UBC Faculty Research and Publications

Periodontal disease and spontaneous preterm birth: a case control study Wood, Stephen; Frydman, Albert; Cox, Stephen; Brant, Rollin; Needoba, Sheilia; Eley, Barry; Sauve, Reg Jul 19, 2006

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-12884_2006_Article_91.pdf [ 825.78kB ]
JSON: 52383-1.0223537.json
JSON-LD: 52383-1.0223537-ld.json
RDF/XML (Pretty): 52383-1.0223537-rdf.xml
RDF/JSON: 52383-1.0223537-rdf.json
Turtle: 52383-1.0223537-turtle.txt
N-Triples: 52383-1.0223537-rdf-ntriples.txt
Original Record: 52383-1.0223537-source.json
Full Text

Full Text

ralssBioMed CentBMC Pregnancy and ChildbirthOpen AcceResearch articlePeriodontal disease and spontaneous preterm birth: a case control studyStephen Wood*1, Albert Frydman2, Stephen Cox3, Rollin Brant4, Sheilia Needoba5, Barry Eley3 and Reg Sauve6Address: 1Department of Obstetrics and Gynecology, Foothills Hospital, 1403 29th St. NW., Calgary, Alberta, T2N 2T9, Canada, 2Suite 403 4600, Crowchild Trail NW, Calgary Alberta, T3A 2L6, Canada, 3Department of Periodontology, King's College, London Dental Institute at Guy's, King's College and St Thomas' Hospitals, London, SE1 9RT, UK, 4Center for Community Child Health Research. 4480 Oak Street, L408 Vancouver, BC, V6H 3V4, Canada, 546 Waterloo Dr SW. Calgary, Alberta, T3C 3E8, Canada and 6Department of Community Health Sciences, Health Science Center 3330 Hospital Dr. NW. Calgary, Alberta, T2N 4N1, CanadaEmail: Stephen Wood* - stephen.wood@calgaryhealthregion.ca; Albert Frydman - info@periosupport.ca; Stephen Cox - stephen.cox@kcl.ac.uk; Rollin Brant - rollin@stat.ubc.ca; Sheilia Needoba - needoba@shaw.ca; Barry Eley - stephen.cox@kcl.ac.uk; Reg Sauve - rsauve@ucalgary.ca* Corresponding author    AbstractBackground: Several studies have suggested an association between periodontal disease andprematurity but this finding has not been consistently observed.Methods: Case control study. Cases (n = 50) were women who had delivered after spontaneouspreterm labor at <35 weeks gestation. Two groups of controls (n = 101) were recruited: womenwho were undelivered but at a preterm gestation and women who delivered at term. A standard,clinical, periodontal examination was performed and gingival crevicular fluid was obtained fromstandardized locations and tested for neutrophil elastase along with the bacterial enzymes gingipainand dipeptidylpeptidase. Data were analyzed with Fisher's exact tests, ANOVA and multivariatelogistic regression.Results: There was no difference in the proportion of sites with significant attachment loss (≥3mm): Cases-3.2%, Controls-2.2% p = 0.21. The gingival crevicular fluid concentrations of elastaseand gingipain were elevated in cases vs. controls 238.8 uU/ul vs. 159.6 uU/ul p = .007 and 2.70 uU/ul vs. 1.56 uU/ul p = .001. On multivariate analysis, the mean log concentration of elastase, but notof gingipain, remained a significant predictor of preterm labor p = .0.015.Conclusion: We found no evidence that clinical periodontal disease is associated withspontaneous preterm birth. Elevated gingival crevicular fluid levels of elastase were associated withpreterm birth but further research is needed before this can be assumed to be a causal relationship.BackgroundPreterm birth remains the most important cause of peri-natal mortality and morbidity. Despite considerableable. Several, recent studies have suggested a relationshipbetween preterm delivery and periodontal disease. Offen-bacher et. al. reported a strong association, OR = 7.9,Published: 19 July 2006BMC Pregnancy and Childbirth 2006, 6:24 doi:10.1186/1471-2393-6-24Received: 30 March 2006Accepted: 19 July 2006This article is available from: http://www.biomedcentral.com/1471-2393/6/24© 2006 Wood et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Page 1 of 8(page number not for citation purposes)research, the pathogenesis of preterm delivery is not wellunderstood and no effective preventative therapy is avail-between periodontal disease and adverse pregnancy out-come defined as either premature delivery or low birthBMC Pregnancy and Childbirth 2006, 6:24 http://www.biomedcentral.com/1471-2393/6/24weight in women in North Carolina [1]. Similar resultswere reported in studies of other U.S. populations by Jeff-coat et. al. and in further work by Offenbacher et al [2,3].However, this association has not been consistentlyobserved. In fact, Davenport et. al. found that the risk ofhaving a preterm infant was actually reduced in womenwith periodontitis in the United Kingdom [4].One concern with these studies is that confounding, espe-cially by socioeconomic status, has not been consistentlycontrolled for. Socioeconomic status has been associatedconsistently with spontaneous preterm birth and may alsobe associated with poor dental care, particularly in juris-dictions where it is not publicly funded. As well, no stud-ies of which we are aware, related to spontaneouspremature birth, have evaluated biochemical measures ofactive periodontal disease in addition to standard clinicalexamination. Since periodontal disease is characterized bya relapsing/remitting pattern, identifying active diseasemay be an important factor in establishing associationswith other disease states such as preterm labor. The popu-lation from which the subjects were recruited is also highrisk for premature delivery. The city where this study wasperformed has one of the highest rates of preterm birth inCanada with 10% of births in 2004 occurring before 37weeks (unpublished local data). The population is pre-dominately Caucasian (81%) with Chinese (6%), SouthAsian (4%), Aboriginal (2%) and Black (1.5%) compris-ing the other main ethnic groups [5]. In addition, and incontrast to the studies published to date, our investigationrestricted the outcome to only spontaneous preterm birth<35 weeks gestation.Periodontitis is an inflammatory process initiated by bac-terial plaque involving the supporting structures of thetooth which include the gingiva, the junctional epithe-lium, root cementum, periodontal ligament, and alveolarbone[6]. These structures are responsible for maintainingthe attachment of the teeth to the upper and lower jaws.Their destruction leads to loss of attachment between thetooth and the alveolar bone and ultimately, to excessmobility, infection, and loss of the tooth. Periodontitis isdiagnosed clinically by measuring a deepening of thespace (pocket) between the root of the tooth and the gin-gival tissue (Figure 1). Attachment loss is an accuratemeasure of disease severity and is defined as the distance,in mm, between the cemento-enamel junction and thebase of the periodontal pocket. As periodontitis is epi-sodic in nature, probing alone cannot determine whetherthe disease is active or quiescent. Several tests have beendeveloped to assess substances in the gingival crevicularfluid (GCF) which can be obtained from periodontalpockets. One of these tests, which measures neutrophilsites with high neutrophil elastase levels are at signifi-cantly greater risk for progressive bone loss over the next6 months [7]. Using the same elastase substrate in fullyquantitative assays, two of the present authors (S.C. andB.E.), in a two year longitudinal study, found that enzymeactivity above a critical value had very high sensitivity andspecificity for future attachment loss [8]. Further testsbased on the two bacterial enzymes dipeptidylpeptidaseand gingipain have proved to be of almost equal diagnos-tic value [9,10].This case control study was designed to assess the possiblerelationship between periodontal disease and spontane-ous preterm delivery using both clinical examination andthe assessment of neutrophil elastase, bacterial gingipainand dipeptidylpeptidase in gingival crevicular fluid.MethodsCases were women with singleton pregnancies who deliv-ered at or before 35 weeks gestation, either after spontane-ous labor or induction of labor for preterm prematurerupture of membranes. The subjects all had access to uni-versally funded prenatal care but not to publicly fundeddental care. Two groups of controls were recruited: agroup of postpartum women who delivered at term andthe other, undelivered women who were assessedbetween 22 and 35 weeks gestation. Women wererecruited in one hospital following delivery and in thehospital antenatal clinics.Consenting subjects had a full periodontal examinationin a standard, well- equipped, dental office maintained byone of the co-investigators, an experienced periodontist(AF). The dental examination was performed between 2days and 28 days following delivery for the post partumsubjects. Clinical examination was carried out by onehighly experienced periodontal hygienist (S.N.). Theintra-person variability in clinical examination by thehygienist was assessed by one of the investigators (AF) ina pilot period, using five volunteers, and was found to beacceptably low (mean .34 mm SD .5 mm). Clinical exam-ination included assessment of oral hygiene with a stand-ardized index (Oral Hygiene Index Simplified, OHI-S)[11] and probing for attachment loss. Probing depth andattachment level were measured with a standard probe(UNC-15) for all teeth in 6 locations (disto-lingual, mid-point lingual, mesio-lingual, disto-buccal, midpoint buc-cal, and mesio-buccal). The number and location ofpoints which bled on probing were noted and a wholemouth bleeding index was calculated as the percentage ofall sites. The hygienist was unaware whether the womenhad delivered prematurely or at term. Samples of gingivalcrevicular fluid for enzyme analysis were obtained in allPage 2 of 8(page number not for citation purposes)elastase, has been shown to be highly predictive of even-tual attachment loss. Armitage et. al. demonstrated thatsubjects except those who had had a course of antibioticslasting more than one week prior to delivery. This groupBMC Pregnancy and Childbirth 2006, 6:24 http://www.biomedcentral.com/1471-2393/6/24was excluded because antibiotics could have affected thesubgingival microflora and reduced the concentrations ofbacterial enzymes. Subjects who had had antibiotics forshort courses only during labor were included. The gingi-val crevicular fluid samples were taken before clinicalassessments with 12 × 1.5 mm Whatman chromatographypaper #1 with markings every 2 mm. The areas sampledwere the typically high risk areas for periodontitis, themesio-buccal gingival crevices of the two molars andpremolars in each quadrant (total of 16 per subject). Thesupragingival plaque was removed and the sites isolatedwith cotton wool rolls and air dried. The strips wereinserted gently and left in place for 30 seconds. The vol-ume of gingival crevicular fluid was assessed by visuallycomparing the degree of strip wetting with that producedby known amounts of serum. Although less accurate thanelectronic measurement [8,9], for the very low volumes offluid expected from non-inflamed gingivae, visual assess-ment was considered adequate for the present investiga-tion. This is because local site data were combined forsubject mean values in the case-control statistical analysisand the great majority of women (117 of 129 sampled)had fluid volumes >0.1 µl at more than half the sites. Eachstrip was labeled and placed into 300 uL of 50 mM 2-(N-0.1% v/v Triton X-100. After 1 hour, with occasional agi-tation at 4°C, the strips were removed and the eluateswere frozen with dry ice. The samples were maintained at-20–70°C and transported for analysis in the laboratoryof two of the co-investigators (B.E., S.C.) Enzyme activitieswere determined using selective peptide substrates forneutrophil elastase (MeOSuc-Ala-Ala-Pro-Val-AFC), bac-terial gingipain (Z-Val-Lys-Lys-Arg-AFC) and dipepti-dylpeptidase (Ala-Pro-AFC). Assay procedures andconditions were as described previously for eachenzyme[8,9] Concentrations of liberated 7-amino-4-trif-luormethyl coumarin (AFC) were measured after 5 hourswith a Perkin Elmer LS30 fluorimeter and enzyme activi-ties were calculated in terms of µUnits pmoles of substratehydrolysed per minute. Laboratory personnel and the co-investigators performing these analyses were not aware ofthe subjects' group.Standardized questionnaires were used to obtain datarelating to medical-dental history and socioeconomic sta-tus.Two control groups were used in this study as a previousinvestigation of periodontal disease in pregnancy hadClinical Examination for Periodontal DiseaseFigure 1Clinical Examination for Periodontal Disease.Page 3 of 8(page number not for citation purposes)morpholino)-ethane-sulphonic acid (MES) buffer, pH5.5, with 0.1 mM dithiothreitol (DTT), 0.15 M NaCl anddocumented decreases in probing depth at term com-pared to preterm gestations [12]. Therefore, comparisonsBMC Pregnancy and Childbirth 2006, 6:24 http://www.biomedcentral.com/1471-2393/6/24between postpartum preterm cases and term controlscould lead to the observation of a spurious difference.Analysis was planned to examine the control groups sep-arately and only combine them if no obvious difference inattachment loss was observed between these two groups.Univariate analysis was planned to evaluate mean attach-ment loss and the frequency of significant attachmentloss, (3 mm), between the cases and controls. Final anal-ysis with logistic regression was planned to include possi-ble confounding factors such as smoking, income, andeducation. Univariate analysis of the log mean enzymeconcentrations was also planned as well as a comparisonof the number of subjects who had concentrations overcritical enzyme levels as this previously had been shownto predict attachment loss [8,9]: neutrophil elastase >400uU/ul, bacterial dipeptidylpeptidase >30 uU/ul and gingi-pain >30 uU/ul. Analysis was performed with Stata ver-sion 8. A sample size of 50 cases and at least 50 controlswas estimated to have at least 80% power to detect anassociation between clinical periodontitis and pretermbirth of the magnitude of an odds ratio equal to 4. Thesample size calculation was also based on an estimate of a13% prevalence of periodontal disease [13].The study was approved by our institutional ethics reviewboard.ResultsDuring the study, we recruited 151 women, 50 cases, 51undelivered controls and 50 postpartum term controls.One of the undelivered controls delivered preterm, twoweeks after her dental examination, so she was reassignedas a case. The remainder of the undelivered controls allsubsequently delivered at term.The 50 cases delivered after spontaneous preterm laborbetween 22 and 35 weeks gestation (mean = 30.8 +/-3.7weeks). The mean gestational age of the undelivered con-trols on examination was 29.2 +/- 4.2 weeks. The post par-tum controls delivered at a mean gestational age of 39.4+/-1.1 weeks. Additional characteristics of the subjects aredescribed in Table 1. The Oral Hygiene Index scores forcalculus and debris as well as the bleeding index (% ofsites with bleeding on probing) were similar amongst thethree groups. (Table 2). There was also no differencebetween the preterm cases and undelivered or postpartumcontrols in mean attachment loss: 0.86 mm, 0.89 mm,and 0.87 mm respectively, p = .93. The mean percentageof sites probed with ≥ 3 mm of attachment loss was simi-lar amongst the three groups: 3.2% of preterm cases, 2.5%of undelivered controls and 1.9% of postpartum controls,p = 0.33. (Table 2).Initial analysis of the attachment loss of the two controlgroups revealed no significant differences so, as planned,for the final analysis, they were combined. The percentageof sites with attachment loss ≥ 3 mm was 2.2% in theTable 1: Characteristics of study population.Preterm Cases n = 50 Undelivered Controls n = 51 Postpartum term Controls n = 50 p valueAge (years) 30.6 +/- 5.9 30.0 +/- 5.2 32.1 +/- 4.0 0.2*Nulliparity 20 (40%) 28 (55%) 24 (48%) 0.3*Post secondary education (years) 2.3 +/- 2.0 4.1 +/- 2.9 4.3 +/- 2.7 <.001*Gross family income ≤ $20,000. 9 (18%) 4 (8%) 3 (6%) 0.15§Smoker during pregnancy 13 (26%) 5 (10%) 3 (6%) 0.01§Last dental cleaning ≤ 6 months 13 (26%) 21 (41%) 21 (42%) 0.12§Value expressed as mean +/- SD or number (percentage)*ANOVA §Fisher's exact.Table 2: Full mouth examination data.Preterm Cases n = 50 Undelivered Controls n = 51 Postpartum term Controls n = 50 p valueProbing depth (mm) 2.11 +/- .33 2.17 +/- .26 2.14 +/- .21 .68†Attachment loss (mm) .86 +/-.32 .89 +/-.26 .87 +/- .18 .93†Attachment loss ≥ 3 mm (mean %) 3.2 +/- .06 2.5 +/- .04 1.9 +/- .02 .33§Extent Severity Score (3,5) * 9 (18%) 10 (20%) 4 (8%) .22§Debris Score 3.6 +/- 2.6 3.0 +/- 2.1 2.9 +/- 1.7 .42†Calculus Score 4.5 +/- 3.6 5.1 +/- 3.7 3.9 +/- 2.8 .19†Bleeding Index (% sites with bleeding) 24 +/- 15 24 +/- 15 20 +/- 11 .41†Page 4 of 8(page number not for citation purposes)Value expressed as mean +/- SD or number (percentage)* Extent Severity Score (3,5) indicates subjects with = 3 mm of attachment loss at 5% or more of sites probed. † ANOVA §Fisher's exact.BMC Pregnancy and Childbirth 2006, 6:24 http://www.biomedcentral.com/1471-2393/6/24combined control group compared to 3.2% in the pre-term cases p = .21. It was originally intended to dichot-omize the subjects with an extent severity score of (3,60),which was used by Offenbacher [1]. This would character-ize all subjects as having periodontal disease if they werefound to have 3 mm or more attachment loss in at least60% of the sites probed. However, we had no subjectswith this degree of periodontal disease. Therefore, as amuch lower incidence of periodontitis was encountered,subjects were dichotomized using an extent severity scoreof (3,5). This characterized the women as having perio-dontitis if they showed attachment loss ≥ 3 mm in at least5% of the sites probed, with the threshold correspondingto a moderate level of disease. This lower threshold wasalso adopted in a recent study [14]. An extent severityscore of (3,5) was not associated with preterm labor in thecrude analysis OR = 1.36 (0.54, 3.41). Univariate analysisalso identified several other factors that were associatedwith preterm birth: gross family income, education andsmoking during pregnancy. Logistic regression analysiswas then performed with a model incorporating variablesfor age, income, smoking during pregnancy, education,dental cleaning within 6 months and periodontal diseasedefined by an extent severity score of (3,5) or greater. Con-trolling for a previous preterm birth was considered sinceit is a well known risk factor for premature birth. How-ever, only one control subject and 7 cases had a previouspreterm birth, so it was neither possible nor meaningful tocontrol for this variable. Again, after adjusted analysis, noassociation between clinical periodontal disease and pre-mature delivery was demonstrated OR = .56 (0.13, 2.37)p = .43. The analysis was repeated for the two controlgroups separately and again, there was no associationwith spontaneous premature labor (data not shown).Gingival crevicular fluid samples were obtained from six-teen standard sites in 40 of 50 cases, 46 of 51 undeliveredcontrols and 43 of 50 postpartum controls. Mean log con-tum controls, p = .018 (Table 3). Of the cases, 33 had atleast one site with a concentration of neutrophil elastaseover the critical value of 400 uU/ul (range 1–11, median =3), compared to 30 undelivered controls (range 1–10,median 2) and 32 postpartum controls (range 1–10,median = 2). Mean log concentrations of gingipain werehigher in the cases compared to undelivered and postpar-tum controls, p = .005 (Table 3). Critical gingipain levels(>30 uU/ul) were observed in 6 preterm cases, three sub-jects with one high enzyme site, one subject with two andtwo subjects with three high enzyme sites. Three undeliv-ered controls and one postpartum control subject had onesite each with high gingipain levels. There was no signifi-cant difference in the mean log concentration of bacterialdipeptidylpeptidase or in the frequency of sites with criti-cal levels between the cases and controls (Table 3). Thefrequency of gingival crevicular fluid enzyme levels overthe critical thresholds for neutrophil elastase and gingi-pain in cases vs. controls are illustrated in Figure 2.Logistic regression analysis was then performed with amodel incorporating terms for mean log enzyme levels,age, gross family income, education, smoking duringpregnancy and dental cleaning in the last 6 months. In theadjusted analysis, the association between neutrophilelastase concentration and spontaneous premature birthremained significant: OR = 1.93(1.13, 3.28) p = 0.015,but the association with gingipain concentration did not:OR= 3.06 (0.68, 13.8) p = .15. The multivariate analysiswas also performed replacing enzyme concentration witha variable for the number of sites with high enzyme levels.The number of high elastase sites was significantly associ-ated with spontaneous preterm birth OR = 1.25 (1.03,1.52), p = 0.021. The number of high gingipain sites wasalso positively associated with premature birth but, again,did not reach statistical significance OR = 1.43(0.36,5.62), p = 0.61. Although the lack of dental cleaningwithin 6 months of delivery had an association with pre-Table 3: Gingival crevicular fluid enzyme levels.Preterm Cases n = 40Undelivered Controls n = 46Postpartum term Controls n = 43p value *Neutrophil elastaseConcentration in GCF. Median (intraquartile range) (uU/ul) 261.0 (160.6, 383.1) 180.8 (81.5, 241.6) 169.7 (117.0, 285.4) 0.018# of subjects with ≥ 1 site with >400 uU/ul 33 30 32GingipainConcentration in GCF. Median (intraquartile range) (uU/ul) 2.24 (1.38, 3.48) 1.73 (1.01, 2.54) 1.60 (1.0, 2.47) 0.005# of subjects with ≥ 1 site with >30 uU/ul 6 3 1Bacterial dipeptidylpeptidaseConcentration in GCF. Median (intraquartile range) (uU/ul) 1.91 (1.17, 3.40) 2.08 (1.15,2.86) 1.16 (.70, 2.40) 0.192# of subjects with ≥ 1 site with >30 uU/ul 5 5 4* ANOVA based on differences in mean log concentrations.Page 5 of 8(page number not for citation purposes)centrations of neutrophil elastase were significantlyhigher in the cases compared to undelivered and postpar-mature birth in univariate analysis, the association was nolonger statistically significant in the final model OR = 1.15BMC Pregnancy and Childbirth 2006, 6:24 http://www.biomedcentral.com/1471-2393/6/24(0.54, 4.26), p = 0.43. Further analysis was then per-formed to determine if high neutrophil elastase was asso-ciated with dental history. It appeared that having a dentalcleaning within 6 months and not smoking during preg-nancy were highly associated with lower gingival crevicu-lar fluid elastase concentrations (ANOVA, p = .02 and p =.025 respectively). Additional regression analysis was alsoperformed to confirm that the time from delivery to den-tal examination was not associated with neutrophilelastase levels p = .28.Discussion and conclusionOur study did not find an association between clinicalperiodontitis, measured by attachment loss, and sponta-neous preterm birth. This is in contrast to several of thestudies to date [1,2,15,16] but is consistent with thereports of Davenport and Moore et al [17,18]. One reasonfor these discrepancies may be our strict definition of pre-term birth to include only those subjects with a pretermbirth after spontaneous labor or induced labor due to pre-term premature rupture of membranes. Several previousstudies have used definitions of outcome that includedintrauterine growth restriction or early second trimestermiscarriage or iatrogenic preterm birth rather than spon-taneous preterm birth alone [1-3,19]. As there is no indi-analysis is methodologically questionable. Differences indisease severity between study population and access toprenatal care may also have lead to different findingsamongst studies. However, this lack of consistency raisesthe possibility that the associations observed in previousstudies are non-causal.Another explanation for a negative finding in any obser-vational study is measurement error. We consider it quiteunlikely that this could be responsible for our findings aswe used only one, highly trained, blinded examiner toassess all the patients. This consistency in examinationshould reduce variability in measurement and any ten-dency to bias the results to a null finding. Finally, limitedpower could be an explanation for our findings. Althoughour study was relatively small, the degree of attachmentloss was almost identical amongst the three groups. Inorder to detect a statistically significant difference, of themagnitude we observed, we would have required over8000 subjects. Our study was also more than adequatelypowered to detect the strong associations that have beenreported in the literature to date, especially if our studyhad observed a similar frequency of severe periodontaldisease as those studies. The fact that we did not find thesame frequency of disease does not invalidate our find-Proportion of subjects by percentage of sites over critical thresholds for gingival crevicular fluid Neutrophil Elastase and Gingi-painFigure 2Proportion of subjects by percentage of sites over critical thresholds for gingival crevicular fluid Neutrophil Elastase and Gingi-pain. Cases vs. combined postpartum and antepartum controls.Page 6 of 8(page number not for citation purposes)cation that these diverse problems have a commonpathogenesis, the amalgamation of these outcomes forings and suggests that an association between clinicallymeasured periodontal disease and prematurity may notBMC Pregnancy and Childbirth 2006, 6:24 http://www.biomedcentral.com/1471-2393/6/24be evident in all populations. However, we do recognizethat, given the relatively small size of our study, inade-quate power is always a possible explanation for a nullfinding.However, we do report an association between a markerof active periodontitis, gingival crevicular fluid neutrophilelastase, and premature delivery. It is difficult to fully eval-uate the significance of this finding, especially as we didnot find an association between premature delivery andstandard clinical measurements of periodontitis. In ourstudy, the gingival crevicular fluid enzymes were meas-ured after delivery. It may be that following a pretermdelivery, women are under increased stress and mayneglect their dental hygiene. As neutrophil elastase mayrise in gingival crevicular fluid with the development ofgingivitis, this may be all that is reflected in our data.However, analysis of the Oral Hygiene Index scoresbetween the case and control groups suggested only aslight, non-significant increase in only the debris scores,but not in the calculus scores, in the preterm group. Thepossibility of selection bias should also be considered inthe evaluation of the results of any case control study.However, none of our patients had symptomatic perio-dontal disease. Therefore, it seems difficult to conceptual-ize how patients or study nurses could have perceivedfactors associated with elevated gingival crevicular fluidenzyme levels and have let this influence the probabilityof enrolment. One explanation for the inconsistencybetween the clinical and enzyme data is that high elastaselevels reflect active periodontal inflammation and this,rather than past attachment loss, is the more importantrisk factor. If so, it may be that, in our minimally diseasedpopulation, the subjects have not had time to developedclinically apparent disease. Based on our data, such a con-clusion is reasonable and would be consistent with arecent publication documenting an increased risk of pre-term birth with progression of attachment loss duringpregnancy [20]. Furthermore, our data suggest that simpletreatments such as dental cleaning and avoidance ofsmoking during pregnancy are associated with lower gin-gival crevicular fluid elastase levels. A reduction in gingi-val crevicular fluid elastase and other enzymes with dentalcleaning has been recently reported by Figueredo [21].Therefore, if a causal relationship between early periodon-tal disease and premature labor is ultimately proven, thismay lead to effective preventative strategies.It is also possible that previous studies and our ownenzyme findings have simply detected a yet to be definedgeneralized susceptibility to infection in women whodeliver preterm. This could place women at increased riskfor a variety of infections including periodontal diseaserisk, cytokine gene polymorphisms, does not appear to bea common factor in preterm birth and periodontitis [14].In conclusion, in contrast to other authors, we did notfind an association between clinical periodontal disease,measured by attachment loss, and spontaneous prema-ture delivery. We did find an association between gingivalcrevicular fluid enzyme levels and premature delivery,possibly indicating an association between active perio-dontal disease and spontaneous premature delivery.However, given the lack of agreement between the analy-sis of the clinical and enzyme data, we feel it would be pre-mature to conclude that a causal relationship exists.Ultimately, because of the difficulty in proving a temporalrelationship between exposure and disease, a case controlstudy can rarely, in and of itself, prove causality. There-fore, at this time a recommendation for screening andtreatment of periodontal disease in pregnant women,with the aim of preventing premature birth, cannot bemade until further research is completed. Of course, it isstill recommended that women who are pregnant or plan-ning to become pregnant continue to maintain optimumperiodontal health with professional cleanings and metic-ulous oral hygiene to prevent periodontal disease. Basedon our results, future investigators in this area should alsoconsider measuring markers of active periodontal diseaseand not rely solely on clinical examination.Competing interestsThe author(s) declare that they have no competing inter-ests.Authors' contributionsAll authors contributed to the design of the study. AF, RSand SW conceived the study. SW supervised the recruit-ment. SN carried out the periodontal examinations underthe supervision of AF. SC and BM carried out the GCFenzyme analysis. RB performed the statistical analysis. Allauthors contributed to and approved the final manu-script.AcknowledgementsThis study was supported primarily by a grant from the Calgary Regional Health Authority Perinatal Funding Competition which is funded by the Ross Division of Abbott Labs. Partial funding was also provided by the Uni-versity of Alberta Fund for DentistryWe would like to thank Ms Heidi Cheung for technical assistance with the gingival crevicular fluid analysis and Deborah Schaab RN for her work in recruiting the patients.References1. Offenbacher S, Katz V, Fertik G, Collins J, Boyd D, Maynor G, McKaigR, Beck J: Periodontal Infection as a Possible Risk Factor forPreterm Low Birth Weight.  J Periodontol 1996, 67:1103-1113.Page 7 of 8(page number not for citation purposes)and ascending chorioamnionitis leading to prematurelabor. On the other hand, one possible basis of shared2. Jeffcoat MK, Geurs NC, Reddy MS, Cliver SP, Goldenberg RL, HauthJC: Periodontal  infection and preterm birth.  Results of aprospective study.  JADA 2001, 132:875-880.Publish with BioMed Central   and  every scientist can read your work free of charge"BioMed Central will be the most significant development for disseminating the results of biomedical research in our lifetime."Sir Paul Nurse, Cancer Research UKYour research papers will be:available free of charge to the entire biomedical communitypeer reviewed and published immediately upon acceptancecited in PubMed and archived on PubMed Central BMC Pregnancy and Childbirth 2006, 6:24 http://www.biomedcentral.com/1471-2393/6/243. Offenbacher S, Lieff S, Boggess KA, Murtha AP, Madianos PN, Cham-pagne CM, McKaig R, Jared HL, Mauriello SM, Auten RL, HerbertWN, Beck JD: Maternal periodontitis and prematurity. Part 1:Obstetric outcome of prematurity and growth restriction.Ann Periodontol 2001, 6:164-174.4. Davenport ES, Williams CE, Sterne JA, Murad S, Sivapathasundram V,Curtis MA: Maternal periodontal disease and preterm lowbirth weight: Case-control study.  J Dent Res 2002, 81:313-318.5. Canada S: 2001 Community Profile, Calgary.  www12 statcan ca2006.6. Carranza FA: Classification of Periodontal Diseases.  In Glick-man's Clinical Periodontology Volume 14. 7th edition. Edited by: Car-ranza FA. Philadelphia,PA., W.B. Saunders Co.; 1990:202-209. 7. Armitage GC, Jeffcoat MK, Chadwick DE, Taggart EJ, Numabe Y,Landis JR, Weaver SL, Sharp TJ: Longitudinal evaluation ofelastase as a marker for the progression of periodontitis.  JPeriodontol 1994, 65:120-128.8. Eley BM, Cox SW: A 2-year longitudinal study of elastase inhuman gingival crevicular fluid and periodontal attachmentloss.  J Clin Periodontal 1996, 23:681-692.9. Eley BM, Cox SW: Correlation between gingivain/gingipain andbacterial dipeptidyl peptidase activity in gingival crevicularfluid and periodontal attachment loss in chronic periodonti-tis patients. A 2-year longitudinal study.  J Periodontol 1996,67:703-716.10. Loesche WJ, Bretz WA, Lopatin D, Stoll J, Rau CF, Hillenburg KL, Kil-loy WJ, Drisko CL, Williams R, Weber HP, Clark W, Magnusson I,Walker C, Hajoel PP: Multi-center clinical evaluation of a chair-side method for detecting certain periodontopathic bacteriain periodontal disease.  J Periodontol 1990, 61:189-196.11. Greene JC, Vermillion JR: "The simplified oral hygiene index.".JADA 1964, 68:7-13.12. Miyazaki H, Yamashita Y, Goto-Kimura K, Shimada N, Sogame A,Takehara T: Periodontal condition of pregnant womenassessed by CPITN.  J Clin Periodontol 1991, 18:751-754.13. J-M B, Payette M, Olivire M, Chabot D, Benigeri M, Williamson S,Lemay A: Etude 1994-95 sur la sante bucco-dentaire des adul-tes quebecois de 35 a 44 ans.  Direction de la sante publique RegieRegionale de la sante et des services sociaux Province of Quebec;1997. 14. Moore S, Ide M, Randhawa M, Walker JJ, Reid JG, Simpson NA: Aninvestigation into the association among preterm birth,cytokine gene polymorphisms and periodontal disease.  BJOG2004, 111:125-132.15. Goepfert AR, Jeffcoat MK, Andrews WW, Faye-Petersen O, CliverSP, Goldenberg RL, Hauth JC: Periodontal disease and uppergenital tract inflammation in early spontaneous pretermbirth.  Obstetrics & Gynecology 2004, 104:777-783.16. Lopez NJ, Smith PC, Gutierrez J: Higher risk of preterm birthand low birth weight in women with periodontal disease.  JDent Res 2002, 81:58-63.17. Davenport E, Williams C, Sterne J, Sivapathasundram V, Femme S,Curtis M: The East London Study of Maternal Chronic Perio-dontal Disease and Preterm Low Birth Weight Infants:Study Design and Prevalence Data.  Ann Periodontol 1998,3:213-221.18. Moore S, Ide M, Coward PY, Randhawa M, Borkowska E, Baylis R,Wilson RF: A prospective study to investigate the relationshipbetween periodontal disease and adverse pregnancy out-come.  Br Dent J 2004, 197:251-258.19. Mitchell-Lewis D, Engebretson SP, Chen J, Lamster IB, Papapanou PN:Periodontal infections and pre-term birth: early findingsfrom a cohort of young minority women in New York.  Euro-pean Journal of Oral Sciences 2001, 109:34-39.20. Offenbacher S, Boggess KA, Murtha AP, Jared HL, Lieff S, McKaig R,Mauriello SM, Moss KL, Beck JD: Progressive Periodontal Dis-ease and Risk of Very Preterm Delivery.  Obstet Gynecol 2006,107:29-36.21. Figueredo CM, Areas A, Miranda LA, Fischer RG, Gustafsson A: Theshort term effectiveness of non-surgical treatment in reduc-ing protease activity in gingival crevicular fluid from chronicperiodontitis patients.  J Clin Periodontol 2004, 31:615-619.Pre-publication historyThe pre-publication history for this paper can be accessedhere:http://www.biomedcentral.com/1471-2393/6/24/prepubyours — you keep the copyrightSubmit your manuscript here:http://www.biomedcentral.com/info/publishing_adv.aspBioMedcentralPage 8 of 8(page number not for citation purposes)


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items