The effect of keratinized mucosa on the severity of peri-implant mucositis differs between periodontally healthy subjects and the general population: a cross-sectional study

This investigation is a mono-centric observational cross-sectional study conducted in 2019. The study from Grischke et al. [35] served as a basis for planning and sample size calculation. All procedures were approved by the ethics committee of Hannover Medical School (no. 8207_BO_S_2018).

In order to ensure the best practice quality of research, the current study adheres to the STROBE guidelines [36] for cross-sectional studies.

The sample size calculation was performed based on the results of the pilot study [35]. To test our hypothesis that supports a significant relation between insufficient KM and increased severity of peri-implant mucositis in periodontally healthy patients, we chose an alpha error rate of 5%. To reach a power of 80% with an estimated effect size of 0.324, a sample size N = 36 study participants for each group was calculated.

All patients attending maintenance care at the Department of Prosthetic Dentistry and Biomedical Materials Science of Hannover Medical School with at least one root-shaped dental implant with more than 1 year of follow-up after loading were asked to participate in the study between February and May 2019. Inclusion criteria were at least one root-shaped dental implant with at least 1 year of follow-up. The exclusion criteria were patients with a history of antibiotic therapy during the last 3 months and expecting or lactating mothers. No further exclusion criteria were defined. Two study participants could not be included in the study because of recent antibiotic therapy (Fig. 1). All study participants were verbally informed prior to study registration. Their oral and written informed consent was given before study participation. The exclusion criteria of the first study are described in detail by the authors Grischke et al. [35]. Patient recruitment process and clinical assessments were performed by one calibrated examiner following a standardized protocol implying that all examinations were carried out with a maximum of consistency in regard to the case history, the order of examination procedures, probing pressure, and visual assessment.

Preceding the clinical examination, detailed information was collected regarding the implant’s observation time, position, loading time, and previous bone augmentation. Further compromised immune system, alcohol or drug abuse, hyperglycemia, oral and maxillofacial tumor with history of head and neck radiation, smoking status, and allergies especially iodine hypersensitivity were enquired. Additionally, a history of periodontitis as well as age and gender was assessed.

Afterwards, the following clinical measurements were taken:

Probing depth (PD) was measured in millimeters at six implant sites (mesiobuccal, buccal, distobuccal, mesiolingual, lingual, distolingual) using a graduated periodontal probe (GY12, DEPPELER SA, Rolle, Switzerland) with an intended force of 0.2 N.

Bleeding on probing or suppuration (BoP/Sup) occurring within 30 s after measurement of PD was recorded.

The local plaque was evaluated by using the modified plaque index (mPl) [37] while the general hygiene behavior was assessed using a full-mouth plaque score approximal space plaque index (API, approximal space plaque index) expressed in percent [38]. In this process, all approximal spaces showing plaque were divided by the number of approximal spaces in both jaws.

For a deepened evaluation of the condition of peri-implant soft-tissues and conceivably inflammation, the mucositis-index (mGI) [39] was used.

The width of peri-implant KM was measured as the distance between the gingival margin and the muco-gingival junction. To improve visibility of the border between the keratinized and non-keratinized mucosa, iodine solution (Lugolsche Lösung 5%, vitalundfit GmBH, Jüchen, Germany) was applied to the mucosa if necessary [40].

For assessing the presence of periodontitis, the Community Periodontal Index (CPI), introduced by the American Academy of Periodontology and the American Dental association in 1992, was applied [41].

Data were collected following a standardized operating procedure and saved electronically.

KM was categorized into either sufficient or insufficient in order to attain analogous and comparable outcomes to the study conducted by Grischke et al. [35]. The dichotomous division of KM was conducted as described before. Briefly, a threshold value of < 2 mm of KM was considered as insufficient and KM ≥ 2 mm as sufficient [35].

In the statistical analysis, we distinguished three different groups: the general study population (i) including periodontally healthy patients or patients with presence and/or history of periodontitis; the first subgroup (ii) including periodontally healthy patients; and the second subgroup (iii) including periodontally healthy patients without a history of periodontitis.

Complying with the references of the Seventh European Consensus Workshop on Periodontology and the statements of the workgroup 4 of the 2017 World Workshop, the key parameter for peri-implant mucositis assumption is the presence of bleeding on gentle probing with or without concomitant suppuration and deepening of peri-implant pockets compared to previous examinations but necessarily with the absence of bone loss beyond crestal bone level changes resulting from initial bone remodeling. However peri-implantitis is defined by changes in the crestal bone level in conjunction with BoP and pus as a common finding in peri-implantitis sites [6, 42].

The general oral hygiene behavior assessed by the API was either categorized as adequate or inadequate with a cutoff value of 35%. An API ≥ 35% was considered inadequate oral hygiene [38]. Non-smokers were defined as never smokers and smokers who quit more than 5 years ago [43].

The variable mGI-BoP was adopted from the study from Grischke et al. [35] for the assessment of the severity of peri-implant mucositis.

The severity score consists of the mGI and BoP being assessed at six sites per implant each. Every site can get up to four points: a maximum of three points for the mGI plus one point when the site shows bleeding on gentle probing. As the severity score is measured at six sites per implant, the variable consequently reaches from 0 to 24 for each implant.

First, a descriptive analysis of potential confounders was performed on patient level by recording frequencies, the mean and standard deviation. For patient-level analysis, only one implant, the worst implant, per patient was considered. The worst implant was assessed by the highest severity score of all implants of a patient. In the first step, the possible confounders were included in a univariate regression model to assess possible relations with the severity score. This was conducted on the patient and implant level.

First, the general population (i) including all patients was analyzed in a multivariate regression model. Second, subgroup analysis was performed (ii) including only periodontally healthy patients and (iii) including only periodontally healthy patients without a history of periodontitis.

A total of 612 root-shaped dental implants in 130 patients were evaluated and examined for signs of peri-implant mucositis and the width of KM during a 4-month study period. The implants investigated were either from Straumann® Dental Implant System (Straumann GmbH, Freiburg, Germany) or Astra Tech Implant System (Dentsply Sirona Deutschland GmbH, Bensheim, Germany). The implant brand showed no significant effect on peri-implant health. The descriptive analysis is represented in Tables 1 and 2 and revealed that in 348 out of 612 implants, the width of KM was < 2 mm. Considering only the worst implant of every patient, 74 out of 130 participants thus 57% of the patient sample shows a lack of KM in the peri-implant tissue. The dataset comprises 71 female and 59 male participants with a mean age of 70 ± 10.32 years with a range between 38 and 92 years. Seventy-one percent of the study population had a history of periodontitis and 55% had, according to the CPI, at the time of examination periodontitis. A lack of general oral hygiene behavior was found in 63% of the patients, whereas only 8% were current smokers. No subjects with compromised immune system, alcohol or drug abuse, hyperglycemia, oral and maxillofacial tumor, or with history of head and neck radiation were identified.

The implants were in function for an average of 10.15 ±6.31 years with a range between 1 and 31 years. The mean PD on the patient level is 3.81 mm. Seventy-one percent of 612 implants are located in the posterior region including the premolar and molar region. Only eleven patients have a healthy peri-implant mucosa showing no presence of BoP at any site.

The distribution of the scores on the patient level is shown in Fig. 2. A score of 18 was found to be the most frequent severity score among all included implants.

Univariate regression analysis revealed that all possible risk indicators were associated with our primary outcome the severity score, so was the exposition of interest insufficient KM (patient level, 95% CI (9.2–15.2), p ≤ 0.001; implant level, 95% CI (7–9.2), p ≤ 0.001). Consequently, all variables were used in the adjusted sensitivity analysis of the primary objective. There was (i) no statistical significance between insufficient KM and the severity score in the general population on patient and implant level analysis, respectively (patient level, 95% CI (− 0.6–2.7), p = 0.195; implant level, 95% CI (− 0.8–0.9), p = 0.962). Multivariate analysis in the general population (i) revealed significantly higher severity scores with increasing PD (patient level, 95% CI (0.9–1.9), p ≤ 0.001; implant level, 95% CI (1.4–2), p ≤ 0.001) and in patients with no residual teeth (patient level, 95% CI (− 8.5–2.8), p ≤ 0.001; implant level, 95% CI (1.9 to − 1), p ≤ 0.001) or in insufficient oral hygiene (API > 35%; patient level, 95% CI (2.5–6.2), p ≤ 0.001; implant level, 95% CI (− 3.7–3.9), p ≤ 0.001) on patient and implant level. Maxillary implants (95% CI (0.1–3.4), p = 0.04) and the number of implants (95% CI (0.3–0.9), p ≤ 0.001) were associated with significantly higher severity scores on the patient level and male gender (95% CI (− 1.8 to − 0.3), p = 0.008) was a risk indicator for higher severity scores on implant level. All results of the univariate and multivariate analysis including all patients are depicted in Table 3. Subgroup analysis showed a significant association between insufficient KM and the severity score in subjects without (ii) presence or (iii) history of periodontitis on implant level and patient level (implant level, 95% CI (0.3–2.9), p = 0.017; patient level 95% CI (2.1–13.6), p = 0.014). All statistical outcomes of groups ii and iii are depicted in Table 4.