Background
Peri-implant diseases are one of the critical complication in implant treatment. However, few studies reported about the meaning of keratinized tissue around dental implant using multivariable analysis. 

Aim
The aim of this present study was to evaluate the necessity of keratinized tissue around dental implant
 for preventing of peri-implant diseases development in a Japanese population with at least 3 years in function.

Material and Methods
Patients were collected from dental university hospital and 6 general dental office. All patients were evaluated by medical history including gender, age, smoking, systematic disease, frequency of dental visit, plaque control record (PCR), history and presence of periodontitis, number and location of implants. Implants examination were consisting of size, length and retention type. Examination of peri-implant tissue were keratinized tissue width (KTW), pocket depth, bleeding on probing and presence of suppuration. In radiographic examination, bone resorption was calculated by comparison of baseline dental x-rays (1 year after prosthesis setting) and follow-up dental X-rays after at least 3 years. Then, each implant were diagnosed into healthy, peri-implant mucositis (presence of bleeding without bone loss), and peri-implantitis (presence of bleeding or/and suppuration with bone loss > 1mm). These data were analyzed by multilevel multinomial logistic regression analyses.

Results
Totally 531 patients (1564 implants) were evaluated in this study. The prevalence of peri‐implant mucositis and peri‐implantitis at the implant level was 27.0% and 9.0%, respectively. Positive correlation was found between peri-implant mucositis and PCR > 20% (odds ratio [OR] = 9.54). Peri-implantits also correlated with PCR > 20% (OR = 5.99), smoking (OR = 3.97), KTW < 2mm (OR = 2.56), insertion in the maxilla (OR = 1.74). There were also significant interaction effects between KTW and PCR (p < 0.01), KTW and smoking (p = 0.01). Therefore, subgroup analyses were conducted to compare the effect of smoking and PCR in subjects with KTW ≥ 2 mm and those with KTW < 2 mm. The subgroup analyses revealed that smoking and PCR > 20% had a significant association with a high OR (OR = 6.4, and 19.4 respectively) for peri-implantitis development in subjects with KTW < 2 mm. However, in the presence of KTW (≥ 2 mm), smoking was not associated with peri‐implantitis development.

Conclusions and Clinical implications
In this present study, it is thought that keratinized tissue around implants is an important factor in preventing the onset of peri-implantitis. Additionally, keratinized tissue width had significant interaction effects with smoking and PCR. Subgroup analyses revealed that smoking and poor oral hygiene became higher risk indicators for peri-implantitis when the keratinized tissue width was insufficient.

Dose really keratinized tissue around implants prevent peri-implant diseases development?