Introduction

Rationale

About every third (35.4%) child aged 8–15 has some kind of malocclusion [1], which can negatively influence quality of life, even to a greater extent than caries [2]. Despite the recent popularity of orthodontic aligners [3], conventional fixed appliance still remains the therapeutic gold standard [4].

The typical orthodontic fixed appliance involves the use of brackets bonded to the tooth’s labial or lingual surface that present complex morphology and favor biofilm retention [5]. Brackets, bands, archwires, and ligatures can make routine oral hygiene challenging [6] by increasing biofilm accumulation and decreasing the physiological self-cleaning action of the saliva and the tongue [5].

Fixed appliance treatment is associated with increased bacterial accumulation, which may alter the oral ecosystem towards pathogenic colonization [7, 8], and increases the risk of caries, periodontal inflammation, and enamel demineralization [6, 9].

Ligation of the archwire within the bracket slot is achieved either with metallic stainless steel ligature wires or elastomeric modules (in the form of single ‘o rings’ or elastic chains of multiple rings), which present substantial differences in their bacterial colonization [11,12,13]. In clinical practice, ligature choice is based on patient preference, esthetic demands, logistic reasons related to appliance interval, or differential clinical performance due the materials’ own characteristics [13]. Elastomeric ligatures have a porous and rough surface as they are composed of organic material, while steel ligatures are made of inorganic metal material, ensuring a smooth and inert surface [14]. During intraoral use, elastomeric ligatures show considerable adsorption and the progressive formation of a proteinaceous biofilm that undergoes partial calcification [15]. Therefore, the use of elastomeric ligatures has been suggested to promote bacterial retention and have a more negative effect on oral hygiene than their metallic counterparts [12,15,16].

Previous studies have provided conflicting evidence on the effects of ligature materials on periodontal health. A recent systematic review [17] on the subject reported that currently no recommendations for one ligation mode over the other are possible and that stainless steel ligatures might be better for biofilm management. However, in that review only two databases were searched up to 2021 and no quantitative data synthesis (meta-analysis) was performed. Another systematic review found that fixed appliances ligated with steel ligatures are associated with increased plaque index scores than self-ligating fixed appliances that have no ligature [18] but did not compare them to elastomeric ligatures.

Objective

The primary aim of this systematic review was to compare the periodontal effects of orthodontic fixed appliances ligated with either elastomeric or stainless steel ligatures.

Materials and methods

Protocol and registration

This systematic review was carried out in adherence to the Cochrane Handbook [19] and its report follows the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) 2020 statement [20]. The study’s protocol was developed a priori and was pre-registered in PROSPERO (CRD42023444383), while any protocol deviations were openly disclosed for transparency reasons (Appendix 1).

Eligibility criteria

Included were studies on patients undergoing orthodontic treatment with fixed labial appliances that are ligated either with elastomeric or stainless steel ligatures. The primary outcome was Pocket Probing Depth (PPD), while secondary outcomes included Plaque Index (PI), Gingival Bleeding Index (GBI) or Gingival Index (GI), total bacterial count, and Streptococcus mutans counts. Included were comparative clinical studies (both randomized and nonrandomized) on humans, while excluded were studies on patients diagnosed with periodontal disease, antibiotic use in the last six months or systemic disease, case reports, case series, animal studies, in vitro/in situ/ex vivo studies, and non-clinical studies.

Information sources and search

Two authors (UH, AC) independently conducted a search of seven databases (MEDLINE via PubMed, Scopus, Web of Science, Embase, Cochrane Database of Systematic Reviews, Cochrane Central Register of Controlled Trials, and Virtual Health Library), using appropriate search terms (Appendix 2), without any restrictions for publication year, language, or type. Furthermore, the reference lists of eligible articles and existing systematic reviews were manually reviewed to identify any potentially relevant studies that might have been missed from the systematic search. Finally, all included studies were checked in Google Scholar using the “Related Articles” option to identify any additional studies.

Selection process

The results of the literature search were imported in Endnote X9 software (Clarivate, Philadelphia, PA) for deduplication and then transferred to electronic spreadsheets. At first, the titles and/or abstracts of all studies identified in the literature were screened and then the remaining full texts were evaluated against the eligibility criteria. Study selection was conducted independently by two authors (UH, SS) and any disagreements were resolved through discussion with a third author (SA).

Data collection process and items

Data collection utilized a pre-defined and piloted extraction form, encompassing the following data: (a) study characteristics, including the primary author with the year of publication, study design, and clinical setting (country); (b) patient characteristics, comprising age and sex; (c) sample size for each intervention; (d) follow-up duration; and (e) measured outcomes. To ensure accuracy and uniformity, two authors (UH, MN) independently performed the data extraction, while any disparities were resolved through discussion with a third author (SS).

Risk of bias of individual studies

The risk of bias of randomized trials was assessed with the Cochrane Risk of Bias (ROB) 2 tool [21] on an intention to treat basis. The risk of bias of non-randomized comparative studies was assessed with the Risk Of Bias In Nonrandomized Studies of Interventions (ROBINS-I) tool [22]. All assessments were conducted by two authors independently (UH, RM), with discrepancies resolved through discussion with a third author (SS).

Effect measures and synthesis measures

The Mean Difference (MD) with its 95% Confidence Interval (CI) was chosen for same outcomes used across studies, while the Standardized Mean Difference (SMD) was used when variations of indices measuring the same outcome (like different PIs) were used. As the periodontal effects of different ligatures were expected to vary among studies (according to different elastomeric materials, level of oral hygiene, and position in mouth of the teeth being measured) a random-effects model was deemed a priori more appropriate to capture this variability and calculate the average distribution of treatment effects across studies [23] and a novel restricted maximum likelihood variance estimator was chosen due to improved performance [24]. Between-study heterogeneity was gauged through forest plot inspection, tau2 (absolute heterogeneity), I2 (relative inconsistency), and uncertainty intervals for all heterogeneity estimates (while also evaluating localization of heterogeneity in the forest plot and existing uncertainty). 95% predictions were calculated to incorporate existing heterogeneity and aid in meta-analytical interpretation by providing a range of possible future effects across the various clinical settings [25]. All analyses were conducted in R 4.2.2. (R Foundation for Statistical Computing, Vienna, Austria) by one person (SNP), with open data provision [26], two-sided P-values, and alpha = 5% (Appendix 1).

Reporting bias assessment and certainty assessment

Hints of reporting biases (including the possibility of publication bias) were planned (Appendix 1) but could ultimately not be assessed. To gauge the certainty of the meta-analytic results, the Grades of Recommendations, Assessment, Development, and Evaluation (GRADE) approach was employed [27] and findings were summarized using a revised table format [28].

Results

Study selection

The initial electronic database search yielded 947 records and seven additional were identified through manual searching (Fig. 1). After eliminating 19 duplicates, 935 records were left for further evaluation and were assessed against the eligibility criteria (Appendix 3). Ultimately, 11 publications, corresponding to 11 distinct clinical studies, were included in the quantitative and qualitative synthesis.

Fig. 1
figure 1

Flowdiagram for the identification and selection of studies for this review

Full size image

Study characteristics

Table 1 shows the characteristics of the 11 studies included in this analysis. Among these studies, the majority (82%; 9/11) were of within-person (clustered) design where both ligation methods were used on different teeth and only 27% (3/11) were randomized trials. These studies were conducted in university clinics of six different countries (Brazil, India, Italy, Pakistan, Sweden, and Turkey). In total 354 patients were included in the 11 studies (median 21 patients per study), who were 42% male (127/301; from the 9 studies reporting sex) and were on average 14.7 years of age (from the 5 studies reporting on age). The majority of included studies (91%; 10/11) used elastomeric ligatures, one study (9%) used elastic chains, and two studies also compared different kinds of elastomeric ligatures. Among the included studies, plaque index was measured in seven studies (64%), gingival index in four studies (36%), probing pocket depth in three studies (27%), bacterial counts in two studies (18%), and Streptococcus mutans counts in five studies (45%).

Table 1 Characteristics of the included studies
Full size table

Risk of bias in studies

Three randomized controlled trials were evaluated using the ROB 2 tool, and all were found to have a high risk of bias due to randomization issues, deviations from intended interventions, and outcome measurement issues (Table 2a). Of the eight non-randomized studies evaluated using ROBINS-I, all were deemed to be in serious risk of bias, primarily attributed to confounding and participant selection. Additionally, a moderate risk of bias was identified in these studies concerning the classification of interventions and the risk of missing data (Table 2b).

Table 2a Risk of bias assessment of included randomized trials with the ROB 2 tool
Full size table
Table 2b Risk of bias assessment of included non-randomized studies with the ROBINS-I tool
Full size table

Data synthesis

A total of five outcomes were assessed in a relatively similar manner from more than one study and were included in meta-analysis (Table 3). Meta-analysis of five studies did not find a statistically significant difference in plaque index between elastomeric and steel ligatures (5 studies; SMD = 0.48; 95% CI = -0.03 to 1.00; P = 0.07; Fig. 2), but great heterogeneity across studies was seen and 4 out of 5 studies pointed towards greater plaque accumulation with the former. Post hoc removal of the single study on the forest plot’s left side, which was responsible for the heterogeneity, led to somewhat larger difference (4 studies; SMD = 0.70; 95% CI = 0.13 to 1.27; P = 0.03), which was now marginally statistically significant. Similarly, no significant differences were found for gingival index (2 studies; MD = 0.01; P = 0.89; Fig. 3) and probing pocket depth (2 studies; MD = 0; P = 0.97; Fig. 4). Meta-analysis of three studies indicated that elastomeric ligatures were associated with increased bacterial load compared to steel ligatures (3 studies; SMD = 0.43; 95% CI = 0.10 to 0.76; P = 0.03; Fig. 5), with the effect being of moderate magnitude. Finally, no statistically significant difference was found in Streptococcus mutans counts between elastomeric and steel ligatures (4 studies; SMD = 0.40; P = 0.21; Fig. 6). Considerable between-study heterogeneity was seen also for this meta-analysis, but as most studies reported minimal differences between compared groups and the overall meta-analysis was similarly not statistically significance, this was deemed to be due to random variation.

Table 3 Meta-analyses comparing elastomeric to stainless steel ligatures
Full size table
Fig. 2
figure 2

Meta-analysis comparing plaque index between fixed appliances ligated with elastomeric and stainless steel ligatures. CI, confidence interval; N number of patients; QH Quigley Hein index; RE random-effects model; SD standard deviation; SiLo Silness & Löe index; SMD standardized mean difference

Full size image
Fig. 3
figure 3

Meta-analysis comparing gingival index between fixed appliances ligated with elastomeric and stainless steel ligatures. CI confidence interval; MD mean difference; N number of patients; RE random-effects model; SD standard deviation

Full size image
Fig. 4
figure 4

Meta-analysis comparing pocket probing depth between fixed appliances ligated with elastomeric and stainless steel ligatures. CI confidence interval; MD mean difference; N number of patients; RE random-effects model; SD standard deviation

Full size image
Fig. 5
figure 5

Meta-analysis comparing bacterial counts between fixed appliances ligated with elastomeric and stainless steel ligatures. CI confidence interval; N number of patients; RE random-effects model; SD standard deviation; SMD standardized mean difference

Full size image
Fig. 6
figure 6

Meta-analysis comparing salivary Streptococcus mutans counts between fixed appliances ligated with elastomeric and stainless steel ligatures. CI confidence interval; N number of patients; RE random-effects model; SD standard deviation; SMD standardized mean difference

Full size image

Table 4 shows the GRADE evaluation of strength of clinical recommendations from performed meta-analyses. In all instances, low quality of evidence was found, due to the inclusion of non-randomized studies that in many instances had methodological limitations that could pose a threat to their internal validity and increase their risk of bias. The results of the single statistically significant meta-analysis were back-translated to elastomeric ligatures having on average 0.30 × 105 CFU/ml greater bacterial counts (95% CI 0.07 to 0.52 × 105 CFU/ml) compared to steel ligatures.

Table 4 Summary of findings table according to the GRADE approach
Full size table

Sensitivity analysis according to the design of the included studies can be found in Appendix 4. The results of the statistically significant meta-analysis on bacterial counts were no different depending on whether randomized or non-randomized studies were used (P = 0.45). In the assessment of Streptococcus mutans counts, significant differences were found between randomized and non-randomized studies, where the latter showed significantly more inflated negative effects (greater Streptococcus mutans counts) for elastomeric ligatures than the former (P < 0.001). As, however, the cumulative meta-analysis was not statistically significant, this does not affect overall recommendations.

Discussion

Result in context

The present review summarizes evidence from 11 studies involving 354 patients undergoing fixed appliance treatment with conventional (not self-ligating) brackets. All included studies compared the impact of metallic versus elastomeric ligatures on periodontal health and is, to our knowledge, the first of its kind.

Periodontal health was assessed among included studies through both clinical and microbiological parameters, including the plaque index, gingival index, probing pocket depth and counts of total bacteria or Streptococcus mutans. Orthodontic treatment with fixed appliances has been previously associated with increased biofilm accumulation and gingival inflammation around both self-ligating and conventionally ligated brackets [29, 30]. Additionally, orthodontic treatment with fixed appliances has been associated with minimal increases in probing pocket depth [31], which are however normalized after appliance removal [32]. Overall, clinical evidence indicates that orthodontic treatment under proper oral hygiene regimens is not associated with clinical attachment loss, even among patients with reduced but healthy periodontium [33,34,35]. At the same time, fixed appliance orthodontic treatment is associated with increased bacterial burden and Streptococcus mutans counts [36, 37], which although transient [32, 38], highlights the importance of fluoride supplementation to minimize the risk for dental caries or enamel demineralization [37].

The increased microbial colonization of elastomeric ligation modules compared to steel ligatures has traditionally been based on either anecdotal data [], data from in vitro studies [39,40,41] or in vivo retrieval studies [15, 42]. In the present review, this was the only consistent difference between ligation methods, with elastomeric ligatures showing moderately higher bacterial counts than steel ligatures (SMD = 0.4; Fig. 5) and with effects ranging from small to large (SMDs 0.10 to 0.8). It is important however to keep in mind that (i) included studies are small, which makes precise effect estimation uncertain; (ii) the wide prediction interval (ranging from − 1.0 to 1.9) indicated that even though the average trend indicates increased counts with elastomeric ligatures, this will not necessarily be the case for each single; (ii) all three studies were in high risk of bias, which indicated that caution is warranted by the interpretation of their findings. Possible explanations include among others that the complex microbial adhesion process is the result of several factors, such as specific lectin-similar reactions, electro-static interactions, and Van der Waals forces between the microorganisms and surface. A recent review [43] reported that bacterial adhesive strength is mainly determined by the amount and nature of contacts between surfaces and macromolecules on the bacterial surfaces, rather than from the physicochemical properties of surface materials. It is commonly believed that higher surface roughness (such as that of elastomeric rings) influences bacterial attachment mainly by increasing the surface area for microbial colonization. Moreover, they are reported to be more difficult to clean compared with smooth surfaces, like those of metallic materials [44,45,46]. However, other in-vivo or in-situ studies contradict this opinion and, according to the present findings, suggest that modification of surface roughness only plays a modest role in altering bacterial adherence and biofilm formation [47, 48]. Indeed, certain bacterial strains increase their adhesion when enhancing the material stiffness, independently from their other physico-chemical properties [49]. A possible explanation for this is that all surfaces exposed to the oral cavity are covered by the acquired pellicle within a short time [41]. This pellicle can level out surface roughness, modulating the physicochemical properties of the materials and, consequently, modifying the bacteria adherence [46, 50]. Another important issue is that periodontal health depends not only on the degree of biofilm formation, but also on the composition of the microbial community [43].

Even though the increased microbial colonization of elastomeric ligatures seems to be true, it does not necessarily translate to worse outcomes of periodontal health for the average patient compared to steel ligatures. Both elastomeric modules and elastic chains present considerable degradation that is accentuated by intraoral ageing [15, 51, 52], which is another reason why they are usually replaced at each appointment. This might partially explain why their increased bacterial load is not reflected on periodontal parameters. It should, however, be noted that these are the average effects of fixed appliance treatment and, as expected, considerable heterogeneity exists. This means that for some patients with suboptimal oral hygiene, the increased microbial load of elastomeric ligatures could contribute to the risk of periodontal inflammation. The existence of a structured oral health promotion protocol during orthodontic treatment is therefore of paramount importance [53] and novel motivation-enhancing interventions could potentially be beneficial in improving oral health and minimizing treatment-related adverse effects [54,55,56,57,58,59].

Ligation choice does not seem to be primarily influenced by differential microbial colonization and periodontal effects, but differences exist in the clinical performance of various ligation methods. Biomechanically, differences in applied force magnitude, torque expression, and frictional resistance have been reported between elastomeric and steel ligatures [60,61,62]. Additionally, unlike elastomeric ligatures, steel ligatures do not suffer from force decay phenomena [63,64,65] and can be left in place for longer periods and therefore be combined with longer intervals between appointments. Finally, elastomeric ligatures present color instability and show considerable staining after intraoral use [66,67,68], which can have a negative impact on esthetics.

Limitations

This review has certain limitations in this review. The most significant is the inclusion of non-randomized studies, which are generally more prone to bias [69]. However, sensitivity analyses according to the study design did not find significant discrepancies between the results from the two study designs. Furthermore, all meta-analyses were informed by few studies with very limited sample sizes that could be biased [70] and present results should be confirmed by future studies with larger samples.

Conclusions

Based on available evidence of low certainty from randomized and non-randomized clinical studies, there might be little clinically relevant differences in the periodontal effects of elastomeric or stainless steel ligatures used for orthodontic fixed appliances. However, the existing studies present serious methodological limitations and more well-designed prospective studies could help formulate robust clinical recommendations.