Osteoporosis and implant failure: An exploratory case-control study

Background: Osteoporosis has been suggested as a risk factor for implant fa ilure, but data supporting such a link are limited. Methods: A case-control study was designed to evaluate the association betw een osteoporosis and dental implant failure. Cases (n = 49) and controls (n = 49) were selected from a private practice. The following measures were c ollected for each patient: 1) peripheral dual-energy x-ray absorptiometry ( pDEXA) bone measurements at the distal and proximal radius and ulna; 2) cla ssification of bone quality and quantity at the time of implant placement; and 3) questionnaire data regarding potentially confounding variables. Gene ralized estimating equations were used to relate the likelihood of having a t least one implant failure in an individual to osteoporosis measures. Results: The results suggested that there was no association between pDEXA scores at the radius and ulna and the risk for implant failure. For every 1 -unit increase in bone density as measured by pDEXA t-score at proximal and distal radius and ulna, the risk for a patient to lose at least one implan t changed by +14% and -6%, respectively (odds ratio, 1.14; 95% confidence i nterval, 0.80-1.62 and odds ratio, 0.94; 95% confidence interval, 0.71-1.23 ). The simple visual assessment of local bone quality had a moderately size d relationship to implant failure. Implants placed in sites with thin corti cal bone increased the chance for a patient to lose at least one implant by 130% when compared to implants placed in a thick cortical layer or compact bone (odds ratio, 2.3; 95% confidence interval, 1.0-5.4). Conclusions: This exploratory study suggests that a simple visual assessmen t of bone quality at the site of implant placement may be more informative regarding implant failure than pDEXA bone density measures obtained at peri pheral bones. In order to determine whether peripheral osteoporosis measure ments have a small or moderate association with implant loss, studies with larger sample sizes will be required.