1 Volume 24 2025 e255233 Original Research Braz J Oral Sci. 2025;24:e255233http://dx.doi.org/10.20396/bjos.v24i00.8675233 1 Curso de Odontologia, Universidade Santo Amaro (UNISA), São Paulo, SP, Brazil. 2 Curso de Odontologia, Fundação Hermínimo Ometto, Araras, SP, Brazil. 3 Faculdade de Odontologia de Piracicaba (FOP), Universidade Estadual de Campinas (UNICAMP), Piracicaba, SP, Brazil. 4 Faculdade de Odontologia, Universidade Brasil, Fernandopolis, SP, Brazil. Corresponding author: Fábio Luiz Mialhe Faculdade de Odontologia de Piracicaba (FOP) Avenida Limeira 901, Bairro Areão Piracicaba, SP, Brazil ZIP: 13414-903 Email: mialhe@unicamp.br Phone number (19) 98226-1515 Editor: Dr. Altair A. Del Bel Cury Received: December 11, 2023 Accepted: October 1, 2024 Oral health among drug addicts and associated variables: a multivariate approach using a canonical correlation analysis Tais Cristina Nascimento Marques1 , Karin Luciana Migliato Sarracini2 , Vanessa Gallego Arias Pecorari3 , Karine Laura Cortellazzi3 , Marcelo de Castro Meneghim3 , Luciana Estevam Simonato4 , Fábio Luiz Mialhe3* , Gláucia Maria Bovi Ambrosano3 Aim: To evaluate the association between oral health conditions with sociodemographic variables and the use of psychoactive substances by drug addicts assisted by Psychosocial Care Centers for Alcohol and Drugs (CAPS-ADs). Methods: This is an analytical cross-sectional study with a sample of 262 drug addicts assisted by CAPS-ADs in three cities in the state of São Paulo, Brazil. Participants were evaluated through intraoral clinical examinations (DMFT and CPI indexes) and they answered a socioeconomic questionnaire. The intensity of use of psychoactive substances was assessed through ASSIST questionnaire. Associations between variables were performed using the canonical correlation analysis method. Results: It was observed that the mean DMFT index was 13.9; 60.2% of the sample used cocaine/crack and 46.5% used marijuana. Regarding the analyses, the canonical correlation coefficient of the highest possible correlation between the combination of variables of oral health problems and the combination of variables of intensity of drug use was 0.4642 (p=0.0005). In addition, the canonical correlation formed by the combination of socioeconomic condition variables and the combination of oral health was statistically significant (canonical correlation coefficient=0.4533; p<0.0001). Thus, it was observed that the high consumption of drugs, mainly cocaine/crack and marijuana, was associated with oral health problems; in addition, worse socioeconomic conditions, mainly lower level of education of the mother and father, were associated with oral health problems in drug users. Conclusion: Both the type of psychoactive substance addiction and sociodemographic variables affected the oral health of drug addicts. Keywords: Oral health. Mouth diseases. Substance-related disorders. Illicit drugs. https://orcid.org/0000-0002-1083-5963 https://orcid.org/0000-0002-5257-5792 https://orcid.org/0000-0002-0300-5697 https://orcid.org/0000-0001-9584-9477 https://orcid.org/0000-0003-2673-3627 https://orcid.org/0000-0002-6413-5479 https://orcid.org/0000-0001-6465-0959 https://orcid.org/0000-0001-5950-6370 2 Marques et al. Braz J Oral Sci. 2025;24:e255233 Introduction According to the World Health Organization (WHO), “psychoactive drugs are sub- stances that, when taken in or administered into one’s system, affect mental pro- cesses, e.g. perception, consciousness, cognition or mood and emotions. Psycho- active drugs belong to a broader category of psychoactive substances, which also includes alcohol and nicotine”1. Psychoactive drugs can be classified as licit drugs, such as alcohol and cigarettes, or illicit drugs, such as marijuana, cocaine, opioids, among others2. There is an estimate projection that 11 per cent of the global population aged 15–64 years will use some kind of drugs by 20302. In Brazil, the National Anti-Drug Policy was created in 2002 and, in 2006, the National System of Policies on Alcohol and Drugs - SISNAD, as well as Law 11,343 that differentiates for the magistrate the drug user or dependent, offer- ing them the opportunity to receive treatment3. The Public Health System in Bra- zil, known as the Unified Health System, provides assistance to users of psycho- active substances through the Psychosocial care Centers for Alcohol and Drugs (CAPS-ADs), with extra-hospital care, thus reducing the need for admissions to psychiatric hospitals4,5. The abuse of psychoactive substances causes great damage to societies, such as health expenditures, rising levels of crime and violence and loss of productivity at school and at work2. However, the public health impact of drug addiction is not limited to the economic impact, but includes other outcomes such as irreparable damage to the user’s physical health, psychological problems, domestic violence, child abuse, family disintegration and high risk of contracting infectious diseases2,6. In the dental field, it was observed that the consumption of alcohol, tobacco and illicit drugs is considered a risk factor for several oral diseases7-11. Drug addicts are at greater risk of having compromised oral health for various reasons, such as socioeconomic conditions, limited access to dental care, poor diet and poor oral hygiene habits10-12. Some studies indicate that drug addicts do not care about per- sonal hygiene due to low self-esteem and lack of motivation, exposing them to greater risks of having oral problems7-9,12. The importance and severity of oral health problems among psychoactive sub- stance users require the availability of comprehensive dental care programs. These programs must be integrated with health services. In addition, programs must adopt multiple approaches involving education, prevention, and treatment. There- fore, it is important to generate data from studies that evaluated the oral health condition of drug addicts and the socioeconomic and addiction-related variables associated with it in order to support the management of public policies aimed at this population. In this study, a multivariate approach was used through canonical correlation analysis. Unlike multiple regression, which predicts a single dependent vari- able from a set of multiple independent variables, canonical correlation simulta- 3 Marques et al. Braz J Oral Sci. 2025;24:e255233 neously predicts multiple dependent variables from multiple independent vari- ables13. Therefore, the use of this analysis methodology enables determining the interrelationships between the group of dependent variables and the group of independent variables. The aim of this study was to evaluate the associations between oral health condi- tions and the intensity of use of psychoactive substances and sociodemographic variables of drug addicts assisted by Psychosocial Care Centers for Alcohol and Drugs (CAPS-ADs). Material and Methods This study received approval from the Research Ethics Committee (CEP) of a Dental School under protocol nº 069/2012. Users of psychoactive substances received an informed consent form containing information regarding the responsibility, method- ology and development of the study. Only after the consent of the participants, they were included in the study. A cross-sectional study was conducted with a sample of adult drug addicts assisted by the three Psychosocial Care Centers for Alcohol and Drugs (CAPS-ADs) located in 3 medium-sized cities (Piracicaba, Limeira and Rio Claro) in the non-metropolitan area of the state of São Paulo – Brazil in 2013. The total population of CAPS amounted to 400 individuals at the time of the survey. A probability sample was calculate considering the situation of higher probability of sampling error (p=0.50) assuming a confidence level of 95% and a sampling error of 5%. In addition, we calculated the sample size for estimating the oral health of par- ticipants, based on previous data with this population in the Brazilian literature that found an overall mean number of decayed, missing and filled teeth (DMFT index) of 14.9 with a standard deviation of 6.3814. When calculating the sample, the power of test of at least 80% with significance level of 5% and minimum significant odds ratio of 1.5 in association with the variables were taken into account, reaching a sample of 262 individuals. An additional percentage of 10% in the initial sample was calculated considering possible losses, reaching a minimum sample number of 286 participants. Convenience sampling was the methodology employed for collect data, that is, as the researcher visited the CAPS-AD, users were invited to participate in the research voluntarily. The participants were clinically examined by one calibrated examiner (as to the presence of decayed, missing, and filled teeth (DMFT index), presence of periodon- tal disease (CPI index), use/need of dental prosthesis and need for dental treatment, all according to the criteria established by the World Health Organization (WHO) for epidemiological surveys in an outdoor setting, under natural light with ball-point probes and mirrors15. During theoretical discussions and calibration exercises it was obtanied an almost perfect level of agreement (Kappa>0.90). During the exper- imental phase, 10% of the volunteers in the sample were re-examined by the same examiner to verify the maintenance of the diagnostic criteria and assessment of intra-examiner sampling error. The mean intraexaminer agreement obtained in this activity was Kappa=0.89. 4 Marques et al. Braz J Oral Sci. 2025;24:e255233 The following socioeconomic and demographic data were collected from the par- ticipants: age, sex, skin color (black, brown, yellow, white), monthly family income (Brazilian minimum wages – BMW), participant’s and parents’ education level, type of housing (own x others) and number of people in the house. The questionnaires were completed in a quiet room in each CAPS-AD and the interviews were conducted face to face and individually. The intensity of use of psychoactive substances was evaluated through the Alco- hol, Smoking and Substance Involvement Screening Test (ASSIST), which has been adapted to Brazilian Portuguese language16,17. It is a structured questionnaire con- taining eight questions about the use of nine classes of psychoactive substances (tobacco, alcohol, marijuana, cocaine, stimulants, sedatives, inhalants, hallucino- gens, and opiates) and ‘other’ drugs. The questions address frequency of use, in life and in the last three months, problems related to use, concerns about use by people close to the user, impairment in performing expected tasks, unsuccessful attempts to stop or reduce use, feeling of compulsion and injection use. Each answer cor- responds to a score, which ranges from 0 to 4, and the total sum can range from 0 to 20. A score ranging from 0 to 3 is considered as indicative of occasional or non-problematic use, from 4 to 15 as indicative of abuse, and ≥16 as suggestive of dependence/addiction16,17. In order to assess the relationship between oral health problems and intensity of use of psychoactive substances, the data set was initially divided into two groups: group I - variables related to oral health (DMFT index; need for dental treatment; CPI index; use and need for dental prosthesis) and group II - variables related to inten- sity of use of psychoactive substances. Multivariate analyses were performed using canonical correlation, measuring the association and contribution of each variable in the respective model. Through canonical correlation analysis, we obtained orthogonal combinations of the variables within each group (I-oral health and II-drug use), which best explain the variability of the data. These combinations of features are called canonical vari- ables. Correlations were tested separately by the approximate F test and together by Wilks’ lambda (Wilks’Lambda), Pillai’s criterion (Pillai’s Trace), Hotelling’s trace (Hotelling-Lawley Trace) and Roy’s Greatest Root. From the two sets, where p is the number of X variables (intensity of use of psychoac- tive substances) and q is the number of Y variables (oral health problems), the canon- ical correlation analysis provided the following combinations: W1=a11x1+a21x2+...+ap1xp v1=b11x1+b21x2+...+bp1xp Thus, the first pair of canonical variables was obtained, with the highest possible correlation (first canonical correlation). The coefficients of the combinations are called canonical coefficients or canonical weights. Next, a second set of canonical variables, uncorrelated with the first pair, with the second highest correlation coeffi- 5 Marques et al. Braz J Oral Sci. 2025;24:e255233 cient was tested. This is the second pair of canonical variables, which contains the maximum remaining information that has not been covered by the explanation of the first pair of canonical variables and is given by: W2=a12x1+a22x2+...+ap2xp v2=b12x1+b22x2+...+bp2xp The construction process of the canonical variables occurred until the number of pairs of canonical variables was equal to the number of variables in the smaller group or until all covariance or correlation information between the two sets was explained by the pairs of selected variables. In addition, multivariate analyses by canonical correlation were also performed to assess the relationship between the set of group I variables (DMFT index; need for treatment; CPI index; use and need for dental prosthesis) and socioeconomic variables, called group III variables (income, living in the same house, psychoac- tive substance user’s level of education, mother’s and father’s level of education, type of housing)18. All analyses were performed according to Ferreira18, using the CANCORR procedure of the SAS program. Results Of the 286 participants initially drawn, 24 did not agree to participate in the survey and the final sample was composed of 262 subjects. Of them, 85.5% were male, 17.9% report 1 or less BMW as their monthly family income, and the majority of the partici- pants (61.1%) had up to 8 years of education. With respect to substance use, 80.9%, 46.5% and 60.3% of the sample were classified by ASSIST as having moderate/high risk of being alcohol, marijuana and cocaine/crack users. In relation to the oral health conditions, it was observed that 71.7% of participants needed dental treatment. Regarding the CPI, 43.1% had dental calculus present and 27.1% had a healthy periodontium at the time of intraoral examination. Regarding the use/need of dental prostheses, 45.8% of participants did not need any type of prosthesis; however, 13.7% needed a complete denture. In relation to the caries expe- rience, the mean DMFT index of the sample was 13.9 (SD±8.9), with a minimum value of 0.0 and a maximum value of 28. In the multivariate analysis of the associations between oral health problems (Group I) and the intensity of use of psychoactive substances (Group II), group I was formed by five variables related to oral health and group II by nine related to alcohol consump- tion. As the number of canonical correlations that can be tested is equal to the num- ber of variables in the smaller group, five independent canonical correlations were tested, that is, five different combinations of oral health and intensity of drug use vari- ables (Table 1). 6 Marques et al. Braz J Oral Sci. 2025;24:e255233 Table 1. Standardized canonical coefficients of the canonical variables intensity of use of psychoactive substances and oral health problems. Variable Standard canonical coefficients Oral health problems (group I) DFMT 0.7908 Need for dental treatment 0.1997 CPI 0.2742 Use of dental prosthesis 0.1539 Need for dental prosthesis 0.1355 Intensity of use of psychoactive substances (group II) Tobacco products 0.1408 Alcoholic beverages 0.0384 Marijuana 0.2458 Cocaine, crack 0.7448 Amphetamine-type stimulants 0.2594 Inhalants 0.5156 Sedatives and sleeping pills 0.2994 Hallucinogens 0.2959 Opioids 0.0335 DMFT: number of Decayed, Missing due to caries, and Filled Teeth in the permanent teeth; CPI: Community Periodontal Index Wilks’Lambda=0.7221 (p=0.0005); Pillai’s Trace=0.2971 (p=0.0015); Hotelling-Lawley Trace=0.3589 (p=0.0002); Roy’s Largest Root=0.2746 (p<0.0001). Only the first canonical correlation (formed by combining the two groups of variables with the highest correlation) was significant (canonical correlation coefficient=0.4642; p=0.0005). This value represents the highest possible correlation between any combi- nation of the oral health problem variables and any combination of the drug use inten- sity variables. The other possible correlation pairs were not significant (p=0.9311; p=0.9543; p=0.9960, and p=0.9737). The bottom of Table 1 presents the multivariate statistics for the null hypothesis that all canonical correlations are zero. The low p-values for these tests (p<0.05) suggest rejecting the null hypothesis that all canonical correlations are zero in the population. The standardized canonical coefficients (Table 1) represent the ‘canonical weights’ and are similar to the coefficients of a multiple regression. These coefficients express how much each variable was correlated with its canonical (set of variables). It was observed that the DMFT variable is the one that is most strongly correlated with the canonical oral health problems, with a 0.7908 correlation coefficient. The sec- ond most influential is the CPI variable, with a 0.2742 correlation coefficient with the canonical variable. The standardized canonical coefficients indicate that the canon- ical variable for the oral health problems group is a weighted combination of DMFT (0.7908), CPI (0.2742), need for treatment (0.1997), denture use (0.1539) and need 7 Marques et al. Braz J Oral Sci. 2025;24:e255233 for dental prosthesis (0.1355) with emphasis on DMFT. The coefficients for the intensity of drug use variables indicate that the canonical variable for the group of these variables was formed by the intensity of use of cocaine/crack (0.7448), inhal- ants (0.5156), hypnotics (0.2994), hallucinogens (0.2959), amphetamines (0.2597), marijuana (0.2458), tobacco (0.1408), beverages (0.0384) and opioids (0.0335), with emphasis on cocaine/crack and inhalants. Table 2 presents the canonical correlation coefficients between the variables related to drug use and the canonical oral health. It can be observed that these two char- acteristics are related (R2=0.2154), that is, the variation in the intensity of drug use explained approximately 22% of the variation in the variables related to oral health. Thus, high consumption of drugs, especially cocaine/crack and marijuana, was asso- ciated with oral health problems. Table 2. Canonical correlation coefficients between variables related to drug use and the canonical of oral health problems Variables related to intensity of use of psychoactive substances Canonical variable related to oral health Tobacco products 0.0290 Alcoholic beverages 0.0071 Marijuana 0.2730 Cocaine, crack 0.3934 Amphetamine-type stimulants 0.1498 Inhalants 0.1419 Sedatives and sleeping pills 0.0269 Hallucinogens 0.0390 Opioids 0.0190 Characteristics of the canonical axis r=0.4641; R2=0.2154; F=1.87; p=0.0005 In the multivariate analysis of the associations between the oral health variables (Group I) and variables related to the users’ living conditions (Group III), group I was formed by the five oral health variables and group III by the six variables related to the user´s socioeconomic conditions. As the number of canonical correlations that can be tested is equal to the number of variables in the group with the smallest number of variables, five independent canonical correlations were tested. Only the first canonical correlation (formed by combining the two groups of variables with the highest correlation) was significant (canonical correlation coefficient=0.4533; p<0.0001). This value represents the highest possible correlation between any combi- nation of the oral health problem variables and any combination of the living condition variables. The other correlation pairs, which are not correlated with the first, were not significant (p=0.3039; p=0.7929; p=0.8945 and p=0.9335) (Table 3). 8 Marques et al. Braz J Oral Sci. 2025;24:e255233 Table 3 presents the multivariate statistics for the null hypothesis that all canonical correlations are zero. The low p-values for these tests (p<0.05) suggest rejecting the null hypothesis that all canonical correlations are zero in the population. Table 3. Standardized Canonical Coefficients of the canonical variables related to socioeconomic condition and oral health problems in drug users. Variables Standardized Canonical Coefficients Oral health problems (group I) DMFT 0.8551 Need of dental treatment 0.1904 CPI 0.3724 Use of dental prosthesis 0.1331 Need of dental prosthesis 0.0529 Socioeconomic conditions (group III) Monthly income 0.2474 Number of people in the house 0.2288 Participant’s education 0.1823 Mother’s education 0.5761 Father’s education 0.3181 Type of housing 0.1143 Wilks’Lambda=0.7226 (p<0.0001); Pillai’s Trace=0.2980 (p<0.00001); Hotelling-LawleyTrace=0.3561 (p<0.0001); Roy’s Largest Root=0.2586 (p<0.0001) The standardized canonical coefficients show that the first canonical variable for the oral health problems group is a weighted combination of DMFT (0.8551), CPI (0.3724), need for treatment (0.1904), denture use (0.1331) and need for prosthesis (0.0529) with emphasis on DMFT and CPI indexes. The coefficients for the socioeconomic condition group variables showed that the canonical variable for the group of these variables was formed by the mother’s educa- tion level, father’s education level, income, number of people in the house, user’s level of education and type of housing, with emphasis on the mother’s (0.5761) and father’s (0.3181) level of education. Table 4 presents the canonical correlation coefficients between the living condition variables and the canonical oral health. It can be seen that these analyzed charac- teristics are related (R2=0.2055), that is, the variation in living conditions explained 20% of the variation in oral health problems in drug users. Worse living conditions, especially lower education levels of mothers and fathers, were associated with oral health problems in addicted patients. 9 Marques et al. Braz J Oral Sci. 2025;24:e255233 Table 4. Canonical correlation coefficients between socioeconomic condition variables and the canonical of oral health problems. Variables related to socioeconomic conditions Canonical variable related to oral health Monthly income 0.2050 Number of people in the house 0.1729 Participant’s education 0.2043 Mother’s education 0.3827 Father’s education 0.3226 Type of housing 0.0187 Characteristics of the canonical axis r=0.4533; R2=0.2055; F=2.84; p<0.0001 Discussion The present study demonstrated, through multivariate analyses, that variables related to intensity of use of psychoactive substances and socioeconomic and demographic variables were associated with oral health problems. Governments and public health systems have intensified their concern to provide spe- cialized care to this population1,2. In the case of Brazil, these individuals are assisted in CAPS-ADs by multiprofessional teams and those with dental problems should be referred for treatment at Family Health Units, as dental professionals are not part of teams working in CAPS-ADs19. This fact may hinder quick dental care of those patients who usually present high rates of oral problems, including dental pain20. This situation of difficulty in accessing oral health services by those patients unfor- tunately is a worldwide phenomenon as pointed out by the scientific literature7,9,19,21. Sheridan et al.22 (2001) evaluated the oral health conditions and access to dental treatment of a sample of drug users and non-users and concluded that drug users reported having greater difficulty in accessing dental treatment and had a significantly higher level of oral health problems. In the present study, 71.7% of patients presented dental treatment needs, corroborating the high prevalence of oral problems in these individuals as in other studies7,9,19-22. The multivariate analysis indicated that there was a significant correlation between the intensity of drug use and oral health problems in the evaluated patients (p<0.05). This fact can be explained by several reasons: users of illicit drugs often neglect their oral hygiene; drugs can harm mouth and teeth by inducing dry mouth, which favor bacterial growth and acid in the mouth. In addition, drugs can induce teeth grinding, acid reflux, loss of blood flow to gums, and greater intake of sugar food or beverages, which damage their teeth9,23,24. In this study, the variation in the intensity of drug consumption explained approxi- mately 22% of the variation in the variables related to oral health. Thus, high con- sumption of drugs, especially cocaine/crack and marijuana, was associated with oral health problems. This fact indicated that cocaine/crack were the worse drugs associated with oral health outcomes. According to Gaio et al.25 (2021), in a study with a sample of 10 Marques et al. Braz J Oral Sci. 2025;24:e255233 7,381 participants in the National Crack Survey, individuals who smoked crack cocaine were polyusers, a fact that can increase the impact of addiction on oral health. Cury et al.26 (2017) compared the oral health of 40 cocaine users with 120 non-users and found a statistically significant association when considering the missing component of the DMFT index. Antoniazzi et al.27 (2021) observed that crack cocaine users had a greater frequency of tooth loss, severity of dental caries and periodontal disease and less use of dental services than the control group. In addition, Toledo et al.28 (2017) describe that cocaine users are the ones with the greatest difficulty in accessing health services. Together, all these factors end up interfering with the oral health of this type of drug user. Regarding the analysis of associations between sociodemographic variables and oral health, information on all exposure variables (living conditions) and the sets of outcome variables (oral health problems), the variation in living conditions explained 20% of the variation in oral health problems. These results confirmed that worse living conditions, mainly low level of education of the mother and father, were associated with oral health problems in drug users. According to Jorge et al.29 (2017), lower socioeconomic status is a variable associated with higher consump- tion of psychoactive substances at same time that it exerts an influence on the health situation. As expected, since these are studies with human beings, the correlation coefficients are not very high, since many other variables are involved in the oral health of each individual. However, this analysis is of great importance for a holistic and expanded interpretation regarding the association between the intensity of illicit substance use and oral health problems. This study presents some limitations. It was a cross-sectional study and a tempo- ral relationship between variables cannot be established. Data collection was car- ried out in 2013 and, therefore, may not represent the current condition of this pop- ulation. Despite that, it is one of the few studies to date that carried out this type of analysis in this population, contributing to new evidence of association between the variables analyzed. It is concluded that high consumption of drugs, mainly cocaine/crack and mari- juana, was associated with oral health problems. In addition, worse living condi- tions, especially lower education of the mother and father, were associated with oral health problems in drug users. Public health and intersectoral interventions aimed at providing holistic care and facilitating access to dental care for this population may improve the quality of life and oral health of these individuals. Acknowledgments The Brazilian National Council for Scientific and Technological Development (CNPq). Data Availability Datasets related to this article will be available upon request to the corresponding author. 11 Marques et al. Braz J Oral Sci. 2025;24:e255233 Conflict of Interest None. Author Contribution Tais Cristina Nascimento Marques: conception and design of the work; acquisition of data for the work; drafting the work or reviewing it critically for important intellectual content. Karin Luciana Migliato Sarracini: acquisition of data for the work; drafting the work or reviewing it critically for important intellectual content. Vanessa Gallego Arias Pecorari: drafting the work or reviewing it critically for important intellectual content. Karine Laura Cortellazzi: conception and design of the work; drafting the work or reviewing it critically for important intellectual content. Marcelo de Castro Meneghim: conception and design of the work; drafting the work or reviewing it critically for important intellectual content. Luciana Estevam Simonato: acquisition of data for the work; drafting the work or reviewing it critically for important intellectual content. Fábio Luiz Mialhe: conception and design of the work; drafting the work or reviewing it critically for important intellectual content. Gláucia Maria Bovi Ambrosano: conception and design of the work; analysis of the data for the work; drafting the work or reviewing it critically for important intellectual content. All authors actively revised and approved the final version of the manuscript. References 1. WHO. World Health Organization. Drugs (psychoactive) [cited 2023 Oct 8]. Available from: https://www.who.int/health-topics/drugs-psychoactive#tab=tab_1. 2. United Nations Office on Drugs and Crime (UNODC). World drug report 2021: Global overview - drug demand and drug supply. Vienna: UNODC; 2021 [cited 2023 Sep 20]. Available from: https://www.unodc.org/res/wdr2021/field/WDR21_Booklet_2.pdf. 3. Brazilian Ministry of Justice. [Legislation and public policies on drugs in Brazil]. Brasília: National Secretariat for Drug Policy; 2011 [cited 2023 Jul 15]. Available from: https://dspace.mj.gov.br/handle/1/11374. 4. Gallassi AD, Nakano EY, Wagner GA, de Oliveira Silva MN, Fischer B. Characteristics of clients using a community-based drug treatment service (‘CAPS-AD’) in Brazil: An exploratory study. Int J Drug Policy. 2016 May;31:99-103. doi: 10.1016/j.drugpo.2016.01.020. 5. Oliveira EN, Vasconcelos Alves PM, Ximenes Neto FR, Teles Andrade A, Vieira Gomes B, Grande AJ. Clinical aspects and care production in a Brazilian psychosocial care center of alcohol and other drugs (CAPS AD). Issues Ment Health Nurs. 2014 May;35(5):356-63. doi: 10.3109/01612840.2013.86928. 6. WHO. World Health Organization. International standards for the treatment of drug use disorders. Geneva: WHO; 2020. 7. D’Amore MM, Cheng DM, Kressin NR, Jones J, Samet JH, Winter M, et al. Oral health of substance-dependent individuals: impact of specific substances. J Subst Abuse Treat. 2011 Sep;41(2):179-85. doi: 10.1016/j.jsat.2011.02.005. 8. Cury PR, Oliveira MG, dos Santos JN. Periodontal status in crack and cocaine addicted men: a cross-sectional study. Environ Sci Pollut Res Int. 2017 Feb;24(4):3423-9. doi: 10.1007/s11356-016-7918-0. Epub 2016 Nov 20. 12 Marques et al. Braz J Oral Sci. 2025;24:e255233 9. Shekarchizadeh H, Khami MR, Mohebbi SZ, Ekhtiari H, Virtanen JI. Oral health of drug abusers: a review of health effects and care. Iran J Public Health. 2013 Sep;42(9):929-40. 10. Bellocchio L, Inchingolo AD, Inchingolo AM, Lorusso F, Malcangi G, Santacroce L, et al. Cannabinoids drugs and oral health-from recreational side-effects to medicinal purposes: a systematic review. Int J Mol Sci. 2021 Aug;22(15):8329. doi: 10.3390/ijms22158329. 11. Teoh L, Moses G, McCullough MJ. Oral manifestations of illicit drug use. Aust Dent J. 2019 Sep;64(3):213-22. doi: 10.1111/adj.12709. 12. Rossow I. Illicit drug use and oral health. Addiction. 2021 Nov;116(11):3235-42. doi: 10.1111/add.15360. 13. Hair Junior JF, Black WC, Babin BJ, Anderson RE, Tatham RL. [Multivariate data analysis]. 6. ed. Porto Alegre: Bookman; 2009. 688p. Portuguese. 14. Ribeiro Edel P, Oliveira JA, Zambolin AP, Lauris JR, Tomita NE. [Integrated approach to the oral health of drug-addicted undergoing rehabilitation]. Pesq Odontol Bras. 2002 Jul-Sep;16(3):239-45. Portuguese. doi: 10.1590/s1517-74912002000300010. 15. WHO. World Health Organization. [Basic health surveys]. 4. ed. São Paulo: Santos; 1999. Portuguese. 16. Henrique IF, De Micheli D, Lacerda RB, Lacerda LA, Formigoni ML. [Validation of the Brazilian version of Alcohol, Smoking and Substance Involvement Screening Test (ASSIST)]. Rev Assoc Med Bras (1992). 2004 Apr-Jun;50(2):199-206. Portuguese. doi: 10.1590/s0104-42302004000200039. 17. WHO ASSIST Working Group. The Alcohol, Smoking and Substance Involvement Screening Test (ASSIST): development, reliability and feasibility. Addiction. 2002 Sep;97(9):1183-94. doi: 10.1046/j.1360-0443.2002.00185.x. 18. Ferreira DF. [Multivariate statistics]. 2. ed. Lavras: Universidade Federal de Lavras; 2011. Portuguese. 19. Falcão CAM, Santos RO, Pereira RMS, Silva TSO, Ferreira RS, Silva FWC, et al. [Oral health in addicts]. Rev Interdiscip Cien Saude 2015;2(3):112-21. Portuguese. 20. Lorencini BV, Bissoli BC, Simões JC, Miotto MHMB, Siqueira MM, Zandonade E. [Dental pain among psychoactive substance users in CAPS AD in Vitoria, Vila Velha and Serra, ES, Brazil]. Cien Saude Colet. 2019 Oct;24(11):4093-102. Portuguese. doi: 10.1590/1413-812320182411.29582017. 21. Mohammadi TM, Hasheminejad N, Salari HR, Rostamizadeh MR, Najafipour H. Association between tooth loss and opium addiction: results of a community-based study on 5900 adult individuals in South East of Iran in 2015. J Int Soc Prev Community Dent. 2017 Jul-Aug;7(4):186-90. doi: 10.4103/jispcd.JISPCD_189_17. 22. Sheridan J, Aggleton M, Carson T. Dental health and access to dental treatment: a comparison of drug users and non-drug users attending community pharmacies. Br Dent J. 2001 Oct;191(8):453-7. doi: 10.1038/sj.bdj.4801206. 23. Cuberos M, Chatah EM, Baquerizo HZ, Weinstein G. Dental management of patients with substance use disorder. Clin Dent Rev. 2020;4(1):14. doi: 10.1007/s41894-020-00078-8. 24. Yazdanian M, Armoon B, Noroozi A, Mohammadi R, Bayat AH, Ahounbar E, et al. Dental caries and periodontal disease among people who use drugs: a systematic review and meta-analysis. BMC Oral Health. 2020 Feb;20(1):44. doi: 10.1186/s12903-020-1010-3. 25. Gaio DC, Bastos FI, Moysés SJ, Moysés ST, Corrêa da Mota J, Nicolau B, et al. Assessing oral health of crack users in Brazil: perceptions and associated factors, findings from a mixed methods study. Glob Public Health. 2021 Apr;16(4):502-16. doi: 10.1080/17441692.2020.1809693. 13 Marques et al. Braz J Oral Sci. 2025;24:e255233 26. Cury PR, Oliveira MG, de Andrade KM, de Freitas MD, Dos Santos JN. Dental health status in crack/cocaine-addicted men: a cross-sectional study. Environ Sci Pollut Res Int. 2017 Mar;24(8):7585-90. doi: 10.1007/s11356-017-8404-z. 27. Antoniazzi RP, Palmeira RV, Schöffer C, Dos Santos BZ, Zanatta FB, Feldens CA. Use of crack cocaine increases tooth loss. Am J Dent. 2021 Dec;34(6):317-21. 28. Toledo L, Góngora A, Bastos FI. [On the sidelines of society: crack use, deviation, criminalization and social exclusion - a narrative review]. Cien Saude Colet. 2017 Jan;22(1):31-42. Portuguese. doi: 10.1590/1413-81232017221.02852016. 29. Jorge KO, Ferreira RC, Ferreira EF, Vale MP, Kawachi I, Zarzar PM. Binge drinking and associated factors among adolescents in a city in southeastern Brazil: a longitudinal study. Cad Saude Publica. 2017 Mar;33(2):e00183115. doi: 10.1590/0102-311X00183115.