Full TextJournal of Periodontology2010, Vol. 81, No. 3, Pages 350-358 , DOI 10.1902/jop.2009.090527(doi:10.1902/jop.2009.0...
birth; 99% of these deaths occur in the worlds least developed countries. This report also indicatesthat a preterm birth (...
preterm or extreme preterm birth by ordinal logistic regression analysis.MATERIALS AND METHODS                            ...
15).Periodontal ExaminationThe periodontal examination was performed at the hospital within 48 hours of parturition by man...
RESULTS                                                       Section:Table 1 summarizes the general descriptive variables...
Table 3. Prevalence of Periodontal Disease in the Sample According to                        Definitions Used in the Study...
95% CI: 2.54 to 15.10), and prior premature birth (OR = 2.39, 95% CI: 1.02 to 5.61). As fordefinition 1, the interactions ...
The very high ORs found by Jeffcoat et al.14 could be the result of the definition used in their cohortof women with perio...
study despite the significant association detected).CONCLUSIONS                                                        Sec...
1.  González R. Administración prenatal de progesterona para la prevención de nacimientos    prematuros: Comentario de la ...
Sample Article
Upcoming SlideShare
Loading in …5

Sample Article


Published on

  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Sample Article

  1. 1. Full TextJournal of Periodontology2010, Vol. 81, No. 3, Pages 350-358 , DOI 10.1902/jop.2009.090527(doi:10.1902/jop.2009.090527)Maternal Periodontal Disease and Preterm or Extreme Preterm Birth: An OrdinalLogistic Regression AnalysisAlessandra Neves Guimarães,* Agustín Silva-Mato,* Luis Otavio Miranda Cota,† FernandaMafra Siqueira,† and Fernando Oliveira Costa†*Department of Health and Sociomedical Sciences, Biostatistics Unit, Faculty of Medicine, AlcaláUniversity, Madrid, Spain.†Department of Periodontology, School of Dentistry, Federal University of Minas Gerais, BeloHorizonte, MG, Brazil.Correspondence: Dr. Alessandra Neves Guimarães, Department of Health and SociomedicalSciences, Alcalá University, Madrid-Barcelona Road, km 33.600, 28871, Alcalá de Henares, Madrid,Spain. Fax: 34-918854874; e-mail: le_ng@hotmail.com.Sections:ABSTRACT Section:Background: Despite previous studies addressing the link between preterm or low birth weightinfants and maternal periodontitis, extreme preterm births have received far less attention. Thisstudy is designed to address the possible association between maternal periodontal disease andpreterm or extreme preterm birth.Methods: Immediately after childbirth, 1,207 women underwent an examination in whichperiodontal disease was assessed according to two alternative definitions: 1) four or more teeth withat least one site showing probing depth (PD) ≥4 mm and clinical attachment loss (AL) ≥3 mm, and2) at least one site showing PD and clinical AL ≥4 mm. For each of these definitions, two types ofmultivariate analysis were conducted: a linear regression analysis for the number of gestationweeks, and a more specific ordinal logistic regression analysis for the ordinal variable gestation timecategorized as normal (term) (n = 1,046 women) or mild–moderate (n = 146 women) or extremepreterm (n = 15 women).Results: Periodontal disease was associated with fewer weeks of gestation by linear regression(definition 1: P = 0.012; definition 2: P <0.001) and with preterm (n = 161; mild–moderate andextreme) or extreme preterm births (n = 15) by ordinal logistic regression (definition 1: odds ratio[OR] = 1.83, 95% confidence interval [CI]: 1.28 to 2.62; definition 2: OR = 2.37, 95% CI: 1.62 to3.46).Conclusion: Our findings suggest that periodontal disease is associated with a premature orextremely premature birth.According to a recent report by the Word Health Organization,1 536,000 women and 3 millionnewborns die each year because of complications during childbirth and immediately after giving
  2. 2. birth; 99% of these deaths occur in the worlds least developed countries. This report also indicatesthat a preterm birth (<37 weeks) is the main cause of neonatal mortality along with both short- andlong-term sequellae for the infant. Prevalence of preterm births range from 6% to 12% in developedcountries and exceed these rates in developing countries. Close to 40% of all preterm births occurbefore 34 weeks of pregnancy, and 20% of these births occur before week 32 of pregnancy(extreme preterm births). Together these premature births account for 50% of all cases of perinatalmorbidity and mortality.1Despite great advances made in perinatal preventive and treatment measures in general, cases ofpreterm births continue to rise throughout the world2 especially due to an increased interruption ofcomplicated pregnancies. There is some evidence that known risk factors, related to thecharacteristics of both mother and fetus, are similar for mild–moderate preterm (32 through 36weeks)3 and extreme preterm births. However, on occasion, the strength of association seems to bestronger for extreme cases.4 The following are among these risk factors: maternal demographiccharacteristics, nutrition status, previous pregnancies, obstetric features related to the currentpregnancy, psychologic characteristics of the mother, adverse environmental factors, uterinecontractions, and biologic and genetic factors and infections, with periodontal disease included in thislast category.2,5,6A possible explanation for this latter association could be that, as a chronic disease of infectiousorigin, periodontal disease induces early delivery through raised systemic levels of pathogenicmicroorganisms or their endotoxins, or directly via inflammatory mediators, especially prostaglandinE2 (PGE2), interleukin (IL)-1, and tumor necrosis factor-alpha. The actions of these mediators affectthe start of labor such that they can consequently trigger early parturition.7-9 In a study byDörtbudak et al.10 conducted in 36 pregnant women in which the early appearance of inflammationmarkers and periodontal disease-causing microorganisms in amniotic fluid were examined, thepresence of IL-6 and -8 and PGE2 during pregnancy weeks 15 to 16 was found to increase the risk ofpreterm birth. In addition, Pitiphat et al.11 were able to confirm that raised plasma C-reactive proteinlevels during early pregnancy (14 weeks) were also positively related to periodontal disease.Since the 1990s, epidemiologic studies8,11-22 linked maternal periodontal disease to preterm birth,although this link was disputed by other authors.23-29 Offenbacher et al.12 were the first to proposesuch a relationship based on a case-control study performed in the United States. In contrast, thelink between periodontal disease and extreme preterm birth was the topic of only a fewinvestigations,14,17,20 all of them showing significant associations. The results and conclusions of all ofthese studies are difficult to compare because of a lack of uniform methods, variation in samplesizes, and the use of different clinical criteria to define periodontal disease.29-31The most widely used statistical analysis method used has been logistic regression because this testallows risk assessment while controlling for other possible factors. This method is appropriate whenthe dependent variable is dichotomous, but is less suitable when the variable under examination hasmore than two categories that follow some kind of hierarchy. This is true of the variable gestationtime, which may be expressed as normal (control) or mild–moderate or extreme preterm. In thistype of situation, the most appropriate model to use seems to be ordinal logistic regression.32Based on the same database used in the present study, a previous report21 used the periodontitisdefinition proposed by López et al.15 This study21 used logistic regression modeling because extremepreterm birth was not considered in the analysis. Thus, to help internal comparisons as well ascomparisons with other studies, the present study was designed to assess the link betweenperiodontal disease, defined according to the two most commonly used clinical definitions,15,33 and
  3. 3. preterm or extreme preterm birth by ordinal logistic regression analysis.MATERIALS AND METHODS Section:Study DesignThe study design was cross-sectional and included 1,686 women aged 14 to 46 years (1,746 womenwere initially recruited, and 60 women refused to participate) attending the public maternity clinic inBelo Horizonte, MG, Brazil, from February 2004 to June 2005 (the same patient cohort studied bySiqueira et al.21). As inclusion criteria, it was established that each mothers pregnancy was single,and the mothers delivered live newborns as the outcome of a pregnancy to term, spontaneouspreterm labor, or premature membrane rupture. Women with multiple pregnancies; congenitalanomalies; in vitro fertilization pregnancies; and prematurity because of pregnancy interruption dueto fetal and/or maternal factors including preeclampsia, heart disease, neuropathy, or placental,cervical, or uterine defects were excluded. After the data collection, all clinical records were reviewedby a gynecology and obstetrics physician (Dr. Ana Marcia Miranda Cota, Mater Dei Hospital, BeloHorizonte, Brazil) to confirm the inclusion and exclusion criteria. Of the initial group of 1,686 womenwho gave their consent to participate, 479 were excluded (for the reasons listed above) to give afinal study sample of 1,207 mothers in good general health who had recently given birth. Thesemothers who were of low socioeconomic status belonged to a heterogeneous ethnic group.The study protocol was approved by the Research Ethics Committee of the Federal University ofMinas Gerais (COEP/UFMG), and the Research Ethics Committee of the Hospital Foundation of MinasGerais State (CEP/FHEMIG), both located in Belo Horizonte, Brazil. Participants were enrolled after asigned informed consent form was obtained.Sociodemographic and Clinical DataFor the collection of sociodemographic data and those related to the current or any previouspregnancy, we used the form issued by the Latin American Center for Perinatology/Woman andReproductive Health (CLAP/WR), Montevideo, Uruguay, which was completed for each participant,along with the clinical records of both mother and newborn.The following data were obtained: of the mother: age (years), education level, marital status,smoking status, alcohol consumption or drug abuse during pregnancy, diabetes, chronic arterialhypertension, number of prenatal visits to the clinic, rate of urinary tract infection, primiparity, priorpremature birth, and prior miscarriage; of the newborn: weight and gestational age, and whetherthe neonate was small for the gestational age.The gestational age was estimated from the last menstrual period. When this date was unavailable,the gestational age was determined in an ultrasound examination undertaken no later thanpregnancy week 14.The study sample was divided into three groups according to the continuous variable gestation timeexpressed as weeks of pregnancy: a control group comprised of mothers whose pregnancy wasnormal (n = 1,046) and therefore lasted ≥37 weeks, a mild–moderate preterm group comprised ofmothers who delivered their babies ≥32 to 36 weeks of pregnancy (n = 146); and an extremepreterm group made up of mothers who underwent <32 weeks of pregnancy before giving birth (n =
  4. 4. 15).Periodontal ExaminationThe periodontal examination was performed at the hospital within 48 hours of parturition by manualcircumferential probing using a millimeter-scale periodontal probe.‡ Two trained periodontists(LOMC, FMS), who were masked to the medical histories of the mothers, were calibrated 3 monthsbefore the study onset. Intra- and interexaminer agreements were determined for the factorsprobing depth (PD) and clinical attachment level. This preliminary study rendered non-weighted κvalues >0.81. Signs of inflammation and destruction of periodontal bone support were alsoassessed, and the following variables recorded: PD, clinical attachment loss (AL), and bleeding onprobing (BOP).All teeth were examined with the exception of third molars, incompletely erupted teeth, teeth inareas with extensive caries lesions, teeth with fractures or iatrogenic damaged restorations, andsurfaces where the cemento-enamel margins could not be distinguished.Periodontal disease was defined as follows: from López et al.:15 definition 1 = the presence of four ormore teeth with one or more sites of PD ≥4 mm and clinical AL ≥3 mm; and from Albandar:33definition 2 = at least one tooth site showing PD and clinical AL ≥4 mm. These definitions wereselected on the grounds that they would be easy to compare among themselves and with the resultsof other studies, because they are definitions most often used across studies.15,16,20-22,26,29,30,34Statistical AnalysesStatistical analyses included the descriptive characterization of the control, mild–moderate preterm,and extreme preterm groups, which were initially assessed in terms of all variables of interestincluding periodontal disease. Each independent variable was subjected to bivariate analysis usingthe Fisher exact test or the Cochran-Armitage linear trend test35 when the variables were categoricand analysis of variance (ANOVA) or the Kruskal-Wallis test when dealing with quantitative variables.For each of the two definitions of periodontal disease, linear regression analysis was performed forthe number of pregnancy weeks followed by a final ordinal regression analysis.In each regression, each variable was first separately analyzed (without correcting for the rest), andthose showing a P value <0.10, along with their interactions, were introduced into the linear andordinal models. During the following steps, non-significant variables (P >0.05) were manuallyremoved, and new variables were added to give the final adjusted full model.Ordinal regression analysis was conducted according to the cumulative proportionality method.36After this analysis, a model fit was assessed using the Brant test of proportionality,37 whichdetermines the extent of parallelism with the first and second cutoff points of the dependentvariable, i.e., for preterm and extreme preterm, such that equality may be assumed among thecoefficients of the variables remaining in the final model, implying a common odds ratio (OR) forboth cutoff points. If the parallel line assumption was violated for any of the independent variablestested, the general method (partial proportional method) was used.All statistical tests were performed using a software program.§
  5. 5. RESULTS Section:Table 1 summarizes the general descriptive variables (categoric and continuous) obtained for thesample of 1,207 women assigned to the control (86.7% of the sample), mild–moderate (12.1%),and extreme preterm (1.2%) groups. Table 1. Characteristics of the Sample According to Categoric (Dichotomous) and Continuous Variables Considered Among the sociodemographic variables examined, maternal age, considered a continuous variable,showed similar means in the mild–moderate, extreme preterm, and control groups. The variableeducation level showed similar frequencies in the three groups, whereas marital status differed inthat the prevalence of single mothers was higher in the mild–moderate and extreme preterm than inthe control group (P = 0.001). No substantial differences in the habits of smoking, alcoholconsumption, and drug abuse during pregnancy were observed among the groups, although thiscould be the outcome of the low proportions recorded in the study sample. This also occurred for thefactors diabetes and chronic hypertension. Among the obstetric variables, a prior miscarriage did notvary significantly among the three groups, whereas the rate of urinary tract infection was a littlehigher in the mild–moderate preterm group, but also not significantly. Finally, the mild–moderateand extreme preterm groups showed fewer prenatal visits with a mean of 5.43 and 3.67,respectively, versus 7.01 in the control group (P <0.0001). This can be explained, in part, by theshorter duration of pregnancy as well as higher proportions of primiparity (P = 0.055) and previouspremature birth (P = 0.060), although in both cases, these differences were non-significant. In thislatter situation, prior preterm births were considered among non-primiparous women.The periodontal disease status of the study sample is shown in Tables 2 and 3. For each of thegroups, control, mild–moderate, and extreme preterm, Table 2 shows mean values obtained for thefollowing factors: the mean number of sites showing PDs and clinical AL levels ≥4, ≥5, and ≥7 mm,and the mean number of sites with BOP. In contrast, Table 3 shows the prevalence of periodontaldisease by definitions 1 or 2. The tables indicate that, overall, the sample showed a high prevalenceof moderate periodontal disease, which increased, as expected, when the second definition wasused, even in the control women. Moreover, the frequency of periodontal disease was clearly higherin the mild–moderate and extreme preterm groups compared to the control group, this being mostevident when the second periodontal disease definition was used. Table 2. Periodontal Status of the Study Sample
  6. 6. Table 3. Prevalence of Periodontal Disease in the Sample According to Definitions Used in the Study Theresults obtained in the linear regression analysis for number of pregnancy weeks are provided inTable 4, and the results of the ordinal regression analysis for duration of pregnancy categorized ascontrol, preterm and extreme preterm are provided in Table 5. In both tables, the results obtainedfor each variable without adjusting for the remaining variables and adjusting for each periodontaldisease definition in the two final models are shown. For each variable, the interpretation of an OR>1 in Table 5 (risk of a preterm or extreme birth) was comparable to that indicated by a negativeregression coefficient in Table 4 (reduction in the mean number of pregnancy weeks). For thisreason, and given the agreement between the two sets of results, we only comment on the dataprovided in Table 5. Table 4. Linear Regression Analysis (unadjusted) of Gestation Time (weeks) and Multiple Linear Regression Analysis (adjusted) of Gestation Time According to Periodontal Disease Definition Table 5. Ordinal Logistic Regression Analysis (unadjusted) of Gestation Time (control, preterm, and extreme preterm) and Multiple Ordinal Logistic Regression Analysis According to Periodontal Disease Definitions Thefinal ordinal regression model (Table 5) for periodontal disease definition 1 identified the followingsignificant risk factors for a preterm or extreme preterm birth: periodontal disease, number ofprimiparity, and prior premature delivery. The interaction effect between primiparity and prenatalvisits (OR = 0.81, 95% CI: 0.70 to 0.93) observed on the risk of a preterm birth indicated that theOR = 6.45 for being primiparous corresponded to women not attending prenatal visits and wasreduced by multiplying by 0.81 for each visit. Finally, the effect of the interaction of urinary tractinfection × single marital status (OR = 3.19; 95% CI: 1.46 to 6.96), whose separate ORs were nonsignificant, indicated that the risk of a preterm delivery in women with a urinary tract infectionbecame significant if they were not married (the OR being 0.96 × 3.19 = 3.06).In the same table (Table 5), a summary is provided of the ordinal logistic regression analysis for thesecond definition of periodontal disease. According to the final model, the following factors emergedas significant risks for a preterm or extreme preterm birth: periodontal disease (OR = 2.37; 95% CI:1.62 to 3.46), number of prenatal visits (OR = 0.85, 95% CI: 0.78 to 0.93), primiparity (OR = 6.20,
  7. 7. 95% CI: 2.54 to 15.10), and prior premature birth (OR = 2.39, 95% CI: 1.02 to 5.61). As fordefinition 1, the interactions between primiparity and prenatal visits (OR = 0.81; 95% CI: 0.71 to0.94) and between urinary tract infection and marital status (OR = 3.28; 95% CI: 1.49 to 7.20)persisted for the final model for definition 2. These effects were interpreted as described fordefinition 1.Although the factor, age, initially emerged as a significant risk factor at its extremes, the data inTables 4 and 5 indicate, when controlling for the remaining variables (including interactions), thissignificance was lost in all models. This can be explained, in part, because age was very related tothe number of prenatal visits, given that women of intermediate age were the ones who mostfrequently attended the maternity clinic. This fact was manifested by the quadratic regressionbetween prenatal visits and age, which indicated that the visits, on average, diminished at the twoextremes of age (R2 = 0.02; weak but highly significant; P <0.0001).Thus, in the linear and ordinal regression analyses, consistent results were obtained for eachdefinition of periodontal disease because the same variables persisted in the final models, with theirR2 and pseudo R2 determination coefficients being similar. However, if the results for eachperiodontal disease definition are compared, although large similarities were detected, definition 2performed better given its higher and more significant ORs. In addition, according to the Brant test,the model for this definition showed the best fit.DISCUSSION Section:The results obtained in this study suggest an association between maternal periodontal disease andspontaneous preterm or extreme preterm birth. Controlling for other known risk factors, this linkwas revealed by a single ordinal logistic regression analysis. Our results indicated that women withperiodontal disease carried an OR = 1.83 of delivering a preterm or extremely preterm baby whenperiodontal disease was defined according to definition 1 and an OR = 2.37 when periodontaldisease defined by definition 2. We can assume that the ORs would be the same when the controlgroup is compared to the preterm group (including cases of extreme preterm) and when the controland mild–moderate -preterm groups are compared to the extreme preterm group, given that theORs were generated by ordinal logistic regression models, where parallelism was confirmed.As far as we are aware, only three previous studies,14,17,20 all conducted in the United States,examined the possible link between extreme preterm birth and maternal periodontal disease.Jeffcoat et al.14 were the first to consider this association in a cohort of 1,313 women. These authorsreported adjusted ORs of 4.45 (95% CI: 2.16 to 9.18) for a preterm birth and 7.07 (95% CI: 1.7 to27.4) for an extreme preterm birth. In a later case-control study performed on 139 women, Goepfertet al.17 obtained an adjusted OR of 3.2 (95% CI: 1.2 to 8.8) for extreme preterm birth. Finally,working with a cohort of 1,020 women, Offenbacher et al.20 confirmed this link and cited adjustedrelative risk values of 2.0 (95% CI: 1.2 to 3.2) for preterm birth and 2.4 (95% CI: 1.1 to 5.2) forextreme preterm birth. In this study, the authors considered periodontal disease progression andincluded labor induction as indicated by the physician.It may be observed that the strengths of association reported by Offenbacher et al.20 are similar tothose detected in our study, whereas those of Goepfert et al.17 and, especially, Jeffcoat et al.14 aresomewhat higher.
  8. 8. The very high ORs found by Jeffcoat et al.14 could be the result of the definition used in their cohortof women with periodontitis. Although the authors considered a loss of attachment level of 3 mm,they were very strict when defining the extension of disease, such that this reading had to berecorded for ≥50 sites. Offenbacher et al.20 similarly attributed these published results (i.e., bothstudies related to extreme preterm birth) to an initially very ill study sample.As in our study, the three previous trials adjusted their final models for the remaining variablesconsidered in each case, whether obstetric or sociodemographic. However, which variables arethought to have a considerable impact on the link between periodontal disease and preterm orextreme birth?As in our study, previous preterm birth was considered by Goepfert et al.17 and Offenbacher et al.20In the latter study, the significant adjusted risks of 2.4 for preterm birth and 3.4 for extremepreterm birth observed were very similar to those ORs noted in the present study (OR = 2.4 for bothdefinitions of periodontitis). Notwithstanding, the results of Goepfert et al.,17 who presented two finaladjusted models, clearly reflect the importance of including or excluding this type of variablebecause the somewhat elevated result attributed to periodontal disease for a preterm delivery wasachieved without considering the factor prior preterm birth; when this was done, they obtained alower and non-significant OR = 2.5 (95% CI: 0.9 to 7.4). Accordingly, this last OR more closelyresembles and is more comparable to our figure and that obtained by Offenbacher et al.;20 its lack ofsignificance is possibly attributable to the lower sample size.With regard to genitourinary infections, Goepfert et al.17 conducted microbiologic tests and biologic-marker tests to compare the occurrence of chorioamnionitis in mothers with and without periodontaldisease but found no significant differences. Subsequent to this, Offenbacher et al.20 included thefactor chorioamnionitis in their logistic model, but independently to periodontal disease, and found itto be significantly correlated with preterm and extreme preterm births. In our study, the variableurinary tract infection showed a significant effect on the outcome measures in the two final modelsbut interacted with marital status.Marital status was the only sociodemographic characteristic that remained in both of our final modelsbut only when its interaction with urinary tract infection was considered. Race, which was consideredin the three United States studies,14,17,20 was not taken into account in the present study because thewomen belonged to a difficult-to-characterize, heterogeneous ethnic group.Our findings reveal, as did those of Offenbacher et al.,20 that smoking, alcohol intake, or drug abuseduring pregnancy did not carry significant risks for preterm or extreme preterm birth. This couldreflect the low proportions of these habits in our study or perhaps that they were underreported bythe women in our sample.In our study, the ordinal logistic regression method enabled us to jointly examine both case groupsin a single model and obtain cumulative ORs for preterm birth and extreme preterm birth for two ofthe most widely used definitions of periodontal disease. The use of these two definitions allowed forinternal comparisons and comparisons with other studies. Effectively, the lack of standardizedcriteria for periodontal disease was described as an important limitation of the different studies.29,30Notwithstanding, our results need to be confirmed in future studies conducted on larger numbers ofwomen, given the low prevalence of extreme prematurity, estimated at 1.0% to 2.0% for thegeneral population1 and at 1.2% in our series (only 15 women; this is the main limitation of our
  9. 9. study despite the significant association detected).CONCLUSIONS Section:The results of our study indicate that periodontal disease in the mother, besides being associatedwith preterm birth, as indicated in most literature reports,8,11-13,15,16,18,19,21,22 is also related to extremepreterm birth. This final conclusion concurs with those of the few previous studies14,17,20 thatexamined periodontal disease and extreme preterm birth.ACKNOWLEDGMENTSThe authors thank the staff of the Odete Valadares Maternity Hospital, Belo Horizonte, MG, Brazil,and the study participants for their collaboration. We also thank Dr. Ana Marcia Miranda Cota, MaterDei Hospital, Belo Horizonte, Brazil, for her help with the clinical data collection process. The authorsreport no conflicts of interest related to this study.REFERENCES Section:
  10. 10. 1. González R. Administración prenatal de progesterona para la prevención de nacimientos prematuros: Comentario de la BSR (última revisión: 3 de marzo de 2008). La Biblioteca de Salud Reproductiva de la OMS; Ginebra: Organización Mundial de la Salud. Available at: http://apps.who.int/rhl/pregnancy_childbirth/complications/preterm_birth/rgcom/es/index.html. Accessed March 20, 2009.2. Goldenberg RL, Culhane JF, Iams JD, Romero R. Epidemiology and causes of preterm birth. Lancet 2008;371:75-84.3. Kramer M S, Demissie K, Yang H, Platt RW, Sauvé R, Liston R. The contribution of mild to moderate preterm birth to infant mortality. JAMA 2000;284:843-849.4. Ancel PY, Cubizolles MJS, Di Renzo GC, Papiernik E, Bréart G. Very and moderate preterm births: Are the risk factors different? Br J Obstet Gynaecol 1999;106:1162-1170.5. Williams CE, Davenport ES, Sterne JA, Sivapathasundaram V, Fearne JM, Curtis MA. Mechanisms of risk in preterm low-birthweight infants. Periodontol 2000 2000;23:142-150.6. Bettiol H, Rona RJ, Chinnn S, Goldani M, Barbieri MA. Factors associated with preterm births in southeast Brazil: A comparison of two cohorts born 15 years apart. Paediatr Perinat Epidemiol 2000;14:30-38.7. Gibbs RS, Romero R, Hillier SL, Eshenbach DA, Sweet RT. A review of premature birth and subclinical infection. Am J Obstet Gynecol 1992;166:1515-1528.8. Davenport ES, Williams CECS, Sterne JAC, Sivapathasundaram V, Fearne JM, Curtis MA. The east London study of maternal chronic periodontal disease and preterm low birth weight infants: Study design and prevalence data. Ann Periodontol 1998;3:213-221.9. Offenbacher S. Maternal periodontal infections, prematurity, and growth restriction. Clin Obstet Gynecol 2004;47:808-821.10. Dörtbudak O, Eberhardt R, Ulm M, Persson GR. Periodontitis, a marker of risk in pregnancy for preterm birth. J Clin Periodontol 2005;32:45-52.11. Pitiphat W, Joshipura KJ, Gillman MW, Williams PL, Douglass CW, Rich-Edwards JW. Maternal periodontitis and adverse pregnancy outcomes. Community Dent Oral Epidemiol 2008;36:3-11.12. Offenbacher S, Katz V, Fertik G, et al. Periodontal infection as a possible risk factor for preterm low birth weight. J Periodontol 1996;67:1103-1113.13. Dasanayake AP. Poor periodontal health of the pregnant woman as a risk factor for low birth weight. Ann Periodontol 1998;3:206-212.14. Jeffcoat MK, Geurs NC, Reddy MS, Cliver SP, Goldenberg RL, Hauth JC. Periodontal infection and preterm birth: Results of a prospective study. J Am Dent Assoc 2001;132:875-880.15. López NJ, Smith PC, Gutiérrez J. Higher risk of preterm birth and low birth weight in women with periodontal disease. J Dent Res 2002;81:58-63.16. López NJ, Smith PC, Gutiérrez J. Periodontal therapy may reduce the risk of preterm low birth weight in women with periodontal disease: A randomized controlled trial. J Periodontol 2002;73:911-924. [Abstract]17. Goepfert AR, Jeffcoat MK, Andrews WW, et al. Periodontal disease and upper genital tract inflammation in early spontaneous preterm birth. Obstet Gynecol 2004;104:777-783.18. Jarjoura K, Devine PC, Perez-Delboy A, Herrera-Abreu M, DAlton M, Papapanou PN. Markers of periodontal infection and preterm birth. Am J Obstet Gynecol 2005;192:513-519.19. López NJ, Da Silva I, Ipinza J, Gutiérrez J. Periodontal therapy reduces the rate of preterm low birth weight in women with pregnancy-associated gingivitis. J Periodontol 2005;76:2144-2153. [Abstract]20. Offenbacher S, Boggess KA, Murtha AP, et al. Progressive periodontal disease and risk of very preterm delivery. Obstet Gynecol 2006;107:29-3621. Siqueira FM, Cota LO, Costa JE, Haddad JP, Lana AM, Costa FO. Intrauterine growth restriction, low birth weight, and preterm birth: Adverse pregnancy outcomes and their association with maternal periodontitis. J Periodontol 2007;78:2266-2276. [Abstract]