Ann Periodontol

Implant Survival in Patients With Type 2 Diabetes: Placement to 36 Months

Harold F. Morris,* Shigeru Ochi,* and Sheldon Winkler
* Dental Implant Clinical Research Group, VA Medical Center, Ann Arbor, MI. Department of Restorative Dentistry, Temple University, School of Dentistry, Philadelphia, PA. University of Otago, School of Dentistry, Dunedin, New Zealand.

Background: Because the life expectancy of individuals continues to increase, dentists providing dental implant treatment can expect to see an increasing number of patients with diabetes mellitus. Today, there are little data available concerning the clinical outcomes involving the use of implant treatment for patients with diabetes mellitus. There are three types of diabetes mellitus: Type 1 (insulin dependent), Type 2 (non-insulin dependent), and gestational. Because of possible complications from patients with diabetes mellitus, they are excluded from participation in most clinical studies of endosseous dental implant survival. Methods: This study attempted to determine if Type 2 diabetes represents a signicant risk factor to the long-term clinical performance of dental implants, using the comprehensive DICRG database. Diabetes was a possible exclusion criterion; however, the nal decision on Type 2 patients was left to the dental implant team at the research center. A total of 2,887 implants (663 patients) were surgically placed, restored, and followed for a period of 36 months. Of these, 2,632 (91%) implants were placed in nondiabetic patients and 255 (8.8%) in Type 2 patients. Failures (survival) were compared using descriptive data. Possible clustering was also studied. Results: A model assuming independence showed that implants in Type 2 patients have signicantly more failures (P = 0.020). However, if correlations among implants within the patient are considered, the signicance level becomes marginal (P = 0.046). The experience of the surgeon did not produce a clinically signicant improvement in implant survival. The use of chlorhexidine rinses following implant placement resulted in a slight improvement (2.5%) in survival in non-Type 2 patients and a greater improvement in Type 2 patients (9.1%); the use of preoperative antibiotics improved survival by 4.5% in non-Type 2 patients and 10.5% in Type 2 patients. The use of HA-coated implants improved survival by 13.2% in Type 2 diabetics. Conclusion: Type 2 diabetic patients tend to have more failures than non-diabetic patients; however, the inuence was marginally signicant. These ndings need to be conrmed by other scientic clinical studies with a larger Type 2 diabetic sample size. Ann Periodontol 2000;5:157-165. KEY WORDS Dental implant failure; dental implants; follow-up studies; diabetes mellitus.
This is U.S. Government-supported research, and there are no restrictions on its use. This paper has been approved for publication by the Dental Implant Clinical Research Group (DICRG), Department of Veterans Affairs. The views expressed in this article are those of the authors and do not reect the views or policies of the U.S. Department of Veterans Affairs, nor imply endorsement by the U.S. Government.

iabetes mellitus is a common endocrine disease that results when the amount of available insulin drops below the value required for normal body functions. Insulin is an anabolic hormone produced in the pancreas. When insufcient amounts of insulin are present, glucose is not efciently transported into cells and adipose tissues, and hyperglycemia occurs. If the fasting glucose blood level reaches or exceeds 126 mg/dl,1 ketoacidosis may occur, along with severe metabolic disturbances. Some patients may exhibit levels of glucose (fasting plasma level <140 mg/dl) that are higher than normal and are considered to have an impaired glucose tolerance. The clinical significance of impaired glucose tolerance is not clear, as most individuals with this condition do not progress to diabetes. Diabetes is prevalent in the United States, with approximately 2% to 4% of the population (nearly 16 million), or
157

Implant Survival in Type 2 Diabetes Mellitus

Volume 5 Number 1 December 2000

one of every 17, having diabetes. Another 3.2% of Americans with a family history of diabetes may experience elevated glucose levels. Diabetics have a 50% chance of having to undergo a surgical procedure at some period in their lifetime.2 CLASSIFICATION There are three type of diabetes: Type 1, Type 2, and gestational. Type 1 diabetes mellitus is characterized by little or no endogenous insulin secretion. Patients with this condition present with acute symptoms, including excessive urination (polyuria), excessive thirst (polydipsia), excessive eating (polyphagia), fatigue, blurred vision, and ketoacidosis. To prevent ketoacidosis and subsequent death, Type 1 diabetics require insulin replacement and become insulin-dependent. Type 2 diabetes mellitus is more common than insulindependent diabetes and usually occurs later in life. Type 2 diabetes is characterized by a gradual decrease in insulin production or the resistance of body cells to the action of insulin. The majority of Type 2 diabetics are overweight. Some Type 2 patients are asymptomatic, with their elevated plasma glucose levels identied during routine laboratory tests. Although insulin levels are decreased, they are not as severely lowered as compared to Type 1 diabetics. Type 2 diabetics do not require insulin replacement and are designated as non-insulin dependent. The etiology of diabetes is unknown, but appears to be a combination of genetics and environmental factors (including viral infections, inadequate diet, and sedentary lifestyle). WOUND HEALING Wound healing is a complex biological event. A group of proteins called growth factors have been identied, which tend to accelerate certain aspects of wound healing. Growth factors have been identied in animal and human saliva. Using a rat model, diabetic animals have been found to produce only about 28% of the platelet-derived growth factor (PDGF) as compared with normal animals.3 Doxey et al.3 demonstrated that metabolic alterations in diabetes-induced animals do not inuence the cytokine levels present, but instead alter the host response by blocking increases in PDGFB and interleukin-1 (IL-1). Research by Meddahi et al.4 studied several heparin-binding growth factors (HBGFs). The data suggested that heparin-like biopolymers may constitute a new family of tissue repair agents for impaired wound healing in diabetic patients. Thaller et al.5 found that an insulin-like growth factor (IGF-1), a somatomedin C, showed promise in accelerating bone repair in streptozotocin-induced diabetic rats. Loder6 believes that the known effects of diabetes on bone and mineral metabolism help explain the prolonged healing rates observed in diabetic patients. Macey et al.7 studied the tensile strength of the femoral
158

bone of normal, insulin treated and untreated streptozotocin-induced diabetic rats 2 weeks after the production of a closed fracture. Untreated, there was a 29% reduction in stiffness compared to the non-diabetic rat. The insulin-treated rats showed an increase in stiffness and tensile values, although these values were not statistically different from the non-diabetic control. ORAL SURGERY FOR DIABETIC PATIENTS Patients with diabetes have a 50% chance of undergoing a surgical procedure at some time during their lives.2 Devlin et al.8 found that the collagenous framework in tooth extraction sites is inhibited in Type 1 diabetics, delaying healing and increasing alveolar bone loss. Iacopino9 reported that hyperglycemia, in conjunction with the elevation of serum low-density lipoproteins and triglycerides, may play a role in the dysregulation of macrophage cytokine production, increased inammatory tissue destruction, and alveolar bone loss. IMPLANTS FOR DIABETIC PATIENTS Shernoff et al.10 reported on 178 implants from three systems placed in 89 Type 2 diabetic patients at 13 Department of Veterans Affairs medical centers. Four failures (2.2%) were found at uncovering. The failure rate increased to 7.3% at the end of 1 year. Initial results suggest that Type 2 diabetic patients can be considered for dental implant therapy. Takeshita et al.11 found 30% less bone-implant contact and 50% less bone thickness with hydroxyapatite (HA)-coated implants in diabetes-induced rats as compared with the non-diabetic group. Nevins et al.12 reported that wound healing and bone-implant contact were signicantly reduced around implants placed in diabetes-induced animals and suggested that osseointegration may be affected by diabetes. The purpose of our study was to examine a subset of data from the Dental Implant Clinical Research Group (DICRG) database to determine the relationship, if any, between Type 2 diabetes and long-term implant survival. The survival data reported cover a period of 3 years after implant placement. MATERIALS AND METHODS In 1991, the DICRG initiated a long-term clinical study in cooperation with the Department of Veterans Affairs to investigate the inuence of implant design, application, and site of placement on clinical success and crestal bone height.13 All patients were screened prior to entry into the study using well-dened criteria. The inclusion criteria were intentionally broad to facilitate the generalization of the results to a large segment of patients seeking implant treatment. The exclusion criteria represented those conditions that represented a signicant

Ann Periodontol

Morris, Ochi, Winkler

risk to either the patient or members of the implant team. They were further divided into: 1) absolute exclusion criteria and 2) possible exclusion criteria. The number of absolute exclusion criteria was kept to a minimum and required that the patient be excluded from the study. Type 1 diabetes (insulin-dependent diabetes mellitus) was considered a significant risk factor and immediately eliminated the patient from further consideration for study participation. The possible exclusion criteria represented those conditions that may or may not introduce a signicant risk factor into the study, and the decision of whether to enter the patient into the study was left to the discretion of the implant team. Type 2 (noninsulin-dependent diabetes mellitus) was considered a possible exclusion criteria. Of the implants placed in 663 patients, 255 (8.8%) were placed in Type 2 diabetics. The implants used were generally representative of the designs and materials available from most implant manufacturers. They included screw, basket, bullet, and ledge designs. The implant materials used were commercially pure titanium, titanium-alloy (Ti-6Al-4V), and hydroxyapatitecoated titanium-alloy. For the purposes of data analysis, the implant cases were placed in five different research strata13 (maxillary and mandibular fully edentulous; maxillary and mandibular partially edentulous; and maxillary anterior single tooth). The implants assigned to each case of the ve strata were determined by randomization. The number of failed implants after 3 years was determined for Type 2 diabetic patients and compared with the failure rate for non-diabetic patients. To provide additional information concerning differences in clinical performance, the treatment procedures were divided into three stages and the stage at which implant failure occurred was recorded. These stages included: 1) placement to uncovering; 2) uncovering to before loading of the prosthesis; and 3) prosthesis loading to 36 months. Specic variables considered for comparison were age, race, gender, experience of the surgeon, use of a chlorhexidine digluconate (0.12%) mouthwash, type of incision, use of preoperative antibiotics, implant coating, and bone density. Bone density was determined through the use of radiographs and tactile sensations observed during preparation of the implant site. Criteria for implant failure included mobility, a periimplant radiolucency, and/or persistent pain, discomfort, or infection attributable to the implant. Implant failures were documented on standardized forms that recorded implant location, type, treatment stage at removal, and reason for removal. All data were forwarded to the DICRG Data Management Center (Ann Arbor, Michigan) for tabulation and analysis. The data were evaluated using descriptive statistics and survival curves. To test the outcomes by taking into

Figure 1.
Implant survival by phase of treatment for all implants placed in Type 2 diabetic and non-diabetic patients. Survival is shown to 36 months.

consideration clustering effects, the data were analyzed with generalized estimating equations (GEE). RESULTS Descriptive Results The results covered the period from implant placement to 36 months. Total survival for non-Type 2 diabetic patients was 93.2% (6.8% failures) with similar survival for the diabetic group, 92.2% (7.8% failures). These results and survival at each phase of implant treatment are shown in Figure 1. Survival for the variable of age, arranged by decades, ranged from 90.7% (40 to 49 year group) to 98.3% (30 year group) for non-diabetic patients (Table 1, Fig. 2). For those patients with Type 2 diabetes, survival ranged from 95.4% (60 to 69 year group) to 83.3% (80 to 89 year group). Patients in the older groups showed lower, but not unexpected, implant survival (Table 1, Fig. 2). The number of patients in these groups (race and gender) was too small for meaningful comparisons (Tables 2 and 3). Only slightly greater survival was recorded for implants placed by the more experienced dentists (Table 4, Fig. 3) in either the non-Type 2 diabetic patient or the Type 2 diabetic patient. There was a large improvement in implant survival in the Type 2 diabetic patients (Table 5, Fig. 4) when chlorhexidine (CHX) (95.6%, 4.4% failures) was used at the time of implant placement, as compared to when CHX was not used (86.5%, 13.5% failures). This difference in survival (9.1%) was large enough to be considered clinically signicant, but was not found in the non-Type 2 diabetic patient. For the non-diabetic group, survival increased only slightly when CHX was used (94.3%, 5.7% failures), compared to when CHX was not used (91.8%, 8.2% failures). It is interesting to note that this improvement in survival persisted at
Spectra System, Core-Vent Corporation, DBA Paragon Company, Encino, CA. Peridex, Procter & Gamble Company, Cincinnati, OH.

159

Implant Survival in Type 2 Diabetes Mellitus

Volume 5 Number 1 December 2000

Table 1.

Type 2 Diabetes by Patient Age (Decades)

No Diabetes Survival Age <30 30-39 40-49 50-59 60-69 70-79 80-89 Total N 59 150 485 407 705 617 24 2,452 % 98.3 97.4 90.7 93.8 93.3 93.2 92.3 93.2 N 1 4 50 27 51 45 2 180 Failure % 1.7 2.6 9.3 6.2 6.7 6.8 7.7 6.8 25 69 83 53 5 235 92.6 89.6 95.4 91.4 83.3 92.2 2 8 4 5 1 20 7.4 10.4 4.6 8.6 16.7 7.8 N Survival % N Diabetes Failure %

Figure 2.
Implant survival by patient age for all implants placed in Type 2 diabetic and nondiabetic patients. Survival is shown to 36 months.

Table 2.

Type 2 Diabetes by Patient Race

No Diabetes Survival Race White Black Hispanic Other Total N 1,986 89 317 60 2,452 % 93.5 90.8 92.4 88.2 93.2 N 137 9 26 8 180 Failure % 6.5 9.2 7.6 11.8 6.8 N 152 19 58 6 235 Survival % 92.1 100 89.2 100 92.2 N 13 0 7 0 20 Diabetes Failure % 7.9 100 10.8 100 7.8

Table 3.

Type 2 Diabetes by Patient Gender

No Diabetes Survival Gender Female Male Total 160 N 154 2,298 2,452 % 93.9 93.1 93.2 N 10 170 180 Failure % 6.1 6.9 6.8 N 7 228 235 Survival % 100 91.9 92.2 N 0 20 20 Diabetes Failure % 100 8.1 7.8

each phase of implant treatment (Table 6) with the use of CHX. The use of either a crestal or remote incision at the time of implant placement did not have a major inuence on survival (Table 7). For those implants placed using a crestal incision in the non-Type 2 diabetic group, survival (93.3%, 6.7% failures) was not clinically different from the implants placed using the remote incision design (94.5%, 5.5% failures). In the Type 2 diabetic group, the remote incision design resulted in slightly lower survival (90.6%, 9.4% failures) than the crestal design (93.6%, 6.4% failures). The advantage to using preoperative antibiotics is shown in Table 8 and Figure 5. Preoperative antibiotic usage tended to provide a favorable improvement in implant survival in both Type 2 diabetics and non-diabetics. For implants in the non-Type 2 diabetic group, survival for those implants placed with preoperative antibiotics was 4.5% higher than implants not provided coverage at placement surgery (Table 9). This improvement in survival was even greater (10.5%) for those in the Type 2 diabetic group. These outcomes are clinically important and should be considered clinically signicant. The advantage to the use of HAcoated implants versus non-HA is evident in Table 10 and Figure 6. In

Ann Periodontol

Morris, Ochi, Winkler

Table 4.

Type 2 Diabetes by Surgeons Experience

No Diabetes Survival Experience Less than 50 implants More than 50 implants Total N 1,086 1,366 % 92.5 93.7 Failure N 88 92 180 % 7.5 6.3 6.8 Diabetes Survival N 194 41 235 % 91.5 95.3 92.2 Failure N 18 2 20 % 8.5 4.7 7.8

2,452 93.2

Figure 3. Table 5.

Type 2 Diabetes by Use of Chlorhexidine Rinses

No Diabetes Survival CHX Use No CHX CHX Total N 1,097 1,373 2,452 % 91.8 94.3 93.2 N 97 83 180 Failure % 8.2 5.7 6.8 N 83 152 235 Survival % 86.5 95.6 92.2 N 13 7 20 Diabetes Failure % 13.5 4.4 7.8

Implant survival by surgeons previous experience for all implants placed in Type 2 diabetic and non-diabetic patients. Survival is shown to 36 months.

Table 6.

Implant Survival (%) in Type 2 Diabetic and Non-Diabetic Patients When Using Chlorhexidine (0.12%) Rinses
Diabetic No CHX Placement-uncovering Before prosthesis 99.0 CHX 96.9 100 Non-Diabetic No CHX 97.0 CHX 97.4

Figure 4.
Implant survival by use of chlorhexidine (0.12%) rinses for all implants placed in Type 2 diabetic and non-diabetic patients. Survival is shown to 36 months.

the non-diabetic patients, survival improved 9.5% just by using an HA-coated implant. In the Type 2 diabetic group, there was a 13.2% (84.7% versus 97.9%) Post-loading 96.9 98.7 98.3 98.5 improvement in survival when HA-coated implants were used. The benefits of HATable 7. coated implants in each phase of implant treatment for the Type 2 Type 2 Diabetes by Type of Incision Used at Time of diabetic group and non-Type 2 Implant Placement group are shown in Table 11. At each phase of treatment, survival of No Diabetes Diabetes the HA-coated implant was slightly higher. These differences ranged Survival Failure Survival Failure from less than 1% to slightly greater Incision Type N % N % N % N % than 10%. These outcomes are clinically important and should also be Crestal 1,622 93.3 117 6.7 160 93.6 11 6.4 considered clinically signicant. Remote 548 94.5 32 5.5 48 90.6 5 9.4 The inuence of bone density on survival to 36 months was mixed in Missing 282 90.1 31 9.9 27 87.1 4 12.9 Type 2 diabetic and non-diabetic Total 2,452 93.2 180 6.8 235 92.2 20 7.8 patients (Table 12, Fig. 7). Overall
90.6 96.4 98.4 161

Implant Survival in Type 2 Diabetes Mellitus

Volume 5 Number 1 December 2000

Table 8.

Type 2 Diabetes by Preoperative Antibiotic Coverage at Placement

No Diabetes Survival Preoperative Antibiotic No Yes Total N 1,028 1,424 2,452 % 90.6 95.1 93.2 Failure N 107 73 180 % 9.4 4.9 6.8 Diabetes Survival N 103 132 235 % 86.6 97.1 92.2 N 16 4 20 Failure % 13.4 2.9 7.8

Figure 5.
Implant survival by use of preoperative antibiotics for all implants placed in Type 2 diabetic and non-diabetic patients. Survival is shown to 36 months.

within hospitals is taken into consideration, a mulImplant Survival (%) at Each Stage of Treatment tiple logistic regression was performed. As shown in Table 13A, patient age, race, and in Type 2 Diabetic and Non-Diabetic Patients gender were not signicant factors. Diabetes was a When Preoperative Antibiotics Were Used signicant factor if the correlation between implants within patients and hospitals was neglected (i.e., Diabetic Non-Diabetic the naive model) at a P = 0.0195 level. With a robust variance estimation method for clustering effects, Pre-Op No Pre-Op Pre-Op No Pre-Op the signicance is just at the P = 0.0498 level. Placement-uncovering 98.5 96.8 98.3 95.8 A chi-square test separately for each stage of implant survival, with clustering taken into considBefore prosthesis 99.3 93.3 98.2 96.6 eration, and considering only a history of diabetes Post-loading 99.3 96.6 98.6 98.1 as a factor, found none of the comparisons of signicance, as shown in Table 13B. However, survival after prosthesis loading determined that diabetes survival for both non-diabetics and Type 2 diabetics was was a possible factor of importance, having a P value less than 0.10. A multiple logistic regression analysis generally similar for bone qualities 2 and 3. Lower sur(Table 13C) with all such factors included reduced down vival was found in bone qualities 1 (95.0% versus 89.3%) to a model with diabetes as a signicant factor (P = and 4 (93.2% versus 77.3%) for implants in the Type 2 0.0074). Thus, diabetes is one of several risk factors for group. implants that had survived to prosthesis loading. Survival Curves The survival curves for patients with Type 2 diabetes DISCUSSION and non-diabetic patients are shown in Figure 8. Diabetes mellitus is a metabolic disease that results Implants in Type 2 diabetic patients failed more frewhen the amount of insulin available drops below the quently and earlier than those without the disease. values required for normal body functions. While no one understands what causes diabetes to occur, it Possible Clustering of Data appears to be a combination of genetics, environmental The comparisons made above were based on descripfactors (including viral infections), poor diet, and tive data which assume that implant failures are independent events and not clustered in individuals or sedentary lifestyle. groups. However, only 20 of the 32 research centers Approximately 2.4% of the total U.S. population are involved in this investigation entered patients with Type diabetic, with another 3.2% having a family history of 2 diabetes. To determine the effects of diabetic hisdiabetes who may exhibit elevated glucose levels. The tory on overall implant survival, i.e., for all failures older age group (over 65) is expected to increase in combined, and the inuence of demographic factors the 21st century. In 1980, 11% of the total U.S. popuwhere the clustering of implants within a patient and lation was over 65; today it is about 12%. By the year Table 9.
162

Ann Periodontol

Morris, Ochi, Winkler

Table 10.

Type 2 Diabetes by Implant Coating (HA-Coated versus Non-HA)

No Diabetes Survival Implant Surface HA-coated Non-HA Total N 1,595 857 2,452 % 96.7 87.2 93.2 N 54 126 180 Failure % 3.3 12.8 6.6 N 141 94 235 Survival % 97.9 84.7 92.2 N 3 17 20 Diabetes Failure % 2.1 15.3 7.8

Figure 6.
Implant survival by use of HA- or nonHAcoated implants for all implants placed in Type 2 diabetic and non-diabetic patients. Survival is shown to 36 months.

Table 11.

Implant Survival (%) in Type 2 Diabetic and Non-Diabetic Patients When Placing HA- or NonHA-Coated Implants
Diabetic HA Placement-uncovering Before prosthesis Post-loading 99.7 100 98.6 Non-HA 95.5 91.9 97.3 Non-Diabetic HA 98.9 99.1 98.8 Non-HA 94.6 94.8 97.8

2030, this number will exceed 21%.14 As implant prosthodontics become more widely accepted as a treatment option for the restoration of missing natural teeth, the group that will benet most is the older age group. Today, a person who is 80 years old can expect to live another 8 years.15 Many of these individuals seeking implant prostheses will have an increased incidence of chronic illness such as cardiovascular diseases, osteoporosis, endocrine abnormalities, liver diseases, collagen disorders, and diabetes mellitus. It is therefore important that the dental implant team understand the risk factors associated with providing dental implant treatment. Wound healing following implant surgery is a highly complex biological event. In view of the limited data from clinical studies, most of the available information about Type 2 diabetics has been obtained from animal studies. It is clear from the available literature that most of

Table 12.

Type 2 Diabetes by Bone Quality

No Diabetes Survival Bone Quality Q-1 Q-2 Q-3 Q-4 Total N 211 1,129 898 207 2,452 % 95.0 93.8 92.0 93.2 93.2 N 11 75 78 15 180 Failure % 5.0 6.2 8.0 6.8 6.8 N 25 108 85 17 235 Survival % 89.3 94.7 93.4 77.3 92.2 N 3 6 6 5 20 Diabetes Failure % 10.7 5.3 6.6 22.7 7.8

Figure 7.
Implant survival by bone quality for all implants placed in Type 2 diabetic and nondiabetic patients. Survival is shown to 36 months.

163

Implant Survival in Type 2 Diabetes Mellitus

Volume 5 Number 1 December 2000

ical centers. To take into consideration any geographic effects, these centers were randomly distributed throughout the U.S. A total of 89 Type 2 patients were identied and screened using well-dened medical and dental criteria. Several different implant manufacturers cooperated in the study, and a total of 178 implants of different designs and materials were placed, restored, and followed. Short-term results reported that 2.2% of the implants failed before or at the time of uncovering and abutment connection. It is important to note that by the end of the rst year, implant failures had dramatically increased to 7.3%. The DICRG clinical study reported here was a much larger, independent scientic clinical study, and entered only Type 2 diabetic patients into the study. The results strongly suggest that Type 2 diabetes is a marginal risk factor for dental implant survival. A greater level Figure 8. of statistical signicance might have been realized had Survival curve to 36 months for implants placed in Type 2 diabetic and the confounding variablesCHX use, preoperative non-diabetic patients. antibiotic use, the use of several different implant designs and materialsnot been present. Dent and colleagues16 rst reported on the signicant benets that can result from the use of preoperwhat is known about the inuence of diabetes on the ative antibiotics. Their comprehensive report docusurvival of endosseous dental implants has been mented the inuence of three different regimens: 1) obtained from well-controlled animal studies. Regardthe use of any dose of antibiotic versus no antibiotic, less of the data obtained from such studies, the nal which showed a significant improvement in implant test must be in well-controlled clinical studies involvsurvival; 2) the use of an American Heart Association ing humans. Diabetes, without question, has been regimen, which provided a more favorable improveimplicated in wound healing complications and periment in survival; and 3) the Petersons regimen that odontal disease; however, little information is availprovided the best improvement in survival. The results able concerning its influence on implant survival. presented here support the ndings of Dent et al., even Shernoff et al.10 published one of the rst controlled, in a Type 2 diabetic population, by improving longscientic clinical studies of patients with Type 2 diaterm survival in Type 2 diabetics by about 10%. betes and implant-supported prostheses. This study, Lambert et al.17 reported that the use of chlorhexlike that of the DICRG study, was a multicenter, mulidine reduced the number of infectious complications tidisciplinary clinical study that involved 13 VA medand possibly improved survival between the Table 13A. time of placement and uncovering. It is interOverall Implant Survival and Demographic Variables esting to note that in our study, the use of Contrast Degrees of Freedom Robust (GEE) Model (Naive) CHX before and after implant placement Wald Chi-Square P Value Wald Chi-Square P Value produced beneficial Overall model 9 758.92 0.0000 1244.85 0.0000 results at all stages of implant treatment for Model minus intercept 8 14.50 0.0696 18.77 0.0161 Type 2 diabetic paDiabetes 1 3.85 0.0498 5.46 0.0195 tients. Morris and Ochi18 Age 2 0.28 0.8681 0.54 0.7640 documented the shortGender 1 0.37 0.5452 1.10 0.2946 term advantages of the use of HA-coated Health status 2 8.47 0.0145 11.03 0.0040 implants to provide Race 2 0.18 0.9123 0.33 0.8470 treatment under the
164

Ann Periodontol

Morris, Ochi, Winkler

Table 13B.

REFERENCES

1. Expert Committee on Diagnosis and Classication of Implant Survival Versus Failure Stage Diabetes Mellitus. Diabetes Care 1998;21:S11. 2. Alberti KG, Thomas DJ. The management of diabetes during surgery. Br J Anaestheol 1979;51:693-710. Chi-Square Degrees of Freedom P Value 3. Doxey DL, Cutler CW, Iacopino AM. Derived growth Stage 1 failures 0.02 1.00 0.88 factor-B and interleukin-1 in a rat model. J Periodontol 1998;69:113-119. Stage 2 failures 0.01 1.00 0.93 4. Meddahi A, Caruelle JP, Gold L, Rosso Y, Barritault D. New concepts in tissue repair: Skin as an examStage 3 failures 0.32 1.00 0.57 ple. Diabetes Metab 1996;22:274-278. 5. Thaller SR, Lee TJ, Armstrong M, Tesluk H, Stern JS. Stage 4 failures 2.72 1.00 0.10 Effect of insulin-like growth factor type-1 (IGF-1) on critical-size defects in diabetic rats. J Craniofac Surg Combined failures 0.21 1.00 0.65 1995;6:218-221. 6. Loder R. The influence of diabetes mellitus on the healing of closed fractures. Clin Orthop 1988;232: 210-216. Table 13C. 7. Macey LR, Kana SM, Jingushi S, Terek RM, BorreStage 4 Failures and Risk Factors tos J, Bolander ME. Defects of early fracture healing in experimental diabetes. J Bone Joint Surg 1989;71:722-733. Degrees of Wald 8. Devlin, Garland H, Sloan P. Healing of tooth extracFreedom Chi-Square P Value tion sockets in experimental diabetes mellitus. J Oral Maxillofac Surg 1996;54:1087-1091. Overall model 10 671.46 0.0000 9. Iacopino A. Diabetic periodontitis: Possible lipidinduced defects in tissue repair through alteration of Model minus intercept 9 60.74 0.0000 macrophage phenotype and function. Oral Dis 1995;1: 214-219. Intercept 10. Shernoff AF, Colwell JA, Bingham, SF. Implants for Diabetes 1 7.18 0.0074 Type-II diabetic patients: Interim report. VA Implants in Diabetes Study Group. Implant Dent 1994;3:183-185. Length 4 20.29 0.0004 11. Takeshita F, Iyama S, Ayukawa Y, Kido MA, Marai K, Suetsugu T. The effects of diabetes on the interface Bone quality 3 11.25 0.0104 between hydroxyapatite implants and bone in rat tibia. J Periodontol 1997;68:180-185. Surgeon experience 1 6.83 0.0089 12. Nevins ML, Karimbux NY, Weber HP, Giannobile WV, Fiorellini JP. Wound healing around endosseous implants in experimental diabetes. Int J Oral Maxillofac Implants 1998;13:620-629. most challenging conditionssmokers, compromised 13. Morris HF, Ochi S. The inuence of implant design, appliASA health status, etc. By simply using an HA-coated cation, and site on clinical performance and crestal bone: implant in Type 2 diabetics, the survival to 36 months A multicenter, multidisciplinary clinical study. Implant improved by 13.2%, which is clinically signicant. These Dent 1992;1:49-55. improvements in survival were evident for each phase 14. Dychtwald K, Flower J. Age Wave: The Challenges and Opportunities of an Aging America. New York: St. Marof implant treatmentplacement to uncovering, uncovtins Press; 1989:56-71. ering and abutment connection, before placement and 15. Casino A. Systemic factors contributing to implant failloading of the prosthesis, as well as following loading ure. Oral Maxillofac Surg Clin North Am 1998;10:177of the prosthesis. 186. 16. Dent CD, Olson JW, Farish SE, et al. Inuence of preCONCLUSIONS operative antibiotics on success of endosseous implants Based on the ndings reported here, we conclude that up to and including stage II surgery: A study of 2,641 implants. J Oral Maxillofac Surg 1997;55:19-24. the use of endosseous dental implants in Type 2 dia17. Lambert PM, Morris HF, Ochi S. Influence of 0.12% betic patients involves a marginal risk to long-term chlorhexidine digluconate rinses on the incidence of implant survival. The survival of implants in Type 2 diainfectious complications and implant success. J Oral betic patients is improved if HA-coated implants are Maxillofac Surg 1997;55:25-30. used and if a chlorhexidine mouthrinse is used at the 18. Morris HF, Ochi S. HA-coated implants: A case for their use. J Oral Maxillofac Surg 1998;56:1303-1311. time of placement. Bone density did not appear to

have a major inuence on implant survival in Type 2 diabetic patients.

Accepted for publication March 30, 2000.

165