QUERIES - jop110299q

[AQ1] Please review the author byline and running heads to ensure that all author surnames are spelled correctly andthat the author order is as meant.

[AQ2] This article and the abstract have been edited for clarity and flow throughout with the intent to retain originalmeaning. This includes text body and headings, all footnotes (affiliation and otherwise), any figures and theirlegends, and any tables and their legends, and any acronyms that have been spelled out per journal style.Please read carefully for completeness and definition, and confirm or amend as needed.

[AQ3] In Key words: Per style, up to 6 key words are allowed. If you would like to add more, please do so. Theymust be MeSH terms.

[AQ4] Please provide the number of males and females in this study and their age ranges.

[AQ5] Per style, the use of the term ‘‘sex’’ is preferred over ‘‘gender.’’ ‘‘gender’’ has been replaced throughout.

[AQ6] For the two graduate prosthodontic residents: Please supply the first and last names of both.

[AQ7] Please verify that figure 2 panels have been named accordingly to match legend descriptions.

[AQ8] In Table 1: Heading correct for ‘‘Parameter Estimate (SE) – 1.96 (95% CI)’’?

[AQ9] Is the column heading ‘‘Percentage’’ correct here in Table 1? ‘‘Range’’ does not seem correct here.

[AQ10] In Table 1, mm correct as unit of measure for row ‘‘Mean distance to distal adjacent structure’’?

[AQ11] Correct that M+D stands for mesial and distal in Tables 1 and 2?

The Effect of Implant Size 5 · 8 mm onCrestal Bone Levels Around Single-Tooth ImplantsRainier A. Urdaneta,* Joseph Leary,† William Lubelski,‡ Kimberly M. Emanuel,*and Sung-Kiang Chuang½AQ1� §

Background: There½AQ2� is substantial evidence supporting thestrain-induced biologic response of bone to mechanicalload. Stress-related factors, such as implant size, have beenassociated with changes in crestal bone levels on dental im-plants. The purpose of this study is to evaluate the effect ofan implant size, specifically 5 mm wide · 8 mm long (5 · 8),on peri-implant bone levels.

Methods: A retrospective cohort study was conducted. Thecohort was composed of patients who had ‡1 plateau root–form implant. Descriptive statistics, univariate and multivari-ate mixed-effects regression models, adjusted for multipleimplants in the same patient were used to evaluate the corre-lation between 5 · 8 implants and other clinically relevant fac-tors on crestal bone levels after insertion of single-toothreplacements.

Results: The cohort was composed of 81 individuals whoreceived 326 implants. The average change in crestal bonelevels (AvBL) for 5 · 8 maxillary and mandibular implantsafter 5.9 years of follow-up were -0.36 and -0.04 mm, re-spectively. In contrast, AvBL for mandibular implants notmeasuring 5 · 8 was -0.51 mm. Of several different local andsystemic factors evaluated, including 10 different implantsizes, 5 · 8 mandibular implants were found to be signifi-cantly less likely to lose bone when compared with mandi-bular implants not measuring 5 · 8 (P = 0.047).

Conclusions: Implants measuring 5 · 8 demonstrated sta-tistically significantly less peri-implant bone loss in the poste-rior mandible. This finding indicated that a specific implantsize may be involved in mandibular bone preservation. How-ever, prospective multicenter clinical trials are needed to val-idate these findings. J Periodontol 2012;83: nnn-nnn.

KEY WORDS

Bone regeneration; dental implants; risk factors½AQ3� .

Minimal or no crestal bone resorp-tion is considered to be an indi-cator of the long-term success of

implant restorations.1,2 A recent meta-analysis3 reported that marginal bonelevels varied significantly (from -0.24up to -0.75 mm) between differentimplant systems after 5 years. Someimplant systems demonstrate no crestalbone loss and/or crestal bone gain afterinsertion of definitive restorations.4-7

Crestal bone gain has been docu-mented around implants with a chemi-cally modified surfacei in both animal8

and human5 models. A 6-year prospec-tive study reported that 43.8% of splintedMorse-taper implants¶ experienced somebone gain.6 Yoo et al.7 reported no boneloss or crestal bone gain in 149 immedi-ately loaded plateau root–form (PRF)#

implants (32.2%) with five implants gain-ing >2 mm of bone. Urdaneta et al.9 eval-uated the effect of >90 different local andsystemic factors, including 10 differentimplant sizes, on crestal bone levels(CBs) around single-tooth PRF implantsand identified five predictors for crestalbone gain after crown insertion. Implantsize of 5 mm wide · 8 mm long (5 · 8)was one of those predictors.

Studies on machine-surfaced implantsreported increased failure rates for wide-diameter comparedwithsmaller-diameter

* Private practice, Boston, MA.† Private practice, Norwood, MA.‡ SoftArtisans, Boston, MA.§ Department of Oral and Maxillofacial Surgery, Massachusetts General Hospital and

Harvard School of Dental Medicine, Boston, MA.

doi: 10.1902/jop.2012.110299

i SLActive, Straumann, Basel, Switzerland.¶ Straumann Dental Implant System, Straumann.# Bicon Dental Implants, Boston, MA.

J Periodontol • October 2012

1

implants.10 In contrast, a recent study on rough-surfaced implants reported higher failure rates forimplants with standard diameters (3.8 mm) thanthose with wider diameters (4.3 to 6 mm wide) andfound that the peri-implant marginal bone resorp-tion was similar around the implants with differentdiameters.11

The purpose of this study is to evaluate the effect ofimplant size, specifically 5 · 8, in peri-implant bonelevels surrounding PRF implants after the insertionof single-tooth replacements. It was hypothesized that5 · 8 implants would be significantly less likely to losebone than non-5 · 8 implants in a multivariate analy-sis with other variables correlated to both 5 · 8 im-plants and changes in CBs.

The specific aim of this study is to evaluate any pos-sible correlation between 5 · 8 implants, several clin-ically relevant variables, specifically stress-relatedvariables, and changes in peri-implant bone levels.

MATERIALS AND METHODS

Study Design and SampleThe present study is designed as a retrospective cohortstudy. The cohort was derived½AQ4� from the population ofpatients who had ‡1 single-tooth PRF implant at the Im-plant Dentistry Center, Boston, Massachusetts, fromJuly 2001 to August 2003. Patients were selected ifthey satisfied the following inclusion criteria: 1) re-stored with ‡1 cementless restoration**; and 2) con-sented to participate after being fully informed of theconditions of the study. Exclusion criteria included in-adequate or unavailable patient charts and/or patientsunwilling or unable to attend the follow-up examination.This study involved the examination of patient records,as well as clinical evaluations during recall appoint-ments from 2004 to 2007. Periapical and panoramicradiographs and clinical photographs were obtained.The characteristics of this cohort have been publishedpreviously9,12,13 and thus will be described only brieflyhere. This study was approved by the Faulkner Hospi-tal Institutional Review Board, Boston, Massachusetts,and the study was conducted in accordance with theHelsinki Declaration of 1975, as revised in 2000.

Study VariablesThe primary predictor variable was the 5 · 8 implant.Variables correlated with changes in peri-implantbone levels in a previous investigation,9 as well as10 implant sizes were grouped into the following cat-egories: 1) biologically relevant variables (sex½AQ5� andage); 2) stress-related variables; and 3) other clini-cally relevant variables. The guidelines for these vari-ables have been described previously9,12,13

Stress-Related VariablesStress is defined as a force divided by the functionalarea over which it is applied. Consequently, stress-

related variables have been divided into two groups:force and area factors. Force factors are dental factorsthat increase or decrease the stress exerted on dentalimplants, and area factors are variables known to in-crease, decrease, or measure the surface area ondental implants.14

Other Clinically Relevant VariablesThe following variables were included: 1) medica-tions; 2) soft tissue parameters; 3) tobacco use; 4)adjacent structures; 5) immediate extraction; and6) presence of interproximal contacts.

Peri-Implant Bone LevelsLinear measurements from the implant–abutment in-terface (IAI) to the highest CB were obtained mesiallyand distally.13 These measurements could be eithera positive number (if the CB was coronal to IAI), zero(when CB was located at IAI), or a negative number (ifCB was located apical to IAI). An average mesiodistalCB (AvCB) was obtained for each implant. Thechanges in peri-implant bone levels were measuredmesially and distally by comparing the AvCB on aperiapical radiograph obtained on the day of theinsertion of the definitive restoration to the AvCBobserved in the most recent radiograph available. Theaverage change in mesiodistal peri-implant bonelevels (AvBL) was obtained for each implant restora-tion. A negative AvBL implied peri-implant bone loss.A positive AvBL suggested an increase in CB over time.

Calibration of Radiographic MeasurementsThe methodology used to calibrate the radiographicmeasurements and reviewers has been publishedpreviously.7,9,13 Periapical radiographs were ob-tained with the use of the long-cone technique anda film-holding system.†† Using a digital extraoral im-aging system,‡‡ digital periapical radiographs werestored. Stored images were displayed on a monitor,§§

and direct measurements were performed in millime-ters by two different examiners13 (two graduate pros-thodontic residents ½AQ6�of the University of MinnesotaSchool of Dentistry, Minneapolis, Minnesota). The ac-tual implant length, available from the manufacturer,was used to obtain a margin of error on each radio-graph evaluated, and measurements were adjustedfor calibration error.13

Statistical AnalysesA database was created using a software packageii

with appropriate checks to identify errors. Descrip-tive statistics were computed for all of the study variab-les. The outcome variable was changes in peri-implant

** Integrated Abutment Crown, Bicon Dental Implants.†† Rinn XCP, Dentsply-Rinn, Elgin, IL.‡‡ Digora PCT, Soredex, Milwaukee, WI.§§ UltraSharp E173FPB, Dell Computer Company, Round Rock, TX.ii Excel 2007, Microsoft, Redmond, WA.

Implant Size and Bone Preservation Volume 83 • Number 10

2

Table 1.

Variables Associated With Changes in Peri-Implant Bone Levels in the UnivariateStatistical Analyses½AQ8�

Variable Number/Mean Percentage

Parameter Estimate (SE)

– 1.96 (95% CI)

Robust

P value

Biologically relevant variablesSex -0.03 (-0.28, 0.22) 0.8

Female 45 55.6Male 36 44.4

Age at implant placement (years) 58.7 Range:½AQ9� 27.8 to 91.8 -0.01 (-0.02, 0.06) 0.4

Stress-related factorsFactors that may affect the amount

of forceAngle Class I 190 59.6 0.27 (-0.12, 0.36) 0.03Angle Class III 53 16.6 -0.45 (-0.76, -0.14) 0.006Opposing structure, tooth 231 72.2 0.4 (0.1, 0.6) 0.001Opposing structure, implant 75 23.4 -0.3 (-0.6, -0.09) 0.007Opposing material, ceramic 114 35.6 -0.17 (-0.38, 0.04) 0.11Opposing material, composite resin 53 16.6 -0.4 (-0.6, -0.08) 0.02Crown cemented on a prefabricated

abutment with a spherical base98 25.2 0.21 (0.02, 0.4) 0.03

Mean distance to distal adjacentstructure (mm)½AQ10�

2.2 Range: 0.2 to 12 0.05 (-0.01, 0.11) 0.12

Factors that may affect/measure thefunctional area

Implant coating, TPS 78 24.1 -0.34 (-0.56, -0.12) 0.003Implant coating, HA 226 69.8 0.28 (0.07, 0.49) 0.009Implant size (in mm; width · length)

6 · 6 13 4 -0.33 (-0.81,0.15) 0.183.5 · 8 3 0.9 *4 · 8 13 4 -0.39 (-0.86, 0.08) 0.104.5 · 8 59 18.3 -0.07 (-0.33, 0.19) 0.605 · 8 97 30 0.21 (0.01, 0.41) 0.046 · 8 4 1.2 *3.5 · 11 40 12.4 0.01 (-0.29, 0.31) 0.944 · 11 47 14.6 0.02 (-0.26, 0.30) 0.884.5 · 11 20 6.2 0.19 (-0.22, 0.60) 0.375 · 11 26 8.1 -0.45 (-0.83, -0.07) 0.02

Mean implant stability¶¶ 0.06 Range: -8 to 9 -0.05 (-0.09, -0.01) 0.01

Other clinically relevant variablesDM 3 3.7 -0.83 (-1.29, -0.37) 0.0008Non-steroidal anti-inflammatory

medications13 16.1 0.39 (0.10, 0.68) 0.01

Diuretics 2 2.5 0.58 (-0.14, 1.3) 0.12Multivitamins 3 3.7 0.74 (-0.07, 1.55) 0.08Tobacco use (n = 81) 8 9.9 -0.09 (-0.46, 0.27) 0.60Mean sulcular bleeding 0.44 Range: 0 to 3 -0.19 (-0.35, 0.03) 0.01Mean GI 0.71 Range: 0 to 3 -0.21 (-0.37, -0.05) 0.006Mean probings M+D 3.19 Range: 1.5 to 6 -0.19 (-0.33, -0.05) 0.01Adjacent to two implants 63 19.6 --0.33 (-0.58, -0.08) 0.009Changes in adjacent structures 44 16.5 -0.27 (-0.5, -0.02) 0.04Immediate extraction 87 27 0.2 (-0.02, 0.42) 0.07Interproximal contact on mesial 206 68.7 0.18 (-0.03, 0.39) 0.10

95% CI = 95% confidence interval; TPS = titanium plasma spray; HA = hydroxyapatite; DM = diabetes mellitus; GI = gingival index; M+D =½AQ11� mesial and distal.* Statistical analysis not performed because of the small number of observations.

¶¶ PerioTest, Medizintechnik Gulden.

J Periodontol • October 2012 Urdaneta, Leary, Lubelski, Emanuel, Chuang

3

bone levels (AvBL), and the main predictor variable wasimplant size: 5 · 8 implants versus implants not mea-suring 5 · 8. The first univariate analysis presentsthe correlation between the outcome variable and allstudy variables (Table 1). The second univariate analysiswas developed to identify possible correlations betweenthe main predictor variable and the study variables (Ta-ble 2). Variables correlated to both 5 · 8 implants andAvBL (P values £0.15 in both univariate analyses) andbiologically relevant variables, such as age and sex, wereentered into a generalized linear mixed-effects regres-sion model or generalized estimating equations regres-sion model that adjusted for clustering observationswithin the same patient15 for dichotomous outcome ina statistical software package.*** In the multivariatemodel (Table 3), the P values £0.05 were considered sta-tistically significant. k statistics were used to compute in-ter-examiner and intra-examiner reliability.

RESULTS

Eighty-one of 326 PRF implants (24.9%) demon-strated varying degrees of apparent peri-implantbone gain overtime. The 5 · 8 implant represented42% (34 of 81) of all implants that gained bone(Figs. 1 through 3) and 30% (97 of 326) of all implantsin this cohort. The AvBL of 5 · 8 maxillary im-plants was -0.36 mm, whereas 5 · 8 mandibular im-plants demonstrated lower average crestal boneloss (-0.04 mm). The AvBL of non-5 · 8 mandibularimplants was -0.51 mm. Non-5 · 8 mandibularimplants included the following: 3.5 · 8, 4 · 8,4.5 · 8, 3.5 · 11, 4 · 11, 4.5 · 11, 5 · 11, 6 ·6, and 6 · 8. A 5 · 8 implant was significantly less

Table 2.

Univariate Models on the Associations Among 5 · 8 Implants and Predictor Variables

Variable Parameter Estimate 95% CI P Values

Biologically relevant variablesSex 0.001 (-0.58, 0.58) 0.99Age at implant placement 0.02 (-0.02, 0.06) 0.27

Stress-related factorsFactors that may affect the amount of force

Angle Class I 0.32 (-0.30, 0.93) 0.32Angle Class III -0.36 (-1.24, 0.51) 0.42Opposing structure, tooth -0.18 (-0.92, 0.56) 0.63Opposing structure, implant 0.004 (-0.74, 0.75) 0.99Opposing material, porcelain 0.26 (-0.39, 0.91) 0.43Opposing material, composite -0.22 (-1.15, 0.71) 0.65Crown cemented on a prefabricated

abutment with spherical base-0.32 (-1.05, 0.40) 0.38

Mean distance to distal adjacent structure 0.07 (-0.06, 0.21) 0.28Factors that may affect the surface area

Implant coating, TPS -0.69 (-1.40, 0.03) 0.06*Implant coating, HA 0.58 (-0.14, 1.30) 0.11*Mean implant stability -0.22 (-0.34, -0.09) 0.0008*

Other clinically relevant variablesDM 0.79 (0.38, 1.21) 0.0002*Non-steroidal anti-inflammatory medications -0.09 (-0.83, 0.65) 0.81Diuretics -0.26 (-1.83, 1.32) 0.75Multivitamins 0.74 (-0.71, 2.2) 0.32Tobacco use 0.03 (-0.70, 0.77) 0.93Mean sulcular bleeding -0.11 (-0.50, 0.27) 0.57Mean GI -0.31 (-0.72, 0.10) 0.14Mean probings M+D 0.27 (-0.12, 0.66) 0.17Two implants -0.25 (-0.88, 0.37) 0.43Changes in adjacent structures -0.04 (-0.77, 0.70) 0.92Immediate extraction 0.08 (-0.52, 0.68) 0.79Interproximal contact on mesial 0.26 (-0.20, 0.71) 0.27

95% CI = 95% confidence interval; TPS = titanium plasma spray; HA = hydroxyapatite; DM = diabetes mellitus; GI = gingival index; M+D = mesial and distal.* P £ 0.15.

*** PROC Mixed for continuous outcome (linear mixed-effects models)and PROC Genmod (generalized estimating equations models withbinomial link function) functions, SAS 2002-2010 v.9.3, SAS, Cary,NC.

Implant Size and Bone Preservation Volume 83 • Number 10

4

likely to lose bone if it was located in the mandible(P = 0.02).

Thirty-eight of 39 mandibular 5 · 8 implants wereplaced in posterior areas. Of mandibular posterior5 · 8 implants, 44.7% gained bone compared with29.3% of maxillary 5 · 8 implants that gained bone.

5 · 8 and Stress-Related Variables5 · 8 mandibular implants with a hydroxyapatite(HA)††† coating demonstrated lower average crestalbone loss (+0.03 mm, indicating bone gain) whencompared with titanium plasma spray (TPS)-coated5 · 8 mandibular implants (-0.22 mm). In contrast,5 · 8 maxillary TPS-coated implants demonstratedlower average crestal bone loss (-0.06 mm) than sim-ilar implants with an HA coating (-0.43 mm). How-ever, an HA coating in the mandible (P = 0.58) anda TPS coating in the maxilla (P = 0.09) were not signif-icantly correlated to AvBL on 5 · 8 implants whenevaluated statistically.

The average crown-to-implant ratio (C/IR) of 5 · 8mandibular implant restorations was 1.7. AvBLof -0.06 and +0.02 mm were observed in 5 · 8 implantrestorations with C/IR < 1.7 and ‡ 1.7, respectively.

The average bucco-lingual crown width (BLCW) of5 · 8 mandibular implant restorations was 9.25 mm.AvBL of -0.18 and +0.02 mm were observed in 5 · 8

implant restorations with BLCW < 9.25 and ‡ 9.25mm, respectively.

Despite positive AvBL, a 5 · 8 mandibular implantwas not significantly more likely to gain bone if it hadlarger C/IR (P = 0.16) or increased BLCW (P = 0.11)when evaluated statistically.

With regard to opposing structures, AvBL of +0.07and -0.41 mm were observed in 5 · 8 mandibularimplant restorations opposing teeth and implants,respectively. A 5 · 8 mandibular implant was signifi-cantly more likely to gain bone if it were opposing nat-ural teeth (P = 0.05; Fig. 1) whereas a 5 · 8 mandibularimplant was not significantly more likely to lose boneif it were opposing implants (P = 0.14).

5 · 8, Covariates, and Peri-Implant Bone LevelsThe correlation between the study variables andchanges in peri-implant bone levels is presented inTable 1. A second univariate analysis was performedto evaluate the possible correlation between 5 · 8implants and other study variables (Table 2). The sta-tistical analysis revealed five variables significantlycorrelated to 5 · 8 implants (P £0.15): 1) diabetesmellitus (DM) (P = 0.0002); 2) HA implant coating(P = 0.11); 3) TPS implant coating (P = 0.06); 4)mean implant stability‡‡‡ (P = 0.0008); and 5) meangingival index (GI) (P = 0.14). Finally, to assess the ef-fect of 5 · 8 implant size on changes in peri-implantbone levels, variables correlated to AvBL identifiedin the first univariate analysis shown in Table 1 thatwere also found to have a significant association withimplant size 5 · 8 in the second univariate analysisshown in Table 2 (P £0.15 in both univariate analyses)were entered into a multivariate model together withthe biologically relevant variables of age and sex. Inthe multivariate model (Table 3), 5 · 8 implantsplaced in the mandible were significantly less likelyto lose bone compared to non-5 · 8 implants (P =0.047 and positive parameter estimate).

DISCUSSION

The purpose of the present investigation is to evaluatethe effect of 5 · 8 implant size on peri-implant bonelevels around single-tooth PRF implants after crowninsertion. The AvBL around 5 · 8 mandibular im-plants was -0.04 mm after an average of 5.9 yearsof loading. In contrast, non-5 · 8 mandibular implantsdemonstrated significantly higher peri-implant boneloss (-0.51 mm). When variables correlated to bothchanges in peri-implant bone levels and 5 · 8 implantsize were entered in a multivariate analysis, 5 · 8 im-plants were found to be significantly less likely to losebone when compared with implants not measuring5 · 8 in the mandible (Figs. 1 through 3).

Table 3.

Multivariate Model to Determine the Effectof 5 · 8 Implant Size and VariablesCorrelated to 5 · 8 Implant Size in Peri-Implant Bone Levels

Variable

Parameter

Estimate

95%

CI

Robust

P value

Biologically relevantfactors

Age at implantplacement

-0.003 (-0.02, 0.02) 0.63

Sex 0.05 (-0.2, 0.3) 0.69

Factors that affect/measure surface area

Implant coating, HA -0.04 (-0.61, 0.53) 0.88Implant coating, TPS -0.30 (-0.88, 0.28) 0.31Implant size, 5 · 8 in

the mandible0.31 (0.01, 0.62) 0.047*

Mean implantstability##

-0.03 (-0.07, 0.02) 0.22

Other factorsDM -0.80 (-1.30, -0.30) 0.003*Mean GI -0.13 (-0.29, 0.03) 0.12

95% CI = 95% confidence interval; HA = hydroxyapatite; TPS = titaniumplasma spray; DM = diabetes mellitus; GI = gingival index.* Statistically significant at P £0.05.

## PerioTest, Medizintechnik Gulden.††† Integra-CP, Bicon Dental Implants.‡‡‡ PerioTest, Medizintechnik Gulden, Modautal, Germany.

J Periodontol • October 2012 Urdaneta, Leary, Lubelski, Emanuel, Chuang

5

Figure 1.Periapical radiographs of two hydroxyapatite-coated PRF implants in the mandibular left posterior area restored with integrated abutment crowns occludingwith natural teeth. Peri-implant bone gain was documented in the 5 · 8 implant restoring the second molar, whereas crestal bone loss was documentedadjacent to the 4.5 · 8 mm implant restoring the first molar. A and C) Radiographs obtained the day of crown insertion, August 22, 2002. B and D)Radiographs obtained the day of last recall, September 15, 2008. E) Bitewing radiograph obtained on September 15, 2008.

Implant Size and Bone Preservation Volume 83 • Number 10

6

The multivariate analysis also identified DM asa significant risk factor for crestal bone loss; however,because the statistical analysis was not designed toevaluate variables correlated to both DM and AvBL,the association between DM and bone loss cannotbe considered conclusive and will not be discussedhere.

5 · 8 implants are shorter and wider than conven-tional implants. In the case of wider implants, shorterlengths have been correlated with significantly lesscrestal bone resorption. In a retrospective evaluation

of 304 wide-diameter implants(5 to 6.5 mm wide), Degidiet al.2 reported that narrowerwidths of 5.0 to 5.5 mm andshorter lengths of <13 mm werecorrelated with a significantlylower rate of crestal bone re-sorption. A randomized clinicaltrial that evaluated implantsplaced in posterior atrophicjaws reported significantly lowerperi-implant bone loss after 1year of loading around 60 short/wide implants (5 mm long · 6mm wide) when compared to68 long/thinner implants (10to 13 mm long · 4 mm wide)placed in augmented bone.16

Neither implant length nordiameter was significantly cor-related with changes in CBs af-ter crown insertion. However,the combination of 5-mm widthand 8-mm length proved signif-icant. This is consistent with thefindings of finite element stressstudies. Petrie and Williams17

proposed that implant diameterand length should be consid-ered together because of theirinteractive effects on crestalbone strain. Finite element ana-lyses have consistently shownthat diameter is more importantthan length for improved stressdistribution around dental im-plants. In a simulated model ofthe posterior mandible, Himm-lova et al.18 reported a 47.9%stress reduction when a 5-mm-wide implant was used insteadof a 3.6-mm-wide implant ofsimilar length (12 mm). In con-trast, only a 7.3% stress reduc-tion was observed around a

17-mm-long implant compared with an 8-mm-longimplant of similar width (3.6 mm). Ding et al.19 eval-uated the combined effect of implant length anddiameter on stress and concluded that increasing bothlength and diameter decreased the stress and strain inthe alveolar crest, but diameter had a more significanteffect than length in relieving stress concentration.

The case supporting the strain-induced biologic re-sponse of bone to mechanical load is compelling.Wolff20 stated that bone changes its external shapeand internal architecture in response to functional

Figure 2.Periapical radiographs of two 5 · 8 hydroxyapatite-coated PRF implants in the mandibular left posteriorarea restored with integrated abutment crowns. Notice the peri-implant bone level changes morethan 4 years after crown insertion. A and B) Radiograph obtained the day of crown insertion, January 21,2003. C and D) Radiograph obtained the day of last recall, October 2, 2007.½AQ7�

J Periodontol • October 2012 Urdaneta, Leary, Lubelski, Emanuel, Chuang

7

pressures.Frost21presented theconceptofaminimumeffective strain (MES) for bone modeling in the rangeof 1,500 to 2,500 microstrains and reported thatstrains above the MES led cortical bone to change itsmass and architecture in ways that keep strains withinor below this threshold. Thus, an increase in mechan-ical usage tends to increase bone deposits, and the sizeof the largest mechanical loads on bone appear moreimportant than their frequency, i.e., a marathon run-ner’s bones are not massive, but a weightlifter’s bonesare.21 Stress in the magnitude of 2.48 · 10 N/mm2 hasbeen shown to cause an increase in bone growth.22

Dental implant research has often focused on waysto reduce stress to prevent progressive bone loss. Thisrationale is supported by previous studies on externalhex-machine-surfaced§§§ (EHMS) implants, splintedimplants, and by a relatively small number of animalstudies. Marginal bone loss has been associated withhigh occlusal stress in several clinical studies usingsplinted EHMS implants.23-27 However, evaluatingthe effect of stress in peri-implant bone on splintedsmooth-surfaced implants has several limitations.First, finite element studies have consistently shownthat splinting implants reduces peri-implant bonestress by distributing the forces between the im-plants.28,29 Second, a significantly higher rate of peri-implant bone loss has been documented aroundEHMS implants when compared with two other com-monly used rough-surfaced implant designs.3 There-fore, the results of these studies should not beextrapolated to current implant designs or single-tooth implants.

Furthermore, clinical studiesusing implant designs withrough-surface texture treat-ments have shown that in-creased stress, as measured byincreased C/IR, is not associatedwith peri-implant bone loss. Ur-daneta et al.13 reported thatC/IR of £4.95 was associatedwith a significant increase infracture and loosening of im-plant abutments but had nosignificant effect on CBs on sin-gle-tooth PRF implants. A studyon Morse-taperiii implants6 (166of 192 of them splinted) reportedthat implants with higher ana-tomic C/IR demonstrated sig-nificantly lower annual crestalbone loss. A study on sinteredporous-surfaced¶¶¶ implants30

reported a decrease in boneloss around shorter (<7 mm)and unsplinted implants com-

pared with longer and/or splinted implants. The au-thors suggested that the bone loss observed aroundlonger/splinted implants could have been caused bystress-shielding-induced bone atrophy. These studiessuggest that a certain amount of stress is necessaryfor crestal bone maintenance.

It is possible that the magnitude of stress producedby masticatory forces in the posterior mandible is dis-tributed around a 5 · 8 PRF implant such that more ofthe bone that is in contact with the implant is loaded at,or higher than, the MES, leading to positive bone re-modeling, observed as gains in both density andcrestal bone height. This theory is supported by thefact that stress multipliers, such as increased C/IRand crown width, were clinically correlated with thepositive changes in bone levels observed around 5 ·8 implant restorations. The fact that 5 · 8 implant res-torations opposing natural teeth were significantlymore likely to gain bone seems to support this theoryas well. Hekimoglu et al.31 reported significantly in-creased strains under higher loads (100 N static load-ing and 75 and 100 N dynamic loading) around animplant opposing a natural tooth compared with twooccluding teeth or two occluding implants.

The fact that mandibular 5 · 8 implants outper-formed similar implants in the maxilla suggests thatthere is an optimum strain environment for bone in dif-ferent areas of the mouth that can be maintained witha specific implant size. This is consistent with the

Figure 3.Periapical radiograph of a 5 · 8 PRF implant restoring a mandibular left first molar. Notice crestal bonemineralization after crown insertion. A) The area below the fins or plateaus (in red) represents its effectivelength, the surface area that actively participates in distributing compressive forces to bone (blue arrow).Radiograph was obtained at crown insertion, May 20, 2002. B) Radiograph obtained at last recallappointment, May 18, 2010.

§§§ Branemark System, Nobel Biocare, Goteborg, Sweden.iii Straumann Dental Implant System, Straumann.¶¶¶ Endopore dental implants, Sybron Implant Solutions, Orange, CA.

Implant Size and Bone Preservation Volume 83 • Number 10

8

results of Clinton and Lanyon,32 who demonstratedthat an optimum strain environment exists for eachspecific anatomic area, and the peak strains innatefor that area should be maintained to optimize the re-sponse of the bone.

The results of the present study indicate that thesurface of a 5 · 8 PRF implant is correlated with bonepreservation in the posterior mandible. However, onlythe functional surface area,33 that is, the area that ac-tively serves to dissipate compressive and tensilenon-shear loads through the implant-to-bone inter-face, should be considered. Of the total surface areaof the 5 · 8 PRF implants (Fig. 3), only the area onthe underside of each of the 10 fins or plateaus/baseof the implant serves to dissipate compressive forceson bone and thus should be considered its effectivelength. It is important to point out that 5 · 8 implantsfrom different implant systems have dissimilar designs(number of threads, screw type versus plateau), whichleads to differences in functional surface areas anddissimilar effective lengths. It is possible that bone maybehave differently around implants of similar size butwith different designs. Thus, the results of this investi-gation should only be applicable to 5 · 8 PRF implants.

The limitations of this study include the following:1) the retrospective design; 2) the sample size; 3)use of a single center for implant placement; 4) inves-tigation of only one implant design; 5) missing or in-complete records; and 6) the use of two-dimensionalradiographs. Retrospective cohort studies rely oncomplete data entered into the patient’s history andhave less validity than randomized prospective clini-cal trials as a result of possible selection bias andconfounding factors. The use of two-dimensional ra-diographs to evaluate a three-dimensional phenome-non, such as changes in peri-implant bone levels,represents another drawback of this study. A largersample would have enhanced the power of the study.Multicenter, prospective, randomized controlled clin-ical trials will be necessary to validate the findings ofthis study.

CONCLUSIONS

The present study evaluates the effect of implant size,specifically 5 · 8, in peri-implant bone levels sur-rounding PRF implants after the insertion of single-tooth replacements. Within the limitations of thisstudy, the following conclusions may be drawn: 1) 5 ·8 PRF implants were found to be statistically signif-icantly less likely to lose bone when compared withPRF implants not measuring 5 · 8 in the mandible(P = 0.047); and 2) 5 · 8 PRF implants were statisti-cally significantly more likely to gain bone if they wereopposing natural teeth (P = 0.05) but not statisticallysignificantly more likely to lose bone if opposingimplants (P = 0.14).

ACKNOWLEDGMENTS

The authors recognize the staff of the Implant Den-tistry Center, Boston, MA, for their cooperation in thisstudy. Drs. Urdaneta, Leary, and Chuang have no fi-nancial interests related to any products involved inthis study. Mrs. Emanuel is a part-time employee ofBicon Dental Implants, and Mr. Lubelski is a formeremployee of Bicon Dental Implants.

REFERENCES1. Smith DE, Zarb GA. Criteria for success of osseointe-

grated endosseous implants. J Prosthet Dent 1989;62:567-572.

2. Degidi M, Piattelli A, Iezzi G, Carinci F. Wide-diameterimplants: Analysis of clinical outcome of 304 fixtures.J Periodontol 2007;78:52-58.

3. Laurell L, Lundgren D. Marginal bone level changesat dental implants after 5 years in function: A meta-analysis. Clin Implant Dent Relat Res 2011;13:19-28.

4. Karlsson U, Gotfredsen K, Olsson C. Single-tooth re-placement by osseointegrated Astra Tech dental implants:A 2-year report. Int J Prosthodont 1997;10:318-324.

5. Ganeles J, Zollner A, Jackowski J, ten BruggenkateC, Beagle J, Guerra F. Immediate and early loadingof Straummann implants with a chemically modifiedsurface (SLActive) in the posterior mandible andmaxilla: 1-year results from a prospective multicen-ter study. Clin Oral Implants Res 2008;19:1119-1128.

6. Blanes RJ, Bernard JP, Blanes ZM, Belser UC. A 10-year prospective study of ITI dental implants placed inthe posterior region. I: Clinical and radiographic re-sults. Clin Oral Implants Res 2007;18:699-706.

7. Yoo RH, Chuang SK, Erakat MS, Weed M, Dodson TB.Changes in crestal bone levels for immediately loadedimplants. Int J Oral Maxillofac Implants 2006;21:253-261.

8. Schwarz F, Herten M, Sager M, Wieland M, Dard M,Becker J. Bone regeneration in dehiscence-type de-fects at chemically modified (SLActive) and conven-tional SLA titanium implants: A pilot study in dogs.J Clin Periodontol 2007;34:78-86.

9. Urdaneta RA, Daher S, Lery J, Emanuel K, ChuangSK. Factors associated with crestal bone gain onsingle-tooth locking-taper implants: The effect of non-steroidal anti-inflammatory drugs. Int J Oral MaxillofacImplants 2011;26:1063-1078.

10. Eckert SE, Meraw SJ, Weaver AL, Lohse CM. Earlyexperience with Wide-Platform Mk II implants. Part I:Implant survival. Part II: Evaluation of risk factorsinvolving implant survival. Int J Oral Maxillofac Im-plants 2001;16:208-216.

11. Krennmair G, Seemann R, Schmidinger S, Ewers R,Piehslinger E. Clinical outcome of root-shaped dentalimplants of various diameters: 5-year results. Int JOral Maxillofac Implants 2010;25:357-366.

12. Urdaneta RA, Marincola M, Weed M, Chuang SK. Ascrewless and cementless technique for the restorationof single-tooth implants: A retrospective cohort study.J Prosthodont 2008;17:562-571.

13. Urdaneta RA, Rodriguez S, McNeil DC, Weed M,Chuang SK. The effect of increased crown-to-implantratio on single-tooth locking-taper implants. Int J OralMaxillofac Implants 2010;25:729-743.

J Periodontol • October 2012 Urdaneta, Leary, Lubelski, Emanuel, Chuang

9

14. Misch CE. Dental evaluation: Factors of stress. In:Misch CE, ed. Contemporary Implant Dentistry, 2nded. St. Louis: Mosby; 1999:123-129.

15. Chuang SK, Wei LJ, Douglass CW, Dodson TB. Riskfactors for dental implant failure: A strategy for theanalysis of clustered failure-time observations. J DentRes 2002;81:572-577.

16. Esposito M, Pellegrino G, Pistilli R, Felice P. Rehabil-itation of posterior atrophic edentulous jaws: Prosthe-ses supported by 5 mm short implants or by longerimplants in augmented bone? One-year results froma pilot randomised clinical trial. Eur J Oral Implantol2011;4:21-30.

17. Petrie CS, Williams JL. Comparative evaluation ofimplant designs: Influence of diameter, length, andtaper on strains in the alveolar crest. A three-dimen-sional finite-element analysis. Clin Oral Implants Res2005;16:486-494.

18. Himmlova L, Dostalova T, Kacovsky A, Konvickova S.Influence of implant length and diameter on stressdistribution: A finite element analysis. J Prosthet Dent2004;91:20-25.

19. Ding X, Liao SH, Zhu XH, Zhang XH, Zhang L. Effect ofdiameter and length on stress distribution of thealveolar crest around immediate loading implants.Clin Implant Dent Relat Res 2009;11:279-287.

20. Wolff J. Das gesetz der transformation der knochen.Berlin: Hirschwald; 1892.

21. Frost HM. The mechanostat: A proposed pathogenicmechanism of osteoporoses and the bone mass effectsof mechanical and nonmechanical agents. Bone Miner1987;2:73-85.

22. Hassler CR, Rybicki EF, Cummings KD, Clark LC.Quantification of bone stresses during remodeling.J Biomech 1980;13:185-190.

23. Quirynen M, Naert I, van Steenberghe D. Fixturedesign and overload influence marginal bone lossand fixture success in the Branemark system. ClinOral Implants Res 1992;3:104-111.

24. Naert I, Quirynen M, van Steenberghe D, Darius P. Astudy of 589 consecutive implants supporting com-plete fixed prostheses. Part II: Prosthetic aspects.J Prosthet Dent 1992;68:949-956.

25. Jemt T, Lekholm U, Adell R. Osseointegrated implantsin the treatment of partially edentulous patients: Apreliminary study on 876 consecutively placed fix-tures. Int J Oral Maxillofac Implants 1989;4:211-217.

26. Lindquist LW, Rockler B, Carlsson GE. Bone resorp-tion around fixtures in edentulous patients treated withmandibular fixed tissue-integrated prostheses. J Pros-thet Dent 1988;59:59-63.

27. Ahlqvist J, Borg K, Gunne J, Nilson H, Olsson M,Astrand P. Osseointegrated implants in edentulousjaws: A 2-year longitudinal study. Int J Oral MaxillofacImplants 1990;5:155-163.

28. Guichet DL, Yoshinobu D, Caputo AA. Effect ofsplinting and interproximal contact tightness on loadtransfer by implant restorations. J Prosthet Dent 2002;87:528-535.

29. Wang TM, Leu LJ, Wang J, Lin LD. Effects ofprosthesis materials and prosthesis splinting on peri-implant bone stress around implants in poor-qualitybone: A numeric analysis. Int J Oral MaxillofacImplants 2002;17:231-237.

30. Rokni S, Todescan R, Watson P, Pharoah M, AdegbemboAO, Deporter D. An assessment of crown-to-root ratioswith short sintered porous-surfaced implants supportingprostheses in partially edentulous patients. Int J OralMaxillofac Implants 2005;20:69-76.

31. Hekimoglu C, Anil N, Cehreli MC. Analysis of strainaround endosseous dental implants opposing naturalteeth or implants. J Prosthet Dent 2004;92:441-446.

32. Clinton T, Lanyon LE. Regulation of bone formation byapplied dynamic loads. J Bone Joint Surg Am 1984;66:397-402.

33. Misch CE, Bidez MW. A scientific rationale for dentalimplant design. In: Misch CE, ed. ContemporaryImplant Dentistry, 2nd ed. St. Louis: Mosby; 1999:334.

Correspondence: Rainier A. Urdaneta, 25 Prairie Ave.,Auburndale, MA 02466. Fax: 617/390-0043; e-mail:rainieru@yahoo.com.

Submitted September 25, 2011; accepted for publicationDecember 16, 2011.

Implant Size and Bone Preservation Volume 83 • Number 10

10