ohpd_2009_02_s0107.pdf - Quintessence Publishing!

Copyrig

ht

by

N

otfor

Qu

in

tessence

Not

forPublication

The Relationship Between Diabetes Mellitusand Destructive Periodontal Disease: A Meta-Analysis

Nilo Guliberto Martins Chávarrya,b/Mario Vianna Vettorec/Carmelo Sansonea/Aubrey Sheihamd

Purpose: The aim of this study was to systematically review the studies on the association between diabetes mellitus(DM) and destructive periodontal disease.

Methods: The methods applied include a literature search strategy, inclusion and exclusion criteria for selecting thestudies, characteristics of the studies, quality assessment and meta-analysis. Data sources included PubMed, EMBASE,SciELO and LILACS. Selected papers were articles relating to human studies investigating whether or not diabetes is arisk factor for periodontitis and if it influences the response to periodontal therapy. Those papers that were publishedbetween January 1980 and June 2007 were retrieved.

Results: Of the 2440 identified studies, 49 cross-sectional and eight longitudinal studies met the inclusion criteria.Twenty-seven of the 49 cross-sectional studies that are included in this review detected more periodontal disease indiabetic subjects compared with non-diabetic subjects. The greater risk of periodontal disease progression wasassociated with type 2 DM, and one study associated DM with response to periodontal therapy. Methodological flaws ofmost of the studies included inadequate control for confounders, insufficient statistical analysis and lack of informationabout sampling design. Random effect model showed a significant association with clinical attachment level (meandifference = 1.00 [CI 95% = 0.15 to 1.84]) and periodontal pocket depth (mean difference = 0.46 [CI 95% = 0.01 to0.91]) between type 2 diabetics and non-diabetics.

Conclusions: Type 2 DM can be considered a risk factor for periodontitis. More studies are needed to confirm theharmful effects of type 1 DM on periodontal disease.

Key words: diabetes mellitus, meta-analysis, periodontal disease, periodontitis, systematic review

Oral Health Prev Dent 2009; 7: 107–127. Submitted for publication: 24.12.07; accepted for publication: 03.03.08.

MModerate chronic periodontal disease is commoninmost countries (Albandar, 2002a, b; Sheiham

and Netuveli, 2002). Poor oral hygiene and bacterial

biofilms composed of anaerobic Gram negativemicroorganisms are the main causative factors forperiodontal breakdown (Christersson, 1993). In addi-tion, the habit of smoking and diabetes mellitus (DM)have been described as the risk factors for periodon-tal disease initiation and progression (Obeid andBercy, 2000; Albandar, 2002a, b). Other putative riskfactors for periodontal disease include psychosocialfactors, socioeconomic conditionsandage (Albandar,2002a, b; Vettore et al, 2003).Most narrative review papers claim an association

between DM and periodontal disease (Belting et al,1964; Campbell, 1967; Hugoson and Jordan,1982; Mealey and Oates, 2006). However, thesereview papers did not follow adequate protocols thatare applied in systematic reviews (Alderson et al,

a Division of Graduate Periodontics, Federal University of Rio deJaneiro, Rio de Janeiro, Brazil.

b Department of Prosthodontics Dentistry, UNIGRANRIO, Duque deCaxias, Brazil.

c Department of Epidemiology and Quantitative Methods in Health,National School of Public Health, Oswaldo Cruz Foundation, Rio deJaneiro, Brazil.

d Department of Epidemiology and Public Health, University CollegeLondon, London, England.

Correspondence: Mario Vianna Vettore, Department of Epidemiologyand Quantitative Methods in Health, National School of Public Health,Oswaldo Cruz Foundation, Rio de Janeiro, Brazil. Tel: +55 2125982620. Fax: +55 21 22706772. Email: [email protected]

ORIGINAL ARTICLE

Vol 7, No 2, 2009 107

Copyrig

ht

by

N

otfor

Qu

in

tessence

Not

forPublication

2004). The need for an adequate methodology forconducting systematic reviews is essential to avoidbias in selecting studies and to appraise the scien-tific literature critically. To overcome the possibleshortcomings of the previous reviews, a systematicreview was performed to assess if DM is a possiblerisk factor for destructive periodontal disease. Thefocused questions are: (i) do type 1 and 2 DMincrease the risk for periodontitis? (ii) do type 1and 2 DM influence the response to periodontaltherapy?

RESEARCH DESIGN AND METHODS

Development of a protocol

A protocol was developed to answer the selected andfocused questions, which included all aspects of thereviewmethodology (Alderson et al, 2004). Themeth-ods applied in this review included a literature searchstrategy, inclusion and exclusion criteria for selectingthe studies, selection process and data extraction,quality assessment and quantitative data synthesis.

Literature search strategy

The data sources included PubMed, EMBASE, SciELOandLILACS. Standardisedmethodological filterswereused to identify analytical studies. First, two themeswere derived that were then combined using the Bool-ean operator ‘AND’. Each theme was created usingthe operator ‘OR’, to search for terms appearing asboth exploded medical subject headings (MeSH)and text words. The first themewas created for ‘diabe-tes mellitus’, or ‘diabetes’, or ‘insulin resistance’, or‘glucose intolerance’, or ‘metabolic control’, or‘impaired glycaemia’ and the second for ‘periodontaldisease’, or ‘periodontitis’, or ‘periodontal’.Manual searching of the identified articles was

also conducted. The search was limited to humanstudies written in English.

Inclusion and exclusion criteria for selectingstudies

Studies were considered for inclusion if they ad-dressed clinical measures of destructive periodontaldisease and/or radiography of periodontal tissues.The analytical studies had to include an estimateof the effect of diabetes on periodontal diseaseand/or the statistical tests for comparison of

groups. The selected studies were published bet-ween January 1980 and June 2007.Case reports describing periodontal conditions in

diabetic subjects, ecological studies, experimentalanimal studies and previous reviews were excluded.In addition, published studies showing repeatedresults from the same original study were excluded.Studies that had inadequate or unclear scoring

systems for the measurement of destructive peri-odontal disease and those in which the type of dia-betes was not reported were excluded. Measuresof destructive periodontal disease were classifiedas adequate when clinical measures were basedon periodontal pocket depths and/or clinical at-tachment loss measures, or alveolar bone lossassessed using radiographs. Scoring measures ofdestructive periodontal disease were consideredinadequate if indices with unsatisfactory reliabilitywere used to assess periodontal status. The indicesincluded community periodontal index of treatmentneeds (CPITN), Ramfjord index and periodontal dis-ease index. Studies that did not report the methodof periodontal assessment and those that did notpresent the results of periodontal measures werescored as unclear.

Selection process and data extraction

Two reviewers (NGMCandCS) independently retrievedand evaluated the articles for eligibility. The extractionof information from studies was conducted by thesame reviewers, and the information included therange of duration of diabetes, type of periodontal dis-ease, definition of periodontitis, type of periodontalprobe used, number of patients, patient characteristicdata, confounders taken into account and the mainresults. Any discrepancies between reviewers werediscussed with a third reviewer (MVV) and resolvedby consensus.

Quality assessment

It was decided a priori that methodological quality forthe studies that were included would be assessedwith a predetermined appraisal form, focusing onthe following issues: sampling calculation, adjust-ment for potential confounders, statistical analysis,sampling design/selection of control group, res-ponse rate and blindness of outcome assessment.Longitudinal studies were also assessed for com-pleteness of follow-up. For quality assessment of clin-ical trials, the Revised CONSORT Statement for

Chávarry et al

108 Oral Health & Preventive Dentistry

Copyrig

ht

by

N

otfor

Qu

in

tessence

Not

forPublication

Reporting Randomized Trials (Moher et al, 2001)was used, which included allocation concealment,method of randomisation, blindness of patients, clini-cian and examiners and completeness of follow-up.The following factors were assessed and the crite-

ria below were used.

Sampling calculationIt was noted if the sample size had been previouslydetermined.

Measures of diabetes

Diabetes was classified as adequate when it wasassessed through glycosylated haemoglobin, fast-ing blood glucose level or oral glucose tolerancetest and as inadequate when it was self-reported.

Were the examiners calibrated for periodontalassessment?• Adequate-level 1: when results on clinical cali-bration were presented.

• Adequate-level 2: when clinical calibration wasperformed, but the results were not reported.

• Inadequate: when clinical calibration was notperformed.

Adjustment for potential confounders• Adequate: when all the following variables weretaken into account: age, smoking, ethnicity,socioeconomic status, previous periodontal ther-apy, number of teeth, immunological and neuro-logical diseases.

• Inadequate: when any of the above variableswere not considered.

• Unclear: when control for confounders was notadequately explained.

Statistical analysis• Adequate: when estimates of association (oddsratio or relative risk) were provided or could becalculated and when multivariate analysis wasconducted.

• Inadequate: when only bivariate analysis wasperformed.

Sampling design/selection of the control group• Was the sample drawn randomly so that it is rep-resentative of the population it reports torepresent?

• Was the control/unexposed group selected fromthe same population that produced the cases/exposed group?

• Was the definition of case/exposed group expli-cit?

• Was the exposure assessed in the same mannerbetween groups?

• Was the outcome assessed in the same mannerbetween groups?

• Was the objective outcome criteria applied tocontrols to exclude cases in the control group?

• Was the objective exposure criteria applied to theunexposed group to exclude exposed persons inthe unexposed group?

Response rate• Was the non-response rate reported?• Were the non-response rates and reasons fornon-response similar in both groups?

Blindness of outcome assessment• Was the outcome assessment blind to exposurestatus?

• Was the level of exposure blind to outcomestatus?

Completeness of follow-up (longitudinal studiesonly)The following criteria were assessed.• Was the number of patients at baseline and atcompletion of the follow-up interval reported forboth groups?

• Were drop-out rates and reasons for drop-outsimilar between groups?

• Was exposure updated in studies with long-termfollow up?

• Was the dose–effect relationship between peri-odontitis and outcome noted?

Quantitative data synthesis and statisticalmethodsFor quantitative data synthesis, studies that pro-vided information on mean differences of periodontalpocket depth (PPD) and/or clinical attachment level(CAL) between diabetics and non-diabetics wereincluded. The data should be presented by the typeof diabetes and periodontal clinical measure. The

Chávarry et al

Vol 7, No 2, 2009 109

Copyrig

ht

by

N

otfor

Qu

in

tessence

Not

forPublication

data were combined to estimate the pooled differ-ence of mean values with confidence interval at95% (CI 95%) using the inverse variance method.Homogeneity among studies was tested by Coch-ran’s Q test. Potential publication bias was testedfor using the rank correlation of Begg’s test (Beggand Berlin, 1994) and Egger’s test (Egger et al,1997). All statistical analyses were performed usingcommercial statistical software, version STATA6(Strata Corp, TX, USA). The significance level thatwas established for analyses was 5% (P � 0.05).

RESULTS

Search results

Of the 2440 identified studies, 57 met the inclusioncriteria. Selected papers included 49 cross-sectional

studies; 17 papers on type 1 DM, 26 on type 2 DMand 6 on type 1 and 2 DM. In addition, eight longitudi-nal studies were included; 3 papers on type 1 DM, 3on type 2 DM and 2 on type 1 and 2 DM (Fig 1).Epidemiological characteristicsandquality assess-

ment results of the analytical studies that wereincluded are presented according to the studydesign: cross-sectional and longitudinal studies, andtype of diabetes: type 1 DM, type 2 DM and type 1and 2 DM.

Cross-sectional studies

The epidemiological characteristics of the cross-sectional studies on the relationship betweendiabetes and periodontal disease are presented inTable 1 (type 1 DM), Table 2 (type 2 DM) and Table 3(type 1 and 2 DM). Of the 49 studies that compared

Potentially relevant publications identified and screened for retrieval: n = 2440

Full text retrieved for more detailed evaluation: n = 129

Studies included in systematic review: n = 57

Selected studies: n = 364

Papers excluded for not addressing different aspects and measures of destructive periodontal disease: n = 2076

Exclusion for other reasons: • No controls (non-diabetics): n = 21 • Periodontal disease considered as a risk factor

for diabetes: n = 18 • Measures of periodontal disease were classified

as inadequate and/or type of diabetes were unclear: n = 72

Papers identified through manual searching: n = 12

Papers excluded based on the exclusion criteria: • Case reports: n = 25 • Comments: n = 4 • Animal studies: n = 9 • Review articles: n = 121 • Duplicated study: n = 23 • Not in English language: n = 63 • Acute necrotising ulcerative gingivitus,

acute necrotising ulcerative periodontitis and endodontic–periodontal lesions: n = 2

Fig 1 Flow chart of stud-ies through the review.

Chávarry et al


Copyrig

ht

by

N

otfor

Qu

in

tessence

Not

forPublication

the parameters of periodontal disease between dia-betic and non-diabetic patients, 27 found significantdifferences between groups in at least one periodon-tal parameter.The range of duration of diabetes was between 1

and 29 years for type 1 DM (seven studies), 1 to17 years for type 2 DM (17 studies) and 4 to17 years for type 1 and 2 DM (three studies). Moststudies considered periodontitis as the periodontaldisease under study. However, the definition of peri-odontitis varied considerably among studies, and ineight studies the criteria used to define periodontitiswere unclear.Six different types of periodontal probeswere used

in studies on type 1 DM. In six studies on type 1 DMand in 12 studies on type 2 DM, there was no infor-mation about the type of periodontal probe used. Ofthe seven studies on type 1 and 2 DM, only onereported on the type of periodontal probe.The sample size of the selected studies varied

from 16 to 15,246. The number of confounders

taken into account varied from 0 to 9, but most stud-ies considered only a few confounders.

Type 1 DMType 1DMwas considered in 17 cross-sectional stud-ies (Table 1). Eight of the 17 studies found an associ-ation between type 1 DM and periodontal disease.Seven studies showed differences in periodontalprobing depth (PPD) measures between groups, andfour studies found significantly more clinical attach-ment loss (CAL) in diabetics compared with non-diabetics. Five studies investigated the relationshipbetween type 1 DM and periodontal disease usingalveolar bone loss. One study detected differencesbetween diabetic and non-diabetic groups using alve-olar bone loss.

Type 2 DMThere were 26 studies on the relationship betweentype 2 DM and periodontal disease (Table 2). Of the

1 CAL

2 PPD

Morton 1995

Alpagot 2001

Yuan 2001

Alley 1993 Firatli 1994 Pinson 1995 Firatli 1996 Firatli 1997 Salvi 1997b

Alpagot 2001 Guthmiller 2001

Yucekal-Tuncer 2003 Aren 2003

Morton 1995

Alpagot 2001

Yuan 2001

Bulut 2001

Orbak 2002

mean difference-1 0 1

Combined0.26 (-0.004–0.533) p = 0.054

Manouchehr-Pour 1981

Alley 1993

Firatli 1994

Pinson 1995

Firatli 1996

Firatli 1997

Alpagot 2001

Guthmiller 2001

Yucekal-Tuncer 2003

-1 0 1

Combined0.11 (-0.03–0.245) p = 0.137

-1 0 1

Combined

1.00 (0.15–1.84) p = 0.021

-1 0 1

Combined

0.46 (0.01–0.91) p = 0.046

1a Type 1 DM 1b Type 2 DM

2a Type 1 DM 2b Type 2 DM

mean difference

mean difference mean difference

Fig 2 Mean differences in CAL and PPD between diabetics and non-diabetics according to the type of diabetes.CAL, clinical attachment loss; DM, diabetes mellitus; PPD, periodontal probing depth.

Chávarry et al

Vol 7, No 2, 2009 111

Copyrig

ht

by

N

otfor

Qu

in

tessence

Not

forPublication

Table1

Cha

racteristics

ofcros

s-se

ctiona

lstud

ieson

type

1DM

andpe

riod

ontaldise

ase

Referen

ces

Ran

geof

duration

ofdiab

etes

Type

sof

period

ontal

dise

ase

Defi

nition

ofpe

riod

ontitis

Type

ofpe

riod

ontal

prob

e

Num

ber

ofpa

tien

tsAge

(yea

rs)

Variables

controlle

dRes

ults

Manouchehr-Pour

etal(1981)

�5.0

years

Adult

periodontitis

Severe

periodontitis:

meanvalueof

CAL

�4mm

and�6teeth

with

>50%ABL

Unclear

TotalN=32,

DsP

n=8,

DmPn=6,

nDsP

n=7,

nDmPn=11

DsP

=49,

DmP=51,

nDsP

=48,

nDmP=31

None

NSDforCAL

Nocorrelation

amongCAL,

FBGL,DOSorDUR

Rylander

etal(1986)*

�10.0

years

Periodontitis

Unclear

Unclear

TotalN=87,

Dn=46,

nDn=41

D=22.1

(19–25),

nD=22.3

(18–26)

Age

SDforCAL

�2mm

NSDforPPD4–5

mm,

PPD>5mmandABL

NocorrelationofDUR,

DOSandHbA

1c

with

periodontal

parameters

Sandholm

etal(1989)

5.2

±3.5

years

Periodontitis

Unclear

Standard

probe

(Hilm

ing,

0.6

mm)

TotalN=170,

Dn=85,

nDn=85

15.1

Ageandgender

NSDforPPD�4mm

Nosubjectswith

PPD�6mmand

ABL�1mm

Hugoson

etal(1989)*

Long

duration

28.9

±10.2;

shortduration

5.2

±1.9

Periodontitis

Based

onclinicaland

radiographical

criteria

modified

from

Hugoson

and

Jordan

(1982)

Unclear

TotalN=231,

Dlong

durationn=82,

Dshort

durationn=72,

nDn=77

20–70

Ageandgender

NSDforPPD�4mm,

PPD4–5

mmandABL

SDforPPD�4mm

insubjects

aged

<45years

SDforPPD�6mm

insubjectswith

long

durationof

diabetes

DePommereau

etal(1992)

1–14years

Unclear

Periodontitis:

loss

ofperiodontal

attachment�1

siteadjacentto2

ormoreteeth

WHO

periodontal

probe

TotalN=123,

Dn=85,

nDn=38

D=15.1,

nD:unclear

Age,durationof

diabetes

anddegree

ofmetaboliccontrol

ofdiabetes

NSDforperiodontitis

prevalence

andABL

Harber

etal(1993)*

Unclear

Periodontitis

Periodontitis:

�1

site,

PPD�5mm,

CAL

�2mm

Michigan

type

‘0’probe

with

Williams

TotalN=227,

Dn=132,

nDn=95

19–40

Pregnancy,

HIV,age,gender

andsm

oking

OR1.8

forformer

smokers,

OR6.9

current

smokers

NSDfornon-sm

okers

Thorstensson

and

Hugoson

(1993)*

>7.0

years

Unclear

Based

onclinicaland

radiographical

criteria

modified

from

Hugoson

and

Jordan

(1982)

Hilm

ingprobe

TotalN=182,

Dn=83,

nDn=99

40–69

Age,gender,duration

ofdiabetes,

numberofteeth

SDforABL

SDforPPD�4mm

and4–5

mm

SDfornumberofsites

PPD�6mmand

ABLin40–49

agegroup

Chávarry et al


Copyrig

ht

by

N

otfor

Qu

in

tessence

Not

forPublication

Table1Cha

racteristics

ofcros

s-se

ctiona

lstud

ieson

type

1DM

andpe

riod

ontaldise

ase(C

ontinu

ed)

Referen

ces

Ran

geof

duration

ofdiab

etes

Type

sof

period

ontal

dise

ase

Defi

nition

ofpe

riod

ontitis

Type

ofpe

riod

ontal

prob

e

Num

ber

ofpa

tien

tsAge

(yea

rs)

Variables

controlle

dRes

ults

Alleyetal

(1993)

Unclear

Moderateto

severe

periodontitis

�4sitesinmultiple

quadrantswith

PPD

andCAL

�5mm,

BOPandABL

Unclear

TotalN=60,

DPn=15,

DnP

n=15,

nDPn=15,

nDnP

n=15

D=35.5,

nD=36.4

Ageandgender,previous

antibiotic

therapy,

previous

periodontal

therapyandrace

NSDforPPDandCAL

Firatlietal

(1994)

Unclear

Periodontitis

andgingivitis

Unclear

Williamsprobe

TotalN=68,

Dn=48,

nDn=20

D=11.6,

nD=12.5

Age

NSDCAL

andPPD

Thereareno

subjects

with

periodontitis

Pinson

etal

(1995)

Mean

6.6

years

Unclear

Unclear

Vine

valleyprobe

TotalN=50,

Dn=26,

nDn=24

D=13.5

±3.4,

nD=13.5

±3.1

Race,age,gender,previous

antibioticstherapy,any

severesystem

icillness,

numberofteethand

previous

ketosis

NSDformeanPPD

andmeanCAL

Nocorrelationof

clinicalvariables

andGHb

Sbordone

etal(1995)

Unclear

Unclear

Unclear

Michiganprobe

TotalN=32,

Dn=16,

nDn=16

D=11.3,

nD=13.2

Age,genderandnumber

ofperiodontalsites

NSDfor

PPDandCAL

Thorstensson

etal(1995)*

>7year/

mean

24.6

years

Periodontitis

ABL�1/3

ofnormal

bone

height

Unclear

TotalN=62,

Dn=28,

nDn=34

Unclear

Ageandgender

SDforPPD

of4and5mm

andPPD�4mm

Firatlietal

(1996)*

48.3

±23.7

months

Periodontitis

and

gingivitis

Unclear

Williamsprobe

TotalN=154,

Dn=77,

nDn=77

D=12.5,

nD=12.6

Age,gender,sm

oking,

ethnicity

andSES

SDforPPDandCAL

Salvietal

(1997)

Unclear

Adult

periodontitis

AAP

Unclear

TotalN=103,

Dn=39,

nDPn=43,

nDn=21

D=48.8,

nD=34.9

Pregnancy,Crohn’sdisease,

previous

antibiotic

therapy,

non-steroidalanti-inflammatory

drug

andsystem

icdisease

NSDforPPD

Guthm

iller

etal(2001)*

Unclear

Unclear

Unclear

Vine

valleyprobe

TotalN=33,

Dn=13,

nDn=20

D=27.0

±7.3,

nD=28.5

±7.1

Gender,ageandgestational

period

SDformeanPPD

andmeanCAL

Yucekal-Tuncer

etal(2003)*

Unclear

Chronic

periodontitis

CAL

�2mm

Borodontic

periodontal

probe

TotalN=16,

Dn=8,

nDn=8

Unclear

Smoking,previous

periodontal

treatmentandantibiotic

therapy

SDforPPD(patient

and

site-based

analysis)

SDforCAL

(site-based

analysis)

Aren

etal(2003)*

4.0

years

Periodontitis

Unclear

Williamsprobe

TotalN=48,

sDn=16,

lDn=16,

nDn=16

sDn=12.8

±5.8,

lDn=12.7

±3.8,

nDn=12.4

±1.9

Age,

gender

andSES

SDforPPD:lD

versus

sDandlD

versus

nD,

NSDforPPD:sD

versus

nD

ABL,alveolar

bone

loss;CEJ,cementoenam

eljunction;

AAP,

American

Academ

yofPeriodontology;D,diabetics;nD,non-diabetics;DsP,diabeticswith

severe

periodontitis;DmP,

diabeticswith

mild

periodontitis;nDsP,non-diabeticswith

severeperiodontitis;nDmP,non-diabeticswith

mild

periodontitis;DP,diabeticswith

periodontitis;sD,shortdurationofdiabetes;lD,long

durationofdiabetes;

nDP,

non-diabetes

with

periodontitis;

nDnP,non-diabetes

withoutperiodontitis;

DnP,diabeticswithoutperiodontitis;

FBGL,fastingbloodglucoselevel;DOS,insulin

dosage;DUR,diabeticduration;

SD,significantdifferencesbetweendiabeticsandnon-diabeticsgroups;NSD,non-significantdifferencesbetweendiabeticsandnon-diabeticsgroups;OR,odds

ratio;WHO,WorldHealth

Organization.

*Studies

with

atleastoneclinicalparameterthatisstatisticallydifferentbetweendiabeticsandnon-diabetics.

Chávarry et al

Vol 7, No 2, 2009 113

Copyrig

ht

by

N

otfor

Qu

in

tessence

Not

forPublication

Table2

Epidem

iologica

lch

arac

teristicsof

cros

s-se

ctiona

lstud

ieson

type

2DM

andpe

riod

ontaldise

ase

Referen

ceRan

geof

diab

etes

duration

Type

ofpe

riod

ontal

dise

ase

Defi

nition

ofpe

riod

ontitis

Type

ofpe

riod

ontal

prob

e

Num

ber

ofpa

tien

tsAge

(yea

rs)

Variables

controlle

dRes

ults

Nelson

etal(1990)

Unclear

CP

<24teethor

�6

teethwith

�25%of

ABLor

�1tooth

with

�50%ABL

Unclear

701

�15

Ageandgender

NSDfor

Periodontal

disease,OR:

2.6

(1.0–6.6)

Emrich

etal(1991)*

Unclear

Periodontitis

CAL

�5mm,

ABL,Score2

(25–49%)Score3

(50–74%)Score

4(�

75%)

Unclear

Total

N=1342,

Dn=254,

Igtn=158,

nDn=930

�15

Ageandgender

SDforCAL

�5mm

OR:2.81SD

forABL�Score

2OR:3.43

Cherry-Peppers

andShip

(1993)

7.7

years

Periodontitis

Unclear

Unclear

TotalN=86,

Dn=11,

IGTn=32,

nDn=43

D=67.9

±11.1,

IGT=60.7

±19.1,

nD=60.2

±18.8

Age,gender,SES,

medicines

andsystem

icdiseases

NSDforPPD

andCAL

Morton

etal(1995)*

>2years/

mean7.9

years

Unclear

Unclear

Williamsprobe

Total

N=48,

Dn=24,

nDn=24

nD45.8

±10.7,

D46.8

±11.9

Ageandany

medicalproblems

SDforCAL

andPPD

Collin

etal(1998)

Unclear

Moderate-to-

advanced

periodontitis

Advanced:

meanABL>50%,

or�2teeth

PPD�6mm

Moderate:mean

ABL25–50%,or

�1

PPD�6mmor

�1

PPD3–6

mm

WHO

periodontal

probe

Total

N=65,

Dn=25,

nDn=40

D=58–76,

nD=59–77

Ageandsm

oking

NSDforABL

andCAL,

%ofsiteswith

PPD3mm

to<6mm

or�6mm

Pwas

seen

in40%Dand13%nD

Kawam

ura

etal(1998)

�1year

Unclear

Unclear

TG-SP5

Total

N=200,

Dn=102,

nDN=98

nD=52±7.9,

D=53±9.3

Cerebrovascular

disease,system

icdisease,taking

medications

which

mightaltercognitive

function,age,gender

andoccupation

NSDforPPD

Cutler

etal(1999)

Unclear

Adult

periodontitis

�4periodontal

pockets�6mm

with

BOPand

radiographical

evidence

of>50%

bone

loss

atthesameteeth

Floridaprobe,

constant

force,

automated

electronic

probe,20g

TotalN=35,

nDUnclear

n=6,

nDPn=7,wDUnclear

n=6,wDPn=5,

poorDUnclear

n=5,

poorDP:

n=6

nDUnclear=28.2

nDP=42.9

wDUnclear=52

wDP=65.8

poorDUnclear

=45,

poorDP=42.5

Tobaccoproducts,

antibiotic

and

system

icdisease

NSDfor

PPDandCAL

Fontana

etal(1999)*

17±4years

Periodontitis

Unclear

WHOperiodontal

probewith

standardised

pressure

of20–25g

TotalN=80,

Dn=40,

nDn=40

D=56±8,

nD=56±6

Age,gender,system

icillness,infections,

inflammation,sm

oking

andcurrentuseofdrugs

SDforABLand

CAL

Parameters

andbone

loss

Chávarry et al


Copyrig

ht

by

N

otfor

Qu

in

tessence

Not

forPublication

Table2Ep

idem

iologica

lch

arac

teristicsof

cros

s-se

ctiona

lstud

ieson

type

2DM

andpe

riod

ontaldise

ase(C

ontinu

ed)

Referen

ceRan

geof

diab

etes

duration

Type

ofpe

riod

ontal

dise

ase

Defi

nition

ofpe

riod

ontitis

Type

ofpe

riod

ontal

prob

e

Num

ber

ofpa

tien

tsAge

(yea

rs)

Variables

controlle

dRes

ults

Kurtis

etal(1999)

Unclear

Adult

periodontitis

CAL

and

PPD>3mm

Williamsprobe

TotalN=72,

DPn=24,

nDPn=24,

nDUnclear

n=24

DP=37–65,

nDP=35–60,

nDUnclear

=17–22

Antim

icrobialagents

andperiodontal

treatmentinthe

previous

6months

NSDfor

PPDandCAL

Collin

etal(2000)

Unclear

Periodontitis

�1sitewith

PPD>4mm

WHO

periodontal

probe

Total

N=122,

Dn=45,

nDn=77

D=67±5.2,

nD=68±5.9

Age

NSDforPPD

Sandberg

etal(2000)*

9.9

years

SD(6.1)

Periodontitis

Advanced:

ABL:3(�

1/3)or

4(angulardefects

andfurcation

involvem

ent)

andPPD>4mm

Unclear

Total

N=204,

Dn=102,

nDn=102

D=64.8,

nD=64.9

Ageandgender

SDfornumberof

PPDof4–5

mm,

NSDforPPD�6mm

SDfornumberof

subjectswith

sites

with

advanced

periodontitis

Bulut

etal(2001)

Unclear

APAP,accordingto

Page

and

Shroeder(1976)

Unclear

TotalN=51,

Dn=17,

nDn=17,

nDUnclear

n=17

nDUnclear

=45.3,

D=52.5,

nDP=42.5

None

NSDformeanPPD

Yuan

etal(2001)

Unclear

Periodontitis

Site

with

PPD>3mm

Williams

Total

N=246,

D=105,

nD=141

D=57.1

±12.0

nD=57.1

±11.7

Ageandgender

NSDforPPD

andCAL

Orbak

etal(2002)*

Unclear

Unclear

Unclear

Unclear

Total

N=60sm

okers

Dn=21,

non-sm

okers

Dn=19,

non-sm

okers

nDn=20

Smokers

D=46±1.4

non-sm

okers

D=43±4.6

non-sm

okers

nD=41±3.9

Age,gender,drugs

thataffectthe

mouthflora,

immunesystem

orinflammatory

response,previous

periodontal

treatmentand

apicalpathology

oftheteeth

SDformean

CAL

NSDfor

meanPPD

Tsai

etal(2002)*

Unclear

Severe

periodontitis

�2siteswith

CAL

�6mm

and�1site

with

PPD�5mm

inoneofthese

sites

Unclear

TotalN=4343

45–90

Age,gender,

education,

race,income,

smokingstatus,

andextent

ofsubgingival

calculus

PoorD

·nD:

OR=2.90

(1.40–6.03)

andbetter

D·nD:

OR=1.56

(0.90–2.28)

Chávarry et al

Vol 7, No 2, 2009 115

Copyrig

ht

by

N

otfor

Qu

in

tessence

Not

forPublication

Table2Ep

idem

iologica

lch

arac

teristicsof

cros

s-se

ctiona

lstud

ieson

type

2DM

andpe

riod

ontaldise

ase(C

ontinu

ed)

Referen

ceRan

geof

diab

etes

duration

Type

ofpe

riod

ontal

dise

ase

Defi

nition

ofpe

riod

ontitis

Type

ofpe

riod

ontal

prob

e

Num

ber

ofpa

tien

tsAge

(yea

rs)

Variables

controlle

dRes

ults

Zielinski

etal(2002)

5–15years

Periodontitis

BOPand

PD>3mm

Severe:

�2

teethwith

PPD�6mm

Moderate:1

site

with

PPD�6mm

or�1site

with

PPD3–6

mm

Michigan

type

‘0’

probe

TotalN=72,

Dn=32,

nDn=40

D=71±7,

nD=74±8

Age,numberofteeth,

severedementia,

anticoagulants,

antibiotic

prophylaxisor

taking

antibiotics

onthedayof

exam

ination,sm

oking,

regularcheckups

anddrymouth

NSDforprevalence

ofsevereand

moderateperiodontitis

Marugam

eetal(2003)

Unclear

Periodontitis

Moderate

ABL:

�1

toothwith

ABS<50%

andno

sites<25%

SevereABL:

�1

site

with

ABS<25%

Notused

Total

N=664

46–57

Unclear

NSDforABL

adjusted

OR=2.55

(0.86–7.54)

Saito

etal(2003)*

Unclear

Periodontitis

Tertilesofmean

ABL(the

ratio

ofthe

distance

ofCEJ

toalveolar

crestby

CEJ

torootapex)

low:10–20.7%,

medium:20.8–28.6%,

high:28.7–55.9%

Notused

Total

N=179

50–54

Gastritis,parotitis,

anti-inflammatory

drugs,

age,sm

oking

history,

systolicblood

pressure,BMI,

triglyceride

andHDL

cholesterol

Correlation:

ABL·CRP

Thehighesttertile

ofABLhada

significant

risk

ofCRPelevation

(�1.3

mg/L)

OR:8.2

(1.6–40.7)P=0.01

Ünlü

etal(2003)*

�5years

Periodontitis

PPD>3mm

Unclear

TotalN=30,

DPn=10,

nDPn=10,

nDUnclear

n=10

D=56,

nDP=49.3,

nDUnclear:

unclear

Systemicdisease,

smoking,nottaking

anymedication,

ageandgender

NSDforPPD

SDforCAL

Güneri

etal(2004)

�5years

Periodontitis

Sites>3mm

Unclear

TotalN=30,

DPn=10,

nDPn=10,

nDUnclear

n=10

DP=56,

nDP=49.3,

nDUnclear=50.3

Smoking,

system

icdiseases

andmedications

NSDforPPD

andCAL

Luand

Yang

(2004)*

Unclear

Periodontitis

Unclear

Unclear

Total

N=164

Dn=72,

nDn=92

D=54.3,

nD=54.9

Ageand

numberof

teeth

SDforCAL

Chávarry et al


Copyrig

ht

by

N

otfor

Qu

in

tessence

Not

forPublication

Table2Ep

idem

iologica

lch

arac

teristicsof

cros

s-se

ctiona

lstud

ieson

type

2DM

andpe

riod

ontaldise

ase(C

ontinu

ed)

Referen

ceRan

geof

diab

etes

duration

Type

ofpe

riod

ontal

dise

ase

Defi

nition

ofpe

riod

ontitis

Type

ofpe

riod

ontal

prob

e

Num

ber

ofpa

tien

tsAge

(yea

rs)

Variables

controlle

dRes

ults

Saito

etal(2004)*

Unclear

Moderate-to-

advanced

periodontitis

MeanPPD

persubjects>2mm:

high

category;

1.3–2.0

mm:

intermediate

category:Mean

CAL

per

subjects>2.5

mm:

high

category;

1.5–2.5

mm:

intermediate

category

Unclear

Total

N=860,

nDn=679,

Dn=181

40–79

Age,gender,BMI,

exercise

frequency,

alcoholconsum

ption,

andsm

okinghabits

MeanPPD

intermediate/

high

OR:1.9

(1.0–3.4)/

OR:2.6

(1.3–5.0),

respectively

MeanCAL

intermediate/

high

OR:1.1

(0.6–2.0)/OR:2.0

(1.0–3.9),

respectively

Cam

pus

etal(2005)

18.2

±5years

Periodontitis

ADAcriteria

(24),

Moderate:PPD4–6

mm,

Advanced:PPD>6mm

Unclear

Total

N=212,

Dn=71,

nDn=141

35–75,

D=61,

nD=59.1

Num

ber

ofteethand

smoking

D·nD

and

gmc·bm

candbm

c·nD

SDfornumber

andpercent

ofPPD>4mm

Mansourand

Abd-Al-Sada

(2005)*

Unclear

Periodontitis

Advanced:�2teeth

with

PPD�5mm

(or30%ofteeth)

or�4teethwith

PPD�4mm

(or60%ofteeth)

Moderate:

�1

teethwith

PPD�5mm

or�2teeth

with

PPD�4mm

(or�30%ofteeth)

PCP-11

Total

N=1593,

Dn=633,

nDn=960

�40,

D=56.7,

nD=56.9

Num

berofteeth,

ageandsm

oking

SDformoderate

periodontitis

(PPD

)

Mattout

etal(2006)*

Unclear

CP

Unclear

Williams

periodontal

probePD

T20gpressure

Total

N=2144,

Dn=71

35–65

Age,sm

oking,

SES

andgender

SDforCAL

NSDforPPD

Borges-Yáñez

etal(2006)

Unclear

Periodontitis

Moderate:at

least2sites

with

CAL

�4mm

Severe:atleast

onesitewith

CAL

�6mm

Michigan

periodontal

probe

TotalN=170

>60

Age,gender,

edentulousness,

smokingand

calculus

NSDforCAL

ABL,alveolarbone

loss;CEJ,cem

entumenam

eljunction;WHO,W

orldHealth

Organisation;DsP,diabeticswith

severeperiodontitis;DmP,diabeticswith

mild

periodontitis;nDsP,non-diabeticswith

severeperiodontitis;

nDmP,non-diabeticswith

mild

periodontitis;D,diabetics;nD,non-diabetics;DP,diabeticswith

periodontitis;sD,shortdurationofdiabetes;lD,longdurationofdiabetes;IGT,impairedglucosetolerance;wDUnclear,well-

controlleddiabetes

withoutperiodontitis;wDP,well-controlleddiabetes

with

periodontitis;poorDUnclear,poorlycontrolleddiabetes

with

periodontitis;PoorDP,poorlycontrolleddiabetes

with

periodontits;nDP,non-diabetes

with

periodontitis;nDUnclear,non-diabeteswithoutperiodontitis;SES,socioeconom

icstatus;FBGL,fastingbloodglucoselevel;DOS,insulindosage;DUR,diabeticduration;SD,significantdifferencesbetweendiabeticand

non-diabeticgroups;N

SD,non-significantdifferencesbetweendiabeticsandnon-diabetics.

*Studies

with


Chávarry et al

Vol 7, No 2, 2009 117

Copyrig

ht

by

N

otfor

Qu

in

tessence

Not

forPublication

Table3

Epidem

iologica

lch

arac

teristicsof

cros

s-se

ctiona

lstud

ieson

type

1an

d2DM

andpe

riod

ontaldise

ase

Referen

ceRan

geof

diab

etes

duration

Type

ofpe

riod

ontal

dise

ase

Defi

nition

ofpe

riod

ontitis

Type

ofprob

eNum

ber

ofpa

tien

tsAge

(yea

rs)

Variables

controlle

dRes

ults

Oliverand

Tervonen

(1993)*

Mean13.8

years

Periodontitis

Unclear

Unclear

GP=15,132,

D=114

20–64,

D=40.6,

GP=40

Age,gender,

rheumatoid

arthritis,AIDS,

malignant

blooddisorders

andlong-term

medications

such

ascortisone,

phenytoin,

non-steroidal

anti-inflammatory

drugsandprevious

immunosuppressives

orantim

icrobialtherapy

SDforprevalence

andextentof

PPD�4mmNSD

forprevalence

and

extentofCAL

�3mm

and�5mmNSDof

PPDbetween1and2

Yavuzyilm

azetal(1996)*

Unclear

Unclear

Unclear

Unclear

TotalN=34,

Dtype

2n=10,

Dtype

2n=7,

nDn=17

nD=23.2

±3.3,

D=54.2

±15.8

Antibiotics,no

periodontaltreatment

withinthepast6months

SDformean

PPDandNSD

forPPD

between1and2

Alpagot

etal(2001)*

Type

110.03±5.73,

type

28.28±3.75

Unclear

Unclear

Unclear

TotalN=177

Dtype

2n=30

Dtype

1n=30

nDn=117

nD=36.0

±11.0,

Dtype

1=39.1

±12.7,

Dtype

2=45.2

±12.1

Unclear

SDformean

PPDandCAL

Persson

etal(2003)*

Unclear

Periodontitis

Siteswith

CAL

�4mm

andPPD�5mm

in>5%of

totalsites

UNCprobe

TotalN=1084,

Dtype

2n=100,

Dtype

1n=31,

nDn=953

60–75

Age,numberof

teeth(�

4),

smokingand

gender

SDfornumber

andproportion

ofsitesof

PPD�5mm

NSDfor

CAL

�4mm

andABL

OR:1.8

(1.1–3.1)

ofperiodontitis

Lalla

etal(2004)*

Unclear

Unclear

Notdefined

Notused

TotalN=300

Dtype

1n=23

Dtype

2n=103,

Unclear=24,

nD=150

nD=55±14.2,

D=56±13.1

Genderandage

SDforABL

Lalla

etal(2007)*

3.96±3.39

Periodontitis

�2teeth

with

�1site

with

CAL

>2mm

and/orbleeding

Unclear

TotalN=700

Dtype

1n=325

Dtype

2n=25

nDn=350

6–18

Age,gender,

ethnicity

andSES

SDformean

and%of

sites>2mm

ofCAL

Unclear,notpresented;GP,generalpopulationofem

ployed

American

adults;SD,significantdifferencesbetweendiabeticandnon-diabeticgroups;N

SD,non-significantdifferencesbetweendiabeticand

non-diabeticgroups;CAL,clinicalattachmentloss;D,diabetics;Ig,immunoglobulin;nD,non-diabetics;ELISA,

enzyme-linkedimmunosorbent

assay;CPITN,community

periodontalindexoftreatment

needs;GCF,gingivalcrevicularfluid;PMN,polym

orphonuclearperipheralblood;ABL,alveolarbone

loss;ABS,alveolarbone

support;WHO,W

orldHealth

Organization;OR,oddsratio;PKC,proteinkinase

C;CRP,C-reactiveprotein.

*Studies

with

atleastoneclinicalparameter

thatisstatisticallydifferentbetweendiabeticsandnon-diabetics.

Chávarry et al


Copyrig

ht

by

N

otfor

Qu

in

tessence

Not

forPublication

16studies that comparedCALmeasuresbetweendia-betics and non-diabetics, seven detected a positiveassociation between DM and periodontal disease.Of the 19 studies that assessed PPD measures, 6on type 2 DM found significant differences of PPDmeasures between diabetics and non-diabetics.Periodontal disease was assessed radiographi-

cally in seven studies that tested the hypothesis thattype 2 DM increased the risk for periodontitis. Threestudies reported more alveolar bone loss in diabet-ics compared with non-diabetics.

Type 1 and 2 DMThere were six studies on the relationship betweentype 1 and 2 DM and periodontal disease (Table 3).Of the four studies that compared PPD measuresbetween diabetics and non-diabetics, all found asso-ciationswith PPD. In four studies, CALmeasureswereused; in two studies significant differences werereported. Alveolar bone loss was compared between

diabetics and non-diabetics in two studies. One studydetected more alveolar bone loss in diabetics.

Quality assessment: results of 49 cross-sectional studies

Sampling calculationOne study on type 2 DM determined the sample size(Lu and Yang, 2004).

Were the examiners trained and calibrated for

periodontal assessment?Thirteen studies were considered adequate for clini-cal calibration of periodontal disease; four type 1DM (Hugoson et al, 1989; Harber et al, 1993; Pinsonet al, 1995; Guthmiller et al, 2001), six type 2 DM(Marugame et al, 2003; Lu and Yang, 2004; Campuset al, 2005;Mansour and Abd-Al-Sada, 2005; Borges-Yáñez et al, 2006;Mattout et al, 2006) and three type

CAL

PPD

Type 1 DM Type 2 DM

stan

dard

ised

eff

ect

precision0 20 40 60

0

2

4

6

Type 1 DM Type 2 DM

stan

dard

ised

eff

ect

precision0 20 40 60 80

-10

0

10

stan

dard

ised

eff

ect

precision0 1 2 3 4

-40

-20

0

20

40

stan

dard

ised

eff

ect

precision0 5 10

-5

0

5

10

Fig 3 Egger’s Publication Bias Plots: mean differences in CALs and PPD in diabetics and non-diabetics, according to the type ofdiabetes. CAL, clinical attachment level; DM, diabetes mellitus; PPD, periodontal probing depth.

Chávarry et al

Vol 7, No 2, 2009 119

Copyrig

ht

by

N

otfor

Qu

in

tessence

Not

forPublication

Table4

Epidem

iologica

lch

arac

teristicsof

long

itud

inal

stud

iesbe

twee

ndiab

etes

andpe

riod

ontaldise

ase

Referen

cePeriodo

ntal

therap

yDefi

nition

ofpe

riod

ontitis

Type

ofpe

riod

ontal

prob

e

Sam

plesize

Age

Variables

controlle

dNum

berof

patien

tsDuration

(mon

ths)

Res

ults

Diabe

tics

Non

-diabe

tics

Ran

geBas

eline

Fina

l

Firatli(1997)§*

No

Unclear

Williams

probe

TotalN=64,

Dn=44,

nDn=20

12.2

±3.9

12.3

±4.3

Unclear

Age,

smoking,

system

icdiseases,

race

and

SES

64

64

60

SDformean

CAL,Significant

correlation:

DURversus

CAL,NSD

formeanPPD

Tervonen

and

Karjalainen

(1997)§

Yes

Unclear

Unclear

Total

N=46,

Dn=36,

nDn=10

29.4

±3.7

30.1

±3.8

24–36

Ageand

number

ofteeth

46

43

36

NSDfor%PPD

and%CAL

Sbodorne

etal(1998)§

No

Unclear

Michigan

type

‘0’probe

Total

N=32,

Dn=16,

nDn=16

9–17

9–17

9–17

RaceandSES

16

16

36

NSDfor

meanCAL

andmeanPPD

Novaes

etal(1996,1997)�*

No

Adult

periodontitis

Michigan

type

‘0’probe

Total

N=60,

Dn=30,

nDn=30

52.3

44.6

30–77

Previous

periodontaltherapy,

antibioticsuse

60

60

12

SDformean

CAL,NSDfor

meanPPD

PoorD

versus

nD:

SDforCAL

andPPD

Taylor

etal(1998)�*

No

Unclear

Unclear

Total

N=362,

Dn=24,

nDn=338

Median=21.2

Median=25.7

15–57

Age,gender,

number

ofteeth

362

362

24

OR=4.2

(1.8–9.9)

Faria-Almeida

etal(2006)�*

Yes

Moderate

chronic

periodontitis:

4–6

mmof

CAL

inall

quadrants

NorthCarolina

periodontal

probe

Total

N=20,

type

2D

n=10,

nDn=10

Unclear

Unclear

35–70

Ageandgender

20

20

6SDforPPD

NSDforCAL

Westfelt

etal(1996)�

Yes

Unclear

Unclear

Total

N=40

type

1D

n=14,

type

2D

n=6,

nDn=20

Unclear

Unclear

46–65

Age,

gender,severity

ofperiodontal

disease,number

ofteeth,DURand

dentalplaque

40

38

60

NSDfor%

PPD�4mm

and�7mm

Chávarry et al


Copyrig

ht

by

N

otfor

Qu

in

tessence

Not

forPublication

1 and 2 DM (Oliver and Tervonen, 1993; Perssonet al, 2003; Lalla et al, 2004).

Adjustment for potential confoundersAll the citations had an inadequate score for theadjustment of potential confounders.

Statistical analysisIn 14 studies, the statistical analysis was consid-ered adequate; 3 type 1 DM (Hugoson et al, 1989,Thorstensson and Hugoson, 1993; Yucekal-Tunceret al, 2003), 10 type 2 DM (Nelson et al, 1990; Em-rich et al, 1991; Cutler et al, 1999; Tsai et al, 2002;Marugame et al, 2003; Saito et al, 2003; Lu andYang, 2004; Saito et al, 2004; Campus et al,2005; Mattout et al, 2006) and 1 type 1 and 2DM (Persson et al, 2003).

Sampling design/selection of control groupThe sample was drawn randomly in five studies ontype 1 DM (Rylander et al, 1986; Hugoson et al,1989; Harber et al, 1993; Pinson et al, 1995; Firatliet al, 1996), in 16 studies on type 2 DM (Nelsonet al, 1990; Morton et al, 1995; Kawamura et al,1998; Kurtis et al, 1999; Sandberg et al, 2000; Bu-lut et al, 2001; Yuan et al, 2001; Tsai et al, 2002;Zielinski et al, 2002; Marugame et al, 2003; Saitoet al, 2003; Güneri et al, 2004; Saito et al, 2004;Mansour and Abd-Al-Sada, 2005; Borges-Yáñezet al, 2006; Mattout et al, 2006) and in three stud-ies on type 1 and 2 DM (Oliver and Tervonen, 1993;Persson et al, 2003; Lalla et al, 2004).The exposed and unexposed groups were from

the same population in seven studies on type 1DM (Rylander et al, 1986; Hugoson et al, 1989; Pin-son et al, 1995; Sbordone et al, 1995; Thorstens-son et al, 1995; Guthmiller et al, 2001; Aren et al,2003), in 19 studies on type 2 DM (Nelson et al,1990; Emrich et al, 1991; Cherry-Peppers and Ship,1993; Kawamura et al, 1998, Cutler et al, 1999;Kurtis et al, 1999; Sandberg et al, 2000; Yuanet al, 2001; Tsai et al, 2002; Zielinski et al, 2002;Marugame et al, 2003; Saito et al, 2003; Ünlüet al, 2003; Güneri et al, 2004; Saito et al, 2004;Campus et al, 2005; Mansour and Abd-Al-Sada,2005; Borges-Yáñez et al, 2006; Mattout et al,2006) and in one study on type 1 and 2 DM (Perssonet al, 2003)The definition of exposed and unexposed groups

was not clearly explained in one study on type 1DM (Harber et al, 1993), in three studies on type2 DM (Morton et al, 1995; Saito et al, 2003; Mat-tout et al, 2006) and in four studies on type 1 andC

hristgau

etal(1998)�

Yes

Moderate-

to-advanced

periodontitis:

�6

teethwith

PPD�4mm

NorthCarolina

periodontalprobe

Total

N=40,

type

1D

n=7,

type

2D

n=13,

nDn=20

Median=

54.5

(30–67)

Median=

50.5

(30–66)

Unclear

Age,gender,

antibioticsuse,

numberofteeth,

smokingand

system

icdiseases

40

40

4NSDfor

median%

PPDandCAL

Unclear,notpresented;

D,diabetics;

nD,non-diabetics;

SES,socioeconomicstatus;DUR,diabeticduration;

SD,significant

differencesbetweendiabeticsandnon-diabetics;

NSD,non-significant

differencesbetweendiabeticsandnon-diabetics;Dtype

1,diabetes

type

1;Dtype

2,diabetes

type

2;PoorD,poorlycontrolleddiabetes.Studies

accordingtotype

ofDM:§type

1;�type

2;�type

1and2.

*Studies

with


Table4Ep

idem

iologica

lch

arac

teristicsof

long

itud

inal

stud

iesbe

twee

ndiab

etes

andpe

riod

ontaldise

ase(C

ontinu

ed)

Chávarry et al

Vol 7, No 2, 2009 121

Copyrig

ht

by

N

otfor

Qu

in

tessence

Not

forPublication

2 DM (Oliver and Tervonen, 1993; Yavuzyilmaz et al,1996; Alpagot et al, 2001; Lalla et al, 2004).In three studies on type 1 DM (Firatli et al, 1994;

Firatli et al, 1996; Aren et al, 2003) and in two stud-ies on type 1 and 2 DM (Persson et al, 2003; Lallaet al, 2004), the exposure was assessed in thesame manner between the groups. Of the 23 studieson type 2 DM, only six assessed the exposure in thesame manner between the groups (Morton et al,1995; Sandberg et al, 2000; Ünlü et al, 2003;Güneri et al, 2004; Borges-Yáñez et al, 2006; Mat-tout et al, 2006).In all studies on type 1 and 2 DM, the outcome

was assessed similarly between the groups, andthe objective outcome criteria that were establishedfor periodontal disease assessment to excludecases in the controls were similar for both thegroups. In two of the five studies on type 1 and 2DM, these criteria were considered unclear (Oliverand Tervonen, 1993; Alpagot et al, 2001).The objective exposure criteria that were applied

to the unexposed group to exclude the exposed inthe unexposed groups was used in five studies ontype 1 DM (Thorstensson and Hugoson, 1993; Firatliet al, 1996; Guthmiller et al, 2001; Aren et al, 2003;Ünlü et al, 2003), in 16 studies on type 2 DM (Nel-son et al, 1990; Emrich et al, 1991; Cherry-Peppersand Ship, 1993; Collin et al, 1998; Kawamura et al,1998; Cutler et al, 1999; Fontana et al, 1999; Kurtiset al, 1999; Yuan et al, 2001; Orbak et al, 2002;Tsai et al, 2002; Marugame et al, 2003; Saitoet al, 2004; Borges-Yáñez et al, 2006; Mattoutet al, 2006) and in one study on type 1 and 2 DM(Alpagot et al, 2001).

Response ratesThe response rates were reported in three studieson type 1 DM (Hugoson et al, 1989; Harber et al,1993; Thorstensson and Hugoson, 1993), in threestudies on type 2 DM (Sandberg et al, 2000; Yuanet al, 2001; Tsai et al, 2002) and in one study ontype 1 and 2 DM (Persson et al, 2003).

Response rates and reasons for non-responsesimilar between groupsOnly one study on type 1 DM reported similarresponse rates and reasons for non-responsebetween groups (Thorstensson and Hugoson,1993).

Blindness of outcome assessmentSix studies on type 1 DM (Thorstensson and Hugo-son, 1993; Pinson et al, 1995; Firatli et al, 1996;

Guthmiller et al, 2001; Aren et al, 2003; Yucekal-Tuncer et al, 2003), seven studies on type 2 DM(Nelson et al, 1990; Emrich et al, 1991; Mortonet al, 1995; Kawamura et al, 1998; Yuan et al,2001; Zielinski et al, 2002; Marugame et al,2003) and no studies on type 1 and 2 DM had theoutcome assessment blinded to the exposurestatus.

Blindness of exposure assessmentFour studies on type 1 DM (Pinson et al, 1995; Firatliet al, 1996; Guthmiller et al, 2001; Aren et al,2003), seven studies on type 2 DM (Nelson et al,1990; Emrich et al, 1991; Morton et al, 1995; Ka-wamura et al, 1998; Yuan et al, 2001; Zielinskiet al, 2002; Marugame et al, 2003) and one studyon type 1 and 2 DM (Oliver and Tervonen, 1993)had the exposure assessment blinded to outcomestatus.

Meta-analysis of cross-sectional studies

Meta-analyses were conducted involving studies inwhich the mean differences of PPD and CAL werereported or could be extracted. The mean differ-ences for PPD and CAL measures between groupswere calculated as the mean value in diabetic groupminus the mean value in non-diabetic group. Positivedifferences between diabetics and non-diabetics forPPD and CAL measures revealed more periodontaldisease in the diabetic groups.The meta-analysis was performed according to the

type of diabetes (Fig 2). For the calculation of thepooled mean difference, each study was assigneda weight consisting of the reciprocal of variance ofdifference between two mean values. Estimates ofthe overall difference in each of the periodontalparameters and their corresponding CI 95% were cal-culated using the random-effect model.Type 1 diabetics were compared with non-diabet-

ics with respect to CAL measures in eight studies(Manouchehr-Pour et al, 1981; Alley et al, 1993; Pin-son et al, 1995; Firatli et al, 1996; Alpagot et al,2001; Guthmiller et al, 2001; Ünlü et al, 2003;Yucekal-Tuncer et al, 2003). The overall differenceof CAL between type 1 diabetics and non-diabeticswas not statistically significant (CAL mean differ-ence = 0.26 (�0.004 to 0.533), P = 0.054) (Fig2.1a). Type 2 diabetics were compared with non-diabetics with respect to CAL measures in threestudies (Morton et al, 1995; Alpagot et al, 2001;Yuan et al, 2001). There were statistically significant

Chávarry et al


Copyrig

ht

by

N

otfor

Qu

in

tessence

Not

forPublication

differences for CAL measures between type 2 diabet-ics and non-diabetics (mean difference = 1.00 (0.15to 1.84), P = 0.021) (Fig 2.1b).The average differences in PPD were compared

between type 1 diabetics and non-diabetics in tenstudies (Alley et al, 1993; Pinson et al, 1995; Firatliet al, 1996; Firatli, 1997; Salvi et al, 1997; Alpagotet al, 2001; Guthmiller et al, 2001; Aren et al, 2003;Ünlü et al, 2003; Yucekal-Tuncer et al, 2003), andtype 2 diabetics and non-diabetics in five studies(Morton et al, 1995; Alpagot et al, 2001; Bulutet al, 2001; Yuan et al, 2001; Orbak et al, 2002).The overall difference in the average of PPDbetween type 1 diabetics and non-diabetics wasnot statistically different (mean difference = 0.11(�0.03 to 0.245), P = 0.137) (Fig 2.2a). The aver-age difference of PPD between type 2 diabeticsand non-diabetics was statistically different (meandifference = 0.46 (0.01 to 0.91, P = 0.046)(Fig 2.2b).Significant statistical heterogeneity was not found

across studies in all analyses: type 1 DM for meanof CAL: Q = 11.526, P = 0.17; type 2 DM for meanof CAL: Q = 2.416, P = 0.2990; type 1 DM for meanof PPD: Q = 7.220, P = 0.614 and type 2 DM formean of PPD: Q = 5.147, P = 0.273.

Publication bias

The results of Begg’s test with continuity correctionand Egger’s test confirmed that there was no evi-dence of publication bias. The results of the Begg’stest for type 1 DM for mean CAL: P = 0.754 and type2 DM for mean CAL: P = 1.000; type 1 DM for meanof PPD: P = 0.371 and type 2 DM for mean of PPD:P = 0.806. The results of Egger’s test for type 1DM for mean CAL: coefficient = 2.52, P = 0.015;type 2 DM for mean CAL: coefficient = �1.30,P = 0.719; type 1 DM for mean of PPD: coeffi-cient = 2.85, P = 0.220 and type 2 DM for meanof PPD: coefficient = 0.99, P = 0.640. Egger’s publi-cation bias plots were used to graphically show noevidence of publication bias (Fig 3). The values ofmean differences were plotted versus variancedifferences.

Longitudinal studies

The epidemiological characteristics of the eight longi-tudinal studies included in this review are describedin Table 4. Two papers were considered to be one

study, as they reported clinical and microbiologicalfindings of the same sample in separate papers(Novaes et al, 1996, 1997). No clinical trial wasidentified in the literature search, as no study ana-lysed the effect of glycaemic control on periodontalstatus. Of the eight selected longitudinal studies,four compared the progression of periodontal dis-ease between diabetics and non-diabetics and fourstudies investigated the effect of diabetes on theresponse to periodontal treatment (Table 4).The duration of diabetes was reported in two stud-

ies on type 1 and 2 DM (Westfelt et al, 1996; Christ-gau et al, 1998) and in one study on type 1 DM(Tervonen and Karjalainen, 1997). The duration ofdiabetes varied from 11 to 16.9 years. Three differ-ent types of periodontal probes were used. In gen-eral, few confounders were considered. Thenumbers ranged from 2 to 6 and the length of thestudies varied from 4 to 60 months.

Type 1 DMOf the three longitudinal studies on type 1 DM, twostudies compared periodontal disease progression(Firatli, 1997; Sbordone et al, 1998), and in onestudy the progression of periodontal disease wasmore pronounced in diabetics (Firatli, 1997). Inone study that investigated the relationship betweentype 1 DM and the response to periodontal therapy,no significant difference in the reduction of periodon-tal measures was observed between diabetics andnon-diabetics (Tervonen and Karjalainen, 1997).

Type 2 DMThree studies were conducted on type 2 DM and onthe progression of periodontal disease (Novaeset al, 1996, 1997; Taylor et al, 1998; Faria-Almeidaet al, 2006). One study found a significant increaseof CAL in diabetics compared with non-diabetics(Novaes et al, 1996). The other study showed thatthe progression of alveolar bone loss over timewas higher in diabetics compared with non-diabetics.The odds ratio was 4.2 (1.8 to 9.9) (Taylor et al,1998). The study that analysed the effect of diabe-tes on periodontal treatment showed a significantincrease of PPD in diabetics compared with non-dia-betics after 6 months of follow up (Faria-Almeidaet al, 2006).

Type 1 and 2 DMTwo studies included subjects with type 1 and 2 DMand analysed the effect of DM on the response toperiodontal therapy (Westfelt et al, 1996; Christgau

Chávarry et al

Vol 7, No 2, 2009 123

Copyrig

ht

by

N

otfor

Qu

in

tessence

Not

forPublication

et al, 1998). No differences were reported in clinicalperiodontal measures between diabetics and non-diabetics after 4 months (Christgau et al, 1998)and after 5 years (Westfelt et al, 1996).

Quality assessment: results of eightlongitudinal studies

Sampling calculationNone of the eight studies determined the samplesize before the study.

Were the examiners trained and calibrated for

periodontal assessment?Of the eight studies, three had examiners who weretrained and calibrated for periodontal assessment(Firatli, 1997; Tervonen and Karjalainen, 1997; Tay-lor et al, 1998).

Adjustment for potential confoundersAll the citations received an inadequate score foradjustment of potential confounders.

Statistical analysisStatistical analysis was scored adequate in onestudy (Taylor et al, 1998).

Sampling design/selection of control groupIn one study, the sample was drawn randomly (Tayloret al, 1998). In five studies, the exposed and unex-posed groups came from the same population (Nov-aes et al, 1996, 1997; Tervonen and Karjalainen,1997; Christgau et al, 1998; Sbordone et al,1998; Taylor et al, 1998). The definition of exposedand unexposed groups was explicit in four citations(Westfelt et al, 1996; Christgau et al, 1998; Sbor-done et al, 1998; Taylor et al, 1998). The assess-ment of exposure was similar between the groupsin one study (Novaes et al, 1996, 1997). The expo-sure was assessed in the same manner in the samegroups in another study (Novaes et al, 1996, 1997).In all longitudinal studies, the outcome wasassessed in the same manner between the groups.The objective outcome criteria that were establishedfor assessed periodontitis for excluded cases in thecontrols were similar for both groups in all citations.Two studies (Novaes et al, 1996, 1997; Taylor et al,1998) had the objective exposure criteria applied tothe unexposed group to exclude exposed in the unex-posed groups.

Response rateFour studies reported the response rate (Firatli,1997; Tervonen and Karjalainen, 1997; Christgauet al, 1998; Taylor et al, 1998).

Blindness of outcome assessmentIn two studies, the examiners who conducted theperiodontal assessment were blinded to the diabeticstatus of the subjects (Tervonen and Karjalainen,1997; Taylor et al, 1998).

Blindness of exposure assessmentIn four studies, the examiners who tested the levelof glycaemic control were blinded for the periodontalstatus of the subjects (Firatli, 1997; Tervonen andKarjalainen, 1997; Sbordone et al, 1998; Tayloret al, 1998).

Updated exposureThe exposure was updated in seven studies (Novaeset al, 1996, 1997; Westfelt et al, 1996; Firatli,1997; Tervonen and Karjalainen, 1997; Sbordoneet al, 1998, Taylor et al, 1998, Faria-Almeida et al,2006).

Dropout ratesTwo studies reported dropouts (Westfelt et al, 1996;Tervonen and Karjalainen, 1997).

Dose–effect relationshipOne study reported dose–effect findings (Taylor et al,1998).

DISCUSSION

This meta-analysis of 57 peer-reviewed studiesleads to the conclusion that type 2 DM is a risk fac-tor for periodontitis. The overall difference for CAL of1.0 mm (CI 95%: 0.15 to 1.84, P < 0.021) was sig-nificant. In addition, longitudinal studies revealedsignificantly more progression of periodontal diseaseamong type 2 diabetics compared with non-diabet-ics. However, the evidence for an associationbetween type 1 DM and periodontitis is insufficient.Pooled differences of clinical measures of periodon-tal disease did not differ significantly between type 1diabetics and non-diabetics, and longitudinal studiesshowed conflicting findings.The lack of association between type 1 DM and

periodontal disease can be explained by the lowmean age of the subjects, namely, between 11

Chávarry et al


Copyrig

ht

by

N

otfor

Qu

in

tessence

Not

forPublication

and 15 years. Even diabetic subjects in this agerange do not frequently develop destructive periodon-tal disease. As the prevalence of destructive peri-odontal disease would be apparent after the age of25 to 30 years (Albandar, 2002a, b), future studieson the relationship between type 1 DM and periodon-tal disease should select adult subjects.Most of the previous review papers on the associ-

ation between DM and periodontal disease did notuse an appropriate protocol for conducting system-atic reviews (Belting et al, 1964; Campbell, 1967;Hugoson and Jordan, 1982; Petersen and Ogawa,2005; Mealey and Oates, 2006; Tan et al, 2006).The limitations in such reviews include the lack ofinformation about the strategies used for searchingthe papers, the criteria used to select the papersand how the selection process and data extractionwere performed. Furthermore, the methodologicalquality of epidemiological studies was not taken intoaccount; for example, in this review, studies in whichthe type of diabetes were not reported and studieswith inadequate measures of periodontal diseasewere excluded. Applying those criteria resulted inthe exclusion of 72 papers and decreased the num-ber of studies considerably. However, the studieswith potential bias of exposure and outcome assess-ment might have been included in other reviews.Only one systematic review on the relationship

between diabetes and periodontal disease has beenconducted using quantitative data analysis (Khaderet al, 2006). Several features distinguish theauthors’ study from the previous meta-analysis byKhader et al (2006). First, the authors’ included farmore studies (57 versus 18). This occurred becausethe meta-analysis was conducted 4 years later andthe search strategy was not restricted to one data-base (MEDLINE). Second, the studies were com-bined according to the type of DM. In Khader et al(2006), the estimates of differences in meta-analy-sis were summarised for type 1 and 2 DM together.As type 1 and 2 DM are heterogeneous conditions,the conflation of the results generates unjustifiedassociations. Third, in this study clear criteria wereused to select studies with adequate assessmentof diabetes and periodontal disease, avoiding theinclusion of poor quality studies. Fourth, potentialpublication bias was assessed through Egger’s testand Begg’s test. There was no evidence of publica-tion bias. Finally, the quality assessment that wasconducted provided an adequate evaluation of themethodology of the selected studies.One limitation of this study is the use of observa-

tional studies in the meta-analysis as randomisedcontrolled trials (RCTs) were not identified. Although

RCTs provide the highest level of evidence, dueto ethical and logistical limitations, they cannotalways be performed. As no RCT was identified inthe authors’ search, observational studies wereincluded in this meta-analysis.The use of pooled estimates from observational

studies is controversial because they do not reachthe level of confidence of meta-analyses of random-ised controlled trials. The result of the meta-analysisin the present review can be considered appropriatebecause some precautions were used to reduce themethodological variability among studies that is fre-quently present in observational studies. Another lim-itation in this review is the inclusionof studieswith fewconfounders taken into account and the variability inthe sampling design and selection of control group.Although thedurationofdiabeteshasbeen reported

in some studies, relevant characteristics of DM suchas medical treatment, body mass index and systemiccomplications were not reported. Future studiesshould collect these data using accurate methods.Most of the reviews on diabetes and periodontal

disease supported the conclusions based on thePima Indian population study, an indigenous commu-nity with a high propensity to diabetes (Nelson et al,1990; Emrich et al, 1991). The conclusions basedon the Pima Indian population are relevant becausethe model minimised the potential confounders. Inaddition, the criteria used for periodontal diseaseassessment was based on alveolar bone loss thatwasassessed radiographically,which isa very reason-able method to detect periodontal disease. However,studies conducted on general populations are alsoimportant because of the differences in their genetic,environmental and behavioural characteristics.In conclusion, there is enough evidence to con-

sider type 2 DM as a risk factor for destructiveperiodontal disease. More studies are needed toassess if type 1 DM is a true risk factor for periodon-tal disease. In future studies on type 1 DM the inclu-sion of population aged 25 years and above isrecommended.

REFERENCES

1. Albandar JM. Periodontal diseases in North America.Periodontol 2000 2002a;29:31–69.

2. Albandar JM. Global risk factors and risk indicators forperiodontal diseases. Periodontol 2000 2002b;29:77–206.

3. Alderson P, Green S, Higgins JPT (eds). Cochrane review-ers’ handbook 4.2.2 (updated March 2004). In: TheCochrane Library, Issue 1. Chichester, UK: John Wiley &Sons, Ltd, 2004.

Chávarry et al

Vol 7, No 2, 2009 125

Copyrig

ht

by

N

otfor

Qu

in

tessence

Not

forPublication

4. Alley CS, Reinhardt RA, Maze CA, DuBois LM, Wahl TO,Duckworth WC et al. HLA-D and T lymphocyte reactivity tospecific periodontal pathogens in type 1 diabetic periodon-titis. J Periodontol 1993;64:974–979.

5. Alpagot T, Silverman S, Lundergan W, Bell C, ChambersDW. Crevicular fluid elastase levels in relation to periodon-titis and metabolic control of diabetes. J Periodontal Res2001;36:169–174.

6. Aren G, Sepet E, Özdemir D, Dinççag N, Güvener B, Firatli E.Periodontal health, salivary status and metabolic control inchildren with type 1 diabetes mellitus. J Periodontol2003;74:1789–1795.

7. Begg CB, Berlin JA. Operating characteristics of a rankcorrelation test for publication bias. Biometrics1994;50:1088–1101.

8. Belting CM, Hiniker JJ, Dummett CO. Influence of diabetesmellitus on the severity of periodontal disease. J Period-ontol 1964;35:476–480.

9. Borges-Yáñez SA, Irigoyen-Camacho ME, Maupome G. Riskfactors and prevalence of periodontitis in community-dwelling elders in Mexico. J Clin Periodontol 2006;33:184–194.

10. Bulut S, Develioglu H, Taner IL, Berker E. Interleukin-1 betalevels in gingival crevicular fluid in type 2 diabetes mellitusand adult periodontitis. J Oral Sci 2001;43:161–177.

11. Campbell MJ. Periodontal disease in the diabetic patientand its treatment. Aust Dent J 1967;12:117–122.

12. Campus G, Salem A, Uzzau S, Baldoni E, Tonolo G.Diabetes and periodontal disease: a case-control study.J Periodontol 2005;76:418–425.

13. Cherry-Peppers G, Ship JA. Oral health in patients with typeII diabetes and impaired glucose tolerance. Diabetes Care1993;16:638–641.

14. Christersson LA. Actinobacillus actinomycetemcomitansand localized juvenile periodontitis, clinical, microbio-logic and histologic studies. Swed Dent J Suppl 1993;90:6–46.

15. Christgau M, Palitzsch K-D, Schmalz G, Kreiner U, Frenzel S.Healing response to non-surgical periodontal therapyin patients with diabetes mellitus: clinical, microbiological,and immunologic results. J Clin Periodontol 1998;25:112–124.

16. Collin HL, Sorsa T, Meurman JH, Niskanen L, Salo T, RonkaH et al. Salivary matrix metalloproteinase (MMP-8)levels and gelatinase (MMP-9) activities in patients withtype 2 diabetes mellitus. J Periodontal Res 2000;35:259–265.

17. Collin H-L, Uusitupa M, Niskanen L, Kontturi Närhi V,Markkanen H, Koivisto A-M et al. Periodontal findings inelderly patients with non-insulin dependent diabetes mel-litus. J Periodontol 1998;69:962–966.

18. Cutler CW, Machen RL, Jotwani R, Iacopino AM. Heightenedgingival inflammation and attachment loss in type 2diabetics with hyperlipidemia. J Periodontol 1999;70:1313–1321.

19. De Pommereau V, Dargent-Paré C, Robert JJ, Brion M.Periodontal status in insulin-dependent diabetic adoles-cents. J Clin Periodontol 1992;19:628–632.

20. Egger M, Smith GD, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997;315:629–634.

21. Emrich LJ, Shlossman M, Genco RJ. Periodontal disease innon-insulin-dependent diabetes mellitus. J Periodontol1991;62:123–130.

22. Faria-Almeida R, Navarro A, Bascones A. Clinical andmetabolic changes after conventional treatment of type 2diabetic patient with chronic periodontitis. J Periodontol2006;77:591–598.

23. Firatli E. The relationship between clinical periodontalstatus and insulin-dependent diabetes mellitus. Resultsafter 5 years. J Periodontol 1997;68:136–140.

24. Firatli E, Ünal T, Saka N, Onan U, Sivas A, Öz H. Serumfructosamine correlates with gingival index in children withinsulin-dependent diabetes mellitus (IDDM). J Clin Period-ontol 1994;21:565–568.

25. Firatli E, Yilmaz O, Onan U. The relationship betweenclinical attachment loss and the duration of insulin-depen-dent diabetes mellitus (IDDM) in children and adolescents.J Clin Periodontol 1996;23:362–366.

26. Fontana G, Lapolla A, Sanzari M, Piva E, Mussap M, DeToni S et al. An immunological evaluation of type II diabeticpatients with periodontal disease. J Diabetes Complica-tions 1999;13:23–30.

27. Güneri P, Unlu F, Yesilbek B, Bayraktar F, Kokuludag A,Hekimgil M et al. Vascular endothelial growth factor ingingival tissues and crevicular fluids of diabetic and healthyperiodontal patients. J Periodontol 2004;75:91–97.

28. Guthmiller JM, Hassebroek-Johnson JR, Weenig DR, John-son GK, Kirchner HL, Kohout FJ et al. Periodontal diseasein pregnancy complicated by type 1 diabetes mellitus. JPeriodontol 2001;72:1485–1490.

29. Harber J, Wattles J, Crowley M, Mandell R, Joshipura K,Kent RL. Evidence for cigarette smoking as a major riskfactor for periodontitis. J Periodontol 1993;64:16–23.

30. Hugoson A, Jordan T. Frequency distribution of individualsaged 20–70 years according to severity of periodontaldisease. Community Dent Oral Epidemiol 1982;10:187–192.

31. Hugoson A, Thorstensson H, Falk H, Kuylenstierna J.Periodontal conditions in insulin-dependent diabetics.J Clin Periodontol 1989;16:215–223.

32. Kawamura M, Fukuda S, Kawabata K, Iwamoto Y. Compar-ison of health behaviour and oral/medical conditions innon-insulin-dependent (type II) diabetics and non-diabetics.Aust Dent J 1998;43:315–320.

33. Khader YS, Kauod AS, El-Qaderi SS, Alkafajei A, BatayhaWQ. Periodontal status of diabetics compared with non-diabetics: a meta-analysis. J Diabetes Complications2006;20:59–68.

34. Kurtis B, Develioglu H, Taner IL, Balos K, Tekin IO. IL-6 levelsin gingival crevicular fluid (GCF) from patients with non-insulin dependent diabetes mellitus (NIDDM), adult peri-odontitis and healthy subjects. J Oral Sci 1999;41:63–67.

35. Lalla E, Cheng B, Lal S, Kaplan S, Softness B, Greenberg Eet al. Diabetes mellitus promotes periodontal destructionin children. J Clin Periodontol 2007;34:294–298.

36. Lalla E, Park DB, Papapanou PN, Lamster IB. Oral diseaseburden in Northern Manhattan patients with diabetesmellitus. Am J Public Health 2004;94:755–758.

37. Lu HK, Yang PC. Cross-sectional analysis of differentvariables of patients with non-insulin dependent diabetesand their periodontal status. Int J Periodontics RestorativeDent 2004;24:71–79.

38. Manouchehr-Pour M, Spagnuolo PJ, Rodman HM, BissadaNF. Comparison of neutrophil chemotactic response indiabetic patients with mild and severe periodontal disease.J Periodontol 1981;52:410–415.

Chávarry et al


Copyrig

ht

by

N

otfor

Qu

in

tessence

Not

forPublication

39. Mansour AA, Abd-Al-Sada N. Periodontal disease amongdiabetics in Iraq. Med Gen Med 2005;7:2.

40. Marugame T, Hayasaki H, Lee K, Eguchi H, Matsumoto S.Alveolar bone loss associated with glucose tolerance inJapanese men. Diabet Med 2003;20:746–751.

41. Mattout C, Bourgeois D, Bouchard P. Type 2 diabetes andperiodontal indicators: epidemiology in France 2002–2003. J Periodontal Res 2006;41:253–258.

42. Mealey BL, Oates TW. Diabetes mellitus and periodontaldisease. J Periodontol 2006;77:1289–1303.

43. Moher D, Schulz KF, Altman D. CONSORT group (Consol-idated Standards of Reporting Trials). The CONSORTstatement: revised recommendations for improving thequality of reports of parallel-group randomized trials. JAMA2001;285:1987–1991.

44. Morton AA, Williams RW, Watts LP. Initial study ofperiodontal status in non-insulin-dependent diabetics inMauritius. J Dent 1995;23:343–345.

45. Nelson RG, Shlossman M, Budding LM, Pettitt DJ, SaadMF, Genco RJ et al. Periodontal disease and NIDDM inPima Indians. Diabetes Care 1990;13:836–840.

46. Novaes AB Jr, Gutierrez FG, de Moraes Grisi MF, NovaesAB. Periodontal disease progression in type II non-insulin-dependent diabetes mellitus patients (NIDDM). Part II –Microbiological analysis using the bana test. Braz Dent J1997;8:27–33.

47. Novaes AB Jr, Gutierrez FG, Novaes AB. Periodontaldisease progression in type II non-insulin-dependent dia-betes mellitus patients (NIDDM). Part I – Probing pocketdepth and clinical attachment. Braz Dent J 1996;7:65–73.

48. Obeid P, Bercy P. Effects of smoking on periodontal healtha review. Adv Ther 2000;17:230–237.

49. Oliver RC, Tervonen T. Periodontitis and tooth loss. J AmDent Assoc 1993;124:71–75.

50. Orbak R, Tezel A, Çanakçi V, Demir T. The influence ofsmoking and non-insulin-dependent diabetes mellitus onperiodontal disease. J Int Med Res 2002;30:116–125.

51. Persson RE, Hollender LF, MacEntee MI, Wyatt CCL, KiyakGA, Persson GR. Assessment of periodontal conditions andsystemic disease in older subjects. J Clin Periodontol2003;30:207–213.

52. Petersen PE, Ogawa H. Strengthening the prevention ofperiodontal disease: the WHO approach. J Periodontol2005;76:2187–2193.

53. Pinson M, Hoffman WH, Garnick JJ, Litaker MS. Periodontaldisease and type I diabetes mellitus in children andadolescents. J Clin Periodontol 1995;22:118–123.

54. Rylander H, Ramberg P, Blohmé G, Lindhe J. Prevalence ofperiodontal disease in young diabetics. J Clin Periodontol1986;14:38–43.

55. Saito T, Murakami M, Shimazaki Y, Oobayashi K, Matsum-oto S, Koga T. Association between alveolar bone loss andelevated serum C-reactive protein in Japanese men. JPeriodontol 2003;74:1741–1746.

56. Saito T, Shimazaki Y, Kiyohara Y, Kato I, Kubo M, Lida Met al. The severity of periodontal disease is associated withthe development of glucose intolerance in non-diabetics:the Hisayama study. J Dent Res 2004;83:485–490.

57. Salvi GE, Yalda B, Collins JG, Jones BH, Smith FW, Arnold RRet al. Inflammatory mediator response as a potential riskmarker for periodontal diseases in insulin-dependent diabe-tes mellitus patients. J Periodontol 1997;68:127–135.

58. Sandberg GE, Sundberg HE, Fjellstrom CA, Wilkblad KF.Type 2 diabetes and oral health. A comparison betweendiabetic and non-diabetic subjects. Diabetes Res Clin Pract2000;50:27–34.

59. Sandholm L, Swanljung O, Rytömaa I, Kaprio EA, MäenpääJ. Periodontal status of Finnish adolescents with insulin-dependent diabetes mellitus. J Clin Periodontol 1989;16:617–620.

60. Sbordone L, Ramaglia L, Barone A, Ciaglia RN, Iacono VJ.Periodontal status and subgingival microbiota of insulin-dependent juvenile diabetics: a 3-year longitudinal study. JPeriodontol 1995;69:120–128.

61. Sbordone L, Ramaglia L, Barone A, Ciaglia RN, Tenore A,Iacono VJ. Periodontal status and selected cultivableanaerobic microflora of insulin-dependent juvenile diabet-ics. J Periodontol 1998;66:452–461.

62. Sheiham A, Netuveli GS. Periodontal diseases in Europe.Periodontol 2000 2002;29:104–121.

63. Tan WC, Tay FB, Lim LP. Diabetes as a risk factor forperiodontal disease: current status and future consider-ations. Ann Acad Med Singapore 2006;35:571–581.

64. Taylor GW, Burt BA, Becker MP, Genco RJ, Sholssman M,Knowler WC et al. Non-insulin dependent diabetes mellitusand alveolar bone loss progression over 2 years. J Period-ontol 1998;69:76–83.

65. Tervonen T, Karjalainen K. Periodotal disease related todiabetic status a pilot study of the response to periodontaltherapy in type 1 diabetes. J Clin Periodontol1997;24:505–510.

66. Thorstensson H, Dahlén G, Hugoson A. Some suspectedperiodontopathogens and serum antibody response inadult long-duration insulin-dependent diabetics. J ClinPeriodontol 1995;22:449–458.

67. Thorstensson H, Hugoson A. Periodontal disease experi-ence in adult long-duration insulin-dependent diabetics. JClin Periodontol 1993;20:352–358.

68. Tsai C, Hayes C, Taylor GW. Glycemic control of type 2diabetes and severe periodontal disease in the US adultpopulation. Community Dent Oral Epidemiol 2002;30:182–192.

69. Ünlü F, Güneri PG, Hekimgil M, Yesilbek B, Boyacioglu H.Expression of vascular endothelial growth factor in humanperiodontal tissues: comparison of healthy and diabeticpatients. J Periodontol 2003;74:181–187.

70. Vettore MV, Leao AT, Monteiro da Silva AM, Quintanilha RS,Lamarca GA. The relationship of stress and anxiety withchronic periodontitis. J Clin Periodontol 2003;30:394–402.

71. Westfelt E, Rylander G, Blohmé, Jonasson P, Lindhe J. Theeffect of periodontal therapy in diabetics results after5 years. J Clin Periodontol 1996;23:92–100.

72. Yavuzyilmaz E, Yumak Ö, Akdoganli T, Yamalik N, Özer N,Ersoy F et al. The alterations of whole saliva constituents inpatients with diabetes mellitus. Aust Dent J 1996;41:193–197.

73. Yucekal-Tuncer B, Uygur C, Firatli E. Gingival crevicular fluidlevels of aspartate amino transferase, sulfide ions and N-benzoyl-DL-arginine-2-naphthylamide in diabetic patientswith chronic periodontitis. J Clin Periodontol 2003;30:1053–1060.

74. Yuan K, Chang CJ, Hsu PC, Sun HS, Tseng CC, Wang JR.Detection of putative periodontal pathogens in non-insulin-dependent diabetes mellitus and non-diabetes mellitus bypolymerase chain reaction. J Periodontal Res 2001;36:18–24.

75. Zielinski MB, Fedele D, Forman LJ, Pomerantz SC. Oralhealth in the elderly with non-insulin-dependent diabetesmellitus. Spec Care Dentist 2002;22:94–98.

Chávarry et al

Vol 7, No 2, 2009 127

ohpd_2009_02_s0107.pdf - Quintessence Publishing!

Documents

Transcript of ohpd_2009_02_s0107.pdf - Quintessence Publishing!