Association Between MaternalPeriodontitis and an Increased Riskof PreeclampsiaLuıs Otavio Miranda Cota,* Alessandra Neves Guimaraes,*Jose Eustaquio Costa,* Telma CamposMedeiros Lorentz,* and Fernando Oliveira Costa*

Background: Periodontal disease has been considered asystemic exposure implicated in a higher risk of adverse preg-nancy outcomes. The aim of the present study was to deter-mine whether maternal periodontitis is associated with anincreased risk of preeclampsia.

Methods: A case-control study was conducted in a publichospital in Belo Horizonte, Brazil. During the study period,588 women, aged 14 to 46 years, were deemed eligible andhad data available for analysis. Maternal demographic andmedical data were collected from medical records. Pre-eclampsia was defined as blood pressure >140/90 mm Hgand ‡1+ proteinuria after 20 weeks of gestation. A periodontalexamination was performed postpartum. Maternal periodonti-tis was defined as the presence of four or more teeth with oneor more sites with a probing depth ‡4 mm and clinical attach-ment loss ‡3 mm at the same site. The effects of maternal age,chronic hypertension, primiparity, smoking, alcohol use, andnumber of prenatal visits were analyzed. Adjusted odds ratios(ORs) for preeclampsia were calculated using multivariatelogistic regression.

Results: The prevalence of periodontitis was 63.9% andpreeclampsia was 18.5%. Variables associated with pre-eclampsia were chronic hypertension (OR = 4.10; 95% confi-dence interval [CI] = 2.0 to 8.4; P = 0.001), primiparity (OR =2.40; 95% CI = 1.5 to 3.9; P = 0.004), maternal age (OR =1.07; 95% CI = 1.0 to 1.1; P = 0.001), and maternal periodon-titis (OR = 1.88; 95% CI = 1.1 to 3.0; P = 0.001).

Conclusion: Maternal periodontitis was determined to beassociated with an increased risk of preeclampsia. J Periodon-tol 2006;77:2063-2069.

KEY WORDS

Case-control study; periodontal disease/adverse effects;periodontitis; preeclampsia; pregnancy outcomes; riskfactors.

Periodontitis is an inflammatory dis-ease affecting supportive tissuesof the teeth, leading to progressive

destruction of connective tissue attach-ment and the alveolar bone. This destruc-tion is characterized by the formation ofa periodontal pocket, defined as the api-cal migration of the junctional epithe-lium and deepening of a gingival crevice.Inflamed periodontal tissues produce sig-nificant amounts of proinflammatory cy-tokines. Current information on etiologyshows bacterial infection, associated witha wider microbial diversity, as the primarycausative agent,1 although an appropri-ately susceptible host is also necessary.2

Because of its chronic inflammatory in-fectious nature, periodontitis has beenconsidered a systemic exposure impli-cated with causative agents in a varietyof systemic illnesses. Recent findingshave suggested that periodontal diseaseis associated with a higher risk of cardio-vascular diseases, atherosclerosis, andadverse pregnancy outcomes, such aspreterm birth and low birth weight.3-6 Aspreeclampsia and atherosclerosis sharesome epidemiologic risk factors, peri-odontal disease has recently been sug-gested as a risk factor of preeclampsia.7-9

Preeclampsia is a multifactorial disor-der affecting ;10% of pregnancies andcontributes significantly to maternal andperinatal morbidity and mortality.7,10

Indeed, preeclampsia is a pregnancy-specific syndrome occurring usually after

* Faculty of Dentistry, Federal University of Minas Gerais, Belo Horizonte, Brazil.

doi: 10.1902/jop.2006.060061

J Periodontol • December 2006

2063

20 weeks of gestation. It is characterized by abnormalvascular response to placentation; reduced organperfusion; vasospasm; activation of the coagula-tion system; inflammatory-like responses; oxidativestress; and some perturbations in volume and bloodpressure control, affecting the placenta, kidney,liver, and brain. Preeclampsia is determined by ma-ternal blood pressure elevation accompanied byproteinuria.11,12

Despite the efforts to understand the etiologic fac-tors of preeclampsia through extensive research in thepast decade, the various pathological mechanismsleading to this disorder are still poorly comprehended.Even with some advances in prenatal care, effectiveprimary preventive measures have been disappoint-ing, and the frequency of preeclampsia has notchanged.12,13 Several etiologies have been proposed,but a common final pathway is likely. Pathologic alter-ations are primarily ischemic in nature,12 and placen-tal dysfunction similar to atherosclerotic vascularchanges have been described previously.14 Indeed,pathophysiological alterations and some risk factorssimilar to those related to coronary disease have beenreported.12 The pathology has been described as theformation of atherosclerotic-like lesions of the spiralarteries. Endothelial damage in the placental vascularbed may be initiated by a number of mechanisms, re-sulting in oxidative and inflammatory vascular dam-age leading to the development of preeclampsia.15

Because placental vascular damage and athero-sclerosis are similar in nature, and because it has beenhypothesized that associations between periodontaldisease and systemic conditions do exist, the objec-tives of the present study were as follows: 1) to assesswhether an association between maternal periodonti-tis and preeclampsia is indeed present; and 2) to eval-uate the association of different demographic andbehavioral risk variables with the development of pre-eclampsia. The existence of the association betweenmaternal periodontitis and preeclampsia brings withit the possibility of identifying higher risk groupsand creating new approaches to the prevention ofpreeclampsia, thus improving prenatal care and thequality of life in addition to increasing the benefits topublic health.

MATERIALS AND METHODS

To assess the association between maternal peri-odontitis and preeclampsia, an unmatched case-control study with simultaneous gathering of dataon exposure and event was conducted.16 The presentstudy received approval from the Federal Universityof Minas Gerais Research Ethics Committee (ETIC005/04), and participants provided a written informedconsent. Eligible women were identified in a publichospital in Belo Horizonte, Minas Gerais, Brazil, within

48 hours of delivery and were invited to participate inthe study. Women were excluded from the study ifthey were <18 years of age without a legal guardian;had had a multiple gestation; had had a spontaneousabortion; had undergone in vitro fertilization; were di-agnosed with pregestational diabetes, heart disease,renal disease, or placental, cervical, and/or uterineabnormalities, or human immunodeficiency virus in-fection; or had any medical condition requiring antibi-otic prophylaxis for dental treatment.

During the 6-month period of data collection (Feb-ruary 2004 to June 2004), 3,381 women receivedmedical obstetric care in the study unit, and 1,115(32.9%) of these women were ineligible (accordingto the exclusion criteria). Of the 2,266 eligible women,778 were invited to participate in the study (accordingto the accessibility and availability of women inthe postpartum hospital routine). From this total, 32(4.1%) women refused to participate, and 158 (20.3%)women were excluded from the analysis because theyhad become ineligible (i.e., referred to another hospi-tal because of the availability of hospital beds or new-born’s postpartum complications). The final samplewas composed of 588 women aged 14 to 46 (meanage: 26.5 years; SD = 6.5 years), from a multiethnicgroup with a low socioeconomic status. Internalgroups were divided as follows: 1) control group =479 non-preeclamptic women; and 2) case group =109 preeclamptic women (Fig. 1).

Preeclampsia was defined as blood pressure >140/90 mm Hg on two separate occasions after week 20 ofgestation and at least 1+ proteinuria.7,11 Blood pres-sure was assessed by a trained medical group fromthe hospital unit under strict conditions to avoid obser-ver and instrumental errors. Assessments were fo-cused on the protocols of the Obstetric High RiskUnit from the hospital, which were based on previousreports.11,17,18 All patients included in this study were

Figure 1.Sampling strategy and study sample.

Maternal Periodontitis and Preeclampsia Volume 77 • Number 12

2064

tested for proteinuria, which was defined as a proteinconcentration ‡30 g/dl, which was equivalent to aurine dipstick value ‡1+ on two separate urine sam-ples taken 6 hours apart. If a positive result was found,24-hour urine specimens were collected to quantifyprotein excretion during the monitoring period incases of conservative routine procedures.

Demographic data, medical history, and detailedinformation on events of the pregnancy and deliverywere systematically obtained from medical records(Latin American Center for Perinatology form). Pa-tient’s medical records were thoroughly examined,and data were confirmed through patient question-naires upon oral examination. All medical data werereviewed by an obstetrician to confirm criteria for in-clusion and exclusion.

A full-mouth periodontal examination was per-formed within 48 hours postpartum in the hospitalbed with an artificial headlamp light. The hospitalbed was adjusted to the sitting position, and teeth werecleaned with sterile gauze when necessary. Twotrained periodontists, blinded to each patient’s iden-tity and medical history, were calibrated at the startof the study and 3 months after using measurementsof probing depth (PD) and clinical attachment loss(CAL). Intra- and interexaminer reliability scores wererecorded until satisfactory agreement was reached.All unweighted k scores were >0.75, and intraclasscorrelation coefficients were ‡0.90. Clinical measure-ments of periodontal status (i.e., PD and CAL) werecollected. PDs and attachment levels were recordedto the nearest millimeters with a University of NorthCarolina (UNC)-15 periodontal probe at six sites pertooth. PD was measured as the distance from the gin-gival margin to the base of the probeable gingivalcrevice. CAL was determined by measuring the dis-tance from the cemento-enamel junction to the baseof the probeable gingival crevice. For the purpose ofthis analysis, maternal periodontitis was defined asthe presence of four or more teeth with one or moresites with PD ‡4 mm and CAL ‡3 mm at the samesite.19

A univariate analysis was performed to evaluate thestatistical significance of the association of variablesof interest between case and control groups usingthe x2 and Fisher exact tests (Table 1). All variables(educational level, chronic hypertension, primiparity,smoking during pregnancy, alcohol use during preg-nancy, maternal age, prenatal visits, and maternalperiodontitis) were entered into a multivariable logis-tic regression model (Table 2) and removed step bystep using the backward elimination procedure.Chronic hypertension was defined as systolic bloodpressure ‡140 mm Hg or diastolic blood pressure‡90 mm Hg, confirmed by multiple measurements,and detected before conception or before gestational

week 20.11 Smoking during pregnancy and alcoholuse were defined as self-reported consumption duringany trimester of pregnancy. This study did not attemptto classify the exposure or patterns of consumptionbecause there was substantial within-person fluctua-tion during pregnancy. The rationale for multivariablemodeling was to control for the effect of potentialconfounders. All variables included in the final multi-variable model (Table 3) were determined to be inde-pendent, assessing the collinearity.20 All analyseswere performed using statistical software.†

RESULTS

Maternal demographic characteristics, medical his-tory, and obstetric data are detailed in Table 1. Signif-icant differences were observed between the case andcontrol groups in relation to chronic hypertensionand smoking during pregnancy. Almost 80% of thewomen in the study sample were aged 18 to 35 years.In both groups, ;70% of the women had undergonemore than six prenatal visits.

Periodontal status is detailed in Table 4. In relationto the number of subjects affected by PD ‡4 mm and

Table 1.

Demographic, Medical, and Obstetric Data

Variable

Group

P

Preeclamptic Non-Preeclamptic

N % N %

Maternal age (years)

<18 7 6.4 48 10.0 0.235

18 to 35 87 79.8 386 80.6

>35 15 13.8 45 9.4

Educational level

None/primary 58 53.2 289 60.5 0.165

Secondary/higher 51 46.8 189 39.5

Chronic hypertension 20 18.3 19 4.0 <0.001

Primiparity 47 43.1 164 34.4 0.086

Smoking duringpregnancy

8 7.3 77 16.1 <0.01

Alcohol use duringpregnancy

6 5.6 33 7.0 0.610

N prenatal visits

0 to 6 32 29.6 136 28.5 0.817

>6 76 70.4 341 71.5

Statistically significant variables are shown in bold.

† Statistical Package for Social Sciences, Version 9.0 for Windows, SPSS,Chicago, IL.

J Periodontol • December 2006 Cota, Costa, Costa, Lorentz, Guimaraes

2065

CAL ‡3 mm, statistically significant differences wereobserved between preeclamptic and non-preeclamp-tic women. Periodontitis was found in 74.3% of thesubjects in the case group and 61.6% of the subjectsin the control group. The number of sites affected byPD and CAL ‡3, ‡4, ‡5, and ‡7 mm was also statisti-cally different between the groups, with higher preva-lence rates in the preeclamptic group.

In the study sample, the prevalence of periodontitiswas 63.9%, whereas the prevalence of preeclampsiawas 18.5%.

The multivariable full model is presented in Table 2.The significant variables in this model were chronichypertension, primiparity, maternal age, and peri-

odontitis. The multivariable final model (Table 3)generated by the backward elimination procedure in-cluded chronic hypertension, primiparity, maternalage, and maternal periodontitis as variables associ-ated with preeclampsia. Women with periodontitishad 1.8 times higher likelihood of having preeclamp-sia than those without periodontitis. Smoking duringpregnancy, despite being significantly associated withnon-preeclamptic women in the univariate analysis,was not retained in the final model (P >0.05). No con-founding effect of smoking during pregnancy was ob-served in the association between periodontitis andpreeclampsia because no meaningful change in theestimated coefficient for periodontitis occurred. In ad-dition, no statistical interaction between smoking dur-ing pregnancy and periodontitis was observed.

DISCUSSION

Our study showed that maternal periodontitis, mea-sured by the presence of PD ‡4 mm and CAL ‡3 mm,is associated with an increased risk for the develop-ment of preeclampsia, regardless of the effects ofmaternal age, educational level, chronic hyperten-sion, primiparity, smoking during pregnancy, alcoholuse during pregnancy, and number of prenatal visits.The multivariable final model also revealed an associ-ation between preeclampsia and chronic hyperten-sion, primiparity, and maternal age.

Similarities between the pathophysiology conse-quences of preeclampsia and atherosclerotic disease

Table 2.

Multivariable Logistic Regression Analysis(full model)

Variable

Estimate

Coefficient Wald Value P

Constant -3.7782 30.7771 0.0001

Educational level -0.1725 0.5044 0.4776

Chronic hypertension 1.3239 12.5373 0.0004

Primiparity 0.7631 8.5110 0.0035

Smoking during pregnancy -0.8362 3.5498 0.0596

Alcohol use during pregnancy 0.3018 0.3232 0.5697

Maternal age 0.0737 14.8067 0.0001

Prenatal visits -0.0530 1.5528 0.2127

Maternal periodontitis 0.6089 5.7308 0.0167

Statistically significant variables are shown in bold.

Table 3.

Multivariable Final Model

Variable

Estimate

Coefficient

Wald

Value P OR 95% CI

Constant -4.3352 50.7789 0.0001

Chronichypertension

1.4184 15.3235 0.0001 4.13 2.03 to 8.40

Primiparity 0.8785 12.7071 0.0004 2.40 1.56 to 3.90

Maternal age 0.0722 15.0089 0.0001 1.07 1.04 to 1.11

Maternalperiodontitis

0.6288 6.3494 0.0117 1.88 1.15 to 3.06

Statistically significant variables are shown in bold.

Table 4.

Periodontal Status

Variable

Group

P

Preeclamptic Non-Preeclamptic

N % N %

Sites

PD ‡3 mm 6,249 58.9 27,097 55.9 <0.001

PD ‡4 mm 1,644 15.5 5,645 11.6 <0.001

PD ‡5 mm 690 6.5 2,657 5.5 <0.001

PD ‡7 mm 104 1.0 321 0.7 <0.001

CAL ‡3 mm 6,368 60.0 27,495 56.7 <0.001

CAL ‡4 mm 1,669 15.7 5,694 11.8 <0.001

CAL ‡5 mm 715 6.7 2,619 5.4 <0.001

CAL ‡7 mm 139 1.3 383 0.8 <0.001

Subjects

PD ‡4 mm* 94 86.2 343 71.6 <0.002

CAL ‡3 mm* 96 88.1 349 72.9 <0.001

Maternal periodontitis 81 74.3 295 61.6 <0.001

Statistically significant variables are shown in bold.* Four or more teeth with one or more sites affected.

Maternal Periodontitis and Preeclampsia Volume 77 • Number 12

2066

has been suggested in the literature,14,15,21,22 andseveral epidemiologic factors predispose the devel-opment of both events. However, despite the similar-ities between atherosclerosis and preeclampsia, littleis known about potential common putative riskfactors.7

Periodontal disease, a novel risk factor associatedwith atherosclerotic events, has been recently re-ported in the development of some adverse preg-nancy outcomes such as low birth weight23 andpreeclampsia.7-9 Some reports have associated peri-odontal disease, which may be categorized as achronic low-grade systemic stressor, with atheroscle-rosis, thromboembolic events,21,24,25 and hypercho-lesterolemia.26 In addition, oral pathogens have beendetected in atherosclerotic plaques where they mayplay a role in the development and progression ofatherosclerosis leading to coronary heart disease.27

Periodontal disease may provide a chronic burdenof endotoxin and inflammatory cytokines, whichserve to initiate and exacerbate atherogenesis andthrombogenesis.4 It is possible that the placentamay be similarly burdened in pregnant women whodevelop preeclampsia. Periodontal disease may serveprimarily as a vascular stressor and bring an addi-tional infectious/inflammatory burden to the placen-tal-fetal unit, thereby increasing the risk of pretermdelivery in preeclamptic women.28

In the hospital unit where this study was under-taken, routine diagnostics and management proce-dures for high-risk pregnancies are standardizedaccording to protocols determined and updated everyyear. On every occasion of such proceedings, themedical group is calibrated to the diagnostic criteriaestablished. It is unlikely that any inconsistencies inthe gathering of information have seriously biasedthe estimates of blood pressure, proteinuria, and med-ical history of the participants.

Quantification of protein excretion over 24 hours isthe gold-standard method to assess proteinuria in aclinical setting.However, the use of thismethod in largeepidemiological studies may not be feasible due to re-sources, time, and cost restraints. In the present study,proteinuria was defined as a protein concentration‡300 mg/l (equivalent to ‡1+ on a dipstick), and itwas assessed on two separate urine samples taken6 hours apart. The accuracy and usefulness of thismethod has been questioned in the literature.29 Never-theless, it has been used in recent epidemiologicalstudies evaluating the relationship between periodon-tal disease and preeclampsia.7,8

The prevalence of preeclamptic women in thisstudy was high (18.5%) compared to other reportsthat showed prevalence rates of 4% to 10%7,11,12,28

in the general population. It must be stated thatthe hospital where this study was undertaken is an

obstetric local referral center. Moreover, differencesrelated to population-specific factors such as socio-economic status, geographic characteristics, ethnic-ity, and the accessibility and efficacy of prenatalcare must be considered and could be responsiblefor higher prevalence rates in some populations, es-pecially in developing countries.30

Based on the criteria of four or more teeth with atleast one site with PD ‡4 mm and CAL ‡3 mm, theprevalence of periodontitis in the study sample washigh. However, the number of sites affected by PDand CAL ‡5 and ‡7 mm was low, showing a slightto moderate form of disease. Louro et al.23 showedsimilar patterns of periodontitis among postpartumBrazilian women. However, it is important to empha-size that previous reports showed prevalence ratesvarying from 50% to 90% according to the criteriaused for periodontitis definition.31,32

The univariate analysis showed that case and con-trol groups were similar in relation to maternal age,educational level, primiparity, alcohol use duringpregnancy, and number of prenatal visits. Differenceswere observed in relation to chronic hypertension andsmoking during pregnancy. Chronic hypertensionwas associated with the preeclamptic group, whichis in accordance with previous reports.12,13 Smok-ing during pregnancy was associated with non-preeclamptic women, showing a protective effectagainst the development of preeclampsia. Indeed,some findings in the literature do exist showing aparadoxical effect of smoking in pregnancies with adecreased risk of preeclampsia.12,33,34 It was hypoth-esized that this beneficial effect might be mediated bynicotine through inhibition of interleukin-2 and tumornecrosis factor production.13 However, it is importantto emphasize that smoking during pregnancy hasmany adverse effects on pregnancy outcomes, lead-ing to many maternal and/or fetal consequences.35

The final multivariable model generated by thisstudy included periodontitis associated with pre-eclampsia with an adjusted odds ratio (OR) = 1.88(95% confidence interval [CI] = 1.15 to 3.06). Canackiet al.8 also demonstrated that preeclamptic womenhad worse periodontal health than normotensivewomen, reporting an OR = 3.47 (95% CI = 1.07 to11.95). The results from Boggess et al.7 providedsome evidence that periodontal disease (OR = 2.4;95% CI = 1.1 to 5.3) and disease progression (OR =2.1; 95% CI = 1.0 to 4.4) are associated with anincreased risk of the development of preeclampsia.Moreover, recent findings from Oettinger-Baraket al.9 showed that clinical and immunological param-eters of periodontal disease were higher amongpreeclamptic women, suggesting a possible role ofperiodontal inflammation in the pathogenesis ofpreeclampsia.

J Periodontol • December 2006 Cota, Costa, Costa, Lorentz, Guimaraes

2067

The multivariable final model also included chronichypertension, primiparity, and maternal age associ-ated with an elevated risk of the development of pre-eclampsia.Previous reports indicate thatpreeclampsiais generally thought of as a disease of first pregnan-cies, and increasing maternal age and chronic hyper-tension are risk factors for preeclampsia.11,12

Additional studies are necessary to substantiateour findings and to clarify the nature of the associationbetween periodontal disease and preeclampsia, espe-cially those with prospective and intervention designsin different populations.

CONCLUSIONS

The findings from this study suggest that maternalperiodontitis is associated with an increased risk ofpreeclampsia. Associations between preeclampsiaand periodontitis should be interpreted with prudence,because the etiology of both events is likely multifac-torial. It is important to emphasize that primary health-care services must be able to diagnose and controlperiodontal disease during pregnancy. Managingperiodontal disease may represent a novel strategyto reduce the incidence and/or complications fromthis pregnancy hypertensive disorder. This would rep-resent a great advance in prenatal care and healthcare policies.

REFERENCES1. Hutter G, Schlagenhauf U, Valenza G, et al. Molecular

analysis of bacteria in periodontitis: Evaluation ofclone libraries, novel phylotypes and putative patho-gens. Microbiology 2003;149:67-75.

2. Kinane DF. Causation and pathogenesis of periodontaldisease. Periodontol 2000 2001;25:8-20.

3. Khader YS, Albashalreh ZSM, Alomari MA. Periodontaldisease and the risk of coronary heart and cerebrovas-cular diseases: A meta-analysis. J Periodontol 2004;75:1046-1053.

4. Beck JD, Offenbacher S. Systemic effects of peri-odontitis: Epidemiology of periodontal disease and car-diovascular disease. J Periodontol 2005;76:2089-2100.

5. Scannapieco FA, Bush RB, Paju S. Periodontal dis-ease as a risk factor for adverse pregnancy outcomes:A systematic review. Ann Periodontol 2003;8:70-78.

6. Lopez R. Periodontal disease, preterm birth and lowbirth weight. Evid Based Dent 2005;6:90-91.

7. Boggess KA, Lieff S, Murtha AP, Moss K, Beck J,Offenbacher S. Maternal periodontal disease is asso-ciated with an increased risk for preeclampsia. ObstetGynecol 2003;101:227-231.

8. Canakci V, Canakci CF, Canakci H, et al. Periodontaldisease as a risk factor for pre-eclampsia: A case controlstudy. Aust N Z J Obstet Gynaecol 2004;44:568-513.

9. Oettinger-Barak O, Barak S, Ohel G, et al. Severepregnancy complication (preeclampsia) is associatedwith greater periodontal destruction. J Periodontol2005;76:134-137.

10. Longo SA, Dola CP, Pridjian G. Preeclampsia andeclampsia revisited. South Med J 2003;96:891-899.

11. Working Group on High Blood Pressure in Pregnancy.Report of the national high blood pressure educationprogram working group on high blood pressure inpregnancy. Am J Obstet Gynecol 2000;183:S1-S22.

12. Sibai B, Dekker G, Kupferminc M. Pre-eclampsia.Lancet 2005;365:785-799.

13. Dekker G, Sibai B. Primary, secondary, and tertiaryprevention of pre-eclampsia. Lancet 2001;357:209-215.

14. Ramos JG, Martins-Costa S, Edelweiss MI, Costa CA.Placental bed lesions and infant birth weight in hyper-tensive pregnant women. Braz J Med Biol Res 1995;28:447-455.

15. Dekker GA, Sibai BM. The immunology of preeclamp-sia. Semin Perinatol 1999;23:24-33.

16. Gordis L. Epidemiology. Philadelphia: W.B. Saunders;2000:140-157.

17. Sibai B. Diagnosis and management of gestationalhypertension and preeclampsia. Obstet Gynecol 2003;102:181-192.

18. Oliveira SMJV, Arcuri EAM, Santos JLF. Cuff widthinfluence on blood pressure measurement duringthe pregnant-puerperal cycle. J Adv Nurs 2002;38:180-189.

19. Lopez NJ, Smith PC, Gutierrez J. Higher risk ofpreterm birth and low birth weight in women withperiodontal disease. J Dent Res 2002;81:58-63.

20. Altman DG. Relation between several variables. In:Practical Statistics for Medical Research, 1st ed. Lon-don: Chapman & Hall; 1991:325-364.

21. Beck JD, Offenbacher S, Williams R, Gibbs P, GarciaR. Periodontitis: A risk factor for coronary heart dis-ease? Ann Periodontol 1998;3:127-141.

22. Mercado F, Marshall RI, Klestov AC, Bartold PM. Isthere a relationship between rheumatoid arthritis andperiodontal disease? J Clin Periodontol 2000;27:267-272.

23. Louro PM, Fiori HH, Filho PL, Steibel J, Fiori RM.Periodontal disease in pregnancy and low birth weight(in Portuguese). J Pediatr (Rio J) 2001;77:23-28.

24. Beck J, Garcia R, Heiss G, Vokonas PS, OffenbacherS. Periodontal disease and cardiovascular disease.J Periodontol 1996;67(Suppl.):1123-1137.

25. Beck JD, Pankow J, Tyroler HA, Offenbacher S. Dentalinfections and atherosclerosis. Am Heart J 1999;138:S528-S533.

26. Katz J, Chaushu G, Sharabi Y. On the associationbetween hypercholesterolemia, cardiovascular dis-ease and severe periodontal disease. J Clin Periodon-tol 2001;28:865-868.

27. Haraszthy VI, Zambon JJ, Trevisan M, Zeid M, GencoRJ. Identification of periodontal pathogens in ather-omatous plaques. J Periodontol 2000;71:1554-1560.

28. Riche EL, Boggess KA, Lieff S, et al. Periodontaldisease increases the risk of preterm delivery amongpreeclamptic women. Ann Periodontol 2002;7:95-101.

29. Waugh JJ, Clark TJ, Divakaran TG, Khan KS, KilbyMD. Accuracy of urinalysis dipstick techniques inpredicting significant proteinuria in pregnancy. ObstetGynecol 2004;103:769-777.

30. Lopez-Jamarillo P, Casas JP, Serrano N. Preeclampsia:From epidemiological observations tomolecular mech-anisms. Braz J Med Biol Res 2001;34:1227-1235.

31. Locker D, Slade GD, Murray H. Epidemiology of peri-odontal disease among older adults: A review. Peri-odontol 2000 1998;16:16-33.

Maternal Periodontitis and Preeclampsia Volume 77 • Number 12

2068

32. Albandar JM, Brunelle JA, Kingman A. Destructiveperiodontal disease in adults 30 years of age and olderin the United States, 1988-1994. J Periodontol 1999;70:13-29.

33. Cnattingius S, Mills JL, Yuen J, Eriksson O,Salonen H. The paradoxical effect of smoking inpreeclamptic pregnancies: Smoking reduces theincidence but increases the rates of perinatal mor-tality, abruption placentae, and intrauterine growthrestriction. Am J Obstet Gynecol 1997;177:156-161.

34. Zhang J, Klebanoff MA, Levine RJ, Puri M, Moyer P.The puzzling association between smoking and hyper-

tension during pregnancy. Am J Obstet Gynecol 1999;181:1407-1413.

35. Billaud N, Lemarie P. Negative effects of maternalsmoking during the course of pregnancy. Arch Pediatr2001;8:875-881.

Correspondence: Dr. Fernando Oliveira Costa, Depart-ment of Periodontology, Faculty of Dentistry, FederalUniversity of Minas Gerais, Antonio Carlos Ave. 6627,Pampulha, Belo Horizonte 31270 901, Brazil. Fax: 55-31-34992430; e-mail: focperio@uol.com.br.

Accepted for publication July 29, 2006.

J Periodontol • December 2006 Cota, Costa, Costa, Lorentz, Guimaraes

2069