Abstract
Background
Dental implant therapy is widely used to restore dentition defects or loss, however, as a complication of it, interproximal contact loss (ICL) has not been adequately studied.
Objective
To explore the prevalence of ICL between implant-supported prostheses and adjacent teeth, and to analyze the risk factors for ICL.
Methods
Two hundred and four participants who received 445 implant-supported prostheses from January 2011 to December 2020 were enrolled in the study. Aluminum strips of different thicknesses were used to determine the interspace between the implant-support prostheses and the adjacent teeth. Medical records of implant prostheses and participants were reviewed. Periodontal condition, occlusion, and additional factors were examined. Statistical analyses were performed to estimate the prevalence rate of ICL and its influential factors.
Results
The prevalence of ICL was 59.8% at the patient level for 10-years function duration and increased over time. For the mesial site, the prevalence of ICL was 40.0% and food impaction (OR = 4.991, P < 0.001), adjacent teeth status (OR = 4.062, P = 0.042), function duration (OR = 1.299, P < 0.001) and bruxism (OR = 2.098, P = 0.034) were independent risk factors. For the distal site, the prevalence of ICL was 24.1% and food impaction (OR = 2.809, P = 0.002), mechanical complications (OR = 7.041, P < 0.001), bruxism (OR = 2.356, P = 0.019) and peri-implantitis (OR = 2.385, P = 0.021) were independent risk factors for ICL.
Conclusions
ICL was found to increase over time and was more common at the mesial site than at the distal site. Food impaction, function duration, adjacent teeth status, implant mechanical complications, bruxism and peri-implantitis were risk factors for ICL.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12903-025-06783-9.
Keywords: Clinical research, Interproximal contact loss, Dental implant, Adjacent teeth, Implant prostheses, Implant complication
Introduction
Dental implant therapy has become a cornerstone for restoring dentition defects due to its high predictability and long-term success. However, its complications are frequently encountered [1], such as peri-implantitis, veneer porcelain chipping, interproximal contact loss (ICL), etc [2–4]. ICL, the deterioration of physiological contact between implant-supported prostheses and adjacent natural teeth or implants, may result in food impaction, increasing the risk of tissue inflammation or caries in the adjacent tooth and reducing the implant longevity [5].
Several studies have investigated the prevalence and risk factors of ICL [6–10]. Data showed ICL prevalence rates ranged from 15.0 to 59.9%, which correlated with age, gender, opposing dentition, functional duration, adjacent tooth vitality, splinted implants, mesial sites, etc. However, significant heterogeneity existed in prevalence rates and risk factors across studies, potentially due to differences in measurement protocols (e.g., dental floss vs. aluminum strips) and analytical frameworks (e.g., patient-level vs. site-level analyses). Notably, all studies were limited by small sample sizes, which undermined the statistical power and generalizability of their findings. Therefore, the epidemiology of ICL after implant restoration should be further explored.
Aimed to provide more evidence-based medicine information of ICL to strengthen the prevention and care of ICL, we conducted this study by investigating the prevalence and risk factors of ICL in 204 patients with implant-supported prostheses.
Methods
Subjects
Participants who received implant-supported prostheses between January 2011 and December 2020 were included in this study, with exclusion criteria defined as: (1) functional time of implant-supported prostheses < 1 year; (2) undocumented proximal contact status at prosthesis delivery; (3) ongoing orthodontic treatment; (4) adjacent teeth exhibiting mobility ≥ Grade II (Miller’s classification); (5) defective crown restorations (e.g. veneer porcelain chipping); (6) edentulous arches; or (7) refusal to participate in follow-up interviews. Power calculations were conducted using the following formula.
 |
Where Z = 1.96 (α = 0.05), P = 0.15, and E = 0.05. This yielded n = 196. The final target sample size was set to 215 for assuming a 10% loss to follow-up.
The study was approved by the IRB of the Henan SATH Hospital of Stomatology, Henan University (HUSOM2020-295). Written informed consent was obtained from each participant before the study.
Medical record review
Data were extracted from electronic medical records and input into a spreadsheet. The data included age, gender, implant location (mandible or maxilla), implant site (molar or premolar), retention type between prostheses and abutment (screw or cement), restoration material (porcelain-fused-metal, resin-porcelain, or porcelain), function duration of prostheses (from the initiation of functional use of the prostheses to the study endpoint) and connection type between abutment and implant (internal or external). Participants’ private information was recorded with a unique identifier for anonymity.
Questionnaire and clinical assessment
The questionnaire was designed for this study and can be found in the supplementary material. It included items on bruxism history, unilateral mastication habits, smoking history, diabetes history, and regular periodontal maintenance. Participants independently completed the questionnaire before the clinical examination.
All clinical examinations were performed by the same examiner under the definite protocol of this study. Periodontal condition, occlusion, and additional factors were examined. The periodontal condition was assessed using a standardized examination chart, which included: (1) implant probing (depth and bleeding) at six sites per prosthesis with a 0.25N force probe; (2) peri-implantitis diagnosis [11], defined as bleeding on probing (BOP) and/or suppuration accompanied by radiographic bone loss > 2 mm beyond physiologic remodeling; (3) gingival papilla recession evaluated via the Jemt Papilla Index (0–4 scale: 0 = no papilla; 4 = complete papilla fill) [12] around implant-supported prostheses. Occlusal patterns were classified as either canine guidance (characterized by disclusion of posterior teeth during lateral excursions) or group function (involving multiple tooth contacts), assessed using 200-µm articulating paper. Occlusal wear severity was graded according to the Tooth Wear Index (0–3 scale: 0 = no wear; 3 = severe dentin exposure), evaluated across opposing teeth and full-mouth dentition. Some factors were also evaluated, including the status of adjacent teeth (implant/natural tooth, implant/crown, and implant/implant), endodontic conditions of the adjacent teeth (vital, nonvital, or absent), food impaction (vertical impaction and horizontal impaction) and mechanical complications (abutment dislocation, screw fracture, implant fracture, loose implant crown).
Radiography
Parallel X-ray technology was used to calculate the bone loss around implant. All participants bit a block when the radiography of the implants was performed. Radiographs were taken at the time of prostheses delivery (baseline) and subsequent follow-up. The bone level of each implant was determined from the implant osseointegrated with alveolar to the bottom of implant. The actual implant length served as the calibration value. The levels of bone loss at the mesial and the distal sites were measured respectively and calculated using the following formula.
 |
Implant Length is the straight-line distance from the implant platform to the apical end of the root.
†Bone Level and Implant Length determined on X-ray in the baseline.
‡ Bone Level and Implant Length determined on X-ray in the following time.
ICL measurement
ICL was measured using aluminum strips by the same examiner. Aluminum strips 20, 30, 40, 50, 60, 70, 80, 90, and 100 μm thick were inserted into the interspace between the implant-support prostheses and adjacent teeth until resistance was encountered, and no more strips could be inserted. Then, the final thickness of the interspace was recorded. If the thickness exceeded 50 μm, it was classified as ICL. Otherwise, it was categorized as proximal contact close. For all the participants, when the prostheses were delivered, the initial interspace between implant-supported prostheses and adjacent teeth would be checked for dental floss to pass through with high resistance. These results were also recorded in their medical records.
Statistical analysis
Data analysis was performed with SPSS 24.0 software (New York, NY, USA). The prevalence of ICL was calculated at three levels (patient, mesial and distal). For each implant-supported prosthesis, the ICL at the mesial and the distal sites were examined separately. Chi-square tests and U tests were used to compare the prevalence of ICL between the mesial and the distal sites and analyze the factors associated with ICL. The level of significance was set at 0.05. Furthermore, Logistic regression analysis was used to investigate independent factors influencing ICL and multivariate linear regression analysis was performed to investigate the association between risk factors and interspace thickness. Variable selection was conducted based on the theory-driven inclusion.
Results
Patient characteristics
A total of 204 participants (93 females and 111 males) with 445 implants were enrolled (Fig. 1). Of these participants, 140 were aged ≤ 60 years and 64 were older than 60, with 93 females and 111 males in total. Among them, 32 individuals had a history of diabetes, 55 had a history of smoking, 23 had bruxism, 84 had a unilateral mastication habit and 83 underwent regular periodontal maintenance. Regarding the 445 implants, 204 were placed in the maxilla while 241 were in the mandible, 396 were in the molar area while 49 were in the premolar area. There were 358 prostheses in total, of which 289 were single crowns and 69 were bridges. Overall, 565 contact sites were counted with 345 mesial sites and 220 distal sites (Table 1).
Fig. 1.
Research process and key points
Table 1.
Description of the study samples
| Item | Number | Percentage |
|---|---|---|
| Patient number | 204 | |
| Gender: Male | 111 | 54.4% |
| Gender: Female | 93 | 45.6% |
| Age ≤ 60 | 140 | 68.6% |
| Age > 60 | 64 | 31.4% |
| Diabetes history | 32 | 15.7% |
| Smoking history | 55 | 27.0% |
| Regular periodontal maintenance | 83 | 40.7% |
| Bruxism | 23 | 11.3% |
| Unilateral mastication | 84 | 41.2% |
| Implant number | 445 | |
| Location: Maxilla | 204 | 45.8% |
| Location: Mandible | 241 | 54.2% |
| Site: Molar | 396 | 89.0% |
| Site: Premolar | 49 | 11.0% |
| Protheses number | 358 | |
| Single crown | 289 | 80.7% |
| Bridge | 69 | 19.3% |
| Contact site | 565 | |
| Mesial | 345 | 61.1% |
| Distal | 220 | 38.9% |
Prevalence of ICL
Finally, among 204 participants who received dental implants within the past ten years, 122 (59.8%) exhibited at least one mesial or distal site identified as ICL. In the analysis of 565 contact sites in participants, 138 of 345 mesial site (40.0%) were identified as ICL, compared to 53 of 220 distal sites (24.1%) (Fig. 2A). The frequency distribution of the interproximal space of 565 contact sites was also analyzed in this study (Fig. 2A) and the thickness of the mesial interproximal space was significantly bigger than that of the distal interproximal space (P < 0.001). Besides, the data demonstrated that the prevalence of ICL exhibited a progressive increase with functional duration of duration, with the mesial site showing significantly greater occurrence compared to the distal site (P < 0.001), as seen in Fig. 2B.
Fig. 2.
Cumulative prevalence of ICL at mesial and distal sites over time. A shows interproximal space distribution, while B illustrates the increase in ICL prevalence with function duration
Potential factors influencing ICL at the patient level
Chi-square tests were used to analyze associate factors of ICL at the patient level. The results indicated that age (P = 0.402), gender (P = 0.453), diabetes history (P = 0.624), smoking history (P = 0.929), regular periodontal maintenance (P = 0.267), bruxism (P = 0.912), and unilateral mastication (P = 0.167) were not significantly associated with ICL at the patient level. Although there was no statistical significance, the participants with unilateral mastication or without regular periodontal maintenance would have higher ICL risks (Table 2).
Table 2.
Univariate analysis of potential prognostic factors for ICL at the participant level
| Item | Contact (n = 82) | Contact loss (n = 122) | P value |
|---|---|---|---|
| Gender (female) | 40 (48.8%) | 53 (43.4%) | 0.453 |
| Age (> 60) | 23 (28.0%) | 41 (33.6%) | 0.402 |
| Diabetes history | 12 (14.6%) | 21 (17.2%) | 0.624 |
| Smoking history | 24 (29.3%) | 35 (28.7%) | 0.929 |
| Regular periodontal maintenance | 38 (46.3%) | 47 (38.5%) | 0.267 |
| Bruxism | 9 (11%) | 14 (11.5%) | 0.912 |
| Unilateral mastication | 29 (35.4%) | 55 (45.1%) | 0.167 |
Nominal data were expressed using numbers (percentages)
Potential factors influencing ICL at the mesial site level
Chi-square tests and U tests were used to analyze the factors between ICL at the mesial site level. The results showed that food impaction (P < 0.001), implant location (P = 0.038), connection type (P = 0.033), adjacent teeth status (P = 0.007), endodontic conditions (P = 0.009) and function duration (P < 0.001) were significantly associated with ICL at the mesial site level. There were no statistically significant differences among the following factors: implant site(P = 0.252), prostheses retained(P = 0.829) and bone loss(P = 0.411) (Table 3).
Table 3.
Univariate analysis of potential prognostic factors for ICL at the mesial site level
| Item | Contact (n = 207) | Contact loss (n = 138) | P value |
|---|---|---|---|
| Food impaction | 50 (24.2%) | 71 (51.4%) | < 0.001*** |
| Implant location (mandible) | 101 (48.8%) | 83 (60.1%) | 0.038* |
| Implant site (molar) | 188 (90.8%) | 130 (94.2%) | 0.252 |
| Adjacent teeth status | - | - | 0.007** |
| Implant/Natural tooth | 149 (72.0%) | 102 (73.9%) | - |
| Implant/Crown | 21 (10.1%) | 25 (18.1%) | - |
| Implant/Implant | 37 (17.9%) | 11 (8.0%) | - |
| Endodontic conditions | - | - | 0.009** |
| Vital | 144 (69.9%) | 99 (71.7%) | - |
| Nonvital | 25 (12.1%) | 28 (20.3) | - |
| Absent | 31 (18%) | 11 (8%) | - |
| Prostheses retained (cement) | 186 (89.9%) | 123 (89.1%) | 0.829 |
| Connection type (internal) | 196 (94.7%) | 122 (88.4%) | 0.033* |
| Function duration (years) | 2.7 ± 1.8 (2) | 3.5 ± 2.2 (3) | < 0.001*** |
| Gingival papilla recession | 70 (33.8%) | 59 (42.8%) | 0.093 |
| Bone loss | 0.21 ± 0.71 (0.00) | 0.24 ± 0.70 (0.00) | 0.411 |
Nominal data were expressed using numbers (percentages); scalar data were expressed using mean ± SD (median)
*P < 0.05, **P < 0.01, ***P < 0.001
Logistic regression was used to reveal the independent factors of mesial ICL as a multivariate analysis method. Function duration (OR = 1.299, 95%CI = 1.161 ~ 1.453, P < 0.001) and adjacent teeth status (P < 0.001) were risk factors for mesial ICL. There was a strong correlation between adjacent teeth status and mesial ICL. Among them, crown (OR = 4.062, 95%CI = 1.053 ~ 15.67, P = 0.042) had a higher risk than natural tooth and implant (OR = 0.314, 95%CI = 0.078 ~ 1.269, P = 0.104) (Table 4). In addition, the use of an external connection between the implant and abutment is more likely to lead to ICL at the mesial site level (OR = 2.066, 95%CI = 0.858 ~ 4.975, P = 0.106). The model goodness-of-fit was assessed using the Hosmer-Lemeshow test, yielding a chi-square value of 3.35 (P = 0.908).
Table 4.
Logistic regression analysis of potential prognostic factors for ICL at the mesial site level
| Item | OR | P value |
|---|---|---|
| Implant location (mandible) | 1.223 (0.763 ~ 1.96) | 0.403 |
| Implant site (molar) | 1.291 (0.777 ~ 2.144) | 0.324 |
| Function duration (years) | 1.299 (1.161 ~ 1.453) | < 0.001*** |
| Status of adjacent teeth | < 0.001*** | |
| Loss | reference | |
| Implant/Natural tooth | 3.169 (0.893 ~ 11.244) | 0.074 |
| Implant/Crown | 4.062 (1.053 ~ 15.67) | 0.042* |
| Implant/Implant | 0.314 (0.078 ~ 1.269) | 0.104 |
| Connection type (external) | 2.066 (0.858 ~ 4.975) | 0.106 |
OR = odds ratio
*P < 0.05, **P < 0.01, ***P < 0.001
Logistic regression analysis was also used to analyze the current data of mesial ICL. When food impaction and gingival papilla recession were included in the regression model, the width of keratinized gingiva (OR = 1.131, 95%CI = 1.024 ~ 1.249, P = 0.015), food impaction (OR = 4.991, 95%CI = 2.809 ~ 8.867, P < 0.001), bruxism (OR = 2.098, 95%CI = 1.058 ~ 4.158, P = 0.034), and full mouth wear (OR = 1.875, 95%CI = 1.028 ~ 3.421, P = 0.04) were significantly correlated with mesial ICL. In addition, the status of adjacent teeth (P < 0.001) was also significantly associated with mesial ICL, crowns (OR = 1.695, 95%CI = 0.437 ~ 6.573, P = 0.446) had a higher risk than natural teeth (OR = 1.344, 95%CI = 0.377 ~ 4.788, P = 0.648) and implants (OR = 0.233, 95%CI = 0.057 ~ 0.947, P = 0.042). The Hosmer-Lemeshow test showed a chi-square value of 10.213 (P = 0.25). After removing the factors of food impaction and gingival papilla recession from the model, the status of adjacent teeth (P < 0.001) and bruxism (OR = 1.905, 95%CI = 0.999 ~ 3.634, P = 0.05) remained significantly associated with mesial ICL. The Hosmer-Lemeshow test showed a chi-square value of 3.856 (P = 0.87) (Table 5).
Table 5.
Logistic regression analysis of associated cross-sectional risk factors for ICL at the mesial site level
| Item | OR | P value | OR | P value |
|---|---|---|---|---|
| Food impaction | 4.991 (2.809 ~ 8.867) | < 0.001*** | - | - |
| Gingival papilla recession | 1.633 (0.627 ~ 4.25) | 0.315 | - | - |
| Periodontitis | - | 0.225 | - | 0.605 |
| Keratinized gingival width | 1.131 (1.024 ~ 1.249) | 0.015* | 1.08 (0.987 ~ 1.182) | 0.093 |
| Keratinized gingival thickness | 1.408 (0.725 ~ 2.733) | 0.312 | 1.388 (0.746 ~ 2.585) | 0.301 |
| Opposing teeth wear | 0.626 (0.326 ~ 1.203) | 0.16 | 0.739 (0.399 ~ 1.367) | 0.335 |
| Full mouth wear | 1.875 (1.028 ~ 3.421) | 0.04* | 1.603 (0.916 ~ 2.806) | 0.098 |
| Adjacent teeth status | - | < 0.001*** | - | < 0.001*** |
| Loss | reference | reference | ||
| Implant/Natural tooth | 1.344 (0.377 ~ 4.788) | 0.648 | 2.1 (0.624 ~ 7.065) | 0.231 |
| Implant/Crown | 1.695 (0.437 ~ 6.573) | 0.446 | 2.89 (0.795 ~ 10.507) | 0.107 |
| Implant/Implant | 0.233 (0.057 ~ 0.947) | 0.042* | 0.304 (0.078 ~ 1.184) | 0.086 |
| Bruxism | 2.098 (1.058 ~ 4.158) | 0.034* | 1.905 (0.999 ~ 3.634) | 0.05 |
OR = odds ratio
*P < 0.05, **P < 0.01, ***P < 0.001
In the multivariate regression model (adjusted R2 = 0.154), mesial ICL was associated with food impaction (B = 17.152, 95%CI = 0.697 ~ 33.607, P = 0.041), function duration (B = 6.870, 95%CI = 3.481 ~ 10.259, P < 0.001), adjacent teeth status (implant/implant) (B=−34.241, 95%CI=−50.538~−17.943, P < 0.001), gingival papilla recession (B = 26.532, 95%CI = 10.759 ~ 42.306, P = 0.001) and connection type (B = 34.927, 95%CI = 11.837 ~ 58.017, P = 0.003) (Table 6).
Table 6.
Multivariate linear regression analyze association between factors and interspace thickness at the mesial site level
| Item | B | P value | VIF |
|---|---|---|---|
| Implant location (mandible) | 6.522 (−7.236 ~ 20.280) | 0.352 | 1.042 |
| Implant site (molar) | 16.735 (−5.722 ~ 39.192) | 0.144 | 1.096 |
| Food impaction | 17.152 (0.697 ~ 33.607) | 0.041* | 1.213 |
| Function duration (years) | 6.870 (3.481 ~ 10.259) | < 0.001*** | 1.049 |
| Adjacent teeth status | - | - | - |
| Implant/Natural tooth | reference | - | - |
| Implant/Crown | 19.534 (−1.629 ~ 40.697) | 0.070 | 1.076 |
| Implant/Implant | −34.241 (−50.538~−17.943) | < 0.001*** | 1.160 |
| Gingival papilla recession | 26.532 (10.759 ~ 42.306) | 0.001*** | 1.161 |
| Connection type (external) | 34.927 (11.837 ~ 58.017) | 0.003** | 1.053 |
VIF = variance inflation factor
*P < 0.05, **P < 0.01, ***P < 0.001
Potential factors influencing ICL at the distal site
Chi-square tests and U tests were used to analyze any factors associated with ICL. The results indicated that food impaction (P = 0.025), mechanical complications (P < 0.001), occlusion (P < 0.001), opposing teeth wear (P = 0.013), full-mouth wear (P = 0.035) and gingival papilla recession (P = 0.039) were significantly associated with ICL at the distal site level. No significant differences were observed among the following factors: restoration material (P = 0.577), probing for depth (P = 0.146), and probing for bleeding (P = 0.678) (Table 7).
Table 7.
Univariate analysis of potential prognostic factors for ICL at the distal site level
| Item | Contact (n = 167) | Contact loss (n = 53) | P value |
|---|---|---|---|
| Food impaction | 27 (16.2%) | 16 (30.2%) | 0.025* |
| Mechanical complications | 6 (3.6%) | 11 (20.8%) | < 0.001*** |
| Restoration material | - | - | 0.577 |
| porcelain-fused-metal | 36 (21.4%) | 12 (22.6%) | - |
| resin-porcelain | 40 (23.8%) | 9 (17%) | - |
| porcelain | 92 (54.8%) | 32 (60.4%) | - |
| Probing for depth | 2.9 ± 1.5 (3) | 3.3 ± 1.8 (3) | 0.146 |
| Probing for bleeding | 72 (42.9%) | 21 (39.6%) | 0.678 |
| Gingival papilla recession | 39 (23.4%) | 20 (37.7%) | 0.039* |
| Function duration (years) | 3.0 ± 1.9(3) | 3.3 ± 2.1(3) | 0.588 |
| Occlusion (group function) | 98 (58.7%) | 46 (86.8%) | < 0.001*** |
| Opposing teeth wear | 40 (24%) | 22 (41.5%) | 0.013* |
| Full mouth wear | 61 (36.5%) | 28 (52.8%) | 0.035* |
| Peri-implantitis | 11 (6.5%) | 8 (15.1%) | 0.053 |
Nominal data were expressed using numbers (percentages); scalar data were expressed using mean ± SD (median)
*P < 0.05, **P < 0.01, ***P < 0.001
Logistic regression was used to identify independent factors of ICL as a multivariate analysis method. Peri-implantitis (OR = 2.385, 95%CI = 1.139 ~ 4.993, P = 0.021) and mechanical complications (OR = 7.041, 95%CI = 3.093 ~ 16.028, P < 0.001) were significantly associated with ICL at the distal site level (Table 8). The model goodness-of-fit was assessed using the Hosmer-Lemeshow test, yielding a chi-square value of 5.199 (P = 0.736).
Table 8.
Logistic regression analysis of potential prognostic factors for ICL at the distal site level
| Item | OR | P value |
|---|---|---|
| Mechanical complications | 7.041 (3.093 ~ 16.028) | < 0.001*** |
| Peri-implantitis | 2.385 (1.139 ~ 4.993) | 0.021* |
| Occlusion | - | 0.235 |
| balanced occlusion | reference | - |
| group function occlusion | 2.063 (0.393 ~ 10.838) | 0.392 |
| cuspid-protected occlusion | 1.084 (0.191 ~ 6.149) | 0.927 |
| Full mouth wear | 1.119 (0.574 ~ 2.179) | 0.742 |
| Opposing teeth wear | 1.108 (0.568 ~ 2.163) | 0.763 |
| Function duration | 1.046 (0.926 ~ 1.182) | 0.466 |
OR = odds ratio
*P < 0.05, **P < 0.01, ***P < 0.001
Logistic regression analysis was further used to examine ICL data at the distal site level. When food impaction and gingival papilla recession were included in the regression model, food impaction (OR = 2.809, 95%CI = 1.444 ~ 5.464, P = 0.002), peri-implantitis (OR = 3.258, 95%CI = 1.484 ~ 7.151, P = 0.003), and bruxism (OR = 2.356, 95%CI = 1.153 ~ 4.814, P = 0.019) were significantly correlated with distal ICL. The Hosmer-Lemeshow test showed a chi-square value of 6.561 (P = 0.585). After removing food impaction and gingival papilla recession from the model, peri-implantitis (OR = 2.991, 95%CI = 1.421 ~ 6.297, P = 0.004) and bruxism (OR = 2.450, 95%CI = 1.258 ~ 4.769, P = 0.008) remained significantly associated with distal ICL. The Hosmer-Lemeshow test showed a chi-square value of 14.309 (P = 0.074) (Table 9).
Table 9.
Logistic regression analysis of associated cross-sectional risk factors for ICL at the distal site level
| Item | OR | P value | OR | P value |
|---|---|---|---|---|
| Food impaction | 2.809 (1.444 ~ 5.464) | 0.002** | - | - |
| Gingival papilla recession | 2.291 (0.821 ~ 6.388) | 0.113 | - | - |
| Periodontitis | - | 0.089 | - | 0.232 |
| Keratinized gingival width | 1.078 (0.959 ~ 1.212) | 0.208 | 1.049 (0.942 ~ 1.167) | 0.385 |
| Keratinized gingival thickness | 0.587 (0.251 ~ 1.374) | 0.22 | 0.619 (0.271 ~ 1.413) | 0.255 |
| Opposing teeth wear | 0.931 (0.448 ~ 1.938) | 0.849 | 0.968 (0.489 ~ 1.913) | 0.924 |
| Full mouth wear | 1.564 (0.79 ~ 3.094) | 0.199 | 1.624 (0.859 ~ 3.073) | 0.136 |
| Adjacent teeth status | - | 0.182 | - | 0.408 |
| Loss | reference | - | reference | - |
| Implant/Natural tooth | 0.852 (0.192 ~ 3.781) | 0.833 | 0.911 (0.217 ~ 3.822) | 0.899 |
| Implant/Crown | 1.188 (0.273 ~ 5.172) | 0.818 | 1.055 (0.252 ~ 4.414) | 0.942 |
| Implant/Implant | 0.507 (0.11 ~ 2.328) | 0.382 | 0.575 (0.131 ~ 2.529) | 0.464 |
| Bruxism | 2.356 (1.153 ~ 4.814) | 0.019* | 2.45 (1.258 ~ 4.769) | 0.008** |
| Peri-implantitis | 3.258 (1.484 ~ 7.151) | 0.003** | 2.991 (1.421 ~ 6.297) | 0.004** |
OR = odds ratio
*P < 0.05, **P < 0.01
In the multivariate linear regression model (adjusted R2 = 0.059), ICL was associated with mechanical complications (B = 39.434, 95%CI = 11.549 ~ 57.318, P = 0.006) and gingival papilla recession (B = 38.150, 95%CI = 17.682 ~ 58.618, P < 0.001) at the distal site level (Table 10).
Table 10.
Multivariate linear regression analyze association between factors and interspace thickness at the distal site level
| Item | B | P value | VIF |
|---|---|---|---|
| Food impaction | 19.273 (−4.748 ~ 43.294) | 0.116 | 1.148 |
| Mechanical complications | 39.434 (11.549 ~ 57.318) | 0.006** | 1.017 |
| Gingival papilla recession | 38.150 (17.682 ~ 58.618) | < 0.001*** | 1.136 |
| Peri-implantitis | 1.662 (−20.873 ~ 24.198) | 0.885 | 1.030 |
| Occlusion (group function) | −10.909 (−27.030 ~ 5.212) | 0.184 | 1.193 |
| Full mouth wear | −1.738 (−20.101 ~ 15.625) | 0.853 | 1.785 |
| Opposing teeth wear | 5.342 (−13.497 ~ 24.180) | 0.578 | 1.561 |
VIF variance inflation factor.
*P < 0.05, **P < 0.01, ***P < 0.001.
Discussion
The prevalence of ICL reported in the previous studies has varied widely. In this study, the prevalence of ICL at the patient level was 59.8%, which is higher than the 43.5% reported in previous meta-analyses [13, 14].This discrepancy likely stems from statistical method variations: whereas not all studies included in those meta-analyses reported the prevalence of ICL at the patient level (several examined the prevalence of ICL only at the site level), this study compensates for this deficiency by reporting the prevalence of ICL at both site and patient levels. When adopting the site-level analysis, our data showed 40.0% mesial and 24.1% distal ICL rates, aligning with the historical range of 34.0%−78.2% [6–8, 15–17]. Methodological variations in ICL quantification also drive this discrepancy. Previous studies employed qualitative tools (dental floss [7, 15, 18, 19], 50 μm metal strips [5, 6, 8, 17], or 38 μm tofflemire matrix bands [16]) with limited sensitivity for detecting the precise interspace, whereas our protocol refined Pang’s method [8] through graded aluminum strips (20–100 μm) to objectively measure incremental space increases.
Mesial ICL risk (40.0%) exceeded distal risk (24.1%) by 1.66-fold, aligning with prior evidence [6–9, 15, 17]. Mechanistically, mesial displacement of natural teeth under occlusal anterior force components [6, 7, 15], a physiological compensation for interproximal attrition [20–22], contrasts with implant ankylosis [23, 24]. This differential mobility permits unrestricted mesial drift of pre-implant teeth while restraining post-implant distal teeth [25]. increasing the prevalence of mesial ICL. Within this paradigm, distal ICL pathogenesis may involve: (1) Flattened proximal surfaces on restorations may accelerate functional wear, particularly in distal site subjected to the anterior component of occlusal forces. (2) The anatomical complexity of distal zones impedes effective oral hygiene, promoting plaque accumulation that drives gingival recession and interproximal caries, thereby indirectly contributing to ICL.
The prevalence of ICL obtained from this study reached 37.1% in 4 years at the mesial site level and 25.3% in 2 years at the distal site level, demonstrating a gradual increase, aligning with the results reported by Pang et al. [8], thus we conclude that the prevalence of ICL increased with function duration. Notably, there was a transient decline in the prevalence of distal ICL in the third year. This pattern may reflect the heightened early-phase recall rates (≤ 2 years post-implantation) and poor oral hygiene adherence, potentially accelerating initial ICL prevalence. However, limited long-term data (n < 30 for > 3-years function duration) suggests selection bias: participants retained beyond 3 years likely have higher oral hygiene adherence, thus potentially mitigating the progression of ICL.
At the mesial proximal contact site level, adjacent teeth status was a significant risk of ICL. Logistic regression analysis revealed that the risk of ICL at mesial site between implant-supported prostheses and crown (OR = 4.062, P = 0.042) was higher than that between two adjacent implant-supported prostheses (OR = 0.314, P = 0.104). This phenomenon may be attributed to the significantly higher hardness of crown compared to natural enamel, which generates hard-hard friction at the implant-crown interface, accelerating proximal wear. Additionally, microleakage at the cementation margin of crown may induce interfacial corrosion, thereby compromising the stability of interproximal contacts. It is worth noting that the wide confidence intervals, ranging from 1.053 to 15.67, make the results less reliable and require more evidence to support them. Multivariate linear regression showed adjacent teeth for implant was a protective factor of mesial interspace thickness (B=−34.241, P < 0.001), possibly because the osseointegrated implant within the alveolar bone behaves more akin to an ankylotic tooth, exhibiting minimal physical drift [26, 27]. Meanwhile, the connection type between abutment and implant influenced the risk of ICL, external connection was the negative factor of mesial interspace thickness (B = 34.927, P = 0.003). We speculate that this is because external connection has more micromotion and mechanical complications. In particular, lateral force leads to instability of joint [28].
At the distal proximal contact site level, the mechanical complication was related to interspace thickness, as well as a significant risk factor of ICL (OR = 7.041, P < 0.001). This phenomenon suggested that ICL may predispose to mechanical failures, while pre-existing mechanical issues could concurrently exacerbate contact deterioration. Notably, occlusion emerged as another key risk factor for distal ICL. Although current clinical protocols advocate reduced occlusal loading and protected occlusion with anterior guidance [29] for implant prostheses, these biomechanical adaptations inadvertently redirect excessive forces to adjacent natural teeth [30]. Crucially, distal implant sites exhibit diminished capacity to dissipate occlusal stresses compared to mesial sites due to reduced adjacent teeth support [6]. Consequently, occlusion emerges as a risk factor for ICL at the distal site. Simultaneously, our data also revealed a higher prevalence of ICL at mesial (OR = 2.098, P = 0.034) and distal sites (OR = 2.450, P = 0.008) among bruxism patients. Bruxism-induced parafunctional habits generate sustained supraphysiological bite forces [31–33]. Given that implants are osseointegrated within the jawbone, patients may lack the sensory feedback to detect excessive bite force. This sensory deficit prevents physiological force modulation, leading to implant mechanical complications and wear of the complete denture [34, 35]. Hence, the prevalence of ICL is increased in patients with bruxism.
There is no definitive evidence to suggest that food impaction and gingival papilla recession lead to ICL or vice versa, the retrospective design of this study also limits causal inference reliability [36]. Here, our hypothesis posits a synergistic relationship where ICL initially manifests and subsequently results in mild food impaction and gingival papilla recession, which then exacerbates the extent of ICL. To investigate this hypothesis, we employed sensitivity analyses by using logistic analytical models: retaining factors of food impaction and gingival papilla recession vs. excluding them. It showed difference in adjusted R² between models retaining and excluding these two factors at the mesial site (pre-exclusion vs. post-exclusion: 0.183 vs. 0.111), and similar finding was observed at the distal site (pre-exclusion vs. post-exclusion: 0.123 vs. 0.061). The observed deterioration in models fit following the exclusion of these two factors suggests their potential role as contributing factors to ICL. Meanwhile, our findings indicate a significant positive correlation between food impaction and ICL at both mesial (OR = 4.991, P < 0.001) and distal sites (OR = 2.809, P = 0.002), corroborating previous studies [7, 8, 15, 16, 18].
As an observational study, this research has several limitations. (1) the measurement method might be disturbed by the mobility of natural teeth and the force applied by the examiner while inserting aluminum strips, potentially leading to measurement error. Although we tried to minimize this error by designating a single examiner to conduct the measurements, a risk of bias remains. (2) the precise moment when ICL occurs was indeterminable, which could lead to its underestimated. Consequently, prospective studies with longitudinal measurements were recommended to establish temporal relationships. (3) Incomprehensive baseline parameters recorded during the implantation (e.g., undocumented floss type/thickness, interproximal space quantification at implantation) precludes precise ICL prevalence calculation. In addition, the presence of food impaction, gingival recession, adjacent dental caries at the time of implantation, and the quality and fit of the implant itself were not documented. These confounders further limit validity of this retrospective study.
Conclusion
This study determined that the prevalence of ICL was 59.8% at the patient level. Notably, ICL was significantly more prevalent at the mesial site level (40.0%) compared to the distal site level (23.4%). Identified risk factors for ICL included food impaction, function duration, the status of adjacent teeth, mechanical complications, bruxism and peri-implantitis.
Supplementary Information
Acknowledgements
The authors thank Yanfang Ren for giving many constructive comments, Yu Wang for analyzing data, Wei Zhang and Juan Kang for collecting data.
Abbreviation
- ICL
Interproximal contact loss
Authors’ contributions
CN, BH: Contributed to conception and design of the study. XC, CZ, YL, GZ, RW: Contributed to data acquisition and analysis. XC, CZ, XL, YZ: Contributed to data analysis, interpretation, and manuscript drafting. All authors critically revised the manuscript and gave final approval for submission.
Funding
This research was supported by grants from the Foundation of Science & Technology Department of Henan Province, China (No. 252102311099), the Foundation of Science & Technology Department of Kaifeng, Henan Province, China (No. 2203023), and the Foundation for Young Scholar in Henan University School of Stomatology (No. HUSSYS2024005).
Data availability
The data that support the findings of this study are available from the corresponding author, upon reasonable request.
Declarations
Ethics approval and consent to participate
This study was conducted in full accordance with the Declaration of Helsinki and approved by the IRB of the Henan SATH Hospital of Stomatology, Henan University (HUSOM2020-295). Written informed consent was obtained from each participant before the study.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Xiaoyi Cui and Chunmiao Zhang are contributed equally.
Contributor Information
Chenguang Niu, Email: asdncg@henu.edu.cn.
Baojie He, Email: hebaojie@sais.cn.
References
- 1.Von Stein-Lausnitz M, et al. Survival rates and complication behaviour of tooth implant-supported, fixed dental prostheses: A systematic review and meta-analysis. J Dent. 2019;88:103167. [DOI] [PubMed] [Google Scholar]
- 2.Adler L, Buhlin K, Jansson L. Survival and complications: A 9- to 15-year retrospective follow-up of dental implant therapy. J Oral Rehabil. 2020;47(1):67–77. [DOI] [PubMed] [Google Scholar]
- 3.Papaspyridakos P, et al. Complications and survival rates of 55 metal-ceramic implant-supported fixed complete-arch prostheses: A cohort study with mean 5-year follow-up. J Prosthet Dent. 2019;122(5):441–9. [DOI] [PubMed] [Google Scholar]
- 4.Papaspyridakos P, et al. Double Full-Arch fixed Implant-Supported prostheses: outcomes and complications after a mean Follow-Up of 5 years. J Prosthodont. 2019;28(4):387–97. [DOI] [PubMed] [Google Scholar]
- 5.Wei H, et al. Implant prostheses and adjacent tooth migration: preliminary retrospective survey using 3-dimensional occlusal analysis. Int J Prosthodont. 2008;21(4):302–4. [PubMed] [Google Scholar]
- 6.Koori H, et al. Statistical analysis of the diachronic loss of interproximal contact between fixed implant prostheses and adjacent teeth. Int J Prosthodont. 2010;23(6):535–40. [PubMed] [Google Scholar]
- 7.Byun SJ, et al. Analysis of proximal contact loss between implant-supported fixed dental prostheses and adjacent teeth in relation to influential factors and effects. A cross-sectional study. Clin Oral Implants Res. 2015;26(6):709–14. [DOI] [PubMed] [Google Scholar]
- 8.Pang NS, et al. Prevalence of proximal contact loss between implant-supported fixed prostheses and adjacent natural teeth and its associated factors: a 7-year prospective study. Clin Oral Implants Res. 2017;28(12):1501–8. [DOI] [PubMed] [Google Scholar]
- 9.Yen JY, et al. Risk assessment of interproximal contact loss between implant-supported fixed prostheses and adjacent teeth: A retrospective radiographic study. J Prosthet Dent. 2020;127(1):86–92. [DOI] [PubMed]
- 10.Bompolaki D, Edmondson SA, Katancik JA. Interproximal contact loss between implant-supported restorations and adjacent natural teeth: A retrospective cross-sectional study of 83 restorations with an up to 10-year follow-up. J Prosthet Dent. 2022;127(3):418–24. [DOI] [PubMed] [Google Scholar]
- 11.Berglundh T, et al. Peri-implant diseases and conditions: consensus report of workgroup 4 of the 2017 world workshop on the classification of periodontal and Peri-Implant diseases and conditions. J Clin Periodontol. 2018;45:S286–91. [DOI] [PubMed] [Google Scholar]
- 12.Jemt T. Regeneration of gingival papillae after single-implant treatment. Int J Periodontics Restor Dent. 1997;17(4):326–33. [PubMed] [Google Scholar]
- 13.Papageorgiou SN, Eliades T, Hämmerle CHF. Frequency of infraposition and missing contact points in implant-supported restorations within natural dentitions over time: A systematic review with meta-analysis. Clin oral implants res, 2018-12. 29 Suppl 18: pp. 309–25. [DOI] [PubMed]
- 14.Sheba M, et al. Interproximal contact loss between implant restorations and adjacent natural teeth: A systematic review and meta-analysis. J Prosthodont. 2024;33(4):313–23. [DOI] [PubMed] [Google Scholar]
- 15.Varthis S, Randi A, Tarnow DP. Prevalence of interproximal open contacts between Single-Implant restorations and adjacent teeth. Int J Oral Maxillofac Implants. 2016;31(5):1089–92. [DOI] [PubMed] [Google Scholar]
- 16.Wong AT, et al. Proximal contact loss between implant-supported prostheses and adjacent natural teeth: a retrospective study. Clin Oral Implants Res. 2015;26(4):e68–71. [DOI] [PubMed] [Google Scholar]
- 17.Ren S, et al. Changes in proximal contact tightness between fixed implant prostheses and adjacent teeth: A 1-year prospective study. J Prosthet Dent. 2016;115(4):437–40. [DOI] [PubMed] [Google Scholar]
- 18.Liang CH, et al. The prevalence and associated factors of proximal contact loss between implant restoration and adjacent tooth after function: A retrospective study. Clin Implant Dent Relat Res. 2020;22(3):351–8. [DOI] [PubMed] [Google Scholar]
- 19.Saber A, et al. Prevalence of interproximal contact loss between Implant-Supported fixed prostheses and adjacent teeth and its impact on marginal bone loss: A retrospective study. Int J Oral Maxillofac Implants. 2020;35(3):625–30. [DOI] [PubMed] [Google Scholar]
- 20.Mijiritsky E, et al. Continuous tooth eruption adjacent to single-implant restorations in the anterior maxilla: aetiology, mechanism and outcomes - A review of the literature. Int Dent J. 2020;70(3):155–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Bartlett D, O’Toole S. Tooth wear and aging. Aust Dent J. 2019;64(1):S59–62. [DOI] [PubMed] [Google Scholar]
- 22.Bishara SE, Treder JE, Jakobsen JR. Facial and dental changes in adulthood. Am J Orthod Dentofac Orthop. 1994;106(2):175–86. [DOI] [PubMed] [Google Scholar]
- 23.Rittel D, Dorogoy A, Shemtov-Yona K. Modeling the effect of osseointegration on dental implant pullout and torque removal tests. Clin Implant Dent Relat Res. 2018;20(5):683–91. [DOI] [PubMed] [Google Scholar]
- 24.Papalexopoulos D, et al. Impact of maxillofacial growth on implants placed in adults: A narrative review. J Esthet Restor Dent. 2023;35(3):467–78. [DOI] [PubMed] [Google Scholar]
- 25.Jo DW, et al. Evaluation of adjacent tooth displacement in the posterior implant restoration with proximal contact loss by superimposition of digital models. J Adv Prosthodont. 2019;11(2):88–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Bosshardt DD, Chappuis V, Buser D. Osseointegration of titanium, titanium alloy and zirconia dental implants: current knowledge and open questions. Periodontol 2000, 2017. 73(1): pp. 22–40. [DOI] [PubMed]
- 27.Pellegrini G, et al. Novel surfaces and osseointegration in implant dentistry. J Investig Clin Dent. 2018;9(4):e12349. [DOI] [PubMed] [Google Scholar]
- 28.Huang Y, Wang J. Mechanism of and factors associated with the loosening of the implant abutment screw: A review. J Esthet Restor Dent. 2019;31(4):338–45. [DOI] [PubMed] [Google Scholar]
- 29.Sheridan RA, et al. The role of occlusion in implant therapy: A comprehensive updated review. Implant Dent. 2016;25(6):829–38. [DOI] [PubMed] [Google Scholar]
- 30.Sarig R, et al. The arrangement of the interproximal interfaces in the human permanent dentition. Clin Oral Investig. 2013;17(3):731–8. [DOI] [PubMed] [Google Scholar]
- 31.Lobbezoo F, Van Der Zaag J, Naeije M. Bruxism: its multiple causes and its effects on dental implants - an updated review. J Oral Rehabil. 2006;33(4):293–300. [DOI] [PubMed] [Google Scholar]
- 32.Chrcanovic BR, et al. Bruxism and dental implant treatment complications: a retrospective comparative study of 98 Bruxer patients and a matched group. Clin Oral Implants Res. 2017;28(7):e1–9. [DOI] [PubMed] [Google Scholar]
- 33.Zhou Y, et al. Does Bruxism contribute to dental implant failure?? A systematic review and Meta-Analysis. Clin Implant Dent Relat Res. 2016;18(2):410–20. [DOI] [PubMed] [Google Scholar]
- 34.Carvalho TS, Lussi A. Age-related morphological, histological and functional changes in teeth. J Oral Rehabil. 2017;44(4):291–8. [DOI] [PubMed] [Google Scholar]
- 35.Wetselaar P, Lobbezoo F. The tooth wear evaluation system: a modular clinical guideline for the diagnosis and management planning of worn dentitions. J Oral Rehabil. 2016;43(1):69–80. [DOI] [PubMed] [Google Scholar]
- 36.Jeong JS, Chang M. Food impaction and Periodontal/Peri-Implant tissue conditions in relation to the embrasure dimensions between Implant-Supported fixed dental prostheses and adjacent teeth: A Cross-Sectional study. J Periodontol. 2015;86(12):1314–20. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are available from the corresponding author, upon reasonable request.