850 nm发光二极管光生物调节对犬牙远中移动过程中疼痛强度及牙本质涎蛋白水平的影响：一项随机对照试验。 机翻

Abstract

Background/purpose

Light-emitting diode-based photobiomodulation (LED PBM) potentially reduces pain and root resorption risk in orthodontics. This study evaluated 850-nm LED PBM effects on pain and dentin sialoprotein (DSP) levels during canine distalization.

Materials and methods

This single-blind, parallel-group randomized controlled trial included 16 orthodontic patients (8 per group). Participants were randomly allocated in a 1:1 ratio to either an active LED group or a sham control group. Pain intensity was assessed using a visual analog scale (VAS) at 0, 24, and 48 h following canine distalization at four consecutive monthly visits (T1–T4) and expressed as the area under the curve (AUC0–48h). DSP concentrations in gingival crevicular fluid were measured at pretreatment (T0), initiation of distalization (T1), and follow-up visits (T2, T3) using sandwich ELISA. Longitudinal changes were analyzed using linear mixed-effects models.

Results

The LED group demonstrated a significantly lower pain intensity compared with the control group, with a significant group-by-time interaction (χ2=9.46, P=0.024). Adjusted mean values of AUC0–48h were significantly reduced in the LED group at T3 and T4, with the greatest difference observed at T3 (P < 0.001). A significant group-by-time interaction was observed for DSP (χ2 = 9.75, P = 0.021), with lower DSP levels in the LED group being reported at T1, T2, and T3 compared with the control group (P < 0.05).

Conclusion

Repeated 850-nm LED PBM was associated with reduced cumulative pain and attenuated DSP increases during canine distalization, suggesting potential benefits for patient comfort and root resorption–related responses.

Keywords: Photobiomodulation, Light-emitting diodes, Orthodontic pain, Root resorption, Orthodontics

Introduction

Photobiomodulation therapy (PBM) is a non-invasive approach that uses light to modulate biological processes in humans.1 At the cellular level, PBM interacts with cytochrome C oxidase, enhancing adenosine triphosphate production, regulating reactive oxygen species, and promoting nitric oxide release.1 These mechanisms improve cellular metabolism, increase antioxidant capacity, and support tissue repair, particularly in hypoxic conditions. Light-emitting diodes (LEDs) are a practical PBM delivery system because they provide broader irradiation fields, stable wavelength emission, and lower cost than lasers.2 LED-based PBM has also demonstrated anti-inflammatory and tissue-reparative effects, including reduced inflammatory cell infiltration, increased fibroblast proliferation, enhanced angiogenesis, and improved collagen synthesis. 3–7 Near-infrared light is often used in photobiomodulation because it penetrates tissue well and interacts with mitochondrial chromophores, especially cytochrome C oxidase. In this range, the 850-nm LED devices could reach deeper into periodontal and alveolar tissues, allowing it to affect cellular activity in areas important for orthodontic tooth movement. Studies have shown that LED PBM at 850-nm wavelengths can inhibit orthodontically induced resorptive activity, reduce pain, and support tissue remodeling, thereby accelerating tooth movement.8–11

Orthodontic pain is a common adverse effect of treatment and may impair compliance and quality of life.12 Although analgesics are commonly used, adverse effects and possible interference with tooth movement have increased interest in PBM as a non-pharmacologic adjunct.13,14 Several studies reported reduced pain intensity after LED-based PBM during orthodontic treatment.8,15

Orthodontically induced inflammatory root resorption (OIIRR) is another important concern during prolonged orthodontic treatment. Experimental and clinical evidence suggest that PBM may reduce the extent or severity of root resorption.8,16,17 However, conventional radiographic diagnosis detects OIIRR only after substantial mineral loss has occurred.18 Molecular biomarkers may therefore offer earlier detection. Dentin sialoprotein (DSP), a non-collagenous acidic protein predominantly found in dentin, can be detected in gingival crevicular fluid (GCF), and has been reported to correlate with the severity and progression of OIIRR.19–22

Despite growing interest in PBM, evidence on its simultaneous effects on orthodontic pain and OIIRR-related changes assessed by molecular biomarkers is lacking. Therefore, this randomized controlled trial evaluated the effects of 850-nm LED photobiomodulation on pain intensity and DSP levels during canine distalization in orthodontic patients.

Materials and methods

Study design

This single-blind, parallel-group randomized controlled trial compared active 850-nm LED photobiomodulation (LED group) with a sham device (control group) during canine distalization. The primary outcome was cumulative pain intensity expressed as AUC0–48h. The secondary outcome was DSP concentration in GCF.

Eligible participants were randomly allocated in a 1:1 ratio to either the LED group or the control group. A total of 20 patients who met the prespecified eligibility criteria were enrolled in the trial. Participants were blinded to group allocation. Group assignment was performed using sequentially numbered, opaque, sealed envelopes prepared by an independent staff member. After each participant selected one envelope, an assisting staff member replaced the letters inside the envelopes without the investigator knowing their arrangement for the subsequent allocation.

Sample size was estimated based on both outcomes, and the larger estimate was adopted. For pain intensity, data from Tortamano et al. indicated that 14 participants (7 per group) were required to achieve 80% power at a significance level of 0.05.23 For DSP levels, data from Kereshanan et al. indicated that 16 participants (8 per group) were required.24 Therefore, the final sample size was set at 16 participants, and 20 were enrolled to allow for attrition.

Participants

Orthodontic patients aged 11–30 years who were consulted at the Faculty of Dentistry, University of Medicine and Pharmacy at Ho Chi Minh City, Vietnam, between September 2022 and October 2025 were screened for eligibility. Eligible participants were those diagnosed with dental crowding, presenting with four fully erupted permanent canines, and indicated for the extraction of all four premolars as part of orthodontic treatment. Exclusion criteria included the presence of active dental caries, gingivitis, or periodontal disease, dental anomalies in shape, number, or size, smoking, systemic diseases, use of medications affecting pain perception or bone remodeling, and a history of previous orthodontic treatment.

The study protocol was approved by the Ethics Committee (Approval No. 393/HDDD-DHYD, code: 22230-DHYD), and registered at ClinicalTrials.gov (Identifier: NCT07276984). This investigator-initiated trial received ethics approval prior to participant recruitment. Trial registration at ClinicalTrials.gov was completed after recruitment had commenced because of an administrative delay. No changes were made to the study protocol, eligibility criteria, intervention, or predefined outcomes after enrollment began. The present study was reported in accordance with the CONSORT 2025 statement. Written informed consent (or parental assent for patients under 18 years) was obtained from all participants prior to enrollment.

Orthodontic procedure and LED intervention

Patients underwent premolar extraction two weeks before the initiation of the alignment and leveling phase. Four molar bands were placed, and 0.022-inch slot self-ligating brackets were bonded to all permanent teeth.

During the alignment and leveling phase, patients underwent treatment with Nickel–Titanium (Ni–Ti) archwires, specifically Ni–Ti 0.012-, 0.014-, 0.016-, and 0.018-inch Ni–Ti archwires. After substantial correction of tooth misalignment, canine distalization was initiated on stainless steel (SS) archwires using sliding mechanics. Nickel-titanium closed-coil springs delivering a force of 150 g were used for canine distalization and were reactivated at each monthly follow-up visit (T1, T2, T3, and T4).

At the initiation of canine distalization (T1), participants in both groups were provided with identical LED devices and standardized instructions for use. In the intervention group, patients received an Orthopulse® device (Biolux Technology GmbH, Vienna, Austria), an LED device emitting near-infrared light at a wavelength of 850 nm. The devices were fully charged and supplied with a charging case. In the control group, patients received a sham Orthopulse® device without an active light source. Participants were instructed to use the device for 10 min daily (5 min per arch) at the same time each day for 4 consecutive months, starting at T1 and continuing throughout the canine distalization period.

To maintain consistency between the two groups, both the active and sham Orthopulse devices were introduced using the same procedure. In addition, all participants followed the same orthodontic treatment schedule and the same activation protocol during canine distalization. The devices were manufactured by Biolux Technology GmbH, Vienna, Austria, and Creation Technologies, Burnaby, BC, Canada (Fig. 1).

[图：CONSORT flow diagram of participants. Abbreviations : LED: light-emitting diode; DSP: dentin sialoprotein; GCF: gingival]

Pain measurements

Pain intensity was assessed using a 100-mm visual analog scale (VAS) presented as a paper-based questionnaire, where 0 indicated no pain, and 100 indicated the worst pain. Before data collection, participants were instructed on how to use the scale and how to indicate the point corresponding to their pain intensity.25–27 Pain assessments were performed at the initiation of canine distalization (T1) and at three subsequent monthly follow-up visits (T2, T3, and T4). At each visit, the immediate post-activation pain score was recorded directly on the VAS. Participants were then provided with a form containing two additional VAS scales for pain assessment at 24 and 48 h after the appointment. The investigator contacted each participant by telephone at these time points and asked them to rate their pain intensity, which was recorded as the corresponding 24-h and 48-h VAS scores. The completed forms were returned at the next visit.

Accordingly, the following scores were obtained at each visit: immediate (0 h), 24-h, and 48-h VAS scores. These three values were used to calculate the area under the curve from 0 to 48 h (AUC0–48h), which was defined as the primary outcome measure of pain intensity over time. AUC0–48h was calculated separately for T1, T2, T3, and T4.

Dentin sialoprotein measurements

DSP concentrations were assessed using GCF samples, which were collected at four time points: pretreatment (T0), initiation of canine distalization (T1), 1 month after distalization (T2), and 2 months after distalization (T3). At each visit, paper strips (Periopaper ®, Oraflow, Sarasota, FL, USA) were inserted into the gingival sulcus at the mesial and distal sites of the maxillary right canine for 30 s at each site. After collection, the samples were immediately processed with phenylmethylsulfonyl fluoride and was centrifuged at 13,200 rpm. The clear supernatant was collected and stored at −80 °C. Repeated freeze–thaw cycles were avoided.

DSP concentrations were determined using a sandwich ELISA kit (Human DSP ELISA Kit, MyBioSource, San Diego, CA, USA, MBS160884). All samples were measured in single measurements. A standard curve was built, and DSP concentrations were determined based on the standard curve, measured in nanograms per milliliter (ng/mL).

Statistical analysis

Changes in AUC0–48h and DSP concentrations over time were analyzed with the use of linear mixed-effects models. Model assumptions were assessed using residual diagnostics. AUC0–48h data showed a normal distribution, whereas DSP levels showed right-skewness. Therefore, restricted maximum likelihood (REML) was used for AUC0–48h values while robust standard errors were applied for DSP analyses. Adjusted marginal means and 95% confidence intervals (CIs) were estimated from the models. This approach accounts for the correlation between patients. The fixed effects were group (LED versus control), time (T1–T4), and their interaction (group × time). Random slopes were not included because each participant had only four time points. In addition, as the study has a limited sample size (n = 16), adding random slopes in preliminary models did not clearly improve model fit. Residual diagnostics showed that using only random intercepts was enough to capture within-subject dependence. We also did not include additional covariates because baseline characteristics were similar between groups.

All statistical analyses were performed using StataMP (version 14.1; StataCorp LLC, College Station, TX, USA). A two-sided P-value <0.05 was considered statistically significant.

Results

A total of twenty patients were enrolled in our randomized, controlled clinical study. Four participants did not complete the study after randomization, with two from the LED group and two from the control group. All four left because of scheduling conflicts and could not attend the required monthly appointments. No one reported any adverse effects or withdrew consent for reasons related to the intervention. Since the number of dropouts was the same in both groups and was due to logistical issues, it is unlikely that this caused any bias between the groups. Therefore, only 16 patients were included in the final data analyses, with 8 in the intervention group and 8 in the control group (Fig. 1). Baseline characteristics were comparable between the LED and control groups (Table 1).

The linear mixed-effects model (REML estimation) of AUC0–48h revealed a significant main effect for the treatment group (P < 0.001) and a significant interaction between group and time (χ2 = 9.46, P = 0.024). Additionally, the Likelihood Ratio (LR) test against a standard linear model was significant (χ2 = 3.15; P = 0.038), confirming that the random-intercept model effectively accounted for inter-individual variability among the 16 participants. Also, the intraclass correlation coefficient (ICC) was 0.213, indicating that 21.3% of the total variance was attributable to between-subject differences, thereby justifying the use of a linear mixed-effects model. The LED group displayed a consistently lower pain intensity than the control group, with a significant group-by-time interaction (P = 0.024). The random-effects parameters indicated significant inter-individual variability in pain perception, with a between-subject variance of 10,023.01 and a residual variance of 36,965.67 (Fig. 2).

[图：Changes in pain intensity (AUC 0–48h ) over time in the LED and control groups. Error bars represent 95% confidence inte]

Adjusted marginal means and 95% CIs of AUC0–48h at each visit are summarized in Table 2. Post-hoc pairwise comparisons revealed that the LED group experienced a significantly lower cumulative pain intensity compared to the control group at T1, T3, and T4 (P < 0.05). At T3, the mean AUC0–48h in the LED group dropped to its lowest value (2029.5 (95% CI: 1879.3–2179.7)) and was significantly lower than that in the control group (2622.0 (95% CI: 2471.8–2772.2); P < 0.001). In contrast, at the second visit (T2), the mean AUC0–48h in the LED group (2521.0 (95% CI: 2371.3–2671.7)) was lower than that of the control group (2710.5 (95% CI: 2560.8–2860.7)), but the between-group difference did not reach statistical significance (P = 0.081).

The LMM analysis demonstrated a significant group-bytime interaction for DSP concentrations (χ2 = 9.75; P = 0.021), indicating differing temporal trends between the LED and control groups. At pretreatment (T0), DSP concentration in the LED group was similar to that in the control group (P = 0.206). Following the initiation of canine distalization (T1), DSP concentrations increased in both groups, with a more pronounced rise observed in the control group (Fig. 3). The LED and control groups diverged sharply following the onset of canine distalization. The LED group demonstrated a significantly smaller increase in DSP levels compared with the control group at T2, with a mean difference of 22.2 units (P < 0.001). At T3, although both groups showed a trend toward stabilization, the LED group continued to exhibit significantly lower DSP levels than the control group, with a mean difference of 10.4 units (P = 0.023) (Table 3).

[图：Changes in dentin sialoprotein concentrations in gingival crevicular fluid over time in the LED group and the control gr]

Discussion

This randomized controlled trial demonstrated that 850-nm LED (PBM) was associated with reduced pain intensity and lower GCF-derived DSP levels compared with the control group during canine distalization, supporting an analgesic effect and a potential reduction in biological responses associated with OIIRR.

In the present study, the LED group showed significantly lower AUC0–48h values for pain intensity than the control group across four consecutive appointments. These findings are consistent with previous studies. Baldo et al. reported the effectiveness of LED PBM in relieving orthodontic pain during anterior retraction over four consecutive appointments.28 Similarly, using the same visual analog scale (VAS), Figueira et al. reported that 850-nm LED PBM reduced orthodontic pain during placement of elastomeric separators.25 Regarding DSP concentrations over time, DSP levels increased in both groups. This finding is consistent with previous studies using the same quantitative method (sandwich ELISA) by Kusumah et al. and Lombardo et al.29,30 This likely reflects the biological response to orthodontic force, as DSP is a dentin-specific protein and its presence in GCF indicates ongoing mineralized tissue breakdown. Notably, longitudinal mixed-effects analysis demonstrated significantly lower DSP concentrations in the LED group than in the control group, with the greatest between-group difference observed at T2 and a sustained difference at T3. These findings suggest a time-dependent and cumulative effect of repeated PBM application. Because DSP levels were comparable at pretreatment (T0), the observed differences at T1 are unlikely to represent baseline imbalance, but rather an early biological response to the intervention. These findings are in agreement with a previous clinical study by Sambevski et al.31 Using the same wavelength of 850 nm, Sambevski et al. reported that root resorption volume in the LED group (0.216 mm3) was 24% lower than that in the control group (0.284 mm3), although the difference was not statistically significant (P = 0.308). Similarly, Sfondrini et al. reported that PBM reduced the number of resorption craters by 32.78%, while crater volume in the LED group was 39.49% lower than that in the control group. In that study, however, laser-based PBM rather than LED devices were used.

Although previous studies have compared pain intensity between the LED and control groups, these comparisons were largely based on isolated intragroup analyses or mean pain scores at individual time points, treating later responses at T3 or T4 as independent from earlier responses at T1 or T2. Such approaches may overlook the temporal relationship between early and later pain experiences. Our study provides a more comprehensive evaluation by incorporating cumulative pain assessment using the area under the curve from 0 to 48 h (AUC0–48h), which better reflects the overall patient experience during canine distalization. Furthermore, by modeling the longitudinal trajectories of pain and DSP levels using linear mixed-effects models rather than relying solely on discrete time-point comparisons, the present study provides a more precise estimate of the effect of 850-nm LED PBM on pain and OIIRR-related biological responses, particularly in a cohort with substantial inter-individual variability in pain perception and DSP levels. To date, no studies have specifically compared pain intensity between LED and control groups during canine distalization, nor have any evaluated OIIRR-related changes between these groups using DSP as a biomarker. In addition, relatively few studies have investigated LED-based PBM, as most previous research has focused on laser-based PBM.

This study has several limitations. The sample size was relatively small, which may limit the generalizability of the results. Using linear mixed-effects models helps make better use of repeated measurements and properly accounts for within-subject correlation. However, this approach cannot overcome the limitations of a small sample size. The wide confidence intervals in several comparisons show real uncertainty in the estimates, and the findings may not apply broadly. These results should be seen as exploratory and preliminary until larger, well-powered studies can confirm them. In addition, DSP concentrations were measured using the sandwich ELISA technique in a single measurement, which may lead to measurement variability. Notably, as DSP is an indirect molecular marker of dentinal matrix breakdown, it should not be considered the same as direct measurements of root resorption. Although higher DSP levels in GCF have been linked to OIIRR severity in earlier studies, the exact relationship between DSP levels and root volume loss seen on radiographs or micro-CT scans is still unclear. These results suggest that 850-nm LED PBM may help slow early biological changes related to OIIRR, but do not directly show a reduction in structural root resorption. More research using radiographic or tomographic methods is needed to confirm this idea. Finally, the lack of investigator blinding, due to the nature of the research, may introduce bias, although blinding was feasible for participants.

In conclusion, 850-nm LED PBM may relieve orthodontic pain and reduce biological responses associated with OIIRR. Mixed-effects models showed a cumulative effect of repeated LED application, with a gradual reduction in pain intensity and lower DSP levels in the LED group than in the control group over time. Further studies with larger sample sizes are needed to confirm these findings and clarify the underlying mechanisms.

Acknowledgments

The authors did not receive any specific grant or funding from public, commercial, or non-profit funding agencies for this research.

Footnotes

[内容请查看原文]

References

[内容请查看原文]