Prevalence, severity, and risk indicators of gingival inflammation in a multi-center study on South American adults: a cross sectional study

ABSTRACT Objectives: The aim of this study is to investigate the prevalence and severity of gingival inflammation and associated risk indicators in South American adults. Material and Methods: Multi-stage samples totaling 1,650 adults from Porto Alegre (Brazil), Tucumán (Argentina), and Santiago (Chile) were assessed. The sampling procedure consisted of a 4-stage process. Examinations were performed in mobile dental units by calibrated examiners. A multivariable logistic regression model was utilized for associating variables as indicators of gingival inflammation (GI) (Gingival Index ≥0.5). Statistical significance was set at 0.05. Results: A total of 96.5% of the adults have GI. Regarding the severity of GI, 22.5% of participants examined have mild GI, 74.0% have moderate GI, and 3.6% have severe GI. The multivariate analyses identify the main risk indicators for GI as adults with higher mean of Calculus Index (OR=18.59); with a Visible Plaque Index ≥30% (OR=14.56); living in Santiago (OR=7.17); having ≤12 years of schooling (OR=2.18), and females (OR=1.93). Conclusions: This study shows a high prevalence and severity of gingival inflammation, being the first one performed in adult populations in three cities of South America.


INTRODUCTION
feature detected in children and adults 1 . It is characterized by swelling, redness, and bleeding modulate the balance between humoral and cellassociated immune responses 28 . This clinical feature is characteristic of both gingivitis and periodontitis. GI is considered to be one major class of periodontal conditions, and is recognized to result from the increase in supragingival plaque and the ensuing host response 26 . Consequently, the prevention of plaque accumulation and early treatment of GI reduce the risks associated with the development of the more destructive periodontal disease 5 , which has also been associated with systemic conditions 8 .
Understanding the epidemiologic pattern of GI is essential for planning appropriate public-health services. It has been clearly demonstrated that plaque-induced GI is prevalent at all ages of the dentate population 2,29 . In recent decades, crosssectional and longitudinal epidemiological studies on periodontitis in adults were performed in Chile 12 and Brazil 14 . Moreover, analytical approaches designed to identify associated factors that could be risk Our main objective in this multi-center, populationbased, cross-sectional and epidemiological study is to investigate the prevalence, severity, and

Study design, sampling, and sample sizes
The present cross-sectional, representative study utilized stratified, multistage probability samples of the civilian, noninstitutionalized adult populations in three South American cities. Data were collected between January and July of 2014.
Our sampling approach considered various age subgroups (18-19; 20-29; 30-39; 40-49; published information that estimated a prevalence of gingivitis of 93.9% 21 (average Gingival Index we determined that a sample size of 550 adults would be appropriate for each of the three cities in the study. To do so, the formula to estimate the prevalence of a population (n=Z 2 1-/2 P(1-P)/e 2 ) was used.
The study participants were selected using a multi-staged probability sampling process. Age groups were formed according to a proportional approach to the base population registries in the total of urban administrative regions in Porto Alegre, Tucumán, and Santiago, according to the last census in each city, and considering differences in gender and age.
The sampling process consisted of four stages: City (primary sampling units -1 st stage); Tract census (2 nd stage); Blocks (3 rd stage); and Individuals within the age group (4 th stage). The cities were chosen according to logistics and interests in the three countries. Using maps of each city, primary census sectors were randomly chosen. The number of sectors in each city was determined according to the city size and census distribution. If the access to a primary census sector was not possible, the next available census sector was chosen. In each census sector, the blocks were randomly chosen. On each block, households were consecutively approached according to the sector starting point, until the number of participants expected for each sector was reached. Places such as nursing homes and commercial establishments were not included. When no potential participants were available for examination in a household, the next household was visited.
Candidates who have expressed an interest in participating in the study were selected based on the following criteria: 18 years of age or older, healthy, and with at least four permanent teeth. Were excluded from the study candidates needing antibiotic prophylaxis prior to dental examination, women who were pregnant or breastfeeding, individuals with fixed orthodontic appliances, or individuals who chronically used nifedipine, cyclosporine, phenytoin, or any prescription medicines that might interfere with the study outcome.
The protocol used for this study is in accordance with the Declaration of Helsinki and was reviewed and approved by the Institutional Review Boards of the University of Chile, Federal University of Rio Grande do Sul, and Maimonides University. All study participants were informed about the aims of the study and signed an informed consent form.

Clinical evaluation and sociodemographic and behavioral data
A sociodemographic and general health interview was conducted and a structured questionnaire, consisting of open and closed questions about demographics, habits, attitudes, and knowledge related to oral health, was designed and administered to all participants. This questionnaire was tested at each of the three study sites and adapted according to the necessities of the local population. Finally, a complete dental examination was performed on all participants in the study.
Prior to the initiation of the study, the principal investigators and examiners met in Porto Alegre in order to standardize diagnostic criteria with the reference examiner (CR). Intra-and inter-examiner Calculus Index, and Gingival Index were above 0.7. In addition, the structured questionnaire was standardized for each of the three study locations.
Each team consisted of one clinical examiner, totaling three dental examiners (RC, AT, FS) and each conducted exams using a manual periodontal probe (UNC-15) and mobile dental units. Good clinical practice standards were used and warranted. Periodontal clinical parameters were evaluated in all teeth, excluding third molars. The parameters evaluated were Visible Plaque Index (VPI), Calculus Index (CI), and Gingival Index (G-Index). Visual plaque assessment was determined using absence (0) or presence (1) of dental plaque according to Silness index was used to evaluate gingival health. Each tooth was divided into six surfaces, three facial and three lingual, as follows: 1) mesio-facial; 2) mid-facial; 3) disto-facial; 4) mesio-lingual; 5) mid-lingual; and 6) disto-lingual. Third molars and those teeth with cervical restorations or prosthetic crowns were excluded from the scoring procedure. Absence (0) or presence (1) of calculus was scored in lower anterior teeth (CI). Each tooth was divided into three lingual surfaces, as follows: disto-lingual, medio-lingual, and mesio-lingual. At the end of clinical examinations, those participants who were 2016;24(5):524-34 diagnosed with periodontal pathologic conditions were provided a written report of their condition and advised to seek an oral health consultation. estimates were isolated for interproximal sites.

Statistical analyses
Continuous data are presented as means ± SDs, and categorical variables are presented as sociodemographic, behavioral, and periodontal values. Chi-square tests were applied to compare distributions of periodontal variables between age groups and centers; to assess differences in the means and percentages, Mann-Whitney or Kruskall-Wallis tests were applied, and statistical analyses were performed using a statistical software package (Stata/IC 13.1). A multi-variable logistic regression model was built to assess the contribution of each variable (age, gender, smoking, and social or cultural factors). The occurrence of gingival addition, variables were analyzed in the model and 10%. VPI was dichotomized at 30%. Odds ratios

RESULTS
A total of 1650 adults from Porto Alegre, Tucumán, and Santiago participated in this study. Females represent 52.5% of the overall study participants, while one third of the participants are  Study participants from Santiago had the highest average gingival index (1.73) compared with those from Tucumán and Porto Alegre, at 1.11 and 1.12, respectively (p<0.001). The mean VPI in interproximal sites is higher than the mean plaque index for all sites examined (81% versus 75%, respectively) for the total study population. and Santiago are found between interproximal VPI and total VPI. The rates of interproximal VPI in Porto Alegre, and Tucumán are 89%, 80%, and 74%, respectively, while the mean interproximal VPI values in healthy adults are 45%, 30%, and 52%, for Santiago, Tucumán, and Porto Alegre, them (p<0.001) ( Table 3). The mean CI rates in Porto Alegre, 74% in Tucumán, and 81% in Santiago CI rates in healthy adults are 52%, 26%, and 41%, respectively (Table 3). GBI is higher in individuals compared with healthy adults (2%, p=0.001), and GBI is higher in Santiago as compared with Porto Alegre and Tucumán (Table 3).
Tables 4a and 4b present mean G-Index as   Table 3  Chi-square test, to assess differences between ages versus severity for each city and total (p-value*), differences between cities by each group (p-value**)  (Table 4b).
The multivariate logistic regression model, which was designed to assess indicators that could that subjects with higher CI mean (OR=18.59); (OR=2.18), and females (OR=1.93) are more likely adjusted for age, presence of diabetes, and selfreported hypertension and smoking (Table 5).

DISCUSSION
The population examined in the present study is comprised of a random sample of individuals aging 18 years or older from Porto Alegre (Brazil), Tucumán (Argentina), and Santiago (Chile). To a representative sample of adult populations from three Latin American cities. The sampling strategy that we employed was successful in achieving a representative and balanced sample of participants, since the individuals examined in each age group for each of the three cities are in the same proportion as in the whole study population for the three cities combined. In this study, 95.6% of the 1650 adults examined from three South American cities present data reported for adults from Jordan, China, and the United States of America, which demonstrated that and 93.9% of their respective populations 1,21,30 . Comparing our results with that from previous studies is somewhat hindered by the use of different nomenclature and diagnostic criteria across studies. The Community Periodontal Index (CPI) was used to report the occurrence of probing pocket Hungarian adult population, and gingival bleeding (CPI=1) was observed in 8% of the population 15 . The National Health and Nutrition Examination Survey III (NHANES III) conducted in the USA between 1988-1994 demonstrated that 50% of the using gingival bleeding as the criterion 3 . A study conducted in Italy, using bleeding on probing (BoP) as the criterion, determined that the prevalence of individuals showing at least one site positive for BoP was 99% 11 . All the aforementioned studies demonstrate that the occurrence of gingival poor gingival conditions. In addition, there is a large discussion in the literature regarding how to develop guidelines to suggest which criteria should be standardized for use in epidemiological studies for gingival conditions 16 . Concerning the severity of 74.0% have moderate gingival inflammation, contrast, representative studies conducted in Europe and Australia demonstrated that the prevalence of higher (17% and 19.7%, respectively) in these populations 15,17 . The multivariate logistic regression models that we employed show that a female with an increased likelihood of presenting gingival of diabetes, and self-reported hypertension and smoking. In this study, to live in Santiago was a risk indicator for presenting GI. It could be because these subjects had lower oral hygiene habits compared with those from the other cities. Indeed, subjects with presence of GI belonging to Santiago had a higher average percentage of VPI, compared with subjects with GI of Tucumán and Porto Alegre. In fact, the plaque index has been associated with worst level of oral hygiene in other populations 9 . There are limited studies at the Latin American level on the prevalence of periodontal conditions 27 , Although subjects belonging to Santiago present of it, is based with those reported for that country, where Chilean adults had a high prevalence of periodontal destruction and indicators of gingival 12 .
The present study also demonstrates a socioeconomic gradient in the oral health of the populations examined. Participants from lower gingival health as compared with participants in 2016;24(5):524-34 higher socioeconomic groups 4 . In this regard, different explanations have been raised in the those in lower socioeconomic situations are also more likely to smoke, and smoking has been found 10 . On the other hand, individuals with higher socioeconomic status are likely to have more positive attitudes regarding oral hygiene and self-care, and better access to available health care options. Therefore, low incomes and low levels of education seem to be variables with good predictability for periodontal diseases 7 in females found in the present study contradicts results from Greece, where women have been shown to have better oral hygiene and gingival status than men 22 . Perhaps this difference is more about culture than gender. It should be noted that only when adjusting the multivariate model by confounding variables, female gender appears as a possible risk indicator for the presence of gingival in females 25 .
Regarding the self-reported diabetes and hypertension among our participants, 7.4% have diabetes and 19.3% have hypertension. Participants with self-reported diabetes and hypertension present a significantly higher prevalence of these conditions. Previous studies have reported comparable results regarding this association between type 2 diabetes and periodontal disease 20 . of our total study participants, and these smokers present a slightly higher prevalence of gingival 25 (2002), in which study participants (soldiers of the German Armed Forces serving between December 1999 and May 2000) who smoked were found to have more prevalent bleeding on probing and from an analysis of data by Ismail, et al. 18 from data collected from 1971 to 1975. An analysis by the National Health and Nutrition Examination Survey in the U.S. (NHANES I) was able to demonstrate an association between smoking and periodontal diseases that was independent of oral hygiene, age, or other probable risk factors. Since then, there has been enough evidence to identify smoking as a risk factor for periodontitis 13   There is a strong need in Latin America to focus on more effective intervention programs to prevent and control periodontal diseases at national levels. It should be emphasized that since periodontitis begins as gingivitis, it is reasonable to conclude both the onset and the progression of periodontal damage caused by periodontitis. While the disability-adjusted life-years (DALYs) due to severe periodontitis and untreated caries have increased since 1990, those due to severe tooth loss have decreased. Oral conditions are all ranked among the top 100 detailed causes of DALYs 23 highlight the challenge in responding to the diversity of urgent oral health needs worldwide, particularly in developing communities. a reversible and easily controlled disease in stage of gingivitis, is found to be highly prevalent among adult study participants from the three cities in Latin America of this study. In addition, these individuals are also more likely to attend preventive or followup visits because socioeconomic characteristics the pattern and type of dental services used 6 . Poor awareness of the importance of periodontal health and the consequences of the disease among the public and even among some general dental practitioners is one of the most common reasons for failure to control and treat periodontal diseases effectively on a population basis 19 . We recommend that effective intervention programs for the prevention and control of periodontal diseases should be implemented at national levels, and the need for such implementation seems to be extremely important in the three countries we studied here. We believe that there is a strong need across Latin America to improve the population's self-awareness about oral health through better oral health education that promotes good oral hygiene and regular dental care.
This study was aimed to detect gingival of either gingivitis or periodontitis in order to determine the risk of gingivitis onset or periodontitis progression in the study population.

CONCLUSIONS
prevalent in the three Latin American cities studied. Overall, 95.6% of the participants aging 18 years that more than two-thirds have moderate gingival inflammation and 3.6% have severe gingival risk indicators such as gender, socioeconomic variables, and the presence of plaque. The present investigation serves as the basis for a longitudinal analysis of oral health in populations of South American adults, and for the development of strategies to improve the health care systems that serve them.