The dental implant has become a favored treatment modality for the restoration of partial and complete edentulism (Messias et al., 2021; Bedrossian and Bedrossian, 2019). High success rates (>90%) and excellent predictability of dental implant treatment have been demonstrated in numerous clinical studies and a multiplicity of indications (Moraschini et al., 2015; Smith Nobrega et al., 2016; Pathak et al., 2023). However, the application of standard diameter implants is limited by factors such as a narrow alveolar ridge, insufficient distance between adjacent natural tooth roots, and limited mesiodistal space in the edentulous area (Buser et al., 2012). As a result, additional treatment like bone augmentation is often required to increase insufficient bone volume. However, these surgical procedures are time-consuming and costly, and require advanced surgical experience.

Therefore, alternative concepts such as narrow-diameter implants (NDIs) are becoming of increasing clinical and scientific interest. Narrow-diameter implants (NDIs) are generally considered to have a diameter ≤3.5 mm and are claimed to be a reasonable alternative to bone augmentation procedures, especially for limited interdental space (Schi et al., 2018; Froum et al., 2007). Moreover, it has been reported that NDIs have competitive treatment outcomes compared to regular-diameter implants (RDIs) in terms of implant survival rate and marginal bone loss (MBL) (Park et al., 2023; Kim et al., 2023). However, compared with RDIs, the smaller implant diameter reduces the bone-to-implant contact area and mechanical strength, which could potentially compromise osseointegration and long-term stability (Ormianer et al., 2012). Therefore, clinicians often monitor the MBL around NDIs via radiography to assess treatment success. This study aimed to further evaluate the clinical performance of NDIs from three different systems by measuring and analyzing radiographic MBL over a 2-year functional loading period.

Bone resorption at the implant margin is a critical factor affecting long-term implant stability. According to the historical success criteria proposed by Albrektsson et al. Albrektsson et al. (1986), MBL should be less than 1.5 mm in the first year of loading, with subsequent annual resorption of less than 0.2 mm. In this study, the mean MBL was 1.49 ± 0.83 mm at the first year and 1.67 ± 0.66 mm at the second year, which approached the upper limit of this historical benchmark. While these values are near the threshold defined by Albrektsson et al. (1986), it is important to note that more recent literature suggests stricter criteria for success. For instance, studies by Galindo-Moreno et al. (2015) and a consensus report by Wang et al. (2025) advocate for MBL ideally not exceeding 0.5 mm–1.0 mm during the first year of function to be considered a truly successful outcome in contemporary implantology. Therefore, while our findings may be considered acceptable by older standards, the observed MBL could be viewed as significant by current, more stringent benchmarks, highlighting an area for ongoing improvement and careful long-term monitoring.

There are obvious differences among implant systems in terms of morphology, material, surface treatment, and connection design (Rahimi et al., 2022). In this study, the Straumann TiZr implants showed the least MBL over the observation period. The Straumann BLT implants used here feature a TiZr alloy (Roxolid^®), known for its high mechanical strength, and a Morse taper internal connection, which is designed to provide a tight seal and minimize micromovement. These features may contribute to enhanced biomechanical stability and a reduced risk of bacterial leakage, potentially leading to less MBL (Bernhard and e t al., 2009). Our findings on the performance of the TiZr alloy implant are consistent with a meta-analysis by Altuna et al. (2016), which also reported low MBL and high survival rates for these implants. In contrast, the Osstem TSIII SA implants, made of Grade IV titanium with an internal conical hex connection, showed the highest MBL. The clinical significance of the statistically different MBL between the implant systems, particularly between Straumann and Osstem, warrants consideration. While a mean difference of approximately 0.6 mm at 2 years may not immediately compromise implant survival, it could be an early indicator of differing long-term susceptibility to peri-implant complications. The superior performance of the TiZr alloy implant with a Morse taper connection in this study may suggest better biomechanical stability and a more effective seal against bacterial leakage, factors known to influence peri-implant bone preservation. Continuous monitoring beyond 2 years is essential to determine if these early patterns in MBL translate to divergent long-term survival and success rates.

With societal aging, the prevalence of osteoporosis increases, particularly among postmenopausal women. The decline in estrogen levels leads to an imbalance in bone metabolism, favoring bone resorption (Cheng et al., 2022). While our study did not directly measure hormonal status or diagnose osteoporosis, the finding of higher MBL in female patients (average age 56 years) aligns with the extensive body of literature suggesting that postmenopausal estrogen deficiency can negatively impact bone metabolism and peri-implant health (Hernández-Vigueras et al., 2016). Thus, the observed gender difference in our cohort may reflect underlying biological factors associated with age and menopause. However, this interpretation should be viewed with caution given the retrospective nature of our study and the lack of hormonal data. Meta-analyses and randomized controlled trials have found that systemic hormone therapy in postmenopausal women can improve bone mineral density, confirming that managing systemic bone health is an effective way to prevent osteoporosis and related fractures (Papadakis et al., 2016; Lindsay et al., 1997; Farquhar et al., 2009). This suggests that for at-risk patient populations, a collaborative approach with a physician regarding systemic bone health may be beneficial for long-term implant success.

NDIs were originally designed for areas with limited mesiodistal space, such as maxillary lateral incisors and mandibular incisors (Schi et al., 2018; Cruz et al., 2021). Their use in the posterior region, particularly for molar replacement, has been debated due to higher occlusal forces. The high survival rate (98.17%) observed in our study is strongly supported by recent literature. For instance, a large systematic review by Valente et al. (2022), analyzing over 1,200 NDIs, reported a comparable cumulative survival rate of 97.80%, confirming the high predictability of this treatment modality in atrophic ridges. In this study, NDIs placed in the posterior (bicuspid) region had a 100% survival rate, though they exhibited greater MBL initially than anterior implants. Other studies support the use of NDIs in the bicuspid region as a reliable alternative to bone augmentation (Papadimitriou et al., 2015). However, biomechanical studies show that stress in the surrounding alveolar bone increases as implant diameter decreases, so caution is still advised for molar replacement (Ormianer et al., 2012). Long-term clinical observation is needed to establish definitive guidelines.

This study has several limitations inherent to its retrospective design. Firstly, the design is susceptible to selection bias, and some patient data, such as smoking status and precise oral hygiene metrics, were inconsistently recorded, precluding robust subgroup analysis of these important confounders. Secondly, a formal sample size calculation was not performed, and the sample size was based on convenience, which may limit the statistical power to detect smaller differences between groups. Thirdly, the 2-year follow-up period provides valuable short-to medium-term data, but longer-term observation is necessary to ascertain the ultimate longevity and MBL patterns of these implants. Finally, the multicenter nature of data collection over a long period may introduce variability in surgical and prosthetic protocols, despite efforts to include cases treated by experienced clinicians. Future prospective, randomized controlled trials with larger sample sizes and standardized protocols are warranted to corroborate these findings.

Within the limitations of this retrospective study, NDIs from all three systems demonstrated a high 2-year survival rate, supporting their clinical utility as a viable treatment option in anatomically restricted sites. The choice of implant system, particularly its material and design, was found to be a significant factor influencing peri-implant MBL, with the TiZr alloy implant system (Straumann^®) exhibiting significantly less bone loss than the Grade IV titanium systems. Furthermore, patient- and site-specific factors, such as female gender and posterior implant location, were associated with greater MBL in the initial healing stages. These findings underscore that NDI treatment outcomes are multifactorial and highlight the importance of careful implant selection and patient evaluation for optimizing long-term success.