size for the primary outcome, pain intensity, was defi ned as a Nerve Research Foundation,
Brain and Mind Research
pooled estimate of mean diff erence in change in mm on 100 mm visual analogue scale.
Institute, University of Sydney,
Sydney, NSW, Australia
Findings We identifi ed 16 randomised controlled trials including a total of 820 patients. In acute neck pain, results of (R T Chow MBBS); Faculty of
two trials showed a relative risk (RR) of 1·69 (95% CI 1·22–2·33) for pain improvement of LLLT versus placebo. Five Health, Leeds Metropolitan
University, Leeds, UK
trials of chronic neck pain reporting categorical data showed an RR for pain improvement of 4·05 (2·74–5·98) of (Prof M I Johnson PhD);
LLLT. Patients in 11 trials reporting changes in visual analogue scale had pain intensity reduced by 19·86 mm Institute of Biomedical
(10·04–29·68). Seven trials provided follow-up data for 1–22 weeks after completion of treatment, with short-term Sciences, Pharmacology
Department, University of
pain relief persisting in the medium term with a reduction of 22·07 mm (17·42–26·72). Side-eff ects from LLLT were São Paulo, São Paulo, Brazil
mild and not diff erent from those of placebo.
(Prof R A B Lopes-Martins PhD);
Faculty of Health and Social
Interpretation We show that LLLT reduces pain immediately after treatment in acute neck pain and up to 22 weeks Science, Institute of
Physiotherapy, Bergen
after completion of treatment in patients with chronic neck pain.
University College, Bergen,
Norway (Prof J M Bjordal PT);
and Section of Physiotherapy
Science, Institute of Public
Health and Primary Health
treatment option,16 possibly because of scepticism about Care, University of Bergen,
Chronic pain is predicted to reach epidemic proportions in its mechanism of action and eff ectiveness.17 Research Bergen, Norway
developed countries with ageing populations in the next from the past decade suggests that LLLT produces anti- (Prof J M Bjordal)
30 years.1 Chronic neck pain is a highly prevalent condition, infl ammatory eff ects,18–21 contributing to pain relief. Correspondence to:
aff ecting 10–24% of the population.2–5 Economic costs of Cochrane reviews of the effi cacy of LLLT in low-back Dr Roberta T Chow, Honorary
Research Associate, Nerve this condition are estimated at hundreds of millions of pain22 and rheumatoid arthritis23 have been unable to Research Foundation, Brain and dollars,2 creating an imperative for evidence-based, cost- make fi rm conclusions because of insuffi cient data or Mind Research Institute, eff ective treatments. Low-level laser therapy (LLLT) uses confl icting fi ndings. However, eff ectiveness depends on University of Sydney, laser to aid tissue repair,6 relieve pain,7 and stimulate factors such as wavelength, site, duration, and dose of 100 Mal ett Street, Sydney, NSW 2050, Australia acupuncture points.8 Laser is light that is generated by LLLT treatment. Adequate dose and appropriate [email protected]
high-intensity electrical stimulation of a medium, which procedural technique are rarely considered in systematic can be a gas, liquid, crystal, dye, or semiconductor.9 The reviews of electrophysical agents. Research into the dose-light produced consists of coherent beams of single response profi le of LLLT suggests that diff erent wavelengths in the visible to infrared spectrum, which can wavelengths have specifi c penetration abilities through be emitted in a continuous wave or pulsed mode. Surgical human skin.17,24,25 Thus, clinical eff ects could vary with applications of laser ablate tissue by intense heat and are depth of target tissue. We have shown the importance of diff erent from LLLT, which uses light energy to modulate accounting for dose and technique in systematic reviews cell and tissue physiology to achieve therapeutic benefi t of transcutaneous electrical nerve stimulation26 and without a macroscopic thermal eff ect (sometimes termed LLLT,11,21 and our approach is an acknowledged means of cold laser). LLLT is non-invasive, painless, and can be easily establishing effi cacy.27 administered in primary-care settings. Incidence of adverse The only systematic review focusing solely on LLLT in eff ects is low and similar to that of placebo, with no reports treatment of neck pain included four randomised of serious events.10,11 controlled trials, and concluded that there was evidence Research into the use of LLLT for pain reduction12,13 and of short-term benefi t of LLLT at infrared wavelengths of tissue repair14,15 spans more than 30 years. However, 780, 810–830, and 904 nm.28 A Cochrane review of reports do not identify this therapy as a potential physical medicine for mechanical neck disorders, since www.thelancet.com Vol 374 December 5, 2009
1982, to July, 2008), the Physiotherapy Evidence Database 490 citations identified by search strategy (January, 1929, to July, 2008), Biosis (January, 1926, to July,
2008), Allied and Complementary Medicine (January,
355 irrelevant or duplicates identified 1985, to July, 2008), and the Cochrane Central Register of through title or abstract review Controlled Trials (second quarter of 2008). Keywords used for neck pain and related conditions were: "neck pain/ 135 articles reviewed in detail strain", "cervical pain/strain/syndrome", "cervical spon-dylosis/itis", "cervicobrachial (pain/disorder/syndrome)", "myofascial (pain/disorder/syndrome)", "trigger points", 97 excluded (case series, mixed conditions, region of pain unrelated "fi bromyalgia", "whiplash/WAD", "osteoarthritis/arthritis", to neck pain, narrative reviews) and "zygaphophyseal/ZG joints". We combined these keywords with synonyms for LLLT: "low-level/low-energy/ 38 potentially relevant RCTs identified low reactive-level/low-intensity/low-incident/low-output/infrared/diode/semiconductor/soft or cold or mid/visible"; "laser therapy", "(ir)radiation", "treatment"; "low- 1 infrared light as a heat source energy photon therapy"; "low output laser"; "LLLT"; 2 sham laser used as placebo-control "LILT"; "LEPT"; "LELT"; "LILI"; "LELI"; "LPLI"; "bio- for another modality 5 no control group stimula tion"; "photobio/stimulation/activation/modu la- 1 retrospective study tion"; "light therapy"; "phototherapy"; "narrow band light 1 dental application only therapy"; "904 nm"; "830 nm"; "632 nm"; "1064 nm"; 2 fibromyalgia treated 1 no pain measure "GaAs"; "GaAlAs"; "HeNe"; and "defocused CO ". We 1 only one patient with neck pain consulted experts and searched reference lists of retrieved 6 cannot separate neck pain data 1 changed laser parameters during trial reports and textbooks for additional references.
Citations were screened and full reports of potentially relevant studies obtained. We applied inclusion and 16 potentially appropriate RCTs to be exclusion criteria, assessed methodological criteria, and included in the meta-analysis extracted data including trial characteristics, demographic data, laser parameters, pain outcome measures, and co- 16 RCTs with usable information by interventions. Non-English language studies were outcome and included in the translated by JMB.
We included randomised or quasi-randomised Figure 1: Selection process
controlled trials of LLLT for acute or chronic neck pain as RCT=randomised controlled trial.
defi ned by trial investigators, and identifi ed by various clinical descriptors included under the term non-specifi c withdrawn because much time had passed without an neck pain.31 These diagnostic labels included neck strain, update, included three LLLT trials, for which outcomes neck sprain, mechanical neck disorders, whiplash, neck did not diff er from those of placebo.29 The same disorders, and neck and shoulder pain. We also used investigators did a meta-analysis30 of 88 randomised surrogate terms for neck pain, such as myofascial pain controlled trials of conservative treatments for acute, and trigger points.32,33 Study participants were restricted subacute, and chronic mechanical neck disorders, which to those aged 16 years and older. We excluded studies in included eight trials using LLLT. They concluded that which specifi c pathological changes could be identifi ed, LLLT has intermediate and long-term benefi ts. such as systemic infl ammatory conditions—eg, These reviews did not identify treatment variables rheumatoid associated with positive outcomes, include non-English fi bromyalgia, neck pain with radiculopathy, and neck language publications, or quantitatively assess data.28,30 pain related to neurological disease. We excluded We have therefore undertaken a new systematic review abstracts and studies for which outcome measures for and meta-analysis of LLLT in neck pain to establish neck pain could not be separated from data for other whether LLLT relieves acute and chronic neck pain and regions of the body. Two reviewers (RTC, JMB) to systematically assess parameters of laser therapy to independently undertook the search of published work, identify treatment protocols and dose ranges (therapeutic screened studies, and extracted data. Any disagreements windows) associated with positive outcomes.
between reviewers were resolved by consensus with other team members acting as arbiters (RABL-M, MIJ).
Investigators had to have used a laser device that Search strategy and selection criteria
delivered irradiation to points in the neck identifi ed by We did a search of published work without language tenderness, local acupuncture points, or on a grid at restriction using Medline (January, 1966, to July, 2008), predetermined points overlying the neck. Control groups Embase (January, 1980, to July, 2008), Cinahl (January, had to have been given either placebo laser in which an www.thelancet.com Vol 374 December 5, 2009
Jadad score Control
Primary pain outcome measure
Cervical myofascial Tender points in neck and distal acupuncture points 60 DB, RCT Cervical Tender points in neck Taverna et al 40 DB, RCT Chronic myofascial Tender points in neck Graded subjective assessment: no change to optimum 39 DB, RCT Cervical pain Site not specifi ed No physical or medical therapy allowed Graded subjective assessment: exacerbation to excellent DB, RCT Acute cervical pain Site not specifi ed No NSAIDs or other medical or physical Graded subjective assessment: exacerbation to excellent DB, RCT Neck pain with Three most painful trigger Simple analgesic drugs allowed as trigger points in needed; NSAIDs, corticosteroids, tricyclic antidepressants excluded; no physical Özdemir et al 60 DB, RCT Neck pain related to 3 Six arbitrary points over neck NR neck osteoarthritis 48 DB, RCT Chronic cervical Local neck points and distal Acupuncture not allowed less than 6 acupuncture points months before inclusion; drug therapy unchanged during trial 62 DB, RCT Neck pain with one 3 One active trigger point in levator scapulae or trapezius DB, RCT Neck pain (non- Multiple tender points in Simple analgesic drugs allowed; no cervical spine and physical therapies 60 DB, RCT Chronic myofascial Up to ten trigger points 40 DB, RCT Myofascial pain Trigger points in upper Simple analgesic drugs as needed; exercise to all groups DB, RCT Cervical myofascial Three trigger points No NSAIDs or analgesic drugs; exercise bilaterally and one trigger and graded assessment point in trapezius Local and distal acupuncture Both groups wore cervical collar; Assessment of subjective pain paracetamol and chlormezanone 90 DB, RCT Non-specifi c neck Local tender points Simple analgesic drugs allowed; no physical therapies Dundar et al 64 DB, RCT Cervical myofascial Three trigger points No NSAIDs or analgesic drugs n=number of patients. DB=double blind. RCT=randomised controlled trial. NR=not reported. VAS=visual analogue scale. NSAIDs=non-steroidal anti-infl ammatory drugs. SB=single blind.
Table 1: Study design and outcome measures
identical laser device had an active operating panel with Assessment of methodological quality and
the laser emission deactivated or an active treatment heterogeneity
control (eg, exercise). We also included trials in which an Reviewers assessed all studies for methodological quality
active control was used as a co-intervention in placebo on the basis of Jadad criteria (maximum score 5).34 Jadad
and real laser groups.
criteria allocate a point each for randomisation, double- To be eligible for inclusion, a study had to compare blind design, and description of dropouts. If pain relief along a 0–100 mm visual analogue scale, a randomisation and double-blind concealment are numerical rating scale, or by patient-reported assured, an additional 2 points are added. If randomisation improvement (eg, categorical report of no change to or double-blind concealment is not assured, a point is complete relief of pain) as a primary outcome measure deducted for each. A trial with a score of 3 or more is before and after laser therapy. Functional measures of regarded as high quality. Data from trials with scores of 3 disability (eg, neck pain disability questionnaire) were or more were grouped and analysed separately from assessed as secondary outcome measures. We also those scoring less than 3. examined adverse events where reported, although did We assessed clinical heterogeneity by considering not specify these a priori. Duration of follow-up was population diff erence in age, sex, duration of symptoms, assessed and defi ned as short term (<1 month), medium- and outcomes. Clinical judgment was used to establish term (1–6 months), and long term (>6 months). whether trials were suffi ciently similar to allow pooling www.thelancet.com Vol 374 December 5, 2009
Method score 3 or above
Soriano et al (1996)39 2·47 (1·60–3·82) Total events: 35 (laser therapy) 13 (placebo control) Test for overal effect: Z=4·09 (p<0·0001) Method score below 3
Aigner et al (2006)40 0·88 (0·52–1·49) Total events: 12 (laser therapy) 13 (placebo control) Test for overal effect: Z=0·47 (p=0·64) Total events: 47 (laser therapy) 26 (placebo control) Test for heterogeneity: χ2=8·86, df=1 (p=0·003), l2=88·7% Test for overal effect: Z=3·15 (p=0·002) Figure 2: Relative risk of improvement in acute neck pain in laser-treated versus control groups in two randomised trials reporting categorical data
RR=relative risk.
of data. The specifi c parameters of laser devices, was a combined outcome measure without units—ie, the application techniques, and treatment protocols were standardised mean diff erence in change between active extracted and tabulated by laser wavelength. Details for laser groups and placebo groups for all included trials, power output, duration of laser irradiation, number of weighted by the inverse of the variance for each study.36points irradiated, and frequency and number of Mean diff erences of change for laser-treated and control treatments were listed. When specifi c details were not groups and their respective SDs were included in the reported, calculations were made from those described statistical pooling. If variance data were not reported as in the report when possible. When crucial parameters SDs, they were calculated from the trial data of sample size were not reported, we contacted manufacturers of laser and other variance data values such as p values, t values, devices and trial investigators to obtain missing SE, or 95% CI. Results were presented as weighted mean information. Not all data were available because of the diff erence between laser-treated and control with 95% CI time elapsed since publication of some studies. in mm on visual analogue scale—ie, as a pooled estimate Heterogeneity was qualitatively assessed for these factors of the mean diff erence in change between the laser-treated by an expert in laser therapy (JMB).
and control groups, weighted by the inverse of the variance We used fi ve levels of evidence to describe whether for each study.37 Statistical heterogeneity was assessed for treatment was benefi cial: strong evidence (consistent signifi cance (p<0·05) with Revman 4.2, and χ² and F values fi ndings in several high-quality randomised controlled greater than 50%. For categorical data, we calculated trials); moderate evidence (fi ndings from one high- combined RRs for improvement, with 95% CI. A fi xed quality randomised controlled trial or consistent fi ndings eff ect model was used unless statistical heterogeneity was in several low-quality trials); limited evidence (one low- signifi cant (p<0·05), after which a random eff ects model quality randomised trial); unclear evidence (inconsistent was used. Publication bias was assessed by graphical plot.38 or contradictory results in several randomised trials); and Revman 4.2 was used for statistical analysis and Microsoft no evidence (no studies identifi ed).35 Excel 2003 (version 11) to plot dose-response curves.
Role of the funding source
Eff ect size for the primary outcome, pain intensity, was There was no funding source for this study. The
defi ned as a pooled estimate of the mean diff erence in corresponding author had full access to all the data in the
change in mm on a 100 mm visual analogue scale study and had fi nal responsibility for the decision to
between the mean of the treatment and the placebo submit for publication.
groups, weighted by the inverse of the SD for every
study—ie, weighted mean diff erence of change between Results
groups. Variance was calculated from the trial data and We identifi ed 16 randomised controlled trials of a possible
given, with 95% CI, in mm on visual analogue scale. For 38 that were suitable for inclusion, and that included
categorical data, reported pain relief was dichotomised 820 patients (fi gure 1). Two trials39,40 provided data for laser
into two categories (improvement or no improvement), therapy of acute neck pain, one treating acute whiplash-
and we calculated relative risk (RR) of improvement, with associated disorders and one treating acute neck pain of no
95% CI. For the secondary outcome, disability, eff ect size defi ned cause. The other 14 trials reported response of
was defi ned as the standardised mean diff erence, which chronic non-specifi c neck pain without radiculopathy to
www.thelancet.com Vol 374 December 5, 2009
Chronic non-specific neck pain method score 3 or above
Taverna et al (1990)52 8·10 (1·13–57·82) Toya et al (1994)53 4·21 (1·67–10·60) Gur et al (2004)46 10·00 (2·56–39·06) Chow et al (2004)42 3·50 (0·95–12·90) Chow et al (2006)13 2·85 (1·76–4·59) Total events: 86 (treatment), 22 (control) Test for heterogeneity: χ2=4·31, df=4 (p=0·37), l2=7·2% Test for overal effect: Z=7·02 (p<0·0001) Total events: 86 (treatment), 22 (control) Test for heterogeneity: χ2=4·31, df=4 (p=0·37), l2=7·2% Test for overal effect: Z=7·02 (p<0·0001) Favours treatment Figure 3: Relative risk of global improvement in laser-treated versus control groups in fi ve trials reporting categorical data for improvement in chronic pain
RR=relative risk.
mean (SD)
mean (SD)
Method quality 3/5 or above
Ceccherelli et al (1989)43 –6·30 (16·50) 43·50 (26·09 to 60·91) Flöter et al (1990)45 11·30 (2·18 to 20·42) Laakso et al (1997)49 (high IR) 14·00 (–5·30 to 33·30) Laakso et al (1997)49 (low IR) 5·00 (–19·43 to 29·43) Seidel et al (2002)51 (30 mW) 1·30 (–18·45 to 21·05) Seidel et al (2002)51 (7 mW) 12·00 (–6·45 to 30·45) Özdemir et al (2001)50 48·00 (39·66 to 56·34) Gur et al (2004)46 32·00 (14·48 to 49·52) Hakgüder et al (2003)47 29·20 (17·76 to 40·64) Chow et al (2004)42 20·00 (4·68 to 35·32) Altan et al (2005)41 4·00 (0·50 to 7·50) Chow et al (2006)13 –3·00 (21·00) 30·00 (21·32 to 38·68) Dundar et al (2006)44 –1·00 (–16·48 to 14·48) Subtotal
19·65 (9·27 to 30·03)
Test for heterogeneity: χ2=136·76, df=12 (p<0·00001), l2=91·2% Test for overal effect: Z=3·71 (p=0·0002) Methodological quality below 3
Ilbuldu et al (2004)48 22·50 (6·54 to 38·46) 22·50 (6·54 to 38·46)
Test for overal effect: Z=2·76 (p=006) 19·86 (10·04 to 29·68)
Test for heterogeneity: χ2=137·76, df=13 (p<0·0001), l2=90·6% Test for overal effect: Z=3·96 (p<0·0001) Favours laser therapy Figure 4: Weighted mean diff erence in chronic pain reduction on 100 mm visual analogue scale between laser-treated and placebo-treated groups from 11 randomised trials grouped
according to Jadad criteria
WMD= weighted mean diff erence. IR=infrared.
laser therapy.13,41–53 Of the studies included, 648 (79%) of the non-steroidal anti-infl ammatory drugs. Four studies sample of patients with chronic neck pain were women, allowed use of simple analgesic drugs as needed. and patients had a mean age of 43 years (SD 9·8), mean Methodological quality assessment values for the trials by symptom duration of 90 months (SD 36·9), and mean Jadad scoring ranged from 0 to 5 (table 1).
baseline pain of 56·9 mm (SD 7·5) on a 100 mm visual Analysis of categorical data for immediate before and analogue scale in any trial. Co-interventions were after LLLT eff ects showed that LLLT groups in the two inconsistently reported (table 1). Ten trials reported trials39,40 of acute neck pain had a signifi cant RR of 1·69 co-interventions, and six studies did not report or limit (95% CI 1·22–2·33) for improvement immediately after co-interventions. Of the studies reporting co-interventions, treatment versus placebo (fi gure 2). Methodological fi ve groups of investigators explicitly excluded use of quality varied between these two studies. Five trials of concurrent physical therapies, and four excluded use of chronic neck pain reported categorical data, and all were www.thelancet.com Vol 374 December 5, 2009
mean (SD)
mean (SD)
Fol ow-up 1–4 weeks after end of treatment
Seidel et al (2002)51 (30 mW) –4·60 (–22·27 to 13·07) Seidel et al (2002)51 (7 mW) 5·50 (–12·81 to 23·81) Gur et al (2004)46 35·90 (19·58 to 52·22) Hakgüder et al (2003)47 26·40 (16·98 to 35·82) Subtotal
20·46 (13·60 to 27·33)
Test for heterogeneity: χ2=15·26, df=3 (p=0·002), l2=80·3% Test for overal effect: Z=5·84 (p=0·0001) Fol ow-up 10–22 weeks after end of treatment
Ceccherelli et al (1989)43 –6·60 (18·20) 44·80 (33·60 to 56·00) Gur et al (2004)46 20·80 (6·33 to 35·27) Ilbuldu et al (2004)48 5·20 (–12·40 to 22·80) Altan et al (2005)41 12·40 (1·84 to 22·96) 23·44 (17·11 to 29·77)
Test for heterogeneity: χ2=22·43, df=3 (p<0·0001), l2=86·6% Test for overal effect: Z=7·26 (p=0·0001) 22·07 (17·42 to 26·72)
Test for heterogeneity: χ2=38·08, df=7 (p<0·0001), l2=81·6% Test for overal effect: Z=9·29 (p<0·0001) Figure 5: Weighted mean diff erence in pain reduction on 100 mm visual analogue scale between placebo-treated and laser-treated groups in seven trials reporting follow-up data
WMD= weighted mean diff erence.
high-quality trials with methodological scores of 3 or pain, symptom duration, age, and sex, and we did not more. RR of pain improvement with LLLT was 4·05 detect any clinical heterogeneity (data not shown). Laser (2·74–5·98) compared with placebo at the end of parameters and application techniques, including treatment (fi gure 3). treatment protocols, were heterogeneous (table 2). Laser Analysis of data from visual analogue scale showed that irradiation was applied to an average of 11 points (range in patients in 13 groups in 11 trials, irrespective of 3–25) in the neck. Energy delivered per point ranged methodological quality, pain intensity was reduced by a from 0·06 to 54·00 J, with irradiation durations of mean value of 19·86 mm (10·04–29·68) compared with 1–600 s. Patterns of treatment ranged from a one-off placebo groups (fi gure 4). Seven trials with eight LLLT treatment to a course of 15 treatments, which were groups provided follow-up data for 1–22 weeks after end administered daily to twice a week. On average, of treatment (fi gure 5). The pain-relieving eff ect in the participants received a course of ten treatments. Visible short term (<1 month) persisted into the medium term (632·8 and 670·0 nm) and infrared (820–830, 780, and (up to 6 months). Five studies provided evidence for 904 nm) wavelengths were used at average power outputs improvement in disability at end the of LLLT treatment ranging from 4 to 450 mW, in pulsed and continuous (fi gure 6). Several questionnaire-based outcome measures wave mode.
were used—specifi cally, the neck pain and disability When trials with signifi cant results in favour of LLLT scale,54 Northwick Park neck pain questionnaire,55 short were subgrouped by wavelength, doses and irradiation form 36,56 Nottingham health profi le,57 and neck disability times seemed fairly homogeneous within narrow ranges index.58 (table 3). We noted a distinct dose-response pattern for Positive publication bias, which tends to exclude each wavelength for which LLLT is eff ective within a negative studies, was not apparent on testing (fi gure 7).38 narrow therapeutic window. For 820–830 nm, mean dose The plot has an aggregation in the lower left quadrant of per point ranged from 0·8 to 9·0 J, with irradiation times several small studies with results showing no or only of 15–180 s. For 904 nm doses, mean dose per point was small changes in visual analogue scale.59 If publication 0·8–4·2 J, with irradiation times of 100–600 s. bias towards only positive studies was present, few Investigators who used doses outside the minimum studies would lie in this position and small studies would (0·075 J and 0·06 J)40,49 and maximum (54 J)44 limits of have exaggerated positive outcomes. The slight these ranges did not show any eff ect of LLLT, lending asymmetry might be partly due to a negative publication further support to a dose-dependent response for LLLT in bias, the small number of studies, and because we have neck pain.
included the most reported studies so far. Signifi cant heterogeneity exists in categorical data for We subgrouped trials according to a-priori protocol in improvement from two studies39,40 of acute neck pain acute and chronic categories for the statistical analyses. (p=0·003, χ2=8·86, I2=88·7%). This fi nding could be Within these categories, we noted small variations attributable to the low dose per point used in one study.40,62 between trials in patient characteristics such as baseline We noted no heterogeneity between trials of chronic neck www.thelancet.com Vol 374 December 5, 2009
mean (SD)
mean (SD)
Özdemir et al (2001)50 4·60 (3·61 to 5·59) Gur et al (2004)46 0·73 (0·21 to 1·25) Chow et al (2006)13 0·91 (0·47 to 1·34) Ilbuldu et al (2004)48 0·78 (0·13 to 1·42) Dundar et al (2006)44 0·29 (–0·20 to 0·78) 1·38 (0·39 to 2·37)
Test for heterogeneity: χ2=59·95, df=4 (p<0·0001), l2=93·3% Test for overal effect: Z=2·74 (p=0·006) Figure 6: Standardised mean diff erence in disability scores between placebo-treated and laser-treated groups from fi ve trials
SMD=standardised mean diff erence.
pain reporting on categorical data (p=0·37, χ2=4·31, I2=7·2%). For continuous data from 100 mm visual analogue scale in chronic neck pain, we detected signifi cant heterogeneity across all wavelengths (p<0·0001, χ2=137·76, I2=90·6%). However, when heterogeneity was addressed separately by wavelengths, most heterogeneity could be accounted for by variations in doses and of patients in 40 application procedures. Removal of the study44 that used a very high dose from the disability analysis eliminated statistical heterogeneity (p=0·31, χ 2=3·61, I2=16·9%). For pain intensity on 100 mm visual analogue scale for 820–830 nm wavelength, this study caused heterogeneity Effect size on 100 mm VAS together with results of a second study50 that showed a
highly signifi cant eff ect, without obvious reasons for Figure 7: Publication bias plot
Plot of eff ect size between placebo and real laser groups within each trial versus their respective sample sizes. Red heterogeneity. After removal of both studies from the circles show one trial. VAS=visual analogue scale.
820–830 nm analysis, statistical heterogeneity was
eliminated (p=0·12, χ2=10·20, I2=41·2%), but the overall Discussion
eff ect remained similar, with narrower confi dence Our results show moderate statistical evidence for effi cacy
intervals after (22·0 mm [14·5–29·6]) than before of LLLT in treatment of acute and chronic neck pain in the
(21·6 mm [10·3–32·9]) removal.
short and medium term. For chronic pain, we recorded an For 904 nm wavelength, statistical heterogeneity was average reduction in visual analogue scale of 19·86 mm evident for analysis of pain intensity on 100 mm visual across all studies, which is a clinically important change.64,65 analogue scale (p=0·00001, χ2=28·37, I2=89·4%). The Categorical data for global improvement also signifi cantly only study in the review using a scanning application favoured LLLT. From our analysis, 820–830 nm doses are procedure in contact with the skin had weaker than most eff ective in the range of 0·8–9·0 J per point, with average results.45 Contrary to other laser application irradiation times of 15–180 s. At 904 nm, doses are slightly procedures, this method irradiates the target area smaller (0·8–4·2 J per point), with slightly longer intermittently. Few studies compare scanning irradiation times (100–600 s) than at 820–830 nm. irradiation with stationary irradiation, and most LLLT Our fi ndings build on those of previous reviews of studies have used a stationary laser beam. Another LLLT28,30 by including non-English language studies, study using 904 nm wavelength41 with non-signifi cant laser acupuncture studies in which local points were results has been criticised for absence of laser testing treated, and a quantitative analysis. Our search strategy and calibration, and the actual dose used remains has identifi ed a greater number of studies than have uncertain.63 Removal of these two trials from the 904 nm previous reviews, and draws attention to the intrinsic analysis of pain reduction on 100 mm visual analogue diffi culties in searching the topic of LLLT. Specifi cally, scale increased the overall eff ect from 20·6 mm no accepted terminology exists for laser therapy. We (95% CI 5·2–36·2) to 37·8 mm (25·4–50·1). have overcome this limitation by using as wide a range 50% of trials did not report side-eff ect data. Side-eff ects of synonyms as possible. reported included tiredness, nausea, headache, and Moreover, many apparently disparate diagnostic increased pain, but were mild and, apart from one study terms are applied to patients presenting with neck pain. in which unusual tiredness occurred more in the laser These terms suggest distinct clinical entities; however, group than in the placebo group (p>0·01),42 did not diff er there is strong evidence that a defi nitive diagnosis of from those of placebo. the causes of neck pain is not possible in a clinical www.thelancet.com Vol 374 December 5, 2009
neck were treated as part of the protocol. Evidence suggests Wavelength Average J per
Frequency of treatment
Number of
that trigger points in the neck coincide with the location of (nm [mode]) output
point (s)
acupuncture points in 70–90% of patients (eg, BL10, GB 20, GB21, and Ah Shi points).69,70 Since trigger points and Three times per week on alternate days for 4 weeks acupuncture points are characterised by tenderness, the Twice per week for 3 weeks treatment eff ect of laser irradiation to tender points, trigger points, or acupuncture points is likely to be the same. We did not distinguish any diff erences in subgroup Six times per week for analyses between these techniques. Thus, when treating neck pain with LLLT, irradiation of known trigger points, One application only acupuncture points, tender points, and symptomatic Five times per week for zygapophyseal joints is advisable. Dose assessment is crucial for interpretation of Three alternate days per outcomes of LLLT studies, for which failure to achieve a week for 1·5 weeks dose in the recommended range has been identifi ed as a Three alternate days per major factor for negative outcomes.71 The direct relation week for 1·5 weeks between positive outcomes of trials with adequate doses Özdemir et al 830 (cw) Five times per week for of laser irradiation for the appropriate condition has been shown in acute injury and soft-tissue infl ammation,21 Twice per week for 4 weeks tendinopathies,72 epicondylitis,11 and osteoarthritis.10 Twice per week for 4 weeks Several crucial parameters of laser devices are needed to assess dose of laser irradiation, but these doses were inconsistently reported in the studies that we reviewed. Hakgüder et al 780 (cw) Five times for week for No study provided all parameters identifi ed as important by the Scientifi c Committee of the World Association of Twice per week for 7 weeks Laser Therapy.74 In neck pain, however, there is little Five times per week for reason to believe that factors other than a plausible anatomical target, dose per point, and irradiation times Three alternate days per are essential for effi cacy of class 3B lasers (5–500 mW). We had suffi cient data relating to each of these Five times per week for components of therapy, when combined with manufacturers' specifi cations, to identify a dose-response Three times per week for pattern for the number of joules per point and wavelength used and positive outcome. Subgrouping of studies by Twice per week for 7 weeks wavelength and ascending doses reduced apparent Five times per week for heterogeneity in treatment protocols and laser parameters, and showed a dose-response pattern with distinct wavelength-specifi c therapeutic windows. Most p=pulsed. cw=continuous wave. IR=infrared. HeNe=helium-neon. NR=not reported.
statistical heterogeneity disappeared when we excluded Table 2: Laser parameters and treatment regimen
trials with small doses or fl aws in treatment procedure from effi cacy analyses. Additionally, a very high dose setting.66,67 By using the term non-specifi c neck pain, (54 J) of 830 nm LLLT used in one trial did not cause which encompasses many descriptors,31 we have benefi cial nor harmful eff ects.44 This fi nding suggests not addressed the clinical reality that patients presenting only that doses of this magnitude are higher than the with neck pain can have several concurrent sources of therapeutic window, but also that LLLT is safe even if pain from joints, muscles, and ligaments.
such an overdose is delivered. Frequency of treatments In addition to aggregating all included studies, varied from daily to twice a week, raising questions about irrespective of diagnostic label, we also combined data optimum treatment frequency. irrespective of the intended rationale for treatment, as Our analysis suggests that the optimum mean dose per long as neck muscles and spinal joints were exposed to point for 820–830 nm was 5·9 J, with an irradiation time laser irradiation. Transcutaneous application results in of 39·8 s, and for 904 nm, 2·2 J delivered with an laser-energy scattering and spreading into a three- irradiation time of 238 s. We recommend a multicentre, dimensional volume of tissue, up to 5 cm for infrared pragmatic trial in an appropriately powered study to test laser.68 Since the same eff ect would be achieved with the eff ectiveness of parameters of this order, with both application of laser energy to acupuncture points, we also pain intensity and functional improvement as outcome included data from studies in which local points in the measures.
www.thelancet.com Vol 374 December 5, 2009
Data from seven trials were available for up to 22 weeks Mean dose per
Mean irradiation time
after the end of treatment, suggesting that positive eff ects point (J)
per point (s)
were maintained for up to 3 months after treatment ended. Trials of knee osteoarthritis,75 tendinopathies,61,76 and low back pain reported similar longlasting eff ects of 820–830 nm13,42,50,53 77,78 These results contrast with those for non- steroidal anti-infl ammatory drugs in arthritis and spinal 904 nm39,41,43,45,46,52 disorders, for which the eff ect ends rapidly when drug Data are mean (SD, when applicable). LLLT=low-level laser therapy.
use is discontinued.71 Reduction of chronic neck pain at the end of treatment of 19·86 mm and at follow-up of Table 3: Mean dose per point and irradiation times for wavelengths of
LLLT used in studies with statistically signifi cant results
23·44 mm on a visual analogue scale of 100 mm represents clinically signifi cant pain relief.64,65 This result compares favourably with those of pharmacological stress and skeletal muscle fatigue with doses similar to therapies that are widely used in treatment of neck pain, those delivering anti-infl ammatory eff ects. This eff ect for which investigators have shown no conclusive has been reported in an animal study85 and in human evidence of benefi t.32 Intake of oral analgesic drugs was studies with biceps humeri contractions and diff erent not systematically reported; however, randomisation wavelengths.86,87 Because muscle fatigue is usually a within trials would keep the confounding eff ect of this precursor of muscle pain, and chronic trapezius myalgia factor to a minimum. is associated with increased electromyograph activity Half the studies obtained data for side-eff ects,39,42,44–46,49,52,53 during contractions and impaired microcirculation,88 with tiredness reported in the laser-treated group in reduction of oxidative stress and muscular fatigue could three studies,42,46,49 which was signifi cant in one study.42 be benefi cial in patients with acute or chronic neck Since LLLT does not generate destructive heat, safety pain.
relates mainly to potential eye damage, dependent on Inhibition of transmission at the neuromuscular class of laser device (classes 1–4), which is defi ned by junction could provide yet another mechanism for LLLT analysis of several parameters. Safety glasses are eff ects on myofascial pain and trigger points.89,90 Such required for classes 3B and 4 to eliminate this risk, and eff ects could mediate the clinical fi nding that LLLT would be required for use in all studies. Systematic decreases tenderness in trigger points within 15 min of reporting of side-eff ects in future studies would also be application.91 Laser-induced neural blockade is a further recommended to clarify short-term and long-term safety potential mechanism for the pain-relieving eff ects of aspects of LLLT.
LLLT.92,93 Selective inhibition of nerve conduction has Mechanisms for LLLT-mediated pain relief are not fully been shown in Aδ and C fi bres, which convey nociceptive understood. Several investigations exploring the stimulation.94,95 These inhibitory eff ects could be mediated pleiomorphic tissue eff ects of laser irradiation provide by disruption to fast axonal fl ow in neurons93 or inhibition plausible explanations for the clinical eff ects of LLLT. of neural enzymes.96 Anti-infl ammatory eff ects of red and infrared laser These tissue eff ects of laser irradiation might account irradiation have been shown by reduction in specifi c for the broad range of conditions that are amenable to infl ammatory markers (prostaglandin E , interleukin 1β, LLLT treatment. Whether specifi c treatment protocols are tumour necrosis factor α), in in-vitro and in-vivo animal necessary to elicit diff erent biological mechanisms is studies and in man.79 In animal studies, the anti- unknown. Heterogeneity of treatment protocols might be infl ammatory eff ects of LLLT are similar to those of due partly to variation in LLLT parameters and protocols,
pharmacological agents such as celecoxib,80 meloxicam,81 eliciting diff erent eff ects. Whatever the mechanism of
diclofenac,82 and dexamethasone.80 Chronic neck pain is action, clinical benefi ts of LLLT occur both when LLLT is
often associated with osteoarthritis of zygapophyseal used as monotherapy13,43 and in the context of a regular
joints,83 which is manifested by pain, swelling, and exercise and stretching programme.46,47 In clinical settings,
restricted movement as clinical markers of local combination with an exercise programme is probably
infl ammation. Laser-mediated anti-infl ammatory eff ects preferable. The results of LLLT in this review compare
at this joint could result in decreased pain and increased favourably with other widely used therapies, and especially
mobility. The distance between skin surface and lateral with pharmacological inter ventions, for which evidence
aspect of the facet joint is typically 1·5–3·0 cm without is sparse and side-eff ects are common.16,32
pressure, and less with contact pressure (measured with Contributors
ultrasonography [unpublished data, JMB]). Since 830 nm RTC participated in the literature search, development of inclusion and and 904 nm lasers penetrate to several centimetres,24,84 exclusion criteria, selection of trials for inclusion in the analysis, anti-infl ammatory eff ects at zygapophyseal joints are a methodological assessment, data extraction and interpretation, and writing of the report. MIJ participated in data analysis and interpretation, plausible mechanism of pain relief. critically reviewed the report with special expertise in pain management, Another possible mechanism of LLLT action on muscle and contributed to writing of the report. RABL-M participated in data tissue is a newly discovered ability to reduce oxidative interpretation and analysis, and critically reviewed the report with respect www.thelancet.com Vol 374 December 5, 2009
to the mechanism of action of laser, and relevance to neck pain. 17 Basford J. Low intensity laser therapy: still not an established JMB participated in development of inclusion and exclusion criteria, clinical tool. Lasers Surg Med 1995; 16: 331–42.
translation of non-English language articles, methodological assessment, 18 Sattayut S, Hughes F, Bradley P. 820nm gallium aluminium data analysis and interpretation, writing of the results section of the arsenide laser modulation of prostaglandin E production in report, and supervised writing of the report as a whole.
interleukin I stimulated myoblasts. Laser Therapy 1999; 11: 88–95.
Confl icts of interest
19 Sakurai Y, Yamaguchi M, Abiko Y. Inhibitory eff ect of low-level laser irradiation on LPS-stimulated Prostaglandin E production and RTC is a member of the World Association for Laser Therapy (WALT), cyclooxygenase-2 in human gingival fi broblasts. Eur J Oral Sci 2000; the Australian Medical Acupuncture College, the British Medical Acupuncture Society, the Australian Pain Society, the Australian 20 Aimbire F, Albertini R, Pacheco MTT, et al. Low-level laser therapy Medical Association, and the Royal Australian College of General induces dose-dependent reduction of TNFα levels in acute Practitioners. MIJ is a member of the International Association of the infl ammation. Photomed Laser Surg 2006; 24: 33–37.
Study of Pain. RABL-M is funded by Fundação de Amparo do Estado 21 Bjordal JM, Johnson MI, Iverson V, Aimbire F, Lopes-Martins RAB. de São Paulo (FAPESP, Brazil) and is scientifi c secretary of WALT, Photoradiation in acute pain: a systematic review of possible from which he has never received funding, grants, or fees. JMB is a mechanisms of action and clinical eff ects in randomized placebo- member of the Norwegian Physiotherapy Association, Norwegian controlled trials. Photomed Laser Surg 2006; 24: 158–68.
Sports Physiotherapy Society, Norwegian Society for Rheumatological 22 Yousefi -Nooraie R, Schonstein E, Heidari K, et al. Low-level laser and Orthopedic Physiotherapy, and has received research awards and therapy for non-specifi c low-back pain. Cochrane Database Syst Rev grants from the Norwegian Manual Therapy Association, the 2007; 2: CD005107.
Norwegian Neck and Back Congress, the Norwegian Research Council, 23 Brosseau L, Robinson V, Wells G, et al. Low-level laser therapy the Norwegian Fund for Postgraduate Training in Physiotherapy, and (classes I, II and III) for treating rheumatoid arthritis. the Grieg Foundation. He is also president of WALT, a position for Cochrane Database Syst Rev 2005; 4: CD002049.
which he has never received funding, grants, or fees.
24 Enwemeka C. Attenuation and penetration of visible 632·8nm and invisible infrared 904nm light in soft tissues. Laser Therapy 2001; Cousins MJ. Pain: the past, present, and future of anesthesiology? 25 Nussbaum EL, Van Zuylen J. Transmission of light through human Anesthesiology 1999; 91: 538–51.
skinfolds: eff ects of physical characteristics, irradiation wavelength Borghouts J, Koes B, Vondeling H, Bouter L. Cost-of-illness of neck and skin-diode coupling relevant to phototherapy. Physiother Can pain in the Netherlands in 1996. Pain 1999; 80: 629–36.
2007; 59: 194–207.
Picavet H, Schouten J. Musculoskeletal pain in the Netherlands: 26 Bjordal J, Johnson M, Ljunggren A. Transcutaneous electrical nerve prevalences, consequences and risk groups, the DMC -study. Pain stimulation (TENS) can reduce postoperative analgesic 2003; 102: 167–78.
consumption by one-third. A meta-analysis with assessment of Webb R, Brammah T, Lunt M, Urwin M, Allison T, Symmons D. optimal treatment parameters. Eur J Pain 2003; 7: 181–88.
Prevalence and predictors of intense, chronic, and disabling neck 27 Li L. What else can I do but take drugs? The future of research in and back pain in the UK general population. Spine (Phila Pa 1976) nonpharmacological treatment in early infl ammatory arthritis. 2003; 28: 1195–202.
J Rheumatol Suppl 2005; 72: 21–24.
Fejer R, Kyvik KO, Hartvigsen J. The prevalence of neck pain in the 28 Chow RT, Barnsley L. A systematic review of the literature of low- world population: a systematic critical review of the literature. level laser therapy (LLLT) in the management of neck pain. Eur Spine J 2006; 15: 834–48.
Lasers Surg Med 2005; 37: 46–52.
Woodruff LD, Bounkeo JM, Brannon WM, et al. The effi cacy of 29 Gross A, Aker P, Goldsmith C, Peloso P. Conservative management laser therapy in wound repair: a meta-analysis of the literature. of mechanical neck disorders: a systematic overview and meta- Photomed Laser Surg 2004; 22: 241–47.
analysis. Online J Curr Clin Trials 1996; 5: 1–116 (withdrawn).
Enwemeka CS, Parker JC, Dowdy DC, Harkness EE, Sanford LE, 30 Gross AR, Goldsmith C, Hoving JL, et al. Conservative Woodruff LD. The effi cacy of low-power lasers in tissue repair and management of mechanical neck disorders: a systematic review. pain control: a meta-analysis study. Photomed Laser Surg 2004; J Rheumatol 2007; 34: 1–20.
31 Jensen I, Harms-Ringdahl K. Neck pain. Siendentopf C, Golaszewski SM, Mottaghy FM, Ruff CC, Felber S, Best Pract Res Clin Rheumatol 2007; 21: 93–108.
Schlager A. Functional magnetic resonance imaging detects 32 Peloso P, Gross A, Haines T, et al. Medicinal and injection therapies activation of the visual association cortex during laser acupuncture for mechanical neck disorders. Cochrane Database Syst Rev 2007; of the foot in humans. Neurosci Lett 2002; 327: 3–56.
3: CD000319.
Tunér J, Hode L. Low level laser therapy—clinical practice and 33 Trinh K, Graham N, Gross A, et al. Acupuncture for neck disorders. scientifi c background. In: Tuner J, Hode L, eds. Low level laser Cochrane Database Syst Rev 2006; 3: CD004870.
therapy—clinical practice and scientifi c background. Sweden AB: Prima Books; 1999: 101–04.
34 Jadad A. Randomised controlled trials—a user's guide. London: BMJ Books, 1998: 97–100.
10 Bjordal J, Johnson MI, Lopes-Martins RA, Bogen B, Chow R, Ljunggren AE. Short-term effi cacy of physical interventions in 35 van Tulder M, Furlan A, Bombardier C, Bouter L, Group. Editorial osteoarthritic knee pain. A systematic review and meta-analysis of Board of the Cochrane Collaboration Back Review Group. Updated randomised placebo-controlled trials. BMC Musculoskelet Disord method guidelines for systematic reviews in the Cochrane 2007; 8: 51.
Collaboration Back Review Group. Spine (Phila Pa 1976) 2003; 11 Bjordal JM, Lopes-Martins RA, Joensen J, et al. A systematic review with procedural assessments and meta-analysis of low level laser 36 Zhang WY, Li WPA. Analgesic effi cacy of paracetamol and of its therapy in lateral elbow tendinopathy (tennis elbow). combination with codeine and caff eine in surgical pain— BMC Musculoskelet Disord 2008; 9: 75.
a meta-analysis. J Clin Pharm Ther 1996; 21: 261–82.
12 Walker J. Relief from chronic pain by low power irradiation. 37 Fleiss J. The statistical basis of meta-analysis. Stat Methods Med Res Neurosci Lett 1983; 43: 339–44.
1993; 2: 121–45.
13 Chow RT, Barnsley LB, Heller GZ. The eff ect of 300mW, 830nm 38 Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta- laser on chronic neck pain: a double-blind, randomized, placebo- analysis detected by a simple, graphical test. BMJ 1997; 315: 629–34.
controlled study. Pain 2006; 124: 201–10.
39 Soriano F, Rios R, Pedrola M, et al. Acute cervical pain is relieved 14 Mester E, Szende B, Spiry T, Scher A. Stimulation of wound healing with Gallium Arsenide (GaAs) laser radiation. A double blind by laser rays. Acta Chir Acad Sci Hung 1972; 13: 315–24.
preliminary study. Laser Therapy 1996; 8: 149–54.
15 Oron U. Photoengineering of tissue repair in skeletal and cardiac 40 Aigner N, Fialka C, Radda C, Vecsei V. Adjuvant laser muscles. Photomed Laser Surg 2006; 24: 111–20.
acupuncture in the treatment of whiplash injuries: a prospective, randomized placebo-controlled trial. Wien Klin Wochenschr 2006; 16 Binder AI. Cervical spondylosis and neck pain. BMJ 2007; www.thelancet.com Vol 374 December 5, 2009
41 Altan L, Bingöl U, Aykaç M, Yurtkuran M. Investigation of the 63 Bjordal JM, Baxter GD. Ineff ective dose and lack of laser output eff ect of GaAs laser therapy on cervical myofascial pain syndrome. testing in laser shoulder and neck studies. Photomed Laser Surg Photomed Laser Surg 2005; 25: 23–27.
2006; 24: 533–34.
42 Chow RT, Barnsley LB, Heller GZ, Siddall PJ. A pilot study of low- 64 Farrar JT, Young JJP, LaMoreaux L, Werth JL, Poole RM. Clinical power laser therapy in the management of chronic neck pain. importance of changes in chronic pain intensity measured on an J Musculoskelet Pain 2004; 12: 71–81.
11-point numerical rating scale. Pain 2001; 94: 149–58.
43 Ceccherelli F, Altafi ni L, Lo CG, Avila A, Ambrosio F, Giron G. 65 Tubach F, Ravaud P, Baron G, et al. Evaluation of clinically relevant Diode laser in cervical myofascial pain: a double blind study changes in patient reported outcomes in knee and hip versus placebo. Clin J Pain 1989; 5: 301–04.
osteoarthritis: the minimal clinically important improvement. 44 Dundar E, Evcik D, Samli F, Pusak H, Kavuncu V. The eff ect of Ann Rheum Dis 2005; 64: 29–33.
gallium arsenide aluminum laser therapy in the management of 66 Bogduk N. The anatomy and pathophysiology of neck pain. cervical myofascial pain syndrome: a double blind, placebo- Phys Med Rehabil Clin N Am 2003; 14: 455–72.
controlled. Clin Rheumatol 2007; 26: 930–34.
67 Barnsley L. Neck pain. In: Hochberg MC, Silman AJ, Smolen JS, 45 Flöter T, Rehfi sch H. Schmerzbehandlung mit laser. Eine Weinblatt ME, Weisman MH, eds. Rheumatology, 3rd edn. doppelblind-studie. Top Medizin 1990; 4: 52–56.
Edinburgh: Mosby, 2003: 567–81.
46 Gur A, Sarac AJ, Cevik R, Altindag O, Sarac S. Effi cacy of 904nm 68 Ohshiro T. The laser apple: a new graphic representation of gallium arsenide low level laser therapy in the management of medical laser applications. Laser Therapy 1996; 8: 185–90.
chronic myofascial pain in the neck: a double-blind and 69 Melzack R, Stillwell D, Fox E. Trigger points and acupuncture randomized-control. Lasers Surg Med 2004; 35: 229–35.
points for pain: correlations and implications. Pain 1977; 3: 3–23.
47 Hakgüder A, Birtane M, Gurcan S, Kokino S, Turan F. Effi cacy of 70 Dorsher PT. Can classical acupuncture points and trigger points be low level laser therapy in myofascial pain syndrome: an compared in the treatment of pain disorders? Birch's analysis algometric and thermographic evaluation. Lasers Surg Med 2003; revisited. J Altern Complement Med 2008; 14: 353–59.
71 Bjordal J, Couppe C, Chow R, Tuner J, Ljunggren A. A systematic 48 Ilbuldu E, Cakmak A, Disci R, Aydin R. Comparison of laser, dry review of low level laser therapy with location-specifi c doses for needling and placebo laser treatments in myofascial pain pain from chronic joint disorders. Aust J Physiother 2003; syndrome. Photomed Laser Surg 2004; 22: 306–11.
49 Laakso E, Richardson C, Cramond T. Pain scores and side eff ects 72 Bjordal J, Couppe C, Ljunggren A. Low-level laser therapy for in response to low level laser therapy (LLLT) for myofascial trigger tendinopathy: Evidence of a dose-response pattern. Phys Ther Rev points. Laser Therapy 1997; 9: 67–72.
2001; 6: 91–99.
50 Özdemir F, Birtane M, Kokino S. The clinical effi cacy of low- 73 Christie A, Jamtvedt G, Dahm K, Moe R, Haavardsholm E, power laser therapy on pain and function in cervical Hagen K. Eff ectiveness of nonpharmacological and nonsurgical osteoarthritis. Clin Rheumatol 2001; 20: 181–84.
interventions for patients with rheumatoid arthritis: an overview of 51 Seidel U, Uhlemann C. A randomised controlled double-blind systematic reviews. Phys Ther 2007; 87: 1697–715.
trial comparing dose laser therapy on acupuncture points and 74 World Association of Laser Therapy. Consensus agreement on the acupuncture for chronic cervical syndrome. Dtsch Z Akupunktur design and conduct of clinical studies with low-level laser therapy 2002; 45: 258–69.
and light therapy for musculoskeletal pain and disorders. 52 Taverna E, Parrini M, Cabitza P. Laserterapia IR versus placebo Photomed Laser Surg 2006; 24: 761–62.
nel trattamento di alcune patologie a carico dell'apparato 75 Gur A, Cosut A, Sarac AS, Cevik R, Nas K, Uyar A. Effi cacy of locomotore. Minerva Ortop Traumatol 1990; 41: 631–36.
diff erent therapy regimes of low-power laser in painful 53 Toya S, Motegi M, Inomata K, Ohshiro T, Maeda T. Report on a osteoarthritis of the knee: A double-blind and randomized- computer-randomised double blind clinical trial to determine the controlled trial. Lasers Surg Med 2003; 33: 330–38.
eff ectiveness of the GaAlAs (830nm) diode laser for pain 76 Stergioulas A. Low-power laser treatment in patients with frozen attenuation in selected pain groups. Laser Therapy 1994; 6: 143–48.
shoulder: preliminary results. Photomed Laser Surg 2008; 26:
54 Wheeler AH, Goolkasian P, Baird AC, Darden BV. Development of the neck pain and disability scale. Item analysis, face and 77 Longo L, Tamburini A, Monti A. Treatment with 904nm and criterion related validity. Spine 1999; 24: 1290–94.
10600nm laser of acute lumbago. J Eur Med Laser Assoc 1991; 55 Leak AM, Cooper J, Dyer S, Williams KA, Turner-Stokes L, 3: 16–19.
Frank AO. The Northwick Park Neck Pain Questionnaire, devised 78 Soriano F, Rios R. Gallium arsenide laser treatment of chronic low to measure neck pain and disability. J Rheumatol 1994; 33: 469–74.
back pain: a prospective randomized and double blind study. 56 McHorney CA, Ware JE, Raczek AE. The MOS 36 Item Short Laser Therapy 1998; 10: 175–80.
Form Health Survey (SF36): 2. Psychometric and clinical tests of 79 Bjordal JM, Lopes-Martins RAB, Iversen VV. A randomised, placebo validity measuring physical and mental health constructs. controlled trial of low level laser therapy for activated achilles Med Care 1993; 31: 247–63.
tendinitis with microdialysis measurement of peritendinous 57 Essink-Bot ML, Krabbe PFM, Bonselt GJ, Aaronson NK. An prostaglandin E concentrations. Br J Sports Med 2006; 40: 76–80.
empirical comparison of four generic health status measures. The 80 Aimbire F, Lopes-Martins R, Albertini R, et al. Eff ect of low-level Nottingham health profi le, the medical outcomes study 36-item laser therapy on haemorrhagic lesions induced by immune complex short-form health survey, the COOP/Wonca charts and the Euro- in rat lungs. Photomed Laser Surg 2007; 25: 112–17.
Qol instrument. Med Care 1997; 35: 522–37.
81 Campana V, Moya M, Gavotto A, et al. The relative eff ects of He-Ne 58 Vernon H, Mior S. The neck disability index: a study of reliability laser and meloxicam on experimentally induced infl ammation. and validity. J Manipulative Physiol Ther 1991; 14: 409–15.
Laser Therapy 1999; 11: 36–42.
59 Begg CB, Berlin JA. Publication bias: a problem in interpreting 82 Albertini R, Aimbire F, Correa FI, et al. Eff ects of diff erent protocol medical data. J R Stat Soc Ser A Stat Soc 1988; 151: 419–63.
doses of low power gallium–aluminum–arsenate (Ga–Al–As) laser 60 Djavid GE, Mehrdad R, Ghasemi M, Hasan-Zadeh H, radiation (650 nm) on carrageenan induced rat paw oedema. Sotoodeh-Manesh A, Pouryaghoub G. In chronic low back pain, J Photochem Photobiol B 2004; 27: 101–07.
low level laser therapy combined with exercise is more benefi cial 83 Bogduk N, Lord SM. Cervical spine disorders. Curr Opin Rheumatol than exercise alone in the long term: a randomised trial. 1998; 10: 110–15.
Aust J Physiother 2007; 53: 155–60.
84 Gursoy B, Bradley P. Penetration studies of low intensity laser 61 Vasseljen O, Hoeg N, Kjeldstad B, Johnsson A, Larsen S. Low therapy (LILT) wavelengths. Laser Therapy 1996; 8: 18.
level laser versus placebo in the treatment of tennis elbow. 85 Lopes-Martins RA, Marcos RL, Leonardo PS, et al. Eff ect of low- Scand J Rehabil Med 1992; 24: 37–42.
level laser (Ga-Al-As 655nm) on skeletal muscle fatigue induced by 62 World Association of Laser Therapy. Recommended anti- electrical stimulation in rats. J Appl Physiol 2006; 101: 283–88.
infl ammatory dosage for low level laser therapy. 2005. http:// 86 Leal Junior EC, Lopes-Martins RA, Vanin AA, et al. Eff ect of 830 nm www.walt.nu/dosage-recommendations.html (accessed Oct 4, low-level laser therapy in exercise-induced skeletal muscle fatigue in humans. Lasers Med Sci 2009; 24: 425–31.
www.thelancet.com Vol 374 December 5, 2009
87 Leal Junior EC, Lopes-Martins RA, Dalan F, et al. Eff ect of 655-nm 93 Chow R, David M, Armati P. 830-nm laser irradiation induces Low-Level Laser Therapy on Exercise-Induced Skeletal Muscle varicosity formation, reduces mitochondrial membrane potential Fatigue in Humans. Photomed Laser Surg 2008; 26: 419–24.
and blocks fast axonal fl ow in small and medium diameter rat 88 Larsson R, Oberg PA, Larsson SE. Changes in trapezius muscle dorsal root ganglion neurons: implications for the analgesic eff ects blood fl ow and electromyography in chronic neck pain due to of 830-nm laser. J Peripher Nerv Syst 2007; 12: 28–39.
trapezius myalgia. Pain 1999; 79: 45–50.
94 Tsuchiya D, Kawatani M, Takeshige C. Laser irradiation abates 89 Nicolau R, Martinez M, Rigau J, Tomas J. Neurotransmitter release neuronal responses to nociceptive stimulation of rat-paw skin. changes induced by low power 830nm diode laser irradiation on the Brain Res Bull 1994; 34: 369–74.
neuromuscular junction. Lasers Surg Med 2004; 35: 236–41.
95 Tsuchiya D, Kawatani M, Takeshige C, Sato T, Matsumoto I. Diode 90 Nicolau RA, Martinez MS, Rigau J, Tomas J. Eff ect of low power laser irradiation selectively diminishes slow component of axonal 655nm diode laser irradiation on the neuromuscular junctions of volleys to dorsal roots from the saphenous nerve in the rat. the mouse diaphragm. Lasers Surg Med 2004; 34: 277–84.
Neurosci Lett 1993; 161: 65–68.
91 Olavi A, Pekka R, Pertti K, Pekka P. Eff ects of the infrared laser 96 Kudoh C, Inomata K, Okajima K, Motegi M, Ohshiro T. Eff ects of therapy at treated and non-treated trigger points. 830nm gallium aluminium arsenide diode laser radiation on rat Acupunct Electrother Res 1989; 14: 9–14.
saphenous nerve sodium-potassium-adenosine triphosphatase 92 Baxter GC, Walsh DM, Allen JM, Lowe AS, Bell AJ. Eff ects of low activity: a possible pain attenuation mechanism examined. intensity infrared laser irradiation upon conduction in the human Laser Therapy 1989; 1: 63–67.
median nerve in vivo. Exp Physiol 1994; 79: 227–34.
www.thelancet.com Vol 374 December 5, 2009

Source: http://www.klinikken-livet.dk/CustomerData/Files/Folders/1-res/26_efficacy2.pdf