Peri-implantitis is a plaque-associated pathologic condition characterized by inflammation of the soft tissues surrounding the implant and progressive bone loss (Berglundh et al. 2018). Clinical signs of peri-implantitis are bleeding on probing (BOP) and radiographic bone loss in comparison to previous bone levels. Redness, swelling, suppuration and deep probing depths (PD ≥6 mm) are commonly observed at implants diagnosed with peri-implantitis (Fig. 1).

There is evidence, from experimental and observational clinical studies, for peri-implant biofilms as the main etiologic factor for peri-implantitis (Berglundh et al. 2018). An imbalance between the peri-implant biofilm and the host response to the biofilm results in dysbiosis and tissue destruction. The onset of peri-implantitis may occur early, within the first 3 years after implant placement, and progresses in an accelerating and non-linear pattern. Peri-implantitis is considered to progress at a faster rate than periodontitis (Berglundh et al. 2018). There are numerous factors, patient-related and local factors, which may influence the host response to biofilm accumulation. This article addresses risk factors for the onset and progression of peri-implantitis.

Peri-implantitis is not uncommon with an estimated patient prevalence of 22% (CI:14-30%) as reported in a systematic review (Derks & Tomasi 2015). Peri-implant mucositis, the precursor to peri-implantitis, has an estimated prevalence of 43% (CI: 32-54%), (Derks & Tomasi 2015). While treatment of peri-implantitis is challenging, clinical improvements obtained following anti-infective treatment may be maintained over a 5-year period in the majority of implants in patients performing high levels of plaque control who attend regular supportive care (Heitz-Mayfield et al. 2018). High survival rates of implants in the medium to long term are documented, however recurrence of disease, disease progression and implant loss may also occur (Heitz-Mayfield et al. 2018; Roccuzzo et al. 2018).

The financial costs of peri-implantitis treatment as well as costs for replacement of an implant or implant-supported prosthesis may be substantial. Furthermore, qualitative interview studies have evaluated the views of patients referred for treatment of peri-implantitis (Abrahamsson et al. 2017; Insua et al. 2017). These studies have found that it is not uncommon for patients to have unrealistically high expectations of their implant therapy as a permanent solution to oral problems. When diagnosed with peri-implantitis their reactions were frequently reported as those of worry and concern, and disappointment that they had not received proper education in oral hygiene procedures. Patients reported that referral and diagnosis of peri-implantitis was seen as a “stressful event with consequences on mental wellbeing and daily life” (Abrahamsson et al. 2017).

With the significant impact peri-implantitis has on patient-reported outcomes, efforts to identify risk factors for the onset and progression of peri-implantitis have become the focus of recent research. Identification of risk factors for disease allows a risk profile for an individual patient to be determined, enabling preventive measures to be implemented.

The definition of a true risk factor has been described as “an environmental, behavioral or biologic factor confirmed by temporal sequence, usually in longitudinal studies, which if present directly increases the probability of a disease occurring and if absent or removed reduces that probability" (Genco 1996). To determine a true risk factor of a disease, an interventional study of longitudinal design is required. The majority of studies available addressing peri-implantitis are cross-sectional or retrospective observational studies and therefore in this article the term "risk" refers to a factor which is associated with peri-implantitis, also termed a risk indicator or potential/putative risk factor.

In the 2017 World Workshop on Classification of Periodontal and Peri-implant Diseases it was concluded that there is strong evidence that there is an increased risk of developing peri-implantitis in patients with a history of periodontitis, poor plaque control and lack of regular supportive care after implant therapy. Data identifying smoking and diabetes as potential risk indicators were considered inconclusive. It was also determined that there is limited, emerging evidence linking peri-implantitis with factors such as presence of submucosal luting cement, positioning of implants resulting in limited access for oral hygiene and absence of keratinized peri-implant mucosa (Berglundh et al. 2018).

In the assessment of risk, analyses should include robust multivariate analyses which adjust for the effects of possible confounders. The majority of studies addressing peri-implantitis risk are derived from cross-sectional studies of small convenience samples of patients referred for peri-implantitis treatment. There are, however, two recent large observational studies which have estimated risk in a random sample of patients who received implant therapy (Derks et al. 2016; Kordbacheh Changi et al. 2019). The first is a study from a Swedish population where a large random sample of patients who were provided with implant restorations in the years 2003/2004 were selected using the national data registry of the Swedish Social Insurance Agency. Patient files of 2,765 patients were collected from more than 800 clinicians. Nine years after treatment, 596 patients attended a clinical examination (Derks et al. 2016). Moderate to severe peri-implantitis (bleeding or suppuration on probing and bone loss >2 mm) was found to be a common condition, diagnosed in 14.5% of the patients. A multi-level analysis identified several patient and implant-related factors associated with moderate to severe peri-implantitis (Derks et al. 2016). Periodontal status (presence of periodontitis) at the 9-year examination, the number of implants placed (≥4) and implants placed in the mandible were associated with an increased risk for peri-implantitis. The study also identified prosthetic delivery by general practitioners as compared to a prosthodontist and the prosthesis restoration margin positioned ≤1.5 mm from the crestal bone at baseline as factors associated with peri-implantitis (Derks et al. 2016).

In the second study Kordbacheh Changi et al. (2019) retrospectively analyzed electronic oral health records from an educational institution in the USA. A validated reference cohort of all patients (2,127 patients, and 6,129 implants) receiving dental implants over a 3.5-year period was used. Electronic oral health records of a random 10% subset of patients were examined for an additional period of at least 2.5 years after implant placement to assess the presence of radiographic bone loss, and electronic health record notes were reviewed to confirm a diagnosis of peri-implantitis at implants with progressive bone loss. The prevalence of peri-implantitis was high with 34% on the patient level. A nested case-control analysis of peri-implantitis-affected implants randomly matched with implants from individuals without peri-implantitis identified potential risk factors. Multiple conditional logistic regression identified two iatrogenic factors (1) ill-fitting fixed prosthesis (OR = 5.9; 95% CI:1.6-21.1) and (2) cement-retained prosthesis (OR = 4.5; 2.1-9.5) in addition to an oral co-morbidity of radiographic evidence of periodontitis (OR = 3.6; 1.7 – 7.6) as statistically associated with peri-implantitis. Cigarette smoking was not identified as a risk factor in this study. The authors suggested possible reasons may include that patients under-report smoking habits in health care delivery settings, and that the role of smoking as an explanatory variable may already be captured by the variable “radiographic presence of periodontitis” due to the correlation between smoking and periodontitis (Kordbacheh Changi et al. 2019).

Table 1 summarizes the potential risk factors and examples of references related to evidence of their association with peri-implantitis. The strength of risk (estimated as low, moderate and high) and the strength of evidence (as determined by the number of studies documenting an association) are indicated for the various patient-, implant-site- and prosthesis-related factors. The factors for which there is substantial evidence are indicated in bold colors while factors with limited, emerging or inconclusive evidence are shown in lighter colors.