We examined whether pregnant women with periodontal disease have an increased risk of preeclampsia, and we empirically evaluated the extent to which associations between periodontal disease and preeclampsia are dependent upon diagnostic criteria used to define periodontal disease operationally.
One hundred and fifty preeclampsia cases and one-fifty normotensive controls who delivered a singleton infant at term were enrolled. Periodontal examinations were performed within 48 h after delivery. Participants’ periodontal health status was classified, a priori, into four categories according to the extent and severity of periodontal disease. Putative risk factors for periodontal disease and preeclampsia were ascertained during in-person postpartum interviews using a structured questionnaire and by medical record abstraction. Logistic regression was used to estimate odds ratios (ORs) and 95% confidence intervals (95% CIs).
No clinically meaningful differences were observed between cases and controls with regard to periodontal parameters. After controlling for known confounders, severe clinical periodontal disease was not associated with an increased risk of preeclampsia (OR = 0.92, 95% CI: 0.26–3.28). In addition, there was no evidence of a linear increase in risk of preeclampsia with increasing severity of periodontal disease (P for trend = 0.65). When different diagnostic criteria previously used in other studies were used, the prevalence of periodontal disease varied substantially. However, the magnitude and direction of associations between periodontal disease and preeclampsia were largely similar regardless of the diagnostic criteria used to define periodontal disease.
This study provides no convincing evidence that periodontal disease is associated with preeclampsia risk among Thai women.
Preeclampsia, a vascular disorder of pregnancy characterized by high blood pressure (BP) and proteinuria, complicates ~5–10% of all pregnancies, and it is a major cause for maternal and perinatal mortality and morbidity worldwide, particularly in developing countries.1 Despite recent progress toward understanding the cause of preeclampsia, the etiology of this serious disorder remains elusive. Putative preeclampsia risk factors include advanced maternal age, multifetal pregnancies, maternal pre-pregnancy obesity, pregestational hypertension, renal disorders, and diabetes mellitus.2,3,4,5 Urinary tract infections, and more recently, periodontal disease, have been identified as potential preeclampsia risk factors.6,7 Available evidence implicates impaired placental implantation, placental hypoxemia, oxidative stress, chronic systemic inflammation, dyslipidemia, insulin resistance, and diffuse endothelial dysfunction in the pathogenesis of preeclampsia.8,9,10 Collectively, available evidence suggests a multifactorial etiology of preeclampsia. Moreover, recent evidence documenting increased risk of atherosclerotic disorders in men and nonpregnant women with periodontal disease11,12 coupled with emerging studies documenting increased preeclampsia risk with periodontal disorders have led investigators to postulate that chronic systemic inflammation, secondary to periodontal disease, may be important in the etiology and pathogenesis of preeclampsia.6
Boggess et al. in 2003 was the first group of investigators to report a link between maternal periodontal disease and risk of preeclampsia. Women with severe periodontitis, defined as ≥15 tooth sites with pocket depths ≥4 mm, had a 2.4-fold increased risk of preeclampsia as compared with periodontally healthy women (odds ratio (OR) = 2.4; 95% confidence interval (CI): 1.1–5.3).6 This early finding has been corroborated by some,13,14,15,16,17,18,19,20,21 though not all investigators.22,23 Some investigators have speculated that heterogeneity in findings across studies may be attributable to differences in the populations studied and heterogeneity in diagnostic criteria used to define periodontal disease. For example, of the 10 published studies6,15,16,17,18,19,20,21,22,23 that focused on preeclampsia risk in relation to maternal periodontal disease, investigators have used at least five different diagnostic criteria6,16,18,23,24 when operationally defining levels of periodontal disease.
On the basis of the available literature, we conducted a case–control study to determine whether maternal periodontal disease is a risk factor for preeclampsia among Thai women. We also conducted a series of sensitivity analyses to evaluate the extent empirically to which associations between periodontal disease and preeclampsia are dependent on the diagnostic criteria used to define periodontal disease operationally.
Study population and selection of cases and controls. A case–control study using one control for each case of preeclampsia was conducted among women who delivered live births at King Chulalongkorn Memorial Hospital, Rajavithi Hospital, and Police General Hospital, Bangkok, Thailand between July 2006 and November 2007. Cases were women with a confirmed diagnosis of preeclampsia and who delivered a singleton infant. Preeclampsia cases were identified by daily monitoring all new admission to antepartum, labor and delivery wards of participating hospitals. Of the 158 eligible cases approached, 154 (97.5%) agreed to participate in the study. Controls were women who delivered a singleton infant at term (≥37 weeks of gestation) without evidence of preeclampsia or any hypertensive disorder during pregnancy and were selected from the same hospital of delivery. An eligible control, delivering immediately after a case patient, was approached and recruited for the study. Of the 161 controls approached, 154 (95.7%) agreed to participate in the study.
Four preeclampsia cases were subsequently excluded along with four matched controls because the cases were founded to have pregestational diabetes mellitus (n = 1) or chronic hypertension (n = 3) during the medical record review. One hundred and fifty preeclampsia cases who delivered a singleton infant and one-fifty normotensive controls who delivered a singleton infant at term (≥37 weeks of gestation) remained for analysis. All participants provided informed consent, and the research protocol was reviewed and approved by ethical committees of the Faculty of Medicine, Chulalongkorn University, Rajavithi Hospital, Police General Hospital and the Institutional Review Boards, Division of Human Subjects Research, University of Washington.
Data collection. After obtaining informed consent, enrolled participants were invited to take part in a 45-min in-person interview in which trained research personnel used a structured questionnaire to elicit information regarding maternal sociodemographic, lifestyle habits, oral health history, and medical and reproductive histories. Participants’ labor and delivery medical records and prenatal medical records were also reviewed by trained obstetric research nurses who used a standardized abstraction form. Information abstracted from medical records included participants’ pre-pregnancy weight, height, BP, pregnancy complications, and condition of the newborn.
Study participants underwent a full-mouth periodontal examination by one of the six trained and calibrated periodontists who were blinded to case–control status. Periodontal examinations were performed at the bedside on the postpartum wards using mouth mirrors, UNC-15 periodontal probes with an external light source within 48 h after delivery. Probing depth (PD) and recession were measured on all teeth except for the third molar in six locations (mesiobuccal, midbuccal, distobuccal, mesiolingual, midlingual, and distolingual). These measurements were made in millimeters and were rounded down to the whole millimeter. Clinical attachment loss (CAL) was calculated from PD and recession, and represented the distance from cementoenamel junction to the base of the periodontal pocket. Plaque and bleeding on probing were recorded dichotomously as either present or absent. Bleeding on probing was determined positive if hemorrhage occurred within 15 s after probing. At the end of periodontal examination, each participant was given instructions regarding dental treatment needs.
Analytical variable specification.
Preeclampsia: The diagnosis of preeclampsia was made using the current American College of Obstetricians and Gynecologists guidelines.25 These guidelines defined preeclampsia as sustained pregnancy-induced hypertension with proteinuria. Hypertension was defined as sustained BP readings of systolic BP ≥ 140 or diastolic BP ≥ 90 mm Hg (with reading taking place ≥6 h apart). The American College of Obstetricians and Gynecologists defined proteinuria as urine protein concentrations of ≥30 mg/dl (or 1+ on a urine dipstick) on ≥2 random specimens collected ≥4 h apart.
Maternal periodontal disease: Participants’ periodontal health status was classified, a priori, into four categories according to the extent and severity of periodontal disease using the following criteria advocated by Albandar:26 severe periodontitis (≥2 nonadjacent teeth with interproximal sites showing ≥6 mm CAL and ≥4 mm PD); moderate periodontitis (≥2 nonadjacent teeth with interproximal sites showing ≥5 mm CAL and ≥4 mm PD); mild periodontitis (≥1 teeth with interproximal sites showing ≥4 mm CAL and ≥4 mm PD). Women not fulfilling any of the above criteria were classified as being periodontal healthy (i.e., women without detectable levels of periodontitis). We also empirically evaluated the extent to which associations between maternal periodontal disease and preeclampsia are dependent upon the various case definitions used to classify women’s periodontal health status. Using a single dataset (150 preeclampsia cases and 150 normotensive controls), we classified participants’ periodontal disease status using case definitions advocated by CDC-AAP working group,27 Canakci et al.,16 Contreras et al.,18 Lopez et al.,24 and Boggess et al.6
Other covariates: Covariates considered in this analysis included maternal age and reproductive and medical histories. Also considered were maternal educational attainment, employment during pregnancy, annual household income, cigarette smoking, and alcohol consumption during pregnancy. Maternal age at the time of interview was expressed in years. Parity was reported as the number of previous pregnancies lasting >22 weeks gestation. Maternal educational attainment was based on self-reports. Pre-pregnancy body mass index was calculated as weight (in kilograms) divided by the square of height (in meters).
Statistical analysis. The distribution of maternal sociodemographic characteristics and medical and reproductive histories according to case and control status was examined. To estimate the relative association between maternal periodontal disease and risk of preeclampsia, conditional logistic regression procedures, taking into account the matching of cases and controls for hospital of delivery, were performed to calculate maximum likelihood estimates of ORs and 95% CIs, adjusted for potential confounding factors.28,29 Confounding was assessed by entering potential covariates into a logistic model one at a time, and then comparing the unadjusted and adjusted ORs. We considered the following covariates as possible confounders in these analyses: maternal age, educational attainment, parity, marital status, annual household income, employment during pregnancy, pre-pregnancy body mass index, onset of prenatal care, cigarette smoking, and alcohol consumption during pregnancy. Final logistic regression models included covariates that altered unadjusted ORs by at least 10%.29 Variables of a priori interest (e.g., maternal age, smoking status, and parity) were forced into final models. All statistical analyses were performed using Stata 10.0 (Stata, College Station, TX) software. All reported P values are two-tailed, and CIs were calculated at the 95% level.
Maternal sociodemographic, medical, and reproductive characteristics of preeclampsia cases and normotensive controls are presented in Table 1. Overall, preeclampsia cases and controls were similar with regards to maternal educational attainment, marital status, annual household income, employment during pregnancy, parity, smoking status, and onset of prenatal care. Preeclampsia cases, however, were older, heavier, and more likely to report a history of hypertensive disorder during previous pregnancies.
As seen in Table 2, no clinically meaningful differences were observed between preeclampsia cases and controls with regard to mean PD, mean CAL, mean number of missing teeth, mean percent of sites with plaque and mean percent of sites exhibiting bleeding on probing. In addition, there were no significant differences in the mean percent of sites with periodontal PD ≥4 mm or CAL ≥4 mm.
Next, participants’ periodontal health status was classified, a priori, into four categories (i.e., periodontal healthy, mild, moderate, and severe periodontitis) according to the extent and severity of periodontal disease.26 We calculated ORs for preeclampsia for each level of clinical periodontal disease. As seen in Table 3, there was a modest but nonstatistically significant association between maternal severe periodontitis and preeclampsia risk (OR = 1.65; 95% CI: 0.58–4.71). However, after adjustment for possible confounding by maternal age, educational attainment, parity, pre-pregnancy body mass index, and smoking status, there was no association between maternal severe periodontitis and risk of preeclampsia (adjusted OR = 0.92; 95% CI: 0.26–3.28). In addition, there was no evidence of a linear increase in risk of preeclampsia with increasing severity of periodontal disease (P for trend = 0.65).
We next evaluated the extent to which various case definitions and criteria for diagnosis of periodontal disease used in previous studies6,16,18,24,27 had an impact on the association between maternal periodontal disease and risk of preeclampsia. The prevalence of periodontal disease varied substantially when different diagnostic criteria were used. When we used the criteria advocated by Albandar,26 25.4% of controls were classified as having moderate or severe periodontal disease. Corresponding prevalence estimates of moderate or severe periodontal disease among controls were 34.0% using the Boggess et al. criteria;6 6.0% Contreras et al. criteria;18 and 76.0% criteria proposed by the CDC-AAP working group.27 However, the magnitude and direction of associations between periodontal disease and preeclampsia were largely similar when different periodontal disease diagnostic criteria were used (Figure 1).
Odds ratios (ORs) and 95% confidence intervals (95% CIs) of preeclampsia in relation to periodontal disease defined using various diagnostic criteria. ORs were adjusted for hospital of delivery, maternal age, educational attainment, parity, pre-pregnancy body mass index, and smoking status during pregnancy.
In this study, we found no evidence of an association between maternal clinical periodontal disease and preeclampsia risk among Thai women after adjustment for possible confounders (adjusted OR = 0.92; 95% CI: 0.26–3.28). In addition, we found no evidence of a linear increase in risk of preeclampsia with increasing severity of periodontal disease (P for trend = 0.65). The findings of this study do not support the hypothesis that periodontal disease is an independent risk factor of preeclampsia.
To the best of our knowledge, there are at least 10 published studies that have focused on quantifying preeclampsia risk in relation to maternal clinical periodontal disease status during pregnancy6,15 or in the early postpartum period.6,16–23 Eight of the ten studies have documented positive associations between maternal periodontal disease and preeclampsia risk.6,15–21 For instance, Boggess et al., in their cohort study of 763 women in North Carolina, reported that women with severe periodontitis at delivery, defined as ≥15 tooth sites with PD ≥4 mm, had a 2.4-fold increased risk of preeclampsia as compared with periodontally healthy women (OR = 2.4; 95% CI: 1.1–5.3). However, periodontal disease at enrollment was not associated with preeclampsia risk.6 Contreras et al. in their study of 130 preeclampsia cases and 243 controls in Cali, Columbia, reported that women with moderate or severe periodontitis, defined as ≥2 tooth site with CAL ≥6 mm, PD ≥4 mm, and bleeding on probing, had a 3.32-fold increased risk of preeclampsia as compared with periodontally healthy women (OR = 3.32; 95% CI: 1.79–6.15).18 Cota et al., in their study of 109 preeclampsia cases and 479 controls in Brazil, reported that maternal periodontal disease, defined as ≥4 teeth with ≥1 sites with PD ≥ 4 mm and CAL ≥ 3 mm, was associated with an increased risk of preeclampsia (OR = 1.88; 95% CI: 1.15–3.06).19 In contrast, Castaldi et al., in their cohort of 1,562 Argentine women, found no association between maternal periodontal disease and preeclampsia risk (OR = 0.99; 95% CI: 0.70–1.40).22 Khader et al., in their study of 115 preeclampsia cases and 230 controls in Irbid, Jordan, also found no statistically significant association between clinical periodontal parameters and preeclampsia risk.23
Reasons for the differences in findings across studies are unclear. Some investigators have speculated that heterogeneity in findings may be attributable to differences in diagnostic criteria used to define periodontal disease. Of the 10 published studies6,15–23 focused on preeclampsia risk in relation to maternal periodontal disease, investigators have used at least five different diagnostic criteria6,16,18,23,24 when operationally defining levels of periodontal disease. However, our study showed that the magnitude and direction of associations between periodontal disease and preeclampsia were largely similar regardless of the diagnostic criteria used to define periodontal disease. Maternal clinical periodontal disease, defined by either a priori diagnostic criteria or others that have been previously used was not associated with preeclampsia risk. In addition, some investigators have speculated that failure to assess effect modification by cigarette smoking may contribute to variation in the results across studies.30,31 For example, Hyman et al. reported an association between periodontal disease and coronary heart disease; however, after stratifying by smoking status, the association was only evident among smokers.32 It is worth noting that our study, conducted in a population where the prevalence of cigarette smoking during pregnancy was very low (i.e., <4%), and the study by Khader et al.,23 conducted among nonsmokers, failed to identify an association between periodontal disease and preeclampsia.
Boggess et al. have suggested that pregnant women with periodontal disease may have transient translocation of periodontal pathogens to the uteroplacental unit, inciting placental inflammation or oxidative stress in pregnancy, which ultimately produced placental damage and the clinical manifestations of preeclampsia.6 Although this hypothesis is biological plausible, causal inferences from their study are limited by the fact that it is not clear whether periodontal disease precedes preeclampsia onset. In fact, the author found no evidence of a positive association between maternal periodontal disease at enrollment (<26 weeks of gestation) and preeclampsia risk later in pregnancy. Preeclampsia, by definition is a disorder with onset after the completion of 20 weeks gestation.25
The absence of consistency in finding across studies, coupled with an ambiguous temporal relationship of an association between maternal periodontal disease and preeclampsia risk raises the possibility that previously observed associations may be noncausal. Alternatively, associations between clinical periodontal disease and preeclampsia may only be evident in some susceptible populations; made so by the presence of other environmental or genetic risk factors. For example, differences in the distribution and virulence of specific periodontal pathogens may contribute to heterogeneity across studies. This thesis is supported by studies documenting differences in the periodontal pathogens detected across population sampled from diverse geographic locations.33–35 Future studies that investigate specific characteristics of the types and virulence of periodontal pathogens may help to move this literature forward.
This study has several strengths including the relatively large sample of preeclampsia cases, the fact that we used only well-trained, calibrated, and blinded periodontists to examine all participants. The high participation rates for cases and controls (97.5 and 95.7%) also served to attenuate concerns about selection bias. Several limitations, however, should be considered when interpreting results from our study. First, our study did not specifically characterize the pathobiology of oral infection (i.e., bacterial counts or antibody levels to periodontal pathogens). Although the full-mouth recording periodontal parameters used in our study has been shown to be more effective in determining periodontal disease, several authors have suggested that clinical measures of periodontitis may not adequately represent the systemic burden of periodontal disease.36,37 Because the clinical signs of periodontal disease are a result of periodontal pathogens interacting with the host’s immune and inflammatory response, it is likely that including measurement of this interaction between microorganisms and maternal host response may provide a means to more fully characterize the exposure that we think of as periodontal disease.38 Second, although we adjusted for many potential confounders, we cannot exclude the possibility that some residual confounding by factors not measured in our study (e.g., urinary tract infection) may have influenced reported risk estimates. Finally, our study was designed to have 80% statistical power to detect a 1.90-fold increase in preeclampsia risk associated with maternal periodontal disease. Although we consider it unlikely that limited power could be responsible for our null findings, we cannot rule out the likelihood of missing weaker associations or associations that may only be present among specific preeclampsia subtypes (i.e., early vs. late onset).
In conclusion, the results of this case–control study do not provide convincing evidence that periodontal disease is associated with preeclampsia risk among Thai women. Future studies that investigate specific characteristics of periodontal pathogens and host’s immune and inflammatory response may help to move this literature forward.
This research was supported by the Rachadapiseksompoj Faculty of Medicine Research Fund (RA 20/49), Chulalongkorn University and the Multidisciplinary International Research Training (MIRT) Program of the University of Washington, School of Public Health and Community Medicine. The MIRT Program is supported by an award from the National Institutes of Health, National Center on Minority Health and Health Disparities (T37-MD001449). We thank the staff nurses at King Chulalongkorn Memorial Hospital, Rajavithi Hospital and Police General Hospital in Bangkok, Thailand for their assistance in data collection. This research was done as partial fulfillment for the requirements of a PhD degree by one of the authors (V.L.) in the Department of Epidemiology, University of Washington, School of Public Health and Community Medicine, Seattle, Washington, USA.
The authors declared no conflict of interest.