Genotoxicity in the oral cells of older people from a Brazilian rural area: a population-based study

Abstract The purpose of this population-based, observational, and cross-sectional study was to evaluate alterations in the oral cells of a population of older people from a Brazilian rural area, using the micronucleus technique to investigate possible associated genotoxic factors. A questionnaire was applied and clinical examination and collection of oral mucosal cells were performed for all older people (≥ 60 years) from a town in southern Brazil. Demographic and socioeconomic variables, deleterious habits (drinking and tobacco use), presence of gastro-oesophageal reflux disease (GERD), and the use of proton pump inhibitors (PPIs) were considered the exposure variables, whereas metanuclear changes (MCs) and the prevalence of cell micronuclei (MN) were considered outcomes. Out of 489 older people, 447 were included in the study, among whom 50.8% were men with a mean age of 70.9 years and 83.9% had a monthly family income greater than US$ 500.00. GERD symptoms were present in 36.2% of the individuals, and 29.1% used PPIs daily, 53.3% consumed alcoholic beverages, and 46.7% used tobacco. The analysis of 1,000 oral mucosal cells per subject showed a MN frequency of 0–2 per individual, and MCs were detected with an average of 15 units per individual (median = 11 per individual). Poisson regression did not show statistical association between the exposure variables and the outcomes (presence of MN and MCs), except for the use of PPIs, which was a protective factor for the prevalence of MN [PR 0.6 (CI 0.3–0,9)]. Age, sex, family income, tobacco use and drinking, and GERD were not associated with the number of MN and MCs in oral mucosal cells of the investigated older people.


Introduction
Genotoxic changes in oral mucosal cells may occur because of several factors and result in DNA damage, leading to the development of dysplastic or tumour processes and, therefore, the constant monitoring of populations is essential. 1 External carcinogenic factors, either physical or chemical, cause abnormalities in genetic processes, accumulating in the cells and leading to their degeneration. 1 Approximately 15% of the Brazilian population lives in rural areas. 2 Despite growing concerns, few studies have focused on the health of Declaration of Interests: The authors certify that they have no commercial or associative interest that represents a conflict of interest in connection with the manuscript. rural populations. 3 It was reported in 2010 that oral health was the fifth most important health topic for residents of rural areas in the United States, outranked by access to good-quality health care, cardiovascular diseases, diabetes, and mental health. 4 Older people in the Brazilian rural context have not been sufficiently investigated, perhaps because they are a minority (estimated at 15.7%) when compared to older people in the urban areas (84.3%) or because of the lack of specific public policies for this population. 5,6 Regarding the health and health care of older people living in rural areas, some difficulties are expected due to characteristics that are inherent to the access to health care (e.g., transportation barriers, poor road conditions or lack of roads, and distance from health centers), to low income, and to the tradition of seeking curative/ therapeutic care rather than preventive care. Because of that, the health status and quality of life of these older people may eventually worsen. [7][8][9] Cancer is one of the most common causes of morbidity and mortality and it results from the interaction of risk factors that affect the monitoring processes of cell proliferation in any region of the body. 10 Drinking and tobacco use are the main risk factors for oral cancer. 11,12 Moreover, studies have suggested that individuals with gastroesophageal reflux disease (GERD) may develop ulcerated lesions in oral tissues. 13,14 The Union for International Cancer Control 15 considered longer life expectancy and estimated the incidence of cancer to be 11 million in 2002, increasing to more than 15 million in 2020. The incidence of oral cancer also rose, representing the sixth most common tumour worldwide, with about 40,000 cases reported every year in the United States, among which 75% are related to the environment and lifestyle 16 . Another study also reported that people's lifestyle and exposure to environmental factors may cause deleterious effects on health, including DNA damage. 17 In addition, relevant biological factors such as age, sex, use of medication, and systemic diseases may have a significant influence on DNA mutations. 18 There are multiple tests to biologically monitor human populations exposed to mutagens. The assessment of genotoxicity using the micronuclei (MN) test in exfoliated cells is an effective and lowcost procedure to investigate epithelial carcinogens and may be used to detect chromosome breakage or mitotic interference, which plays a significant role in carcinogenesis. 19,20 MN originate from chromosomes or from their fragments and remain in the anaphase during nuclear division. The correlation between the frequency of MN/MC and the severity of this genotoxic damage has been shown. 19,20 Therefore, it may be argued that MN and MC changes in exfoliated oral epithelial cells represent important signs of early genotoxic events induced by carcinogenic agents. The present study evaluated oral cellular alterations in a population of older people from a rural area in southern Brazil, using the micronucleus technique to investigate possible associated genotoxic factors. It was hypothesized that deleterious habits (drinking and tobacco use), the presence of GERD, and continuous use of some medications -proton pump inhibitors (PPIs) -affect the number of MN and MCs in the oral epithelial cells of this population.

Methodology
The present study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines. 21

Study design and ethical considerations
This is a population-based, observational, and cross-sectional study. The research project was approved by the local Research Ethics Committee and registered in the Brazilian Platform (number 3.315.733). The older people aged over 60 years were invited to participate and were included in the study after signing an informed consent form.

Context and participants
The present study was performed in a rural town with 1,941 inhabitants located in the northeastern region of the state of Rio Grande do Sul, in southern Brazil. Approximately 25% of the population is composed of older people (age ≥ 60 years), and agriculture is the main economic activity. The data were collected between May and September 2019.
The local Health Department provided the list of inhabitants with their respective ages and addresses. The town's entire older population (≥ 60 years) was invited to participate. Data were collected from all participants at their homes on a previously scheduled date. Individuals who were not at home after two visits were regarded as losses. Moreover, individuals with mental or physical disorders who did not collaborate with the application of the questionnaires or clinical examination were excluded. Samples with less than 1,000 cells were also excluded from the study. Losses and refusals were constantly assessed so that the study would maintain a uniform and homogeneous distribution of participants per microregion. To optimize such distribution and the participation of the older people, visits were scheduled to the best possible time for the participant, avoiding any conflict with any appointment they had already scheduled.

Data sources and variables
The study variables were divided into three blocks: a. A: demographic and socioeconomic variables, including personal data, birth date, sex, and family income; b. B: behavioural factors, including tobacco use and drinking, which were investigated using the WHO ASSIST instrument; 22 c. C: general health status, the use of PPIs, and gastric problems investigated by the GERD-SQ questionnaire. 23 The present study used the following as exposure variables: information on sex (male or female), age (60 to 65 years, 66 to 74 years, and ≥ 75 years), and socioeconomic level as per the monthly family income (< US$ 500.00 and ≥ US$ 500.00). The amount (US$ 500.00) corresponds to the income of a couple receiving social security benefits such as retirement, and it is equivalent to approximately two minimum wages.
Tobacco use and drinking were measured with the Alcohol, Smoking, and Substance Involvement Screening Test (ASSIST) questionnaire. 22 This eightquestion instrument assesses the consumption pattern of toxic substances, such as alcohol and tobacco throughout life and especially over the last three months, classifying the individual as having low, moderate, or high risk for each substance.
GERD was also investigated as an exposure variable. It was deemed present when the participant reported a medical diagnosis of the disease or had a score ≥ 2 in questions 1 (heartburn) or 3 (regurgitation) in the GERD-SQ questionnaire. 23 This variable was categorized as either presence or absence. The daily use of PPIs was also considered an exposure variable (yes/no). In addition, the sum of three relevant factors (drinking, tobacco use, and presence of GERD) was assessed as follows: 0 = presenting none of these factors; 1 = presenting at least one of these factors; 2 = presenting at least two of these factors; 3 = presenting the three factors.
Finally, an intraoral examination was performed to collect oral mucosal cells by scraping the right and left buccal mucosa using a disposable wooden spatula. The collected material was placed in flasks containing a fixative solution of acetic acid and methanol (3:1) until processing of the sample. 24 Prior to cell collection, the participants rinsed their mouths with 10 mL of distilled water for 1 min to remove possible food residues or dead cell debris from the buccal mucosa, thus avoiding any interference with the interpretation of the results.

Cell material processing
Collected oral cell samples were transferred to Falcon tubes and centrifuged at 1,000 rpm for 10 min (Ls4 centrifuge, CELM™, São José dos Campos, SP, Brazil). The process was repeated until the material was colourless and free of residues. The clear samples were then seeded onto plates stained with 10% Giemsa Wright (Renylab Química™, Barbacena, Brazil). One thousand oral mucosal cells per individual were analyzed under a light microscope (Olympus Bx50, Tokyo, Japan; magnification: x1000). An experienced, trained, and blinded examiner performed all the analyses. The kappa coefficient was used to assess the intraexaminer level of agreement (k = 0.90).
Thomas et al. 25 described the criteria used to report on the presence of MN and MCs. The results were presented as the absolute number of cells with MN and MCs found in every 1,000 cells. Cell abnormalities were considered by assessing the intensity of color, texture, and focal plane of the nucleus. Normal cells were those that presented an intact and relatively homogeneous cytoplasm, little or no contact with adjacent cells, and an intact and homogeneous nucleus with a smooth and distinct nuclear perimeter. 24,26,27 The number of MN and MCs was recorded. 24,26-28 Micronucleated cells were characterized by the presence of the main nucleus and a smaller one, called micronucleus (MN), resulting from chromosome breakage caused by genotoxicity. [26][27][28] MN evaluation also considered the presence of a) a regular round or an elliptical contour within the cytoplasm; b) similar colour to that of the main nucleus; c) less than 1/3 of the nucleus diameter; and d) complete separation from the nucleus, allowing clear identification of the nucleus and the limits of the MN. Overlapped cells were excluded from the evaluation.
Karyorrhectic cells were characterized by a more extensive chromatid aggregation, indicating nuclear fragmentation and disintegration, in an advanced stage of cell death by apoptosis. Pyknotic cells were characterized by a small nucleus with condensed chromatin and intense color. The nuclear diameter is 1/3 to 2/3 smaller than that of differentiated cells and indicates an advanced stage of cell death by necrosis. Karyolytic cells present a slightly stained chromatin, difficult to analyse under a light microscope, and indicate a more advanced stage of cell death by necrosis. Binucleated cells were characterized by the presence of two nuclei with similar characteristics to those of differentiated cells. The presence of binucleation is indicative of failure by cytotoxic action in the cytokinetic process during cell reproduction. Button cells show the main and accessory nuclei close to each other and connected by thin chromatin threads. The accessory nucleus has the same morphological characteristics and color as those of the main nucleus. However, it has a smaller diameter (1/4 of the nucleus). This type of morphology might originate from the presence of dicentric chromosomes with abnormal anaphase behaviour during segregation. 24,[26][27][28] The MN per 1,000 cells and the MCs per 1,000 cells were considered outcome variables, categorized as follows: MN (NO = 0; YES = 1-2) and MC dichotomized by the median (≤ 11 MC and > 11 MC).

Data collection
The application of the WHO-ASSIST questionnaire, the clinical examination, and oral cell collection were performed at the participant's home. The participant remained seated in a comfortable chair with a headrest for the oral examination and cell collection. A headlamp was used by the examiner to illuminate the oral cavity. Adequate disposable personal protective equipment (PPElaboratory coat, goggles, gloves, mask, face shield, and cap) was used by the professionals, in addition to disposable wooden spatulas and sterile flasks containing acetic acid and methanol (3:1). The data were collected by two professionals: a dentist (examiner) and an oral health assistant (annotator). Oral cell samples were processed and evaluated by a calibrated microbiologist.

Sample size
All town residents (≥ 60 years) were invited to participate. The exclusion criteria were applied during the first visit for data collection.
Calibration, training, and pilot study A 16-hour training lesson was used to instruct, train, and calibrate the research team. A pilot study was then performed with 20 individuals from a longstay institution for older people in another southern Brazilian town. The pilot study helped evaluate the research protocol, adjusting the data collection instruments and method and identifying the most suitable approach for the study population. The intraexaminer agreement was obtained during the pilot study (k = 0.90).

Statistical analysis
A descriptive analysis was initially performed to determine the relative and absolute frequencies of exposure and outcome variables. Poisson regression analysis was used to evaluate factors associated with the presence of MN and the presence of more than 11 MCs. Prevalence ratios were obtained for the exposure variables and their respective 95% confidence intervals. All analyses were performed using the Stata software (StataCorp. 2015, Stata Statistical Software: Release 14. College Station, USA: StataCorp LP).

Results
Among those individuals who were eligible for data collection (489), 42 were excluded because they were either not at home at the scheduled time for the visits (34) or the oral mucosal cell samples did not have the minimum amount of 1,000 cells (8). Thus, 447 individuals participated in the study, with a response rate of 91.4%. The mean age of the participants was 70.9 years, 50.5% were male, and most of them had a monthly income greater than US$ 500.00 (83.7%). Table 1 presents the distribution of covariates according to the presence of MN. Almost 20% of the participants showed at least one MN change per 1,000 cells. No participant showed more than two MN changes. Poisson regression did not show statistical association between exposure variables and outcomes, except for the use of PPIs, which was a protective factor for the prevalence of MN [PR 0.6 Considering the small number of smokers and drinkers in some categories, they were grouped as follows: tobacco use (YES = low and moderate risks; NO = never used), and drinking (YES = low, moderate, and high risks; NO= never used). In addition, three relevant risk factors (GERD, drinking, and tobacco use) for oral cell DNA mutation were grouped and associated with the presence of MN, showing no statistical significance (Table 1).
The mean MC found for all collected cell samples was 15/1,000 cells and the median was 11/1,000 cells. For the sake of dichotomization, the median was considered the parameter (≤ 11 MCs and > 11 MCs) and 49.2% of the participants showed more than 11 MCs. Table 2 shows that the prevalence ratio of MCs and the respective confidence intervals did not present statistical significance (p > 0.05) when associated with any of the variables studied. Although the absolute values with >11 MCs were almost 10% higher for drinking and 2% for tobacco use (Table 2), no statistical significance was observed in these associations (p = 0.06 and p = 0.68, respectively). So, age, sex, family income, tobacco use and drinking, PPI use, or GERD were not associated with the number of MCs in the oral mucosal cells of the investigated older participants ( Table 2).

Discussion
The present study, conducted with a population of older people from a rural area, partially rejected the experimental hypothesis because it showed that factors such as sex, deleterious habits, and the presence of GERD did not significantly affect the number of MN and MCs in the oral epithelial cells of this population. However, there was a significant influence of PPI use on MN.
A study 28 compared 80 healthy young individuals (aged 19 to 29 years) with older people (≥ 60 years), associating sex and age with the frequency of MN and MCs. After applying a questionnaire and collecting epithelial cells from the oral mucosa of the participants, the authors concluded that the occurrence of MN was higher among the older people, irrespective of sex. Another study 29 evaluated 245 individuals with a mean age of 60.5 years by collecting mucosal cells from the mid-oesophagus (approximately 30 cm from the upper dental arch) and concluded that the frequency of MN did not show significant differences (p > 0.05) regarding sex and place of residence (rural or urban area) of the individuals. Similarly, the present study did not show significant differences in the number of MN and MCs between men and women. The age range of the older individuals also did not affect the number of MN and MCs, suggesting homogeneous exposure of the study population to the genotoxic factors throughout life. Tobacco use has been described as the main risk factor for the development of malignant and carcinogenic oral lesions, and the combination of deleterious habits such as drinking and tobacco use may increase the prevalence of oral cancer. 30 A greater aggressiveness of oral carcinoma in drinkers and smokers may be due to the increased permeability of the cell membrane caused by ethanol and the consequent exposure of intracellular content to tobacco as a carcinogen. 31 A study 32 evaluated the genotoxic effects of tobacco and noted a significant increase in the frequency of MN in individuals that either smoked or chewed tobacco, or smoked and chewed tobacco concomitantly, when compared to the control group. These studies suggested that drinking and tobacco use are risk factors for the development of precancerous lesions.
The number of MN in normal oral mucosal cells may reach 2 MN/1,000 cells, 33 which is in agreement with the cell analyses of the present study, suggesting absence of genotoxicity in the oral mucosal cells of the evaluated population. The present study also did not find a significant difference for the number of MN and MCs among tobacco users or drinkers or among those who used both substances, although the absolute values in individuals with > 11 MCs were almost 10% greater for drinkers and 2% higher for tobacco users (Table 2). This result may be explained by the fact that tobacco use and drinking were moderate in the evaluated population, probably not reaching the thresholds for the microscopic diagnosis of genotoxicity, which consists of a minimum consumption of 30 cigarettes/day. 33,34 A study 1 evaluated 120 Brazilian farmers for their exposure to tobacco and pesticides, warning about the importance of biomonitoring studies in populations exposed to genotoxic agents, particularly rural workers. The authors also reported that the MN test is easy to perform, minimally invasive, and inexpensive, and it may be used as a tool for epidemiological and genotoxic investigations in individuals with harmful health habits, thus justifying the use of such methodology in the present study.
PPIs are used for the long-term treatment of gastroesophageal disorders, including acid reflux. However, there are growing concerns about the improper and excessive use of these inhibitors. Yang et al. 35 assessed 21 Wistar rats, simulating the long-term use of omeprazole. After euthanizing the animals, morphological changes were found in the bile duct, including ductal epithelial proliferation, micropapillary growth of the biliary epithelium, focal narrowing, and bile duct obstruction. These are characteristics of precancerous lesions and indicate a high-fat diet. Although omeprazole has been subjected to a wide range of genotoxicity tests, which were all negative, the ability of this compound to interact with DNA and induce non-programmed DNA synthesis in the gastroesophageal mucosa has been discussed. In this sense, Mereto et al. 36 assessed the gastric mucosa of rats that received 100 mg/kg/ day of omeprazole for 14 days, but they did not find a significant increase in the number of MN. Similarly, Sinués et al. 37 performed a clinical trial with 33 healthy volunteers who received 20 mg/day of omeprazole for 14 days. Although the number of MN slightly increased, omeprazole did not present genotoxic effects after 14 days of treatment. In the present study, PPI use was a protective factor for the prevalence of MN (p = 0.04), but with no clinical significance, given that the frequency of up to 2 MN/1,000 cells is considered normal. [34][35][36][37][38] Considering oral cancer, it is essential to look at the moment of screening, as it plays a fundamental role in understanding the prognosis of the disease. Critical signs and symptoms that can be identified during the initial screening can increase a patient's chances of survival. Reports suggest that socioeconomic factors, lack of public awareness, and delays in primary health care centers are some of the main parameters that contribute to patient mortality and morbidity. The conventional visual examination of oral lesions can effectively monitor patient mortality in the presence of risk factors. Yet, some disadvantages limit the clinical use of this method. Thus, screenings that efficiently differentiate benign from malignant lesions, as well as provide information about the early stage of cancer, facilitate the detection of complications associated with the diagnosis of oral cancer. 39 In addition, previous studies 40, 41 have reported that diseases such as oral submucous fibrosis, mucosal ulcerations, and xerostomia are significantly more common in patients with GERD. In a recent study, Gilligan et al. 42 claimed that chronic traumatic oral lesions (caused by acids, for example) may eventually predispose to oral carcinomas. Thus, considering that subclinical cellular alterations, such as the number of micronuclei, can also precede oral carcinomas, 43 the present study performed a population-based screening test, correlating the frequency of MN and MCs with risk factors such as the presence of GERD, smoking, and drinking. In addition, a systematic review 44 indicated that some PPIs might induce genomic instability and increase the risk of certain types of cancer, suggesting caution with long-term therapeutic strategies and self-medication with PPIs.
GERD is a highly prevalent disorder in the global population. According to Mamede et al., 45 clinical evidence indicates that GERD may cause changes to tongue tissues. For Lipan et al., 46 the reflux that advances to the laryngopharynx and later to other head and neck regions, including the oral cavity, may result in relevant problems. Kuo et al. 14 analyzed 39,845 GERD patients and found 98 cases of head and neck cancer, in which oropharyngeal and hypopharyngeal cancers in men were statistically associated with GERD.
A study 47 on mild and severe esophagitis and cancer cases with the comet assay and biopsy of the distal third of the esophagus reported damage to the DNA of esophageal mucosal cells. Moreover, the study proved that most DNA changes were directly connected to the level of inflammation in the region. Nevertheless, the relationship between GERD and extraesophageal cancers is controversial. 48 In this sense, the results of the present study indicate that the assessed population did not show a relationship between GERD and increased MN and MCs in the oral mucosal cells, suggesting that the oral mucosa is not deleteriously affected by the presence of GERD in the population studied. Yet, GERD patients have multiple changes in pH and salivary microbiome when compared to healthy patients. 49 A higher prevalence of dental erosion and caries was reported for patients with GERD as compared to healthy individuals, and the pathophysiological mechanisms of GERD involve changes in saliva physiology. 50 As the present study also evaluated deleterious habits (smoking and drinking) directly related to the oral mucosa, and saliva is present in all regions of the oral cavity, the oral cells were collected from the right and left buccal mucosae, which are easily accessible sites, avoiding nausea and discomfort for the older people and also optimizing cell collection.
Note that the instrument used to assess the presence of GERD was subjective (a validated questionnaire), which is less robust than objective assessments such as endoscopy and pH monitoring, and this may be considered a limitation of this study. In addition, other factors such as the time for the diagnosis of GERD, the length of PPI use, and exposure to toxic products such as fertilizers and pesticides were not considered in the present study. Thus, future studies should focus on different populations, comparing residents of urban and rural areas, young adults and older people, and inhabitants of distinct geographical regions.

Conclusion
Considering the methodology used and the evaluated population (older individuals from a rural area), it can be concluded that factors such as sex, age greater than 60 years, tobacco use and drinking, and the presence of GERD did not affect the number of MN and MCs in the epithelial cells of the oral mucosa.