Table 10. MIC and zone diameter breakpoints for staphylococci Comments 1-3 relate to urinary tract infections (UTI) only. 1 These recommendations are for organisms associated with uncomplicated urinary tract infections only. For complicated infections and infections caused by Staphylococcus aureus and Staphylococcus epidermidis, which are associated with more serious infections, systemic recommendations should be used.
Jop090532 982.99Volume 81 • Number 7 Periodontal Disease Activity Measuredby the Benzoyl-DL-Arginine-Naphthylamide Test Is AssociatedWith Preterm BirthsHui-Chen Chan,* Chen-Tsai Wu,† Kathleen B. Welch,‡ and Walter J. Loesche§ Background: Infection is a risk factor for preterm birth. This study was conducted in the field and addressed the link be-tween periodontal pathogens measured with the benzoyl-DL-arginine-naphthylamide (BANA) test and preterm birth.
Methods: This prospective study was performed in Chan- ghua, Taiwan. Periodontal examinations included the plaqueindex, papillary bleeding scores, and measurement of theBANA enzyme in plaque samples at the second and third tri- The prevalence of preterm birth (PB) is often associated with in- mesters. Independent variables included maternal demo- graphic characteristics, previous pregnancy histories, risk because of the activation of the immune factors, plaque and gingivitis scores, and current pregnancy system, by microorganisms that trigger the release of inflammatory cytokines Results: There were 19 (7%) preterm deliveries among the such as interleukin-8, interleukin-1, and 268 subjects. A history of a previous preterm birth and low tumor necrosis factor-alpha.1 This in- birth weight, frequency of prenatal visits, preterm uterine con- flammatory cascade and microbial en- tractions, antepartum hemorrhages, placenta previae, and dotoxins derived from invading bacteria preterm premature rupture of membranes were significantly stimulate the production of prostaglan- related to preterm birth (P = 0.035, 0.027, <0.001, 0.025, dins,2 which can increase uterine con- 0.006, 0.014, and <0.001, respectively). Maternal weight tractions that result in preterm labor gain was higher with a normal term delivery (P = 0.003). Mul- causing a PB.3 Among the infections tivariable logistic regression analyses showed that the number that have been associated with PBs is of BANA-infected sites in the third trimester (odds ratio [OR]: periodontal disease.1 5.89; 95% confidence interval [CI]: 1.5 to 31.6), maternal Periodontal disease is an inflamma- weight gain (OR: 0.78; 95% CI: 0.65 to 0.91), antepartum tory process in the periodontal tissue that hemorrhages (OR: 10.0; 95% CI: 2.2 to 46.9), and preterm is initiated by the microorganisms in the premature rupture of membranes (OR: 12.6; 95% CI: 3.97 to subgingival plaque.4 Although the mi- 42.71) had significant influences on preterm-birth outcomes.
crobiota of the subgingival plaque is Conclusions: BANA-positive plaque in the third trimester complex, there appears to be a selection was associated with preterm births after controlling for other for anaerobes when there is an active risk factors. The BANA test can be used to screen pregnant disease process, especially for the women at chairside and/or bedside to apply suitable interven- Gram-negative, proteolytic species Por- tion tactics. J Periodontol 2010;81:982-991.
phyromonas gingivalis, Tannerella for-sythia (previously T. forsythensis), and Treponema denticola.5 P. gingivalis, T.
BANA; periodontal disease; preterm birth.
forsythia, and T. denticola were foundin higher levels in subgingival plaques * Department of Orthodontic and Pediatric Dentistry, School of Dentistry, University of of women who delivered PB or low–birth Michigan, Ann Arbor, MI.
† Dental Department, Show Chawn Memorial Hospital, Changhua, Taiwan.
‡ Center for Statistical Consultation and Research, University of Michigan.
women who delivered normal term § Marcus Ward Professor Emeritus School of Dentistry, Professor Emeritus of Microbiology & Immunology, School of Medicine, University of Michigan.
(NT) babies.6,7 This association was con-comitant with an increase in maternal J Periodontol • July 2010 Chan, Wu, Welch, Loesche serum inflammatory mediators and antibodies.6-8 P.
of the University of Michigan (Hum00000327) and gingivalis was also detected in the amniotic cavity Hospital Institutional Research Board of the Show of pregnant women with threatened premature la- Chawn Memorial Hospital (940607). Subjects were bor7-9 and along with other periodontal species in recruited between September 2005 and June 2006 the placentas of women with preeclampsia.10 during their second-trimester visit to the OBGYN clinic.
Efforts to detect these periodontal pathogens in Experimental Protocol dental plaque have included DNA probes, cultural Pregnant women with a singleton gestation in the sec- procedures, microscopic measures, and immuno- ond trimester (gestational age: 13 to 28 weeks) were logic reagents. P. gingivalis, T. forsythia, and T. den- recruited, whereas subjects with multiple gestations ticola possess a trypsin-like enzyme that can who were undergoing fertility treatment were ex- hydrolyze the synthetic trypsin substrate benzoyl- cluded. The subjects signed consent forms after the DL-arginine-naphthylamide (BANA).11 The presence format, purpose, and nature of the study were pre- of these organisms in subgingival plaque can be de- sented to them. Information regarding maternal char- termined by the ability of the plaque to hydrolyze acteristics, which included demographic information BANA12 using a 5-minute chairside assay. In the de- data, previous pregnancy histories, and risk factors tection of P. gingivalis, T. forsythia, and T. denticola, (history of smoking and alcohol consumption), was the BANA test had a 92% sensitivity and a 70% spec- collected via a written questionnaire. At the second- ificity compared to DNA probes and polyclonal im- trimester appointment, dental measurements, in- munologic reagents.13 When the BANA test was cluding the plaque index score (PI)20 and papillary compared to checkerboard DNA–DNA hybridization bleeding score (PBS)21 were made, and measurement using highly specific, whole-genomic DNA probes of the BANA enzyme in plaque samples using the to P. gingivalis, T. forsythia, and T. denticola, it had BANA testi was performed. The PI, PBS, and BANA a sensitivity of 95% and was most effective for the de- test were repeated when the subjects returned for their tection of these organisms when their levels in subgin- third-trimester visit (gestational age: 25 to 40 weeks).
gival plaque were high, i.e., in the initial diagnosis of Pregnancy outcomes, which were collected from hos- chronic periodontitis.14 The results suggest that the pital records, included the patient's status during BANA test could be used as a surrogate for DNA pregnancy (gestational age at first prenatal visit, fre- probes in the detection of these bacterial species in quency of prenatal visits, prepregnancy weight and plaque samples.
height, and weight and height at the last prenatal Probing measurement of pocket depth and attach- visit), complications during pregnancy, type of deliv- ment often require a separate visit to a dental clinic.
ery, gender of the infant, gestational age, and birth The BANA test can be obtained at chairside and/or weight at delivery. The participants did not have to bedside15 and was shown to be significantly associ- make any visits other than their regularly scheduled ated with probing depth16,17 and to predict future at- prenatal visits.
tachment loss after initial treatment.18 The BANA testwas used in epidemiologic studies where it was found Dental Measurements to be an important explanatory variable for attach- Because periodontal disease frequently begins sub- ment loss in seniors19 and to correlate with the Com- gingivally in the interdental papilla around the poste- munity Periodontal Index of Treatment Needs index.16 rior teeth, plaque samples from the four interdental This suggested that the BANA test could be used sites between the first and second molars of each under field conditions, such as the waiting room of quadrant were collected. If teeth were missing, the a hospital, to obtain information concerning the peri- plaque sample was removed from the mesial or distal odontal status of the patient. The purpose of this pro- side of the remaining tooth.
spective study was to investigate the relationship The PI score was recorded on a scale of 0 to 3 at the between the presence of BANA-positive species in interdental site from which the plaque sample would subgingival plaque during the second and third tri- be obtained for the BANA test. No plaque disclosing mesters with the subsequent development of PB un- solution was used. After the PI was measured, the der field conditions.
supragingival plaque was removed from the site,and a soft wooden toothpick¶ was inserted between MATERIALS AND METHODS the first and second molars in each quadrant. When the toothpick was removed, the PBS21 was recorded This study was conducted in the Obstetric and Gyne- at the interdental papilla on a scale from 0 to 5. The cological (OBGYN) clinics in Show Chawn Memorial subgingival plaque adherent to the toothpick was Hospital and San Ann Hospital in Changhua, Taiwan,after all procedures were reviewed and approved i BANAMet LLC, Ann Arbor, MI.
by the Health Sciences Institutional Research Board ¶ STIM-U-DENT, Johnson & Johnson, New Brunswick, NJ.
Periodontal BANA Scores and Preterm Births Volume 81 • Number 7 used for the BANA test. A separate toothpick was categoric variable (e.g., educational level and ethnic- used to obtain each plaque sample, and both sides ity). Chi-square tests were calculated to assess the of each toothpick were wiped onto the lower strip of relationship of categoric variables (e.g., ethnicity) the BANA card. The upper strip of the BANA card with PB (yes or no). Independent samples t tests were was moistened with distilled water, and the card was carried out to determine whether there was a signifi- folded at the perforation mark so that the lower and cant difference in the means of continuous variables upper reagent strips met. The folded card was incu- (e.g., time of dental measurement) for PB versus NT bated in a special BANA test–designed heater at outcomes. The potential independent variables for 55C for 5 minutes. After incubation, the lower re- PB included in the multivariable logistic regression agent strip containing plaque was discarded in a man- model were antepartum hemorrhage, a preterm pre- ner appropriate for contaminated material. The color mature rupture of the membrane, infected second on the upper strip was recorded by the consensus and third BANA plaque samples, ethnicity, and weight of two examiners (HC and Natalie Grossman, gain during pregnancy. The statistical significance of BANAMet, Ann Arbor, MI) with no blue = negative, each variable and odds ratio (OR) were calculated us- faint blue = weakly positive, and blue = positive. The ing the Firth bias correction25 for a small sample size.
intraexaminer k agreement was 0.92, and the interexa- The various predictors were entered in different com- miner k agreement was 0.90. For statistical analyses, binations into the models, and the model with the weakly positive and positive results were recorded highest Nagelkerke maximum rescaled R2 value26 as positive. A woman was defined as being BANA in- and lowest Akaike's information criterion (AIC)27 fected when plaques from ‡2 of the four sample sites value was selected. An a level of 0.05 was used for were either weakly or strongly BANA positive.
all statistical tests. Statistical analyses were carriedout using a software package.** Definition of Pregnancy OutcomesThe primary outcome was PB, which was defined as a birth occurring <37 weeks (<259 days) of gestation There were 317 women who consented to participate and referred to the time that elapsed between the first in this study, of whom 13 were seen in the Show day of the last menstrual period and the day of deliv- Chawn Memorial Hospital and 304 were seen in the ery.22 The obstetricians (Hong-Chen Chang, Jinn- San Ann Hospital. Eighteen subjects were excluded Fa Bai, and Biau Hsiung Chen, San Ann Hospital; for the following reasons: multiple gestations (n = 1), and Hui-Yin Chiu, Show Chawn Memorial Hospital, early pregnancy termination (n = 2), and no informa- Changhua, Taiwan) were masked to the subjects par- tion on the pregnancy outcome (n = 15). As a result, ticipating in the study and had no knowledge of the 299 subjects had BANA test results obtained in the women's periodontal data.
second and/or third trimesters. To address the effect Statistical Analyses of BANA-infected plaque on pregnancy outcomes, Power calculations were made prior to initiation of the the 268 subjects who had BANA test results in the study, assuming that the total PB rate in Taiwan is second and third trimesters and pregnancy out- ;7%23 and that 60% of our participants would be comes were included in the statistical analysis. Of BANA positive.24 We expected to be able to recruit these 268 subjects, 194 (72.4%) subjects had >10 300 patients in the time frame of the study and calcu- prenatal appointments. A total of 207 women lated that a two-group x2 test with a 0.05 one-sided (77.2%) delivered their babies vaginally, and 61 sub- significance level would have 76% power to detect jects (22.8%) underwent cesarean section deliveries.
the difference between a 10% incidence of PB babies Nineteen women had a PB outcome (7%). There was in the BANA-positive group and 3% PB babies in the no significant difference in gender between PB and NT BANA-negative group, assuming sample sizes in infants (P >0.999).
the two groups of 180 and 120, respectively. Power Maternal Demographic Characteristics calculations were performed using statistic software.# The ages of subjects ranged from 16 to 43 years with Independent variables included maternal demo- a mean – SD of 27.2 – 4.3 years. Sixty-seven women graphic characteristics (age, occupational level, edu- were ‡30 years of age. There was no significant differ- cational level, ethnicity, and previous medical ence in maternal age, occupational level, educational history), pregnancy history (previous pregnancies, level, and previous medical history among subjects previous PBs, previous LBW infants, and abortion his- who had PBs and subjects who had NT births (Table tory), risk factors for PB (prepregnancy body mass 1). Thirteen of 268 subjects were non-Taiwanese, index [BMI], smoking, and alcohol consumption),complications during pregnancy, second- and third-trimester dental measurements (the PI, PBS, and # nQuery Advisor version 7.0, Statistical Solutions, Cork, Ireland.
** PASW Statistics 18, SPSS Inc, Chicago, IL; SAS for Windows, release BANA test). Frequency data were generated for each 9.2, 2002-2008, SAS Institute, Cary, NC.
J Periodontol • July 2010 Chan, Wu, Welch, Loesche Relationship of Maternal Characteristics to PB for 268 Subjects Pregnancy Outcome Normal (n = 249) (n [%]) PB (n = 19) (n [%]) Less than senior high school Senior high school or further Previous medical history Medical condition Previous pregnancy history First-time pregnancy Number of previous pregnancies Previous spontaneous abortion Previous artificial abortion Prepregnancy BMI (kg/m2) First prenatal care Periodontal BANA Scores and Preterm Births Volume 81 • Number 7 Table 1. (continued ) Relationship of Maternal Characteristics to PB for 268 Subjects Pregnancy Outcome Normal (n = 249) (n [%]) PB (n = 19) (n [%]) * Fisher exact test.
and these women had a higher proportion of PBs the second and third trimesters (P <0.001 and than the Taiwanese women (P = 0.055). None of these P <0.001, respectively) than women who were BANA non-Taiwanese women had graduated from senior negative (fewer than two of the four sampling sites) high school, whereas ;50% of the Taiwanese women (data not shown). There was a tendency for the prev- had schooling beyond high school.
alence of a BANA infection to increase from the sec-ond to third trimesters (Table 3).
Previous Pregnancy History (Table 1) There were no significant differences in the gesta- A total of 119 women (44.4%) were pregnant for the tional age at the time of the examination among the first time, and they were no more likely than the multi- women destined to have an NT birth or a PB (Table parious women to have a PB (>0.999; Table 1). Preg- 4). In the second and third trimesters, there was no nant women who had a previous PB and previous LBW significant difference among women who had PBs infants were more likely to have a PB (P = 0.035 and and women who had NT births in the adequacy of oral hygiene procedures or in the presence of gingivitis, al- Risk Factors and Pregnancy Complications though there was a slight tendency for the prevalence There was no significant relationship between a PB of gingivitis to be higher in the second trimester in and prepregnancy BMI, smoking, alcohol consump- women destined to have a PB (Table 4). In the third tion during pregnancy, and time of the first prenatal trimester, but not the second trimester, there was visit (P >0.05 for all comparisons; Table 1). The BMI a tendency for women with ‡2 BANA-positive sites at the first prenatal visit for the women destined to to be in the PB group compared to women with <2 be in the NT group (mean: 21.04; SD: 3.57) and BANA-positive sites (P = 0.08) (Table 4). Although women destined to be in the PB group (mean: the BANA results for each individual did not change 20.61; SD: 2.33) was comparable (P = 0.604). There- between the second and third trimesters for the major- after, women in the NT group gained significantly ity of the subjects, there was a tendency for a BANA more weight than the women in the PB group (P = infection to decrease in the NT subjects compared 0.003) (data not shown). Ninety-eight percent of to the PB subjects in the third trimester. Twelve per- the women never smoked and the four women who cent of the NT subjects had a decrease in BANA infec- were current smokers were in the NT group. Only tions, whereas only 5% of PB subjects had a decrease.
two women reported consuming alcoholic beverages Twenty-six percent of the PB subjects showed an in- (Table 1). There was a highly significant increase in crease in BANA infections, whereas 17% of the NT PB for women who had a preterm uterine contraction, subjects showed an increase in BANA infections (data antepartum hemorrhage, placenta previa, and pre- not shown). The sample size was too small to show term premature rupture of membrane (P = 0.025, 0.006, 0.014, and <0.001, respectively) (Table 2).
Logistic regression analysis was used to model the occurrence of PBs based on the values of the various Dental Measurements explanatory variables. The following variables were Approximately 80% of the women had good or fair oral not significant in any of the models: age, education hygiene, and there was no change in oral hygiene as and occupation, smoking status, alcohol consumption, they progressed from the second to third trimesters oral hygiene status, presence of gingivitis, second- (Table 3). Sixty-six percent of the women had gingi- trimester BANA results, prepregnancy BMI, number vitis in the second trimester, and this prevalence in- of previous pregnancies, and whether the pregnancy creased significantly to 78% in the third trimester.
was the subject's first pregnancy. The model with Women who were BANA positive (i.e., ‡2 sites of the highest maximum square value and lowest AIC the four sampling sites were positive or weakly posi- is shown in Table 5. BANA-infected plaque samples tive) were more likely to have gingival bleeding in in the third trimester (OR: 5.9), ethnicity (OR: 5.6), J Periodontol • July 2010 Chan, Wu, Welch, Loesche Relationship of Pregnancy Complications to PB for 268 Subjects Normal (n = 249) (n [%]) PB (n = 19) (n [%]) Genitourinary infection Hemorrhage at <28 weeks Preterm uterine contraction Antepartum hemorrhage PPROM = Preterm premature rupture of membrane; PIH = pregnancy-induced hypertension; PGD = pregnancy gestational diabetes.
* Fisher exact test.
antepartum hemorrhage (OR: 10.0), and the preterm periodontal disease with PB and/or LBW infants.28 premature rupture of membranes (OR: 12.6) were sig- This variability in disease definition has lead to contra- nificantly positive predictors of PBs, whereas maternal dictory results with regard to the role of periodontal dis- weight gain (OR: 0.78) was a significantly negative ease in adverse pregnancy outcomes.28 Although predictor of PBs, after adjusting for other potential risk periodontal disease is regarded as an infection, only factors. Although BANA-infected plaque samples in occasionally are markers of infection used to recognize the second trimester were not significantly related to this infection, and then it is usually the host response, PBs, this variable was kept in the model because of as noted by bleeding, the visual appearance of tissue its contrast with the third-trimester BANA results.
inflammation, and the presence of inflammatorymarkers. This current prospective study analyzed peri- odontal pathogens in dental plaque, albeit indirectly by Periodontal disease is traditionally defined on the basis an enzyme test, to address the association between of clinical morbidity about the teeth, such as probing periodontal disease and adverse pregnancy outcomes.
depth, clinical attachment loss, and radiographic bone Our population-based study demonstrated a link be- loss. There is no consensus as to what constitutes the tween BANA-positive dental plaques in the third tri- threshold for periodontal disease as documented by mester and PBs.
the fact that at least 14 different definitions and 50 Periodontal disease, as a source of persistent infec- different measurements have been used to associate tion, has been indicated by increased serum C-reactive Periodontal BANA Scores and Preterm Births Volume 81 • Number 7 protein (CRP) levels.29 In pregnant women, elevated with an increased risk for preterm delivery7 and were CRP levels were associated with periodontal disease detected at higher levels in women who delivered pre- in African American women30 and with an increased term LBW infants6 and in the placentas of women risk of preeclampsia.31 In this regard, the value of mi- with preeclampsis.10 crobial tests to diagnose a periodontal infection would DNA probes were used to establish the connection seem worthwhile. Although the subgingival plaque among P. gingivalis, T. denticola, and T. forsythia with flora is bacteriologically complex, P. gingivalis, T. for- adverse pregnancy outcomes in the cited studies.7,32 sythia, and T. denticola have emerged as major peri- They were also used to show that periodontal treat- odontal pathogens.5 These species were associated ment in the second trimester could significantly reducethe levels of these species in plaque samples.32 The use of DNA probes is a laboratory-based procedurethat requires equipment and resources that were not Comparison of Categoric Variables for available for the present study, whereas the BANA test Dental Measurements Between the could be performed at chairside and lended itself to Second and Third Trimesters the type of field study described. The BANA test ap-pears to be a reliable surrogate for the use of DNA probes in the detection of P. gingivalis, T. denticola, and T. forsythia in plaque samples.14 Bayingana33 showed that the BANA test was more reflective of gin-gival conditions during pregnancy than were DNA Periodontal disease, as a response to a chronic in- fection, shares risk factors with a PB.1 The logistic re- gression model with the Firth correction for a small sample size was performed to control for the possible confounding effects of other predictors. This model in- dicated that the odds of having a PB were 5.9 times higher for women in the third trimester with ‡ 2 infected BANA-positive or weakly BANA-positive sites com- pared to women with fewer than two BANA-positive * McNemar test.
or weakly BANA-positive sites after controlling for † PI score ‡2 in <50% of sites.
other variables. Other investigators6,7 obtained bacte- ‡ Two or more sites bled after measurement by the toothpicks.
§ Two or more sites that were BANA positive or weakly positive.
riologic data from women in the second trimester and Relationship of Dental Measurements to PB Second Trimester (n = 268) Third Trimester (n = 268) Gestational age at sampling (n [mean days])* Oral hygiene (n [%])† Gingivitis (n [%])† BANA-infected plaque samples (n [%])† * Independent samples t test.
† Fisher exact test.
‡ PI score ‡2 in <50% of sites.
§ Two or more sites bled after measurement by the toothpicks.
i Two or more sites were BANA positive or weakly positive.
J Periodontol • July 2010 Chan, Wu, Welch, Loesche Relationships of Predictors to PB Based on Logistic Regression Analysis for 268 SubjectsWho Had Plaque Sampled for the BANA Test in the Second and Third Trimesters* Explanatory Variables Antepartum hemorrhage Premature rupture of membranes Second-trimester BANA infection§ Third-trimester BANA infection§ df = degrees of freedom; CI = confidence interval.
* All variables were included in the model simultaneously. Age, gender, oral hygiene, gingivitis, history of PBs, history of LBW, and premature contraction were not significant in the model. Firth bias correction was used for the analysis Nagelkerke R2 = 0.412; AIC = 94.87.
§ BANA-infection mean plaque removed from ‡2 sites were BANA positive or weakly positive.
found that the levels of eight plaque bacteria, including In one intervention study,35 an average of 1.3 ses- P. gingivalis, T. forsythia, and T. denticola, tended to be sions of scaling and root planing was associated with higher in the second trimester in mothers who deliv- a ‡2-mm loss of attachment at ‡4 sites in 41% of the ered preterm babies than in mothers who delivered women. Scaling and root planing often causes bac- term babies.7 Our results show that 58% of the women teremia, the intensity of which increases with the se- in the second trimester had a BANA infection, but that verity of periodontal disease,38-40 and increases the no connection between a BANA infection and PB could level of interleukin-6,41,42 which has been indicated be shown until the third trimester. This could have im- as a risk for PB.43 It is possible that mechanical de- portant implications for the timing of treatment.
bridement without a concurrent usage of an antimi- Because periodontal disease is preventable and crobial agent may cause a bacteremia or incite an treatable, treating periodontal disease during preg- acute inflammatory response.44 In this regard, the nancy should improve pregnancy outcomes. Two strategy for reducing PBs by periodontal intervention large, well controlled, intervention studies34,35 that might consider restricting the treatments to women used debridement procedures (i.e., scaling and root who have a periodontal infection in the early third tri- planing was delivered in the second trimester) were mester, include the use of antimicrobial agents, and unsuccessful in reducing PBs. A study36 that began provide intervention at the gestational age of 28 to in the second trimester and continued into the third trimester and included an antimicrobial agent (i.e., There is evidence that suggests that ethnicity might a 0.12% chlorhexidine rinse) in addition to debride- play a role in PBs.1 In our study, 4.9% of the subjects ment was successful in reducing PBs. This difference were non-Taiwanese and they were more likely to in outcomes suggests that scaling and root planing in have a higher rate of an adverse pregnancy outcome the second trimester was not enough to reduce PBs.
compared to the Taiwanese women (P = 0.055). The The importance of timing of the treatment was shown non-Taiwanese women tended to have a lower social by a study37 in which women who had periodontal economic status as indicated by the educational level disease and were hospitalized with a threatening PB compared to the Taiwanese women (Table 1). Also, were randomly assigned to a treatment group or to married immigrants faced problems of adaptation, a non-treatment group.37 The treatment was pro- communication difficulties, a lack of family support vided in the third trimester, at ;32 weeks, and con- from their home town, and barriers to healthcare sys- sisted of oral hygiene instructions, scaling and root tem use at the beginning of their lives in Taiwan.45 planning, and polishing of teeth with a fluoride paste.
Smoking exhibits a dose-dependent relationship The babies of the treated women were delivered at with PBs, as does a very high consumption of alco- 37.5 weeks and weighed 3,079 g, which was signifi- hol.1,46 In Taiwan, women rarely smoke or use alco- cantly more than the 2,602-g infants who were born hol or drugs, and especially do not do so during at 36.1 weeks to the women in the comparable con- pregnancy. The self-reported data regarding smoking and alcohol consumption were only 1.5% and 0.7%, Periodontal BANA Scores and Preterm Births Volume 81 • Number 7 respectively. As a result, our study provided informa- School of Dentistry, University of Michigan, for advice tion on the association between periodontal disease and guidance, and to BANAMet, Ann Arbor, Michigan, and PBs without these confounding factors. In Taiwan, for their BANA tests and incubators. This research was the National Health Insurance provides for 10 paid funded by a gift from Oralife, Toronto, Ontario, to the prenatal examinations during pregnancy. Almost all School of Dentistry, University of Michigan. Dr. Loesche (i.e., 98%) of our participants had their first prenatal has received financial support from BANAMet and is consultation at <12 weeks gestational age, which a partner of BANAMet, the manufacturer of the BANA would eliminate inadequate prenatal care as a risk test. Drs. Chan, Wu, and Welch report no conflicts of factor for a PB.47,48 interest related to this study.
A history of a PB or a history of LBW was not signif- icant in the adjusted models. Both histories are usually risk factors for PBs.1 Possible reasons for the lack of 1. Goldenberg RL, Culhane JF, Iams JD, Romero R.
significance include the small number of PBs or that Epidemiology and causes of preterm birth. Lancet2008;371:75-84.
the association of these factors with PBs is through 2. Romero R, Hobbins JC, Mitchell MD. Endotoxin stim- an underlying BANA infection so that adjusting for ulates prostaglandin E2 production by human amnion.
BANA infection removes the association. We tested Obstet Gynecol 1988;71:227-228.
for the latter possibility by removing the BANA data 3. Goldenberg RL, Hauth JC, Andrews WW. Intrauterine from the models and still found no association of a his- infection and preterm delivery. N Engl J Med 2000; tory of a PB or LBW to be significant predictors.
4. Pihlstrom BL, Michalowicz BS, Johnson NW. Periodon- In our population, women with £10 prenatal visits tal diseases. Lancet 2005;366:1809-1820.
had more PBs compared to women who had >10 pre- 5. Loesche WJ, Grossman NS. Periodontal disease as natal visits (P <0.001). This result is consistent with a specific, albeit chronic, infection: diagnosis and the shorter gestational age and lower weight gain ob- treatment. Clin Microbiol Rev 2001;14:727-752.
served in the PB women.
6. Offenbacher S, Jared HL, O'Reilly PG, et al. Potential pathogenic mechanisms of periodontitis associatedpregnancy complications. Ann Periodontol 1998;3: Traditional measurements for diagnosing periodontal 7. Lin D, Moss K, Beck JD, Hefti A, Offenbacher S.
disease focused on clinical morbidity and often resulted Persistently high levels of periodontal pathogens as- in an inconsistent diagnosis and the inability to recog- sociated with preterm pregnancy outcome. J Peri- nize active disease.28 In this regard, studying the anaer- 8. Hasegawa K, Furuichi Y, Shimotsu A, et al. Associa- obic bacterial burden and the inflammatory response tions between systemic status, periodontal status, may be more critical than measuring probing depths.
serum cytokine levels, and delivery outcomes in To our knowledge, our study provides a new insight pregnant women with a diagnosis of threatened pre- by addressing the infectivity progression with BANA- mature labor. J Periodontol 2003;74:1764-1770.
associated periodontal pathogens and suggests that 9. Leo´n R, Silva N, Ovalle A, et al. Detection of Porphyr- the third-trimester bacterial status of the subgingival omonas gingivalis in the amniotic fluid in pregnantwomen with a diagnosis of threatened premature plaque may be an important predictor of PBs. The abil- labor. J Periodontol 2007;78:1249-1255.
ity of the BANA test to detect anaerobic periodontal 10. Barak S, Oettinger-Barak O, Machtei EE, Sprecher H, pathogens makes it a useful tool for chairside screening Ohel G. Evidence of periopathogenic microorganisms of at-risk populations such as pregnant women.
in placentas of women with preeclampsia. J Periodon-tol 2007;78:670-676.
11. Loesche WJ. The identification of bacteria associated with periodontal disease and dental caries by enzy- This research was performed by Dr. Hui-Chen Chan in matic methods. Oral Microbiol Immunol 1986;1:65-72.
partial fulfillment of the requirements for a master's 12. Loesche WJ, Syed SA, Stoll J. Trypsin-like activity in thesis in Pediatric Dentistry. The authors thank Dr.
subgingival plaque. A diagnostic marker for spiro- Hui-Yin Chiu at Show Chawn Memorial Hospital and chetes and periodontal disease? J Periodontol 1987; Dr. Jinn-Fa Bai, Dr. Hong-Chen Chang and Dr. Biau 13. Loesche WJ, Lopatin DE, Giordano J, Alcoforado G, Hsiung Chen, San Ann Hospital, for providing access Hujoel P. Comparison of the benzoyl-DL-arginine- to clinic patients and clinical facilities. Also, thanks to naphthylamide (BANA) test, DNA probes, and immu- Mrs. Natalie Grossman, consultant for BANAMet, for nological reagents for ability to detect anaerobic her participation in helping to read BANA results, and periodontal infections due to Porphyromonas gingiva- to all of the subjects for participating in this research.
lis, Treponema denticola, and Bacteroides forsythus. J The authors gratefully acknowledge Dr. Fenno, De- Clin Microbiol 1992;30:427-433.
14. de Andrade JA, Feres M, de Figueiredo LC, Salvador partment of Biologic & Materials Sciences, School SL, Cortelli SC. The ability of the BANA Test to detect of Dentistry, University of Michigan, and Dr. Taichman, different levels of P. gingivalis, T. denticola and T.
Department of Periodontics and Oral Medicine, forsythia. Brazilian Oral Research 2010; in press.
J Periodontol • July 2010 Chan, Wu, Welch, Loesche 15. Rubenfire M, Grossman NS, Kaciroti N, Apsey DJ, 33. Bayingana C. The prevelance of members of the ‘‘red Loesche WJ. Anaerobic dental flora and the acute complex'' in pregnant women as revealed by PCR and coronary syndrome. Coron Artery Dis 2007;18:111- BANA hydrolysis. [Thesis]. Bellville, South Africa: University of the Western Cape; 2005;1-93.
16. Grisi MF, Novaes AB, Ito IY, Salvador SL. Relationship 34. Michalowicz BS, Hodges JS, DiAngelis AJ, et al.
between clinical probing depth and reactivity to the Treatment of periodontal disease and the risk of BANA test of samples of subgingival microbiota from preterm birth. N Engl J Med 2006;355:1885-1894.
patients with periodontitis. Braz Dent J 1998;9:77-84.
35. Offenbacher S, Beck JD, Jared HL, et al. Effects of 17. Apsey DJ, Kaciroti N, Loesche WJ. The diagnosis of periodontal therapy on rate of preterm delivery: periodontal disease in private practice. J Periodontol a randomized controlled trial. Obstet Gynecol 2009; 18. Loesche WJ, Giordano J, Hujoel PP. The utility of the 36. Lopez NJ, Da Silva I, Ipinza J, Gutierrez J. Periodontal BANA test for monitoring anaerobic infections due to therapy reduces the rate of preterm low birth weight in spirochetes (Treponema denticola) in periodontal dis- women with pregnancy-associated gingivitis. J Peri- ease. J Dent Res 1990;69:1696-1702.
19. Elter JR, Beck JD, Slade GD, Offenbacher S. Etiologic 37. Radnai M, Pal A, Novak T, Urban E, Eller J, Gorzo I.
models for incident periodontal attachment loss inolder adults. J Clin Periodontol 1999;26:113-123.
Benefits of periodontal therapy when preterm birth 20. Silness, J. and H. Lo ¨e. Periodontal disease in preg- threatens. J Dent Res 2009;88:280-284.
nancy. II. Correlation between oral hygiene and 38. Forner L, Larsen T, Kilian M, Holmstrup P. Incidence of periodontal condition. Acta Odontol Scand. 1964; bacteremia after chewing, tooth brushing and scaling in individuals with periodontal inflammation. J Clin 21. Loesche WJ. Clinical and microbiological aspects of chemotherapeutic agents used according to the spe- 39. Lafaurie GI, Mayorga-Fayad I, Torres MF, et al. Peri- cific plaque hypothesis. J Dent Res 1979;58:2404- odontopathic microorganisms in peripheric blood after scaling and root planing. J Clin Periodontol 2007;34: 22. Berg AT. Menstrual cycle length and the calculation of gestational age. Am J Epidemiol 1991;133:585-589.
40. Kinane DF, Riggio MP, Walker KF, MacKenzie D, 23. Ko YL, Wu YC, Chang PC. Physical and social pre- Shearer B. Bacteraemia following periodontal proce- dictors for pre-term births and low birth weight infants dures. J Clin Periodontol 2005;32:708-713.
in Taiwan. J Nurs Res 2002;10:83-89.
41. Forner L, Nielsen CH, Bendtzen K, Larsen T, 24. Lee Y, Tchaou WS, Welch KB, Loesche WJ. The Holmstrup P. Increased plasma levels of IL-6 in transmission of BANA-positive periodontal bacterial bacteremic periodontis patients after scaling. J Clin species from caregivers to children. J Am Dent Assoc 42. Ide M, Jagdev D, Coward PY, Crook M, Barclay GR, 25. Firth, D. Bias Reduction of Maximum Likelihood Wilson RF. The short-term effects of treatment of Estimates. Biometrika 1993;80:27-38.
chronic periodontitis on circulating levels of endo- 26. Nagelkerke, N. J. D. A note on a general definition of toxin, C-reactive protein, tumor necrosis factor-alpha, the coefficient of determination. Biometrika 1991;78: and interleukin-6. J Periodontol 2004;75:420-428.
43. Krupa FG, Faltin D, Cecatti JG, Surita FG, Souza JP.
27. Akaike, H. ‘‘A new look at the statistical model Predictors of preterm birth. Int J Gynaecol Obstet identification''. IEEE Transactions on Automatic Con- 44. Tonetti MS, D'Aiuto F, Nibali L, et al. Treatment of 28. Manau C, Echeverria A, Agueda A, Guerrero A, periodontitis and endothelial function. N Engl J Med Echeverria JJ. Periodontal disease definition maydetermine the association between periodontitis and pregnancy outcomes. J Clin Periodontol 2008;35:385- 45. Yang YM, Wang HH. Life and health concerns of Indonesian women in transnational marriages in Tai- 29. Slade GD, Offenbacher S, Beck JD, Heiss G, Pankow wan. J Nurs Res 2003;11:167-176.
JS. Acute-phase inflammatory response to periodon- 46. Goffinet F. Primary predictors of preterm labour. BJOG tal disease in the US population. J Dent Res 2000;79: 30. Horton AL, Boggess KA, Moss KL, Jared HL, Beck J, Sivapathasundaram V, Fearne JM, Curtis MA. Mech- Offenbacher S. Periodontal disease early in pregnancy anisms of risk in preterm low-birthweight infants.
is associated with maternal systemic inflammation Periodontol 2000 2000;23:142-150.
among African American women. J Periodontol 2008; 48. Blondel B, Dutilh P, Delour M, Uzan S. Poor antenatal care and pregnancy outcome. Eur J Obstet Gynecol 31. Ruma M, Boggess K, Moss K, et al. Maternal peri- Reprod Biol 1993;50:191-196.
odontal disease, systemic inflammation, and risk forpreeclampsia. Am J Obstet Gynecol 2008;198:389.
Correspondence: Dr. Hui-Chen Chan, North University Avenue, Department of Orthodontic and Pediatric Dentistry, 32. Novak MJ, Novak KF, Hodges JS, et al. Periodontal School of Dentistry, University of Michigan, Ann Arbor, MI bacterial profiles in pregnant women: response to 48109-1078. E-mail: [email protected].
treatment and associations with birth outcomes inthe obstetrics and periodontal therapy (OPT) study.
Submitted September 13, 2009; accepted for publication J Periodontol 2008;79:1870-1879.
February 25, 2010.
Target cell availability and the successful suppression of HIV by hydroxyurea and didanosine Rob J. De Boer AIDS 1998, 12:1567–1570 Keywords: Hydroxyurea, immunosuppression, target cell availability, 72 weeks of ddI–HU treatment, three out of sixpatients had no detectable plasma virus, and that there