The Burning Mouth Syndrome and LLLT – a useful concept?

By Jan Tunér

Burning Mouth Syndrome (BMS) is a common disease but still a diagnostic and therapeutic challenge for clinicians. Despite many studie, its nature remains obscure and controversial; nowadays there is no consensus about definition, diagnosis and classification. BMS is characterized clinically by burning sensations in the tongue or other oral sites, often without clinical and laboratory findings. According to the etiology, BMS cases should be subdivided into three subtypes: BMS by local factors (lfBMS), BMS by systemic factors (sfBMS) and neurological BMS (nBMS), the most frequent, in which the symptom is caused by central or peripheral neurological malfunctions affecting in particular the taste pathway. To establish the type of BMS, both anamnesis and clinical examination, including laboratory tests, are necessary; nBMS cases will be recognized by exclusion of any other type. In case of lfBMS or sfBMS, the treatment of the main pathology will be resolutive; in nBMS cases many authors proposed different pharmacological trials without satisfactory results and the current opinion is that a multidisciplinary approach is required to keep the condition under control. (From Romeo et al, 4)

Therapies for BMS are scarce and not very effective, probably because the origin is not well-known. LLLT has been documented as a promising therapy in four studies. The laser parameters have varied considerably. The highest power used is in the study by Yang et al. (1), using an 800 nm 1.5 W laser, applying 105 J to each cm^2 of the affected area. The clinical effect was very satisfying. Still, the paper is criticised by Vukoja et al. (2) in a comment, suggesting that the effect is merely a placebo effect. This opinion is based upon a very old study (6), showing no result of LLLT. In a reply, the Yang group points out that they were using much higher energies than the one referred to by the Vukoja group. So here is another blunt example of the lack of parameter analysis. That triggered me to write to the Yang group, as follows:

“I have read your original article on BMS and the recent criticism in PMLS. Looking back at the involved papers, I can only say that you are completely right and that the criticism is unjustified. Vukoja et al. refer the old paper by Hansen and Thoroe. Let us look at the figures:

Hansen used a pulsed 904 nm 30 mW laser. Initially 1.8 J was applied to each site and if there was no effect, latter sessions used 3.6 J. Since the laser had an unusual type of pulsing, with reduction of output over the various pulse repetition rates, the actual energy delivered to each site was estimated to be 0.9 and 1.8 J per point. This is clearly stated in the paper.

It is known that pain reduction can be achieved either by influencing the inflammatory process or inducing an inhibitory effect on nerve transmission. In the first case low - medium energies are required and in the latter high energies. 1.8 J per point is possibly able to mediate a reduction of the inflammatory process, thereby indirectly influencing the pain (and is inflammation involved in BMS?). However, for immediate pain reduction, certainly much higher energies are needed. When the Hansen paper was published (1990), little was known about the dose requirements and about the effects of LLLT in general. 20 years later this is a forgiving aspect of the ambitious but negative paper by Hansen.

In your study 105 J was used per area (cm^2), so about 50 times more energy - sufficient to elicit an inhibitory effect. Whether or not this inhibitory effect has a long term effect on the condition itself is not clear, but then again, very little is known about the aetiology of this condition. Pain reduction is for the time being good enough.

As you can see, I am discussing joules (energy) rather than dose (J/cm^2). For analgesia, the energy is important and I believe too many readers do not know the difference between J and J/cm^2, so both should always be reported.

For two decades I have been irritated about careless citations of old studies such as the one by Hansen and about the lack of respect for the laser parameters. Too many apples and bananas are mixed. Too few peer reviewers are competent in the LLLT field, unfortunately - even in the specialized journals.”

The abstract of the Yang study can be read here (link will open in a new window): PubMed: Yang - Burning Mouth.

The abstract of the Hansen study can be found here (link will open in a new window):
PubMed: Hansen - Oro-Facial Pain.

What is there to learn from other studies?

In the study by Kato et al. (3), 790 nm, 120 mW, 1.2 J per point (3-9 points), 6 J/cm^2 was applied. This was a successful approach, too. The abstract reads (link will open in a new window):
PubMed: Kato - Burning Mouth.

Romeo et al. (4) used a combination of 650/910 nm, 0.53 J/cm^2, for 15 minutes, twice a week for four weeks. 25 patients were treated, 68% reported significant relief of symptoms. The long treatment time and frequency of sessions is probably essential in this study. The abstract is found here (link will open in a new window):
PubMed: Romeo - Burning Mouth.

Dos Santos et al. (5) on the other hand, used 660 nm, just 40 mW, 20 J/cm^2 and only 0.8 J per point 10 sessions. 0.8 J is a small energy per point, but a higher energy density, depending on the use of a small probe. Overall pain reduction was 58%. Abstract can be read here (link will open in a new window):
PubMed: dos Santos - Burning Mouth.

Bottom line

After reading the available literature, it is obvious that LLLT is a promising method for BMS patients, especially taken the low effectiveness of other modalities and the lack of side effects. An evaluation of the best parameters is not possible due to the great variety of parameters used in the four discussed studies. Very high energies such as in (1) have been shown to be effective and are possibly working through inhibition of neural activity. The remaining three studies have been using considerably lower energies and the mechanisms here ought to be different. But the analysis takes us back to the start – why did Hansen not find any effect when modern studies using energies in the same energy range succeed? The lack of effect in this old study is not, however, any ground for the assumption that LLLT for BMS is a placebo effect.


1. Yang HW, Huang YF. Treatment of burning mouth syndrome with a low-level energy diode laser. Photomed Laser Surg. 2011;29(2):123-5.

2.Vukoja D, Alajbeg I, Vučićević Boras V, Brailo V, Alajbeg IZ, Andabak Rogulj A. Is effect of low-level laser therapy in patients with burning mouth syndrome result of a placebo? Photomed Laser Surg. 2011;29(9):647-8; discussion 648, 651. Comment on Photomed Laser Surg. 2011 Feb;29(2):123-5.

3. Kato IT, Pellegrini VD, Prates RA, Ribeiro MS, Wetter NU, Sugaya NN. Low-level laser therapy in burning mouth syndrome patients: a pilot study. Photomed Laser Surg. 2010;28(6):835-9.

4. Romeo U, Del Vecchio A, Capocci M, Maggiore C, Ripari M. The low level laser therapy in the management of neurological burning mouth syndrome. A pilot study. Ann Stomatol (Roma). 2010 Jan;1(1):14-8.

5. dos Santos Lde F, Carvalho Ade A, Leão JC, Cruz Perez DE, Castro JF. Effect of low-level laser therapy in the treatment of burning mouth syndrome: a case series. Photomed Laser Surg. 2011;29(12):793-6.

6. Hansen HJ, Thorøe U. Low power laser biostimulation of chronic oro-facial pain. A double-blind placebo controlled cross-over study in 40 patients. Pain. 1990;43(2):169-79.