Determinants of periodontal health in pregnant women and association with infants’ anthropometric status: a prospective cohort study from Eastern Uganda

One in four adults suffers from periodontal disease and pregnancy is associated with increased susceptibility [1‐5]. Pregnancy affects a woman’s hormonal balance and acts as a modifying factor of the pathogenesis of periodontal disease [6]. Although pregnancy does not cause periodontal diseases, numerous studies have confirmed increased gingival inflammation occurring between the second and eight months of gestation [6]. Thus, pregnant women have a higher incidence of gingivitis compared with their non-pregnant counterparts and the prevalence rates vary between 36% and 100% [7, 8]. A recent study of pregnant women in Brazil revealed a prevalence of periodontal disease of 47% indicating a strong need for initiating oral care during early pregnancy [2]. Lieff et al. [9], observed an increase in attachment loss between the moment women enrolled into their study at <26 weeks gestation and at delivery. Similar results have been reported from other studies conducted in developed as well as developing countries [10‐17]. Nevertheless, a number of case-control studies reported no difference in periodontal health between pregnant subjects and non- pregnant controls, leaving findings in this area inconsistent and conflicting.

Apart from the effects of hormonal changes, other factors such as HIV infection, lack of dental care, poor oral hygiene, smoking, low educational level, low employment status, increased age and ethnicity contribute to a worsened periodontal condition during pregnancy [2, 18]. Lieff et al. [9] observed that black women were more likely than white women to have periodontal disease both at enrollment into their study and delivery. Taani et al. [11] reported more aggravated at periodontal condition in pregnant women compared with non-pregnant controls even after adjusting for the effect of increasing age, long standing poor personal hygiene and changed psychological status. Unemployment, reflecting low socio-economic status, inaccessibility to dental care and unawareness of oral hygiene was significantly associated with increased gingival inflammation scores. There is a lack of research on socio-behavioral and clinical determinants of pregnant women’s periodontal status and oral hygiene, particularly in developing countries.

Recent meta-analytical reviews have reported a positive association between maternal periodontitis and adverse pregnancy outcomes, such as prematurity, low birth weight infants and preeclampsia [19‐22]. Both case control- and prospective studies have shown that periodontitis precedes preterm birth [23, 24]. Severe periodontal disease in pregnant mothers has also been reported to be linked with low birth weight at full term delivery in low-and middle income populations [23, 24]. A recent survey of pregnant women in Tanzania did not confirm the hypothesis of periodontal disease as a risk factor for preterm low birth weight [17]. Studies on treatment of periodontal disease during pregnancy are conflicting in that some have shown a reduction in the rate of preterm births, whereas others have not [24]. Low birth weight infant delivery, defined by the World Health Organization as a birth weight of <2500 gram presents a major health problem in developing countries. Low-birth weight babies are at increased risk of serious health problems, lasting disabilities and even death. Preterm-low birth weight is the second leading cause of infant death and its incidence continues to rise [5, 25, 26]. In the Ugandan population, 24% of neonatal deaths are purported to be due to preterm births [27].

Participating women were members of a multi-center cluster-randomized behavioral intervention trial: Safety and Efficacy of Exclusive Breast feeding Promotion in the era of HIV in Sub Saharan Africa –PROMISE EBF (Id NCT00397150 at http://​clinicaltrials.​gov) conducted in Uganda, Burkina Faso, Zambia and South Africa. The aim of PROMISE EBF was to develop and test an intervention to promote exclusive breastfeeding, to assess its impact on infants in African contexts and to strengthen the evidence base regarding optimal duration of exclusive breastfeeding [28]. In Uganda, the Mbale district was purposively selected as the intervention site. The units for randomization were clusters made up of 1–2 villages with an average of 1000 inhabitants (35 infants per year given a birth rate of 3.5%). All pregnant women in 24 clusters (18 rural and 6 urban), were eligible for the study. Clusters were selected according to accessibility in terms of being on a main road with reasonable standard especially during the rainy season, access to church, school, trading center and water from the village well. The women were recruited into the PROMISE EBF study between January 2006 and June 2008. There were a total of 6 interviews and 1 oral examination scheduled for each participant: a recruitment interview and a clinical oral examination at 7 months of gestational age, followed by interviews of mothers and anthropometric measurement of infants at 3-, 6-, 12-, and 24 weeks post- partum. Women who did not intend to breastfeed and infants born with serious diseases or deformities that prevented breastfeeding were excluded from participation.

A total of 886 pregnant women were eligible for the study and information was obtained from 877 of the participants (participation rate 98.9%) (mean age 25.6, SD 6.4). Of these, 713 women participated in the oral interview and clinical oral examination conducted between 2006 and 2008. The number of participants satisfied a sample size of 800 pregnant women calculated for the oral sub study, assuming a prevalence of tooth loss (i.e. at least one tooth lost) of 50%, a precision of 0.05 and a design effect of 2. At three weeks post- partum, 635 mothers were re-interviewed in their homes and 593 of their infants were assessed for anthropometric status. The procedures of recruitment and participation are detailed in the PROMISE EBF study profile [18, 29‐31]. Ethical Clearance was obtained from the Ethical Review Board, Faculty of Medicine, Makerere University. Written consent was obtained from all participants in the study and prior to each examination and interview a verbal consent was obtained.

Periodontal disease during pregnancy was influenced by greater maternal age, single marital status, being primiparous, having received oral health information and using mosquito bed nets. Wasting status at 3 weeks postpartum was worse in infants of mothers with periodontal problems. Studies of periodontal condition during pregnancy vary with respect to ethnicity and socio-demographic features of the populations investigated. Definitions of disease parameters, such as gingival bleeding, probing pocket depths and attachment and alveolar bone loss and the number of sites per tooth and number of teeth examined per individual also vary across studies [2, 5, 9]. These inconsistencies influence the results and limits valid comparisons between studies [37]. Numerous studies have shown that estimates of prevalence and severity as well as distributional characteristics of periodontal condition vary according to the method used for recording [37]. In this study, the CPI, [15, 16] was deemed an appropriate screening system since the clinical examinations were carried out at household settings. Although CPI scores 1 and 2 primarily reflect gingival inflammation, bleeding and calculus, conditions that do not necessarily progress to periodontal destruction, this index has been used extensively to reflect periodontal health in various populations in Europe, Kenya, Tanzania and Ethiopia [3, 4].

The overall findings of this study, characterized by low prevalence of bleeding, high prevalence of calculus and infrequent occurrence of shallow pockets, is consistent with findings in African populations in general [3, 4]. Studies using the CPITN have indicated that the prevalence of severe periodontal disease is low in the general populations of sub Saharan Africa [3, 4, 13, 14]. Moreover, the oral hygiene has been described as poor with plaque and calculus accumulating with increasing age [3]. The low prevalence of periodontal condition revealed by the present study was supported by a moderate prevalence of women showing poor oral hygiene status (12.1%), but was at odds with the substantial rate of self- reported periodontal symptoms (65.4%). Considering the limitations associated with the CPI scoring system and that this method does not provide a complete measure of periodontal disease, the present study demonstrated shallow pockets of 4–5 mm amounting to 1.7% in the urban and 0.2% in the rural participants, whereas 67% presented with any sign of periodontal disease (CPI ≥ 1). Miyazaki et al. [7] found a prevalence of 31% shallow pockets in pregnant Japanese women, whereas the prevalence of any periodontal disease amounted to 97%. Compared to an earlier study of Ghanaian pregnant women with mean number of bleeding sextants amounting to 3.2 and 1.9 in second and third trimester [14], the corresponding figure of 1.2 observed in this study was moderate. Consistent with the present results, other studies from sub- Saharan Africa have reported on shallow pockets among postpartum women in all age groups, whereas in pregnant women shallow pockets have been observed in 25-45-year-olds during the second trimester and in 35–45 year -olds during the third trimester [17]. Some studies have documented no worsening of clinical attachment loss during pregnancy, but a progressive increase in plaque and bleeding scores with a peak level occurring at seven months of gestational ages [10‐12]. Since only one clinical examination was performed, it was not possible to affirm any exacerbation of pre-existing or early periodontal disease among the pregnant women investigated in the present study. Moreover, using the number of women with bleeding, calculus and pockets instead of attachment loss as periodontal outcome variables focuses the extent of the infection at the time of the survey more than the consequences of previous disease processes.

Several determinants of maternal oral health indicators were identified, suggesting that subgroups of women may be at increased risk of deteriorating oral health during pregnancy. A pattern of positive associations between periodontal disease and age has been found in numerous studies globally and among Ugandan students where similar age distributions were observed regarding both aggressive- as well as chronic periodontitis [38, 39]. Moreover, the present findings accord with a strong social gradient in periodontal disease reported previously [2, 18]. In this study, use of mosquito bed nets and marital status were used as surrogates of socio-economic status [40]. Previous studies with the cohort of Mbale women revealed that perinatal deaths were associated with preterm birth, malaria and complicated deliveries [27]. In this study, pregnant women who reported use of mosquito bed nets were less likely to have periodontal disease while single mothers were more likely to have poor OHI-S status. In accordance with some previous investigations but at odds with others, parity was positively associated with periodontal disease. This association was maintained even after adjusting for relevant covariates such as the possible effect of increasing age which is often associated with multiple pregnancies [11, 39]. Scheutz and Baelum [13] reported no association between parity and tooth loss, whereas attachment loss was found to be more pronounced in multiparous than in primiparous women in Tanzania. A positive association between parity and periodontal disease might be attributed to accumulated tissue destruction across time rather than an intrinsic parity- related abnormality.

Although about one third of the pregnant women investigated had their last dental visit within the previous 6 months, primiparous women were more likely to document such a visit than their multiparous counterparts. The limited frequency of dental visits corroborates findings pertaining to pregnant women from other cultures [41, 42]. One possible explanation is that pregnant women tend to avoid dental care because they believe poor oral health to be a routine consequence of pregnancy and because of fear that dental care would harm the fetus [5]. Although The American Academy of Periodontology recommends maintenance of oral hygiene during pregnancy, only about 20-40% of pregnant US women receive dental care [42]. In this study, higher proportions of women educated above primary school level and having received information on oral health had visited a dentist. Thompson et al. [39] observed that women with a low level education were less likely than their higher educated counterparts to report tooth cleaning during pregnancy. Participants of the present study came from an area with low economic resources and where the access to dental care is poor. In such settings prevention and treatment of oral diseases has not been strongly advocated [43]. Low rates of dental attendance might be attributed to few available dental clinics and a low priority given to this service in the Ugandan socio-cultural context where prevailing levels of communicable diseases, poverty and restructuring of health systems are more important concerns.

Height- for- age scores at 3 weeks postpartum were significantly lower in infants of mothers having CPI scores > 0 and poor oral hygiene during pregnancy. Under the assumption that poor anthropometric status at 3 weeks post- partum may be one possible consequence of preterm low birth weight, the present results may add to the evidence of a positive relationship between periodontal problems and poor pregnancy outcomes [19‐24]. However, this finding should be interpreted with caution as many unmeasured factors, such as feeding practice immediately after birth and infectious diseases, might have confounded the observed association in this study. Moreover, it cannot be ruled out that unaccounted factors such as smoking, maternal weight, anemia and other medical problems may have confounded the estimates presented, either positively or negatively. Future studies involving populations with a low risk of severe periodontitis should enroll larger samples. As the present study was based on recruited women in PROMISE-EBF, it was limited by the sampling methodology used. Thus, the results of this study should not be generalized to the whole population of Ugandan women or to the population of pregnant women in Mbale. However, prospective and interventional studies provide the most important evidence related to a possible relationship between maternal oral health status and infants’ post- partum anthropometric status, and the major strength of this present study is its longitudinal design.

Socio-demographic factors and oral health information were independently associated with periodontal status, oral hygiene and use of dental care in pregnant women in Mbale Eastern Uganda. Second, the height- for- age status at 3 weeks postpartum was significantly worse in infants of mothers having CPI scores > 0 and poor oral hygiene. Efforts to prevent oral diseases during pregnancy should be part of the local state and national health policy agenda and should focus on modifiable barriers to oral health care. When pregnant women have access to dental care and are motivated to use dental care on a regular basis both mother and child may experience benefits.

MW PhD is a researcher employed at Makarere University, Uganda. ANÅ is a professor of odontology employed at the University of Bergen, Norway. IO is a researcher employed at Makarere, JT is a medical doctor and professor at Makarere University, Uganda.

The study was part of the EU-funded project PROMISE-EBF (contract no INCO-CT 2004-003660, http://​www.​promiseresearch.​org). It was also financially supported by the Norwegian Research Council funded project Oral health in a global perspective (project number 156744)

Steering Committee:

Thorkild Tylleskär, Philippe Van de Perre, Eva-Charlotte Ekström, Nicolas Meda, James K. Tumwine, Chipepo Kankasa, Debra Jackson.

Participating countries and investigators:

(Country PI first, others in alphabetical order of surname)