The aims of this study were to retrospectively investigate the long-term development of gingival recession in a cohort of orthodontic patients and to compare the prevalence of gingival recession in orthodontically treated patients 10 to 15 years posttreatment to that of untreated subjects with malocclusion.
The sample included 88 patients with mean ages of 12.1 years (SD, 2.4 years) at pretreatment, 15.1 years (SD, 2.4 years) at posttreatment, and 27.9 years (SD, 2.5 years) 10 to 15 years posttreatment. The control group comprised 102 untreated patients seeking orthodontic treatment with a mean age of 28.7 years (SD, 3.1 years). Gingival recession was evaluated on study models.
The prevalence of both labial/buccal and lingual/palatal gingival recession increased during orthodontic treatment with further increases during the long-term posttreatment period; 98.9% of the orthodontically treated participants had at least 1 labial/buccal recession, and 85.2% of the patients had at least 1 lingual/palatal recession 10 to 15 years posttreatment. In addition, the proportion of patients with multiple labial/buccal or lingual/palatal recession sites increased considerably in the same time period. The prevalences of labial/buccal gingival recession were similar in the orthodontically treated patients 10 to 15 years posttreatment and the untreated controls. Study group patients with a crossbite before treatment showed 2.73 more recessions (95% CI, 0.28-5.17; P = 0.029) than did those without a transverse discrepancy. Untreated subjects with crowding greater than 3 mm per arch had 3.29 more recessions (95% CI, 0.73-5.68; P = 0.012) to 4.92 more recession sites (95% CI, 1.70-8.15; P = 0.003) than did those with mild or no crowding.
Within the limitations of this study, it seems that, in regard to the prevalence of gingival recession, orthodontically treated patients are not compromised in the long term compared with those with malocclusion that was untreated for many years.
Long-term development of gingival recession in orthodontic patients was studied.
Prevalence of gingival recession increases during and after orthodontic treatment.
Recession was similar in patients and untreated controls 15 years later.
Orthodontic patients seem not have worse long-term outcomes in regard to the prevalence of gingival recession than untreated controls seeking orthodontic treatment.
Gingival recession refers to exposure of the root surface caused by displacement of the gingival margin apically to the cementoenamel junction. Gingival recession may occur in many populations. Localized or generalized, it can create an esthetic problem and lead to dentin hypersensitivity and root caries.
Recession of the marginal gingiva can be associated with anatomic or pathologic factors. It can develop as a result of periodontitis or natural remission of the tissue through aging, which is the most common type of gingival recession. In these patients, recession develops more commonly on the buccal than on the lingual surfaces, and the periodontal structures show no sign of inflammation. Traumatic tooth brushing, tobacco, and intraoral and perioral piercings, aberrant tooth morphology or tooth eruption, and anatomically reduced thickness of the bone or thin gingival biotype overlying the roots have been also associated with gingival recession.
Active orthodontic treatment can induce gingival recession when teeth are moved outside the alveolar bone. Difficulty with tooth brushing and plaque accumulation during orthodontic treatment and fixed retention can lead to inflammation-related recession. According to Pandis et al, there is no clear connection between the presence of fixed retainers and the development of gingival recession. However, long-term fixed bonded retainers on proclined mandibular incisors could cause attachment loss. The evidence on the prevalence of gingival recession in orthodontically treated and untreated patients is contradictory. Thomson found no influence of orthodontic treatment on the development of gingival recession, whereas Slutzkey and Levin and Renkema et al reported that gingival recession was more prevalent in orthodontically treated patients than in untreated controls. These conflicting findings could be related to the difference in length of the observation periods or the different characteristics of the control groups in those studies.
Therefore, the aim of this study was to retrospectively analyze the long-term development of gingival recession in orthodontically treated patients from pretreatment to 10 to 15 years posttreatment. Moreover, we studied whether the prevalence of gingival recession in the orthodontically treated patients 10 to 15 years after treatment was different from that of untreated subjects with malocclusion seeking orthodontic treatment matched for age and sex.
Material and methods
This retrospective cohort study was approved by the ethics committee of Bern, Nordwest and Central Switzerland (EKNZ 2015-349, HVF, Kat A), and every patient signed informed consent to participate. For the structure of this article, the recommendation of the STROBE Statement were followed.
Two groups were selected: (1) a study group comprising orthodontically treated patients whose treatment was finished 10 to 15 years before this study and (2) a control group of untreated subjects with different types of malocclusions seeking orthodontic treatment.
The study group was selected from patients at a private orthodontic practice in Grenchen, Switzerland. In this practice, it was routine to keep pretreatment (initial, T1) and posttreatment records (final, T2) for at least 10 years after the last retainer follow-up visit, which was usually performed between 1 and 4 years posttreatment. To study patients at least 10 years posttreatment (T3), a recall was established for those who had their last retainer follow-up appointment in 2005, 2006, and 2007 (thus, their T2 records were taken between 2001 and 2006). Three hundred ninety-four consecutive patients who met the following inclusion criteria were eligible to be included in the study: (1) treated with fixed appliances, (2) treated by the same orthodontist, (3) maxillary and mandibular retainers bonded immediately after active orthodontic treatment, and (4) nonsyndromic. They were invited for a recall assessment. No age restriction was applied during the sample selection; 118 patients agreed to participate; of those, 14 were “no shows.” Eventually, 104 patients were evaluated. After the recall appointment, the following additional exclusion criteria were applied: (1) orthodontic retreatment, (2) debonding more or less than 10 to 15 years ago, (3) T1 and T2 casts missing or of poor quality in the region of the gingival margin, (4) periodontal disease other than gingival inflammation, and (5) nonwhite patients. Finally, 88 patients were included. The flow chart ( Fig 1 ) shows in detail the procedure of patient selection.
The control group was selected from 969 consecutive adults who came for consultation to the Department of Orthodontics and Dentofacial Orthopedics at the University of Bern in Switzerland between 2006 and 2016. The participants were selected if they met the following inclusion criteria: (1) availability of a full set of pretreatment records (photographs, casts, radiographs), (2) no history of prior orthodontic treatment, and (3) between 22 and 37 years of age at the assessment; the age criterion was applied to match participants from the study cohort at T3. Exclusion criteria were (1) nonwhite, (2) syndromic, (3) casts of poor quality in the region of the gingival margin, (4) periodontal disease other than gingival inflammation; 143 subjects were initially considered for inclusion in the control group ( Fig 1 ). Thirty-three of them had previous orthodontic treatment, and 8 were nonwhite and were subsequently excluded. Finally, 102 subjects were included in the control group.
The primary outcome was the presence of gingival recession in buccal or lingual sites of the teeth, and they were scored as “yes” if the cementoenamel junction in the center of the buccal or labial aspect was clearly exposed; otherwise, it was scored as “no.”
The participants were requested to complete a questionnaire regarding hygiene, habits, and any complications regarding their retainer. Demographic data such as sex, age, hygiene technique, type of toothbrush, and type of retainer were recorded during the clinical assessment at T3.
In the study group, plaster models made at T1, T2, and T3 were scored, using the yes/no scoring system. One rater (M.M.) rated all patients clinically at T3 and scored all plaster models at each time point (T1, T2, and T3). When judgment was difficult on the plaster models, intraoral photographs were additionally used. During the clinical evaluation (T3, 10-15 years after orthodontic treatment), oral hygiene was recorded and rated on a 3-level scale (good, average, or poor). For the hygiene technique, the participants were requested to perform their brushing method and were asked what kind of toothbrush (soft, middle, hard) they generally use. Gingival recession was scored buccally and labially using the 4-point classification system of Miller. If calculus was present around the retainer, it was removed with an ultrasonic scaler about 5 minutes before scoring the recession and taking the impression. The results of this clinical assessment of gingival recession were used to evaluate the kappa level between clinical and cast scores. The condition of fixed retainers was assessed, and any breakage or detachment was noted. Finally, intraoral photographs and alginate impressions were taken. These records made during this appointment were designated as T3.
In the control group, scoring was performed retrospectively on plaster or digital models made from scanned plaster models (approximately in 50% of the subjects). The rater (M.M.) scored the plaster models manually and the digital models on the computer monitor using Onyx Ceph 3 software (Chemnitz, Germany). When the judgment was difficult to make, intraoral photographs were viewed.
In both groups, the pretreatment malocclusion in the sagittal, vertical, and transverse planes and the amount of crowding were assessed. In the sagittal plane, the Angle classification was used; in the vertical plane, the amount of overbite was registered, and 3 subgroups were created: (1) large overbite, greater than 4 mm; (2) normal overbite, between 1 and 4 mm; and (3) reduced overbite, less than 1 mm. In the transverse plane, the presence of unilateral or bilateral crossbite was registered. Crowding was categorized in each dental arch as mild (space deficiency, <3 mm), moderate (space deficiency, 3-5 mm), or severe (space deficiency, >5 mm). When the extent of crowding differed in the maxillary and mandibular dental arches, the arch with more severe crowding was used to categorize the participant.
Information on smoking habits, plaque accumulation, and bleeding on probing of gingival pockets was not available for any participant in either group.
For the clinical assessment, 2 raters (M.M. and M.G.) evaluated together 10 participants from the study group to familiarize themselves with the scoring system. Subsequently, both raters evaluated 26 participants twice and independently from each other.
For the assessment on plaster models, 4 months after the clinical assessment, both raters calibrated themselves by scoring recession sites on the plaster models of 7 subjects who did not participate in this investigation. Any disagreements between the raters were discussed and resolved after inspection of the corresponding intraoral photographs. Subsequently, both raters scored 25 plaster models independently from each other to assess interrater agreement.
To evaluate the agreement between scores of plaster and digital models, 35 plaster and 35 digital models of the same participants were scored independently by the 2 raters.
For the assessment on digital models only, 3 months after scoring the 35 digital models, both raters independently rescored the same models to assess interrater agreement within and between raters.
Method reliability was assessed with the kappa statistic (linearly weighted kappa). The interpretation of the kappas was according to the method of Altman : less than 0.20, poor; 0.21-0.40, fair; 0.41-0.60, moderate; 0.61-0.80, good; and 0.81-1.00, very good.
Descriptive statistics were calculated for patient characteristics and the distribution of buccal and lingual recessions.
Linear regression analysis was performed to assess possible associations between age, sex, oral hygiene (level, technique, type of toothbrush), retainer type, malocclusion, treatment duration, and the sum of recessions at the patient level for the study and control groups. Analyses were done separately for lingual and buccal recessions.
All statistical analyses were conducted using a statistical package (version 14.2; StataCorp, College Station, Tex).
The proportion of women was slightly higher, but not statistically significant, in the treated group than in the untreated group (71.6% vs 61.7%, respectively; Table I ). Both groups were well matched regarding age: the mean ages at T3 were 27.9 years (SD, 2.5 years; range, 23.8-37.1 years) in the study group, and 28.7 years (SD, 3.1 years; range, 22.6-37.1 years) in the control group. The mean time since debonding was 12.9 years (SD, 1.2; range, 10.4-15.9 years). All orthodontically treated participants received bonded retainers in the maxilla (bonded to all incisors) and mandible (bonded to either canines only or all 6 anterior teeth). In participants who had a mandibular retainer bonded only to the canines (44.3%), a round 0.027-in beta-titanium alloy wire was used; the remaining participants (55.6%) had a 0.016 × 0.022-in braided steel wire (D-Rect; Ormco, Glendora, Calif).
|Study group||Control group|
|T1 (n = 88)
|T2 (n = 88)
|T3 (n = 88)
|Total group (n = 102)
|Subgroup (n = 20) ∗
|Age (y)||12.1 (2.4)||15.1 (2.4)||27.9 (2.5)||28.7 (3.1)||28.7 (2.8)|
|Male||25 (28.4%)||25 (28.4%)||25 (28.4%)||39 (38.2%)||10 (50%)|
|Female||63 (71.6%)||63 (71.6%)||63 (71.6%)||63 (61.7%)||10 (50%)|
|Treatment (T1–T2)||2.9 (1.3)|
|Retention (T2–T3)||12.9 (1.2)|
|Type of mandibular retainer|
|Bonded to all 6 anterior teeth||–||–||49 (55.6%)||–||–|
|Bonded to canines only||–||–||39 (44.3%)||–||–|
In both groups, the malocclusion at T1 was analyzed, and the distribution of the malocclusion characteristics is shown in Table II .
|Study group||Control group||Study group||Control group||Study group||Control group|
|Sagittal||Angle Class I||Angle Class II||Angle Class III|
|Vertical||Overbite 1-4 mm||Overbite >4 mm||Overbite <1 mm|
|Crowding||Mild (<3 mm)||Moderate (3-5 mm)||Severe (>5 mm)|
Table III gives intrarater and interrater agreement for the clinical examinations, the plaster and digital model scoring in the different tooth groups (incisors, canines, premolars, and molars), and the different locations of recession (buccal and lingual). In general, the agreement in scoring recession was good to very good (kappa, >0.6). Only the interrater agreement for clinical assessment of lingual recession in the region of incisors and canines was fair to moderate (kappa, 0.27 and 0.48, respectively), and agreement for clinical and cast assessments of buccal recession in the molars was moderate (kappa, 0.59).
|Casts vs clinical, M.M.||0.785 (0.731-0.838)||0.665 (0.521-0.809)||0.898 (0.848-0.949)||0.684 (0.618-0.751)||0.741 (0.693-0.789)||0.751 (0.679-0.823)||0.826 (0.788-0.864)||0.596 (0.540-0.653)|
|Casts vs digimodels, M.M.||0.841 (0.778-0.905)||0.814 (0.760-0.959)||0.885 (0.795-0.974)||0.778 (0.673-0.883)||0.814 (0.760-0.959)||0.749 (0.644-0.853)||0.820 (0.756-0.885)||0.794 (0.719-0.870)|
|Casts vs digimodels, M.G.||0.843 (0.777-0.910)||0.794 (0.674-0.914)||0.848 (0.742-0.953)||0.749 (0.627-0.871)||0.810 (0.741-0.878)||0.856 (0.775-0.936)||0.797 (0.729-0.865)||0.906 (0.854-0.958)|
|Casts intrarater, M.M.||0.827 (0.732-0.922)||0.878 (0.712-1.000)||0.967 (0.937-0.997)||0.842 (0.751-0.934)||0.773 (0.686-0.861)||0.864 (0.759-0.969)||0.938 (0.903-0.974)||0.816 (0.737-0.896)|
|Casts intrarater, M.G.||0.812 (0.720-0.903)||0.735 (0.514-0.955)||0.967 (0.936-0.997)||0.896 (0.820-0.972)||0.804 (0.722-0.886)||0.795 (0.667-0.922)||0.908 (0.864-0.952)||0.846 (0.773-0.919)|
|Clinical interrater, M.M. vs M.G.||0.487 (0.318-0.656)||0.274 (–0.153-0.701)||0.987 (0.968-1.000)||0.703 (0.655-0.832)||0.806 (0.723-0.890)||0.644 (0.461-0.827)||0.930 (0.892-0.968)||0.744 (0.655-0.832)|
|Digimodels, interrater, M.M. vs M.G.||0.672 (0.585-0.759)||0.734 (0.514-0.955)||0.802 (0.675-0.929)||0.613 (0.474-0.752)||0.810 (0.742-0.878)||0.808 (0.716-0.901)||0.796 (0.728-0.865)||0.729 (0.644-0.813)|
|Casts, interrater, M.M. vs M.G.||0.751 (0.672-0.830)||0.745 (0.611-0.878)||0.828 (0.720-0.936)||0.705 (0.581-0.828)||0.803 (0.733-0.872)||0.796 (0.701-0.891)||0.822 (0.758-0.885)||0.753 (0.670-0.835)|
The prevalence of labial/buccal gingival recession increased during orthodontic treatment: 54.5% of the participants had at least 1 recession site at T1, and 85.2% had at least 1 recession site at T2 ( Table IV ). After treatment, there was a further increase of the prevalence of recession: 98.9% of the orthodontically treated participants had at least 1 recession site at T3. From T1 to T3, also the proportion of patients with multiple recession sites (>5) increased from 10.2% (T1), to 27.3% (T2), and to 86.3% (T3).
|Study group (n = 88)||Control (n = 102)|
|Labial/buccal recessions per patient|
|≥ 25 recessions||0||0.0||0||0.0||0||0.0||3||2.9|
|Palatal/lingual recessions per patient|
The prevalence values of recession in the control subjects and the orthodontically treated participants at T3 were similar: almost no one was recession-free. Approximately 10% of the participants had 1 to 4 recession sites, and 53% had 5 to 14 recession sites. However, 37.2% of the controls had 15 recessions or more, whereas only 26.1% of the orthodontically treated participants had 15 or more recession sites ( Table IV ; Fig 2 ).
The prevalence of lingual/palatal gingival recession increased during the orthodontic treatment: 18.2% of the participants had at least 1 recession site at T1, and 59.1% had at least 1 recession site at T2 ( Table IV ). After treatment, there was a further increase of the prevalence of recession: 85.2% of the orthodontically treated participants had at least 1 recession site at T3. From T1 to T3, the proportion of patients with multiple recession sites (>5) increased from 0% (T1), to 9.1% (T2), and to 38.6% (T3).
The prevalence values of lingual/palatal recession in the control subjects and the orthodontically treated participants at T3 were different. In general, the treated participants had fewer recession sites than did the controls; 14.8% of the treated participants were recession-free, whereas all controls had at least 1 recession site; 46.6% of the treated and 39.3% of the untreated participants had 1 to 4 recession sites; 34.1% of the treated and 52.0% of the untreated participants had 5 to 9 recession sites ( Table IV ; Fig 2 ).
The regression analysis assessing the effect of malocclusion on the sum of recessions is shown in Table V . In the study group, the participants with a crossbite at T1 showed 2.73 (95% CI, 0.28-5.17; P = 0.029) more buccal recessions at T3 than did those with no discrepancy. There was a weak association between recession and crowding and vertical malocclusion at T1 ( Table V ). For the palatal/lingual recession, patients with a reduced overbite showed 2.19 (95% CI, 0.15-4.22; P = 0.035) recessions more than did patients with a deep or normal overbite before treatment. There was a weak association between recession and Class III malocclusion at T1 ( Table V ).
|Predictor||Study group (n = 88)||Control group (n = 102)|
|β coefficient||95% CI||P value||β coefficient||95% CI||P value|
|Class II||−0.53||−3.43, 2.37||0.718||−0.12||−2.61, 2.36||0.922|
|Class III||1.60||−3.65, 6.85||0.718||−1.59||−6.21, 3.03||0.497|
|Malocclusion vertical plane|
|Overbite 1-4 mm||Reference|
|Overbite >4 mm||−2.28||−4.85, 0.28||0.080||−3.53||−6.05, −1.01||0.007|
|Overbite <1 mm||−0.21||−3.85, 3.42||0.907||−1.66||−5.05, 1.72||0.331|
|Malocclusion transverse plane|
|Crossbite||2.73||0.28, 5.17||0.029||0.65||−2.32, 3.62||0.663|
|Mild 0-3 mm||Reference|
|Moderate 3-5 mm||2.51||−0.11, 5.13||0.060||3.29||0.73, 5.68||0.012|
|Severe >5 mm||1.17||−2.17, 4.53||0.487||4.92||1.70, 8.15||0.003|
|Class II||0.00||−1.60, 1.59||0.990||1.06||−0.22, 2.33||0.103|
|Class III||2.86||−0.03, 5.76||0.052||−0.5||−2.87, 1.87||0.676|
|Malocclusion vertical plane|
|Overbite 1-4 mm||Reference|
|Overbite >4 mm||0.12||−1.31, 1.55||0.866||−0.59||−1.94, 0.77||0.393|
|Overbite <1 mm||2.19||0.15, 4.22||0.035||0.26||−1.56, 2.08||0.777|
|Malocclusion transverse plane|
|Crossbite||0.57||−0.84, 1.98||0.428||0.98||−0.55, 2.52||0.207|
|Mild 0-3 mm||Reference|
|Moderate 3-5 mm||0.92||−0.56, 2.42||0.219||0.91||−0.48, 2.30||0.198|
|Severe >5 mm||0.09||−1.81, 2.01||0.918||1.151||−0.60, 2.90||0.194|