Color Change of Resin-based Composites After In Vitro Bleaching Protocols: A Systematic Review and Meta-analysis
To systematically review the literature on color stability of resin-based composites (RBC) after in vitro bleaching protocols and to assess the influence of bleaching protocols by meta-regression analysis on RBC color stability, and the association with clinical and experimental characteristics. The electronic search was conducted in MEDLINE/PubMed, Scopus, and Web of Science databases and included English language studies that evaluated and reported color differences (CIELAB values) of RBC after in vitro bleaching procedures using hydrogen peroxide and/or carbamide peroxide. Database search for color change of RBC after bleaching retrieved 1335 eligible papers after removing duplicates. After initial screening, 66 articles were assessed for full-text reading with final inclusion of 23 published papers. A meta-regression analysis showed that storage time (p≤0.01), color measuring device (p≤0.01), and background color (p≤0.01) had influenced on color changes of RBC. The bleaching protocol (bleaching agent and time of application) did not influence on color changes of RBC (p>0.01). There is evidence that RBC change color after bleaching, but the change is not clinically significant.SUMMARY
Objectives
Methods
Results
Conclusions
INTRODUCTION
Resin-based composites (RBC) were introduced in the 1960s. Since then, RBC experienced several improvements of the organic and inorganic phases, such as the progressive reduction of filler particle size,1 the development and inclusion of new monomers,2 and monomer combinations.3 As a result, RBC have shown good mechanical properties1 and long-term clinical performance.4 Furthermore, the increasing dissemination of minimally invasive dentistry,5 concerns about amalgam toxicity,6 and esthetic demands7 promoted RBC to the most used dental material for direct restorations. Yet, the replacement of anterior RBC restorations because of esthetic failures, including lack of color stability, have been reported.8,9
Considering esthetic materials in dentistry, color stability is the material’s ability to maintain its color after aging (in service), staining, or bleaching.10 The CIELAB (ΔE*ab) metric is recommended to calculate the color stability.11,12 As color is a psychophysical property, the observers′ interpretation should be considered when evaluating color stability.10 In addition, the color difference (ΔE*ab) alone has little to no clinical meaning, and, therefore, the color stability should be evaluated through the perceptibility (PT) and acceptability (AT) thresholds for CIELAB color differences (ΔE*ab).11,13 Further, the CIELAB color space is a tridimensional space that considers changes in the color coordinates (L*, a*, and b*), which makes it impossible to determine how much “whiter” an object has become.14 For this reason, some studies proposed whiteness indexes such as WI,15 WIC,16 or WID.17
Dental bleaching is one of the most popular esthetic procedures in dentistry,18,19 with three possible protocols: in-office (performed by a professional); at-home (prescribed by a professional but performed by the patient at home); and over the counter (no professional prescription or follow-up).20 As a result, different application times and concentrations of carbamide peroxide and hydrogen peroxide agents can be used for in-office and at-home bleaching protocols.10,21,22
Patients submitted to dental bleaching often have restored teeth, and the effect of bleaching agents on RBC is not completely understood. Studies reported on oxidation of amine compounds and pigments23 and chemical bond degradation24 by bleaching agents that may alter the color perception of RBC21 and leading to esthetic complaints from patients. Nevertheless, the color change of RBC after bleaching is a controversial subject,10,25,26 and the influence of the RBC type and the bleaching protocol are inconclusive.8,10,27 Therefore, the aim of this study was to systematically review the literature on color stability of RBC after in vitro bleaching protocols and to assess by meta-regression analysis the influence of bleaching protocols on RBC color stability and the association with different covariates (RBC type, storage time, background color, and color measuring device). The hypothesis tested was that bleaching results in color alteration on RBC below the AT.
METHODS
This review was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.28 The research question was: Is color change of RBC after bleaching clinically significant?
Search Strategy
Electronic searches were conducted in three different databases (MEDLINE/PubMed, Scopus, and Web of Science), with no restriction for publication date with the last search performed on March 15th, 2020. The search strategy was designed to find articles written in English that evaluated in vitro color changes in RBC after bleaching, as described in Table 1.
Study Selection
Search results were duplicated using Mendeley software. Two trained reviewers (SBNB and MLV, both PhD students) independently selected the studies by title and abstract, according to the eligibility criteria. Records meeting the criteria or classified as unclear were retrieved for full-text analysis, which was performed independently by the same reviewers. Whenever necessary, screening discrepancies were resolved with the assistance of a third senior reviewer (ÁDB). During full-text reading, exclusion reasons were recorded.
Eligibility Criteria
The present study included published in vitro studies written in the English language that evaluated the color change of RBC after different bleaching procedures. In contrast, the exclusion criteria were as follows:
Type of study: case report, technical report, literature review, questionnaire-based studies, animal studies, and in vivo studies.
Materials: studies that only evaluated nonmethacrylate-based composite resins (silorane and ormocer), studies that did not use hydrogen peroxide and/or carbamide peroxide bleaching agents.
Methodology: studies that used colored solutions; studies that stored specimens for more than 16 days to perform the final color measurement or used artificial accelerated aging; studies that did not polish the specimens; studies that did not calculate color changes or did not use the CIELAB color difference equation.
Outcome: studies that did not report color difference and standard deviation values.
Data Collection
Articles meeting the inclusion criteria were subjected to critical appraisal, which was carried out by two reviewers (SBNB and MLV) independently. A standardized data extraction form was created using Excel software (Microsoft Corporation, Redmond, WA, EUA), collecting the following data:
Study characteristics: Author, publication year, objectives.
Materials characteristics: Resin composite type, bleaching agent type, and concentration.
Methods characteristics: Storage time and medium, bleaching agent protocol, color measuring device, background, color difference formula, color difference (ΔE) values, standard deviation, and sample size (n).
These data were collected and categorized for meta-regression analysis. The primary outcome was the CIELAB color difference (ΔE*ab) after application of bleaching agents (hydrogen peroxide and/or carbamide peroxide) onto the RBC, stored in a colorless substance for up to 16 days. The secondary outcomes included the influence of covariates such as the RBC filler portion, storage time, bleaching protocol, background color, and color measuring instrument.
The following experimental groups from selected studies were excluded from the present review: control group using no bleaching protocols, dry storage, storage longer than 16 days, storage in coloring substances, and color changes performed using CIEDE2000 color difference formula. Missing data from published papers were requested up to three times to the corresponding author. If data was not informed, the article was excluded from analysis.
Data Synthesis
A descriptive analysis of the findings summarized the data. Mean and standard deviation values, and sample size (n) were used to obtain the 95% confidence intervals (95% CI) for each group. Mean color difference (ΔE*ab) values and 95% CI from all groups of the included studies were used in a linear meta-analysis of random effects.29 A reference category for each variable was arbitrarily selected and used for comparison against other categories. The presence of heterogeneity was tested using I2 statistic and Chi-square tests, with Chi-square p-value < 0.05 or I2 > 50%, indicating high heterogeneity.30 As heterogeneity was high throughout all analyses, random effects models were used. Outliers and normality of residuals were checked by diagnostic procedures.
Meta-regression was performed to verify the impact of collected characteristics on color change. Bleaching procedure and further covariates (RBC, storage, color measuring device, background color, and bleaching protocol) were entered into the model, and a backward stepwise approach was applied for variables selection, keeping variables with a p-value ≤ 0.2. Analyses were performed using Stata 14 (StataCorp LP, College Station, TX, USA).
RESULTS
The search resulted in 1335 studies after removing duplicates. After the evaluation of titles and abstracts, 66 full-text studies were assessed for eligibility, and 24 papers were selected for the meta-analysis. Additional information on reasons for exclusion is shown in Figure 1.
Citation: Operative Dentistry 47, 2; 10.2341/20-234-LIT
To improve the accuracy of the model, the residuals were analyzed, and one study with outliers was excluded.31 Thus, 23 papers and 126 experimental groups were eligible, and data was collected. Experimental groups that did not match the inclusion criteria were excluded for the following reasons: nonbleached groups,22,26,32–40 dry-stored samples,41 nonmethacrylate resins (silorane or ormocer),33,34,39,42 nonhydrogen peroxide, or noncarbamide peroxide bleaching agents,35 color changes calculated with CIEDE2000 metric,12 and samples submitted to accelerated artificial aging.43
Table 2 shows the included studies after systematic search and the collected variables of interest. They were organized by alphabetical order and type of background. Most articles evaluated the following type of RBC: nanofilled (32%), microhybrid (27%), and nanohybrid (15%). The majority of the studies used a spectrophotometer (87%) as the color measuring instrument, and the samples were placed on a white background (48%). Carbamide peroxide (57%) and hydrogen peroxide (43%) were the bleaching products reported in the studies. Studies on in-office bleaching reported 20–60 minutes (52%) and 90–180 minutes (48%) of bleaching agent application time, while studies on at-home bleaching reported 7–21 hours (43%) and 28–147 hours (57%) of bleaching agent application time. Most studies reported a storage time of 1–2 days (75%). These data are summarized in Figure 2.
Citation: Operative Dentistry 47, 2; 10.2341/20-234-LIT
CIELAB metric, and its ΔE*ab equation, has been the most prevalent approach to calculate color difference values in dentistry. Thus, all studies included in the present review used CIELAB metric.21,22,25,26,32–50Figure 3 shows the mean values of color differences and 95% confidence intervals (95% CI) for all experimental groups included in the present review. The overall color difference value was 2.02 ΔE*ab units that is between PT (ΔE*ab=1.22) and AT (ΔE*ab=2.66) values, which means an acceptable color match.
Citation: Operative Dentistry 47, 2; 10.2341/20-234-LIT
Results from the meta-regression analysis are presented in Table 3. Nanohybrid composites showed greater color changes after bleaching procedures when compared to nanofilled composites (p=0.004). Storage time (p£0.01), color measuring device (p≤0.01), and background (p≤0.01) also influenced color changes. Only the bleaching protocol (bleaching agent and time of application) did not influence color changes (p>0.01).
DISCUSSION
The present systematic review and meta-analysis was designed to examine the controversial issue of RBC color changes after bleaching and its clinical significance. The results confirmed the study hypothesis that RBC color changes resulting from bleaching are below the AT.
As the human eye does not perceive small color differences,14 a single analysis of such differences may not be clinically significant.13 Therefore, the PT and AT were introduced to bring clinical relevance to visual color assessments.51 Thus, if the color difference is at or below PT, it represents an excellent match between color before and after a procedure, such as bleaching. If the color difference is between PT and AT values, as found in this meta-analysis (ΔE*ab=2.02), it represents an acceptable color match. A color difference above AT represents an unacceptable color match.11,12 As some studies,13,51,52 using different methodologies, proposed different ΔE values for PT and AT, the ISO11 published, in 2016, the threshold values (PT=1.22 ΔE*ab units, and AT=2.66 ΔE*ab units) for tooth-colored dental materials based on a multicenter study with different groups of observers.13 The use of color thresholds is widespread and well accepted in dentistry. From the 23 articles included in the meta-analysis, three of them did not use any visual thresholds to evaluate the results.33,34,40 Most papers published up to 2016 used AT = 3.3 ΔE*ab units.a The present systematic review analyzed all included data based on the visual thresholds published in the ISO/TR 28642:2016.11
Although ΔE*ab is commonly used to evaluate color after dental bleaching, there are more specific and useful parameters to evaluate “whiteness,” such as the whiteness index (WI),15 the CIE whiteness index (WIC),16 and the whiteness index for dentistry (WID)17 that is based on the CIELAB color space, and it was especially developed for dentistry. Only one study included in this review used a whiteness index.25 In addition, a recent study53 reported visual thresholds for WID [0.61 units for whiteness perceptibility threshold (WPT) and 2.90 units for whiteness acceptability threshold (WAT)]. Future studies on color changes after dental bleaching should consider complementing their data analysis for “whiteness” changes of teeth or dental materials using WID, WPT, and WAT.14
Previous studies explored the reasons for RBC changing color after dental bleaching. Some studies reported that free radicals available from high-concentration hydrogen peroxide can diffuse into the polymer,39,42,54 others related color changes with longer bleaching exposure time, allowing for free radical infiltration into the polymer.8,47 Nevertheless, the present review found an overall color difference value of 2.02 ΔE*ab units after bleaching of RBC. Previous clinical studies on tooth bleaching efficacy reported values between 4.3–4.655 and 7.1–10.6 ΔE*ab units,56 with no significant color changes after 1 year.57 Some studies suggested that RBC do not get whiter after bleaching,10,25 as teeth do.58 At the outset, one may sense that RBC color stability after bleaching is a desirable characteristic. However, as RBC and teeth show different gradients of color changes after bleaching, the color difference between RBC restorations and teeth may become clinically unacceptable, and patients may demand new restorations after dental bleaching. Such clinical dilemmas should be further investigated to support evidence-based information to clinicians and patients.
Due to the methodological variability among the selected studies, a high heterogeneity (99.9%) was observed (Figure 3). Nevertheless, in addition to the primary outcome, the present study was able to evaluate the following variables: bleaching protocol, composite type, storage time, color measuring device, and background color. Thus, two bleaching protocols were considered: in-office bleaching and at-home bleaching. For the purposes of the present study, over-the-counter products were considered as at-home bleaching. It was not possible to determine any standardization of bleaching agent application protocol. There was a great variation on the total application time of the bleaching agent among the selected studies: in-office bleaching protocols were applied from 20 minutes to 180 minutes,22,25,33,34,42,44,46–49 and at-home bleaching protocols were applied from 7 hours to 147 hours.b Yet, some studies used different concentrations of the bleaching agent for the same application time.35,46,47,50 This meta-analysis shows insufficient evidence to infer that the type of bleaching protocol influences on RBC color change, which is probably associated to the great variability of dental bleaching protocols.
The association between filler content and color changes of RBC after dental bleaching is controversial. While some studies reported that the filler portion does not influence RBC color changes,25,26,33,34,36,38 other studies showed the opposite.21,32,42 In the nanohybrid composites the filler ratio is variable,59 combining nanometric and larger particle size fillers.1 Their morphology and size are product dependent,59 and the nanoparticle concept is different from the nanofilled RBC.60 Such heterogeneity in the filler portion can produce variable properties, especially related to the solvent stability.59 Despite a similar chemical basis, variations in monomers61 and photoinitiator systems62 can change the color stability of RBC. The meta-regression showed that the filler portion influenced the outcome with a different behavior for nanohybrid RBC compared to other types of RBC. However, this color change is not clinically significant when considered by the color difference thresholds.
Studies have reported on several methods to store and age RBC samples, including dry22 and relative humidity environments,63 immersion in a dye solution64 and colorless substances,25,26 and artificial accelerated aging.10,43,65 To standardize storage conditions, this review only included studies that used water as a storage medium and a maximum of 16 days for the storage time. In the present study, time of storage influenced the outcome, that is, storing for 14–16 days was significantly different from shorter storage times, which is not surprising since literature shows that RBC color changes are time related, even in colorless media.66 This information is clinically relevant, although such covariates (storage time and media) should be carefully evaluated in in vitro studies, because none of the storage and aging methods can mimic the complex dynamics that occur in the oral environment.
Several factors can influence color measurements,14 such as illuminant,67 color measuring device,68 color difference metric,25,69,70 and color background.71 The variables “color background” and “color measuring device” significantly influenced the outcome of this study. As RBC can exhibit different levels of translucency,72 the background color can affect the scattering, absorption, and reflectance of the material.14 On a white background the light can be reflected, and on a dark background the reflection is reduced.73 Most studies included in this review used white background,c followed by the studies that did not report the background.21,22,32,39,43,48–50 Studies that used white background showed a different behavior from those using a dentin shade35,40 and black25 backgrounds. Due to the variety of color backgrounds found in this systematic review along with the information that 87% of the studies reported using AT and PT, which were obtained from different studies over a single background, it is an urgent need for further studies on the influence of background color on visual thresholds.
Despite numerous types of instruments to evaluate and measure color in dentistry, the included studies used either colorimeters or spectrophotometers. Colorimeters measure tristimulus values of the light reflection after the light source passes through filters but do not measure the spectral reflectance of an object74 as spectrophotometers do.75,76 The real color of an object cannot be determined, since there is no gold standard for a correct evaluation, due to the nature of color.77 Thus, a way to measure the instrument trustworthiness is focused on the repeatability (when the same method, operator, and/or instrument is used) and reproducibility (different method, operator, and/or instrument).68,74 Studies have shown that when compared to colorimeters, spectrophotometers have better repeatability and reproducibility, providing better results.68,78,79 Regarding the amount of measurements and measuring moment, the present review showed that, in addition to before and after color measurements, some studies also measured the color difference during the bleaching protocol.21,25,41,44 Considering the abovementioned information on the time dependency (aging) of RBC color changes, additional measurements would be relevant in studies using medium- to long-term aging.
This meta-analysis was designed to investigate the color change of RBC after bleaching and the clinical significance of it. Reports using different bleaching protocols, experimental methods, and study objectives were included, making for very heterogeneous data. Therefore, associated factors such as longer storage/aging periods, material degree of conversion, and surface polish aspects could not be considered. Lastly, as the collected data is from in vitro studies, the influence of clinical factors such as tooth brushing, staining food, and masticatory dynamics were not taken into account. Therefore, the above-mentioned factors are limitations of the present study.
CONCLUSIONS
Within the limitations of the present review and meta-analysis, it is concluded that RBC experience color change after bleaching, but it is clinically acceptable when considering the dental color thresholds. The type of bleaching protocol did not influence the color change of resin-based composite. Nanohybrid RBC showed a different color stability behavior. As methodological variables (background color, color measuring device, and storage time) influenced on color changes of RBC, it is an urgent need for standardization of experimental variables in laboratory studies. As several articles were excluded because of insufficient reported data, authors should be more careful to provide enough information in future publications so that clinical decisions could be based on scientific evidence.
Flow diagram of the systematic review.
Some of the covariates investigated and the data collected.
Forest plot showing the mean color changes (ΔE*ab) reported in the experimental groups included in the present review. Dotted line represents the overall color difference value (2.02 ΔE*ab units). Values for the perceptibility (PT) (1.22) and acceptability (AT) (2.66) thresholds are indicated.13
Contributor Notes
Clinical Relevance
Dental bleaching alters the color of resin-based composites, but the color change is clinically acceptable. Different bleaching protocols produce similar color changes on resin-based composites.