Revista Ibero-Americana de Humanidades, Ciências e Educação — REASE doi.org/10.51891/rease.v12i9.30592 IMPACTS OF THE COVID-19 PANDEMIC ON METHODS AND RECURRENCE OF SELF-HARM AMONG BRAZILIAN ADOLESCENTS IMPACTOS DA PANDEMIA DE COVID-19 NOS MÉTODOS E NA RECORRÊNCIA DE AUTOLESÃO ENTRE ADOLESCENTES BRASILEIROS IMPACTOS DE LA PANDEMIA DE COVID-19 EN LOS MÉTODOS Y LA RECURRENCIA DE AUTOLESIONES ENTRE ADOLESCENTES BRASILEÑOS Leticia Costa Leopoldino Jacobs1 Vanessa Barbosa Romera Leme2 Clarissa Pinto Pizarro de Freitas3 Milene Maria Xavier Veloso4 Júlia Rebeca da Silva Magalhães Barros5 Luísa Fernanda Habigzang6 ABSTRACT: This study analyzed predictors of recurrence and methods of self-harm among Brazilian ISSN: 2675-3375 2 Revista Ibero-Americana de Humanidades, Ciências e Educação — REASE Violence, as defined by the WHO (2002), refers to the intentional use of physical force or power against oneself, another person, or a community, with a high likelihood of resulting in injury, death, psychological harm, or impaired development. Among the different manifestations of violence, self-harm, focus of the present study includes suicidal behaviors such as ideation, planning, and attempts, as well as self-injurious acts with or without suicidal intent (WHO, 2002). It is estimated that for every suicide death there are more than 20 attempts, highlighting the magnitude and urgency of addressing this phenomenon (ALVES FJO, et al., 2024). Although suicide and non-suicidal self-injury are distinct phenomena, both share trajectories of psychological suffering along a continuum of mental health risk in adolescence (NOCK MK, 2016). Self-injury refers to intentional behaviors directed against oneself that may or may not result in tissue damage, ranging from milder forms such as scratching or interfering with wound healing to potentially lethal methods, including cutting, hanging, and poisoning (NOCK MK, 2009; NOCK MK, 2010; VELOSO-BESIO C, et al., 2023). These behaviors often function as maladaptive emotion regulation strategies in response to adverse experiences such as intense distress, family conflict, interpersonal difficulties, and everyday stressors (NOCK MK, 2010). Many adolescents report engaging in self-injury to obtain temporary emotional relief, a dynamic described as addictive due to the reinforcing cycle of pain and relief (NOCK MK, 2010). In Brazil, self-inflicted injuries have been subject to compulsory reporting in the Notifiable Health Conditions Information System (SINAN) since 2011, although the system does not distinguish suicidal intent from self-harm (BRASIL, 2024). Despite this limitation, SINAN provides important evidence on the epidemiological profile of self-harm among Brazilian adolescents. In 2021, 114,159 cases of self-directed violence were recorded, 70.3% involving females. The most frequent self-harm methods was poisoning/intoxication (67.1%), followed by sharp objects (17.9%) and hanging (6.6%). Men had a 3.3-fold higher risk of selfharm by hanging and a 7.2-fold higher risk by firearm compared with women, while Indigenous populations presented higher proportions of these methods (BRASIL, 2024). Over recent decades, self-harm among adolescents has become a growing public health concern. The COVID-19 pandemic worsened mental health indicators worldwide, and infectious disease outbreaks are recognized as factors that intensify suicidal and self-injurious behaviors, particularly in contexts of socioeconomic vulnerability (ALVES FJO, et al., 2024; Revista Ibero-Americana de Humanidades, Ciências e Educação. São Paulo, v. 12, n. 9, set. 2026. ISSN: 2675-3375 3 Revista Ibero-Americana de Humanidades, Ciências e Educação — REASE SOARES FC, et al., 2022; WHO, 2025a). Globally, suicide rates declined by 36% between 2000 and 2019; however, they increased by 17% in the Americas, with Brazil showing a rise of 43% (ALVES FJO, et al., 2024). During the pandemic, although overall suicide rates in Brazil remained relatively stable, increases were observed among adolescents, especially girls, Black youth, and residents of peripheral areas (ORNELL F, et al., 2022). The Brazilian pandemic context was characterized by worsening social vulnerabilities. Poverty and food insecurity increased, disproportionately affecting populations already experiencing socioeconomic inequalities (NASSIF PIRES L, et al., 2021). Prolonged school closures widened educational disparities and negatively affected the social and emotional development of children and adolescents (NASSIF PIRES LN, et al., 2021). At the same time, domestic violence and unemployment increased, revealing the fragility of social protection systems and the limited reach of public policies during the crisis (SOUZA LJ e FARIAS RCP, 2022). Furthermore, the absence of effective violence-prevention strategies deepened ethnoracial and gender inequalities, disproportionately affecting women, Black populations, and socially vulnerable groups (ALLOATTI MN e MATOS OAL, 2023; DANTAS ESO, et al., 2023). These dynamics highlight the intersection between social determinants of health and historical inequalities in Brazil. Although there is consensus that the COVID-19 pandemic affected the mental health of children and adolescents, evidence regarding its impact on self-harm remains mixed. In Brazil, some studies reported reductions in hospitalization rates combined with proportional increases in suicide deaths during the pandemic (CABRAL S, et al., 2023; GOMES FILHO CH, et al., 2022), while others found no significant changes in notifications among children and adolescents (SGOBBI FM, et al., 2022). Internationally, a meta-analysis identified increases in suicidal ideation and suicide attempts but not in suicide mortality rates (YAN Y, et al., 2023). Consequently, recent research has sought to characterize the epidemiological profile of selfharm among Brazilian adolescents, focusing on recurrence, methods used, victim characteristics, and associated contextual factors (BRITO FAMD, et al., 2021; CABRAL S, et al., 2023; CIVIERO JCZ e TORRES JRP, 2024; FATTAH N e LIMA MS, 2020; GOMES FILHO CH, et al., 2022; LEITE FMC, et al., 2022; SGOBBI FM, et al., 2022). The recurrence of self-inflicted injuries has emerged as an important indicator of persistent psychological distress among adolescents. Evidence shows that a substantial proportion of reported cases involves individuals with a previous history of self-harm. In Revista Ibero-Americana de Humanidades, Ciências e Educação. São Paulo, v. 12, n. 9, set. 2026. ISSN: 2675-3375 4 Revista Ibero-Americana de Humanidades, Ciências e Educação — REASE Brazil, analyses of SINAN indicate that recurrent cases account for approximately four in every ten notifications of self-directed violence (BRASIL, 2024). Similarly, Fattah N and Lima MS (2020) found that 41% of notifications in a southern state involved individuals with previous episodes of self-harm or suicide attempts, with higher rates among female adolescents aged 15–19 years. In Espírito Santo, Leite FMC, et al. (2022) observed recurrent violence in 46.4% of adolescent cases, with higher prevalence among girls (PR = 1.26), adolescents aged 10–14 (PR = 1.20), and those with disabilities or mental disorders (PR = 1.52). The home was the main setting of repeated episodes, emphasizing the domestic environment as an important context of vulnerability. These findings indicate that recurrence reflects ongoing exposure to cumulative risk factors, including interpersonal violence, structural inequalities, and limited access to psychosocial care (HOLLAND KM, et al., 2017; UDDIN R, et al., 2019; YAN Y, et al., 2023). Regarding methods of self-harm, Brito FAMD, et al. (2021) analyzed notifications between 2009 and 2016 and identified a substantial increase in recorded cases from 2.1 to 25.7 per 100,000 adolescents. Notifications were more frequent among girls aged 15–19 years and occurred predominantly in the home environment. Poisoning was the most frequent self-harm methods, followed by sharp objects, with a marked increase among female adolescents. The authors also highlighted limitations in the quality of notifications and in the training of health professionals to identify and report cases, emphasizing the need to strengthen prevention and follow-up strategies. Other Brazilian studies have described consistent patterns of inequality in self-harm notifications, particularly regarding gender, race/color, and geographic context. Female adolescents aged 15–19 and those identified as Black or Brown are among the most affected groups, with poisoning and cutting as the most common methods and the home as the main place of occurrence (CIVIERO JCZ e TORRES JRP, 2024; FATTAH N e LIMA MS, 2020; LEITE FMC, et al., 2022). These patterns reinforce the role of structural inequalities and intersectionality in shaping psychosocial vulnerability, highlighting the importance of public policies that explicitly address gender identity, ethnic-racial dimensions, and sexual diversity (WEAVER ND, et al., 2025). Evidence also demonstrates strong associations between self-injurious behavior, exposure to interpersonal violence, mental disorders, and social inequalities (LEITE FMC, et al., 2022). A retrospective cohort study conducted in Taiwan found that adolescents exposed to Revista Ibero-Americana de Humanidades, Ciências e Educação. São Paulo, v. 12, n. 9, set. 2026. ISSN: 2675-3375 5 Revista Ibero-Americana de Humanidades, Ciências e Educação — REASE violence had significantly higher risks of suicide and mental disorders than those without such experiences (SUNG C, et al., 2025). Over a 15-year follow-up period, suicide rates were significantly higher among adolescents who had experienced violence, even after adjusting for sex, age, area of residence, and psychiatric comorbidities. Among victimized adolescents, the most frequent suicide methods were poisoning, hanging, firearm use, and sharp instruments (SUNG C, et al., 2025). These findings highlight the importance of understanding how victimization experiences affect adolescent mental health and emphasize the need for targeted psychological interventions and prevention policies. Despite advances in the literature, important gaps remain regarding the temporal dynamics of self-harm and the effects of critical events such as the COVID-19 pandemic on recurrence and methods used (YAN Y, et al., 2023). In addition, the interaction between sociodemographic characteristics, exposure to other forms of violence, and contextual factors of occurrence remains insufficiently explored. Repeated exposure to multiple forms of violence, known as polyvictimization, has been associated with greater psychological vulnerability and increased risk of revictimization (SUNG C, et al., 2025). In this context, the present study aimed to test predictive models for recurrence and for the methods employed in adolescent selfharm (hanging, blunt objects, sharp instruments, hot substances/thermal injury, and poisoning/intoxication, as well as firearms), taking into account the pandemic period (pre, during, and post-COVID-19), exposure to other forms of violence, location and area of occurrence, and demographic variables (sex, age, race/ethnicity identification, schooling, and the presence of disability or mental disorder). By identifying associated patterns and factors, we seek to provide empirical evidence to inform improved prevention strategies, early intervention, and the improvement of public policies targeted at adolescent mental health. METHODS This is an ecological study with an epidemiological survey design that used notifications of self-directed violence among adolescents aged 10 - 19 years recorded in Brazil’s Notifiable Health Conditions Information System (SINAN) of the Unified Health System (SUS). Data on notifications of self-harm among adolescents were collected from SINAN via the State Health Surveillance Secretariats-specifically, the Noncommunicable Diseases and Injuries Surveillance Units-in seven Brazilian states (Amazonas, Pará, Ceará, the Federal District, São Paulo, Rio de Janeiro, and Rio Grande do Sul), spread across the country’s five macro-regions. Revista Ibero-Americana de Humanidades, Ciências e Educação. São Paulo, v. 12, n. 9, set. 2026. ISSN: 2675-3375 6 Revista Ibero-Americana de Humanidades, Ciências e Educação — REASE Each state provided a database containing all Individual Notification Forms (FIN-SINAN) filed by health professionals for cases of self-directed violence against adolescents between 2018 and 2022. The data collection is part of a multicenter, multimethod study on family violence in Brazil that includes secondary analyses of official records, surveys on individual and family indicators, and qualitative interviews. The chosen time frame encompasses a recent period of notifications of self-harm in the Brazilian context, expanding the available evidence beyond a prior national study that described notifications recorded by health professionals between 2018 and 2021 (SGOBBI FM, et al., 2022). Data analysis was conducted in 2023. Initially, suspected or confirmed cases of violence are recorded by professionals in public or private health services using the FIN-SINAN form, which comprises 69 items distributed across eight categories: general data; individual notification; household information; information about the person assisted; characteristics of the occurrence; types of violence; information on the probable perpetrator; and referral information. Forms are completed based on psychosocial and clinical assessments conducted by professionals, following the specific guidance in the FIN-SINAN manual (BRASIL, 2016). For each variable, professionals selected from predefined response options in addition to “unknown” and “not applicable.” The “unknown” option was used when information was unavailable (e.g., unconscious victims or absence of an accompanying person), whereas “not applicable” was used when the question did not pertain to the victim (e.g., pregnancy items for male individuals). Importantly, these alternatives were not considered missing data. Subsequently, municipal, state, and federal surveillance intermediaries transfer the data to SINAN Net, an application developed by DataSUS/MS. SINAN Net is used to collect, transmit, and consolidate data obtained daily by the epidemiological surveillance system for conditions of compulsory notification, providing essential information for decision-making and for analyzing the population’s morbidity profile within SUS (BRASIL, 2016). As part of evidence-informed decision-making practices, some of the collected information is publicly available at https://datasus.saude.gov.br/informacoes-de-saude-tabnet/. In Brazil, compulsory notification of self-harm and other forms of violence against adolescents is regulated by Ordinance No. 104/2011 and is mandatory for professionals in both the public and private health networks. Data collection follows the guidelines of the “Instruction Sheet for Reporting Interpersonal and Self-Inflicted Violence” (BRASIL, 2016), which standardizes completion and Revista Ibero-Americana de Humanidades, Ciências e Educação. São Paulo, v. 12, n. 9, set. 2026. ISSN: 2675-3375 7 Revista Ibero-Americana de Humanidades, Ciências e Educação — REASE reporting procedures. Access to the full database was authorized by the Research Ethics Committee (CAAE: [OMITTED FOR BLIND REVIEW]). The sample included 78,658 suspected or confirmed cases of self-harm among adolescents aged 10 to 19 years, recorded between 2018 and 2022. The mean age was 15.8 years (SD = 2.2). Most cases involved female (78.7%), white (51.7%) or brown (30.8%) individuals, and those without physical, auditory, or visual disabilities (78.1%). A total of 1.5% of the sample reported mental disabilities, while 21.7% presented with a mental disorder. Regarding education, adolescents were predominantly enrolled in primary (31.0%) or secondary (35.8%) education. Most cases occur in urban zones (92.8%), with residences being the primary place of occurrence (84.4%). Reports were distributed across different periods of social isolation: before (46.3%), during (11.7%), and after (42.0%). Exposure to other forms of violence was also reported, including physical violence (27.2%), psychological violence (5.5%), negligence (1.3%), sexual violence (0.2%), and torture (0.5%). Various means of aggression were identified, most frequently poisoning/intoxication (56.0%) and sharp objects (28.5%), followed by hanging (4.2%), blunt objects (2.0%), hot substances/objects (0.7%), and firearms (0.3%). Recurrence of self-harm was observed in nearly half of the cases (47.7%). Sociodemographic and contextual 8 information on the reported cases is summarized in (Table 1). Table 1. Sociodemographic characteristics of the sample (N = 78,658). Variable Gender Male Female Ignored Missing Ager 10 11 12 13 14 15 16 17 18 19 Race/ethnicity identification White Black Yellow Brown Indigenous N (%) 16748 61895 15 (21,3) (78,7) (0,0) 0 609 1714 4053 7139 9585 11254 12048 11545 10426 10285 (0,7) (2,1) (5,1) (9,0) (12,1) (14,3) (15,3) (14,6) (13,2) (13,0) 40557 5190 470 24175 511 (51,7) (6,6) (0,6) (30,8) (0,7) Revista Ibero-Americana de Humanidades, Ciências e Educação. São Paulo, v. 12, n. 9, set. 2026. ISSN: 2675-3375 Revista Ibero-Americana de Humanidades, Ciências e Educação — REASE Ignored Missing Physical Auditory and Visual Disability Yes No Ignored Missing Mental Disability Yes No Ignored Missing Mental Disorder Yes No Ignored Missing Schooling Primary Education Secondary Education Incomplete Higher Education Ignored Missing Occurrence Zone Rural Urban Peri-urban Ignored Missing Place of occurrence Bar or similar Commercial/Services School Collective housing Industry/Construction Sports facility Residence Public road Other Ignored Missing Period of social isolation Before During After Missing Exposure to other forms of violence Physical Yes No Ignored Missing Psychological Yes No Ignored Missing 7523 232 (9,6) 399 60476 16519 1264 (0,5) (78,1) (21,3) 1178 59721 16519 1240 (1,5) (77,1) (21,3) 16918 44609 16517 614 (21,7) (57,2) (21,2) 23437 27062 1373 23721 3065 (31,0) (35,8) (1,8) (31,4) 3662 69059 408 1319 4210 (4,9) (92,8) (0,5) (1,8) 142 235 2095 774 24 63 66319 2546 1621 4803 36 (0,2) (0,3) (2,7) (1,0) (0,0) (0,1) (84,4) (3,2) (2,1) (6,1) 36415 9177 33066 0 (46,3) (11,7) (42,0) 9 12409 32641 465 77 (27,2) (71,7) (1,0) 2522 42180 732 158 (5,5) (92,8) (1,6) Revista Ibero-Americana de Humanidades, Ciências e Educação. São Paulo, v. 12, n. 9, set. 2026. ISSN: 2675-3375 Revista Ibero-Americana de Humanidades, Ciências e Educação — REASE Neglect Yes No Ignored Missing Sexual Yes No Ignored Missing Torture Yes No Ignored Missing Means of Aggression – Hanging Yes No Ignored Missing Means of Aggression – Blunt object Yes No Ignored Missing Means of Aggression – Sharp object Yes No Ignored Missing Means of Aggression – Hot substance/object Yes No Ignored Missing Means of Aggression – Poisoning/Intoxication Yes No Ignored Missing Means of Aggression – Firearm Yes No Ignored Missing Recurrence Yes No Ignored Missing 611 44047 752 182 (1,3) (96,9) (1,6) 117 44537 748 190 (0,2) (98,0) (1,6) 264 44396 741 191 (0,5) (97,7) (1,6) 3274 73012 1282 1090 (4,2) (94,1) (1,7) 1544 (2,0) 74652 (96,3) 1302 (1,7) 1160 22166 54527 1120 845 (28,5) (70,1) (1,4) 546 75703 1290 1119 (0,7) (97,6) (1,7) 43678 33461 1165 603 (56,0) (42,9) (1,2) 197 76017 1294 1150 (0,3) (98,1) (1,7) 37386 25944 15120 208 (47,7) (33,1) (19,3) Fonte: Dados SINAN . Nota: M = Mean; SD = Standard Deviation. The analysis focused on reports of self-harm among adolescents across all regions of the country, with emphasis on the seven Brazilian states noted above. The variables analyzed Revista Ibero-Americana de Humanidades, Ciências e Educação. São Paulo, v. 12, n. 9, set. 2026. ISSN: 2675-3375 10 Revista Ibero-Americana de Humanidades, Ciências e Educação — REASE included: sex (female, male); age (10-19 years); race/color (White, Black, Brown, Yellow/Asian, Indigenous); method of self-injury (hanging, blunt object, sharp instrument, hot substance/object, poisoning/intoxication, and firearm); area of occurrence (urban, periurban, rural); place of occurrence (residence, public thoroughfare, public areas); and recurrence (SINAN form item 53: “occurred other times”). In addition, we examined the number of notifications across three COVID-19 - related periods: (a) before social distancing (January 1, 2018 to March 19, 2020); (b) during mandated social distancing (March 20, 2020 to January 17, 2021); and (c) after social distancing (January 18, 2021 to December 31, 2022). Binary logistic regression analyses were conducted to test two predictive models. The first model used recurrence as the outcome, with the following predictors: demographic variables (age, sex, race/color, schooling, and disability/mental disorder), area of occurrence, place of occurrence, isolation period (pre-, during-, and post-COVID-19), and exposure to other forms of violence (physical, psychological, neglect, sexual, and torture). The second model used method of self-harm as the outcome (hanging, blunt object, sharp instrument, hot substance/thermal injury, poisoning/intoxication, and firearm). For each method, the predictors tested were: demographic variables (age, sex, race/color, schooling, and disability/mental disorder), area of occurrence, place of occurrence, isolation period (pre-, during, and post-COVID-19), exposure to other forms of violence (physical, psychological, neglect, sexual, and torture), and recurrence. For reasons of parsimony and to improve the clarity of presentation, only the most frequent self-harm methods (poisoning/intoxication and sharp instrument) are presented in tables in the Results section. The tables presenting the results for the less frequent self-harm methods are available from the authors upon request. Odds ratios (ORs) were estimated for the predictors, and model explanatory power was assessed using Tjur’s R². To enhance parsimony, cases with sex recorded as “unknown” were excluded from inferential analyses. For the logistic regressions, similar categories were combined and variables were coded as follows: (1) age as a continuous variable; (2) sex: male = 1, female = 0; (3) race/color: White = 0, ethnoracial minority = 1 (Black, Brown, Indigenous, or Yellow/Asian); (4) schooling: elementary or less = 0 (illiterate or elementary school), secondary = 1 (high school or incomplete higher education); (5) disability/mental disorder: none = 0, presence of any condition = 1 (physical disability, mental disorder, or psychological condition); (6) area of occurrence: urban = 0, peri-urban or rural = 1; (7) place of occurrence: residence = 0, public roads/spaces = 1; (8) isolation period: pre = 0, during = 1, post = 2; (9) Revista Ibero-Americana de Humanidades, Ciências e Educação. São Paulo, v. 12, n. 9, set. 2026. ISSN: 2675-3375 11 Revista Ibero-Americana de Humanidades, Ciências e Educação — REASE exposure to physical, psychological, neglect, sexual violence, and torture: absent = 0, present = 1; and (10) recurrence: absent = 0, present = 1. Statistical analyses were performed using RStudio (2020), an integrated development environment for R. RESULTS Recurrence of self-harm was more frequent among male adolescents (OR = 1.392; ≈39% higher than females), among those with any disability or mental disorder (OR = 2.713), and among victims who had experienced sexual violence (OR = 1.650) or torture (OR = 1.502). Lower odds of recurrence were observed in peri-urban/rural areas compared with urban settings (OR = 0.803) and among adolescents who had experienced neglect (OR = 0.748). Tjur’s R² was 0.052, indicating that the model explained a small proportion of the variance in recurrence (Table 2). Table 2. Coefficients of the multiple logistic regression model including all possible independent variables recurrence and methods of self-harm behavior. Recurrence b SE df p-value OR Exp (B) 95% CI for OR Upper Lower Constant -0.100 0.079 14 0.204 0.905 0.775 1.056 Age -0.015 0.005 14 0.002 0.985 0.976 0.994 Sex (1) 0.331 0.025 14 0.000 1.392 1.325 1.461 Race/ethnicity identification (1) 0.126 0.020 14 0.000 1.134 1.090 1.181 Disability (1) 0.998 0.023 14 0.000 2.713 2.595 2.836 Schooling (1) -0.019 0.013 14 0.150 0.981 0.956 1.007 Zone (1) -0.220 0.042 14 0.000 0.803 0.740 0.872 Place (1) 0.135 0.036 14 0.000 1.145 1.067 1.228 Isolation – during (1) 0.158 0.052 14 0.002 1.171 1.058 1.297 Isolation – after (2) 0.144 0.029 14 0.000 1.154 1.091 1.221 Physical violence (1) 0.145 0.023 14 0.000 1.156 1.105 1.210 Psychological violence (1) 0.049 0.044 14 0.258 1.051 0.965 1.145 Neglect (1) -0.290 0.107 14 0.007 0.748 0.607 0.923 R Tjur 0.052 Revista Ibero-Americana de Humanidades, Ciências e Educação. São Paulo, v. 12, n. 9, set. 2026. ISSN: 2675-3375 12 Revista Ibero-Americana de Humanidades, Ciências e Educação — REASE Sexual violence (1) 0.500 0.207 14 0.016 1.650 1.109 2.504 Torture (1) 0.407 0.137 14 0.003 1.502 1.152 1.974 Fonte: Dados do SINAN. Nota: b = logistic regression coefficient; SE = standard error; df = degrees of freedom; p-value = significance level; OR Exp (B) = odds ratio; 95% CI for OR = 95% confidence interval for the odds ratio; R Tjur = effect size measure based on model fit. Most Frequent Self-Harm Methods 1. Sharp object Analyses of self-harm by sharp instrument indicated that age was a protective factor (OR = 0.844), such that each additional year of age was associated with an approximately 15.6% decrease in the odds of the outcome, with statistical significance. Belonging to an ethnoracial minority (OR = 1.098) and having a disability (OR = 1.110) significantly increased the odds. With respect to the isolation periods, both during (OR = 1.274) and after (OR = 1.190) were associated with higher odds, statistically significant in both cases. Among violence types, physical violence (OR = 1.836), sexual violence (OR = 1.746), and especially torture (OR = 3.134) substantially increased the odds. In contrast, psychological violence (OR = 0.724) and neglect (OR = 0.455) significantly reduced the odds. Finally, recurrence showed the largest effect (OR = 3.782), indicating nearly a fourfold increase in the likelihood of the outcome, with strong statistical significance. Tjur’s R² was 0.118, indicating modest explanatory power for this model (Table 3). Table 3. Coefficients of the multiple logistic regression model including all possible independent variables methods of self-harm behavior – Sharp object. Sharp object b SE df p-value Constant Age Sex (1) Race/ethnicity identification (1) 0.749 -0.170 -0.035 0.093 0.087 0.005 0.028 0.022 15 15 15 15 0.000 0.000 0.213 0.000 OR Exp (B) 2.116 0.844 0.965 1.098 Disability (1) Schooling (1) Zone (1) Place (1) Isolation–during (1) Isolation – after (2) Physical violence (1) Psychological violence (1) 0.104 0.008 -0.026 -0.051 0.242 0.174 0.608 -0.323 0.024 0.015 0.047 0.039 0.056 0.031 0.025 0.049 15 15 15 15 15 15 15 15 0.000 0.606 0.582 0.190 0.000 0.000 0.000 0.000 1.110 1.008 0.974 0.950 1.274 1.190 1.836 0.724 95% CI for OR Upper Lower 1.783 2.510 0.835 0.853 0.913 1.021 1.050 1.147 1.059 0.979 0.888 0.880 1.141 1.120 1.748 0.658 R Tjur 0.118 1.163 1.038 1.069 1.025 1.422 1.264 1.929 0.796 Revista Ibero-Americana de Humanidades, Ciências e Educação. São Paulo, v. 12, n. 9, set. 2026. ISSN: 2675-3375 13 Revista Ibero-Americana de Humanidades, Ciências e Educação — REASE Neglect (1) Sexual violence (1) Torture (1) Recurrence (1) -0.787 0.557 1.142 1.330 0.137 0.201 0.138 0.025 15 15 15 15 0.000 0.006 0.000 0.000 0.455 1.746 3.134 3.782 0.346 1.177 2.396 3.605 0.593 2.593 4.120 3.970 Fonte: Dados do SINAN 2. Poisoning/Intoxication Self-harm by poisoning/intoxication was more likely among males (OR = 1.502) and occurred more frequently at home (OR = 0.399). The presence of physical violence (OR = 0.549), psychological violence (OR = 1.268), neglect (OR = 0.574), sexual violence (OR = 0.404), or torture (OR = 0.397) reduced the odds of poisoning. Tjur’s R² was 0.099, indicating low explanatory power of the model (Table 4). Table 4. Coefficients of the multiple logistic regression model including all possible independent variables methods of self-harm behavior – Poisoning/Intoxication. Poisoning/ Intoxication Constance Age Sex (1) Race/ethnicity Identification (1) Disability (1) Schooling (1) Zone (1) Place (1) Isolation – during (1) Isolation – after (2) Physical violence (1) Psychological violence (1) Neglect (1) Sexual violence (1) Torture (1) Recurrence (1) b SE df p-value OR Exp (B) 95% CI for OR Upper Lower 0.206 0.283 1.131 1.153 1.428 1.579 0.912 0.990 -1.423 0.132 0.407 - 0.051 0.081 0.005 0.026 0.021 15 15 15 15 0.000 0.000 0.000 0.014 0.241 1.142 1.502 0.950 -0.020 -0.003 -0.473 -0.918 -0.166 0.022 0.014 0.044 0.038 0.052 15 15 15 15 15 0.369 0.843 0.000 0.000 0.001 0.980 0.997 0.623 0.399 0.847 0.938 0.971 0.572 0.371 0.765 1.024 1.024 0.679 0.430 0.939 0.012 0.029 15 0.677 1.012 0.956 1.071 -0.600 0.024 15 0.000 0.549 0.524 0.575 0.238 0.045 15 0.000 1.268 1.162 1.385 -0.555 -0.907 0.116 0.228 15 15 0.000 0.000 0.574 0.404 0.456 0.254 0.719 0.623 -0.923 -0.868 0.151 0.021 15 15 0.000 0.000 0.397 0.420 0.294 0.403 0.531 0.437 R Tjur 0.099 Fonte: Dados do SINAN DISCUSSION Logistic regression analyses revealed that self-harm recurred in 47.7% of the sample. Prior studies reported similar figures (41% [FATTAH N e LIMA MS, 2020; 46.4% LEITE Revista Ibero-Americana de Humanidades, Ciências e Educação. São Paulo, v. 12, n. 9, set. 2026. ISSN: 2675-3375 14 Revista Ibero-Americana de Humanidades, Ciências e Educação — REASE FMC, et al., 2022]), underscoring a clinically relevant phenomenon that reflects the persistence of psychological distress and the fragility of institutional prevention and care responses. This high proportion indicates that nearly half of the notified adolescents had previously selfinjured, revealing a pattern of repetition and chronic vulnerability. The most salient predictor was the presence of disability or mental disorder (OR = 2.713), suggesting that adolescents with these conditions are nearly three times more likely to reengage in self-harm, consistent with literature highlighting the strong comorbidity between psychiatric disorders and selfdestructive behaviors (BRASIL, 2024; LEITE FMC, et al., 2022; NOCK MK, 2010; SUNG C, et al., 2025; WEAVER ND, et al., 2025). Sexual violence was also a relevant predictor, increasing the likelihood of recurrence by 65%. The literature indicates that early sexual victimization is associated not only with the onset but also with the repetition of self-harm, due to persistent feelings of guilt, shame, and helplessness that foster self-destructive coping strategies (ALVES FJO, et al., 2024; JEGLIC EL, et al., 2025; WHO, 2025b). Similarly, the experience of torture showed a significant association, underscoring the cumulative burden of severe trauma and its relationship with the chronicity of psychological distress (LIMA RFF, HABIGZANG LF e MORAIS NA, 2025; WHO, 2025b). These findings confirm the overlap between multiple forms of victimization and the risk of revictimization through self-harm, a phenomenon already discussed in studies on polyvictimization in the Brazilian context (LEITE FMC, et al., 2022; LIMA RFF, HABIGZANG LF e MORAIS NA, 2025; SUNG C, et al., 2025). Demographic factors also exerted a meaningful influence. Male adolescents (OR = 1.392) and those residing in urban areas had a higher probability of recurrence compared with female adolescents and those living in rural zones. These results differ from prior findings (FATTAH N e LIMA MS, 2020; LEITE FMC, et al., 2022). Leite FMC, et al. (2022) identified greater recurrence among girls exposed to psychological violence and neglect (30%), which may reflect temporal and geographic differences, given that their study was conducted in a single Brazilian state and prior to the COVID-19 pandemic. The magnitude of recurrence observed in the present study-coupled with multiple individual, contextual, and relational vulnerabilitiessuggests structural failures in the early detection and ongoing follow-up of adolescents experiencing psychological distress. Taken together, these findings indicate that the repetition of self-harm reflects both discontinuities in mental health care and the persistence of social and territorial inequalities, underscoring the urgency of intersectoral public policies that connect schools, families, and health and social assistance services to ensure effective monitoring and Revista Ibero-Americana de Humanidades, Ciências e Educação. São Paulo, v. 12, n. 9, set. 2026. ISSN: 2675-3375 15 Revista Ibero-Americana de Humanidades, Ciências e Educação — REASE support for adolescents living in conditions of social vulnerability (ORNELL F, et al., 2022; WEAVER ND, et al., 2025; WHO, 2025b) Self-harm by sharp objects-present in more than 28% of the sample-was significantly associated with prior exposure to multiple forms of violence, including physical and psychological violence, neglect, sexual violence, and torture. These predictors were also identified by Sung C, et al. (2025), reinforcing the link between victimization and the choice of self-injury method. According to Civiero JCZ e Torres JRP (2024), adolescents tend to use sharp objects as an emotion-regulation strategy because such methods are generally less lethal yet provide immediate relief from psychological pain. This association can also be understood through the lens of polyvictimization, in which different forms of violence converge to intensify subjective suffering and foster the adoption of self-harming practices. In the specific case of torture, the repeated use of cutting instruments by aggressors-to inflict pain, exert control, humiliate, or instill psychological terror-may be internalized by victims as a form of self-punishment or as an attempt to retain control over suffering that finds no space for elaboration (VELOSO-BESIO C, et al., 2023). In this context, self-harm ceases to be merely an individual response to distress and comes to represent the reproduction of experienced patterns of violence, often chronically and in silence. Therefore, self-harm by sharp objects should be interpreted not only through clinical or psychological dimensions but also as a reflection of the interpersonal and structural violence to which adolescents are subjected (LEITE FMC, et al., 2022; WEAVER ND, et al., 2025; WHO, 2025b). This perspective broadens the scope of intervention, underscoring the importance of interdisciplinary and intersectoral approaches that take into account the imprint of trauma and its ramifications for the selection of self-harm methods. Poisoning/intoxication was the most prevalent method of self-harm among the adolescents analyzed (56%), exhibiting a risk profile distinct from other means. The model showed that male sex and exposure to psychological violence were significantly associated with a higher likelihood of this behavior, in contrast to national and international evidence suggesting higher poisoning rates among girls (SUNG C, et al., 2025; WHO, 2025b; YAN Y, et al., 2023). This divergence points to cultural and contextual specificities in Brazil, related both to access to toxic substances in domestic settings and to patterns of expressing psychological distress among adolescents, which may vary by gender, territory, and social vulnerability (ALVES FJO, et al., 2024; BRASIL, 2024; LEITE FMC, et al., 2022). The positive Revista Ibero-Americana de Humanidades, Ciências e Educação. São Paulo, v. 12, n. 9, set. 2026. ISSN: 2675-3375 16 Revista Ibero-Americana de Humanidades, Ciências e Educação — REASE association with psychological violence reinforces the hypothesis that chronic interpersonal stressors-such as humiliation, threats, and persistent invalidation-can trigger less impulsive and more planned forms of self-harm, mediated by feelings of hopelessness, worthlessness, and loss of control (NOCK MK, 2010). On one hand, poisoning may represent a silent strategy of self-destruction, often occurring in contexts of emotional isolation and internalized suffering where adolescents lack safe spaces to express their pain (ORNELL F, et al., 2022; WHO, 2025b). On the other hand, the protective effects observed for physical violence, sexual violence, and neglect suggest that these forms of victimization tend to be associated with more direct and lethal methods of selfharm, such as hanging or the use of blunt objects-a pattern consistent with the model of severity escalation and cumulative vulnerability identified in previous studies (LEITE FMC, et al., 2022). Moreover, the domestic environment concentrated most poisoning incidents, indicating that episodes take place in intimate and invisible spaces where suffering is rarely detected until it reaches critical levels (WEAVER ND, et al., 2025). Taken together, these findings underscore that poisoning, as a form of self-harm, cannot be understood solely through a clinical lens; it must be interpreted as an expression of relational and structural suffering among adolescents exposed to psychological violence and social isolation. Consequently, it is essential to strengthen intersectoral public policies aimed at preventing psychological violence and promoting mental health, prioritizing the development of protective bonds in family and school settings and expanding access to safe, culturally sensitive emotion-regulation strategies (ALVES FJO, et al., 2024; ORNELL F, et al., 2022; WHO, 2025b). FINAL CONSIDERATIONS This study contributes to global efforts to understand and prevent adolescent self-harm by mapping factors associated with both recurrence and the methods used an approach still underexplored in Brazil and Latin America, where suicide rates continue to rise despite global declines. To our knowledge, this is the first investigation to examine predictive models addressing both outcomes in a large sample of Brazilian adolescents across pre-, during-, and post-pandemic periods. The findings underscore the heterogeneity of self-harm, marked by multiple coexisting methods and a high recurrence rate (47.7%), revealing distinct risk profiles. Vulnerability was greater among female and White adolescents, those with disabilities or mental disorders, and those exposed to physical or sexual violence or torture, who showed Revista Ibero-Americana de Humanidades, Ciências e Educação. São Paulo, v. 12, n. 9, set. 2026. ISSN: 2675-3375 17 Revista Ibero-Americana de Humanidades, Ciências e Educação — REASE higher probabilities of recurrence and use of methods such as poisoning and sharp objects. The pandemic period intensified these patterns, with an increase in lethal methods, particularly hanging, especially in peri-urban and rural areas. Despite its contributions, the study has limitations. The models showed low explanatory power, indicating the need to integrate environmental, psychosocial, and family variables capable of capturing subjective dimensions of self-harm. The use of secondary SINAN data may involve underreporting and inconsistencies, especially for less visible forms such as psychological violence. Although the sample covered all macro-regions, comparisons across states were not feasible. Notification quality depends on the training and perceptions of health professionals, potentially underrepresenting certain forms of violence and methods. Future research should integrate epidemiological and qualitative approaches to deepen the understanding of subjective meanings and social contexts of self-injury. Distinguishing notifications from public and private services may also help elucidate class-related disparities intersecting with gender, sexual orientation, and ethnoracial factors. REFERENCES 1. ALLOATTI MN, MATOS DE OLIVEIRA AL. Deepening and widening the gap: the impacts of the COVID-19 pandemic on gender and racial inequalities in Brazil. Gender, Work & Organization, 2023; 30(1): 329-344. 2. ALVES FJO, et al. The rising trends of self-harm in Brazil: an ecological analysis of notifications, hospitalisations, and mortality between 2011 and 2022. The Lancet Regional Health–Americas, 2024; 31: 100691. 3. BRASIL. Ministério da Saúde. Panorama dos suicídios e lesões autoprovocadas no Brasil de 2010 a 2021: Boletim Epidemiológico. Brasília: Ministério da Saúde, 2024; 55(4). 4. BRITO FAM, et al. Violência autoprovocada em adolescentes no Brasil, segundo os meios utilizados. Cogitare Enfermagem, 2021; 26: e76261. 5. CABRAL S, et al. Decrease in suicide rates in Brazil during the COVID-19 pandemic. Psychiatry Research, 2023; 329: 115443. 6. CIVIERO JCZ, TORRES JRP. Estudo dos casos notificados de lesões autoprovocadas no Paraná entre 2017 e 2021. Revista Multidisciplinar do Nordeste Mineiro, 2024; 3(3). 7. DANTAS ESO, et al. Suicide among women in Brazil: a necessary discussion from a gender perspective. Ciência & Saúde Coletiva, 2023; 28(5): 1469-1477. Revista Ibero-Americana de Humanidades, Ciências e Educação. São Paulo, v. 12, n. 9, set. 2026. ISSN: 2675-3375 18 Revista Ibero-Americana de Humanidades, Ciências e Educação — REASE 8. FATTAH N, LIMA MS. Perfil epidemiológico das notificações de violência autoprovocada de 2010 a 2019 em um estado do sul do Brasil. SMAD: Revista Eletrônica Saúde Mental Álcool e Drogas, 2020; 16(4): 65-74. 9. FINKELHOR D, et al. The victimization of children and youth: a comprehensive, national survey. Child Maltreatment, 2005; 10(1): 5-25. 10. GOMES FILHO CH, et al. Estudo sobre a correlação entre taxas de suicídio e a pandemia de COVID-19. Saúde, Ética & Justiça, 2022; 27(1): 9-17. 11. HOLLAND KM, et al. Antecedents of suicide among youth aged 11–15: a multistate mixed methods analysis. Journal of Youth and Adolescence, 2017; 46(7): 1598-1610. 12. JEGLIC EL, et al. The psychological impact of experiencing sexual abuse revictimization by a different perpetrator in childhood. Children, 2025; 12(8): 1070. 13. LEITE FMC, et al. Recurring violence against adolescents: an analysis of notifications. Revista Latino-Americana de Enfermagem, 2022; 30: e3682. 14. LIMA RFF, et al. Exploring subtypes of child maltreatment in a large Brazilian sample: cooccurrence patterns and associated characteristics. Child Abuse & Neglect, 2025; 165: 107495. 15. NOCK MK. Recent and needed advances in the understanding, prediction, and prevention of suicidal behavior. Depression and Anxiety, 2016; 33(6): 460-463. 16. NOCK MK. Self-injury. Annual Review of Clinical Psychology, 2010; 6: 339-363. 17. NOCK MK, FAVAZZA AR. Nonsuicidal self-injury: definition and classification. In: NOCK MK. Understanding nonsuicidal self-injury: origins, assessment, and treatment. Washington: American Psychological Association, 2009; 9-18. 18. ORNELL F, et al. Differential impact on suicide mortality during the COVID-19 pandemic in Brazil. Brazilian Journal of Psychiatry, 2022; 44(6): 628-634. 19. NASSIF PIRES L, et al. Multi-dimensional inequality and COVID-19 in Brazil. Investigación Económica, 2021; 80(315): 33-58. 20. SGOBBI FM, et al. Self-inflicted injury in children and adolescents during the COVID-19 pandemic: epidemiological analysis. Saúde, Ética & Justiça, 2022; 27(2): 60-66. 21. SOARES FC, et al. Tendência de suicídio no Brasil de 2011 a 2020: foco especial na pandemia de covid-19. Revista Panamericana de Salud Pública, 2022; 46: e212. 22. SOUZA LJ, FARIAS RCP. Violência doméstica no contexto de isolamento social pela pandemia de COVID-19. Serviço Social & Sociedade, 2022; 144: 213-232. 23. SUNG C, et al. Association between adolescent violence exposure and the risk of suicide: a 15-year study in Taiwan. Children, 2025; 12(1): 10. Revista Ibero-Americana de Humanidades, Ciências e Educação. São Paulo, v. 12, n. 9, set. 2026. ISSN: 2675-3375 19 Revista Ibero-Americana de Humanidades, Ciências e Educação — REASE 24. UDDIN R, et al. Suicidal ideation, suicide planning, and suicide attempts among adolescents in 59 low-income and middle-income countries: a population-based study. The Lancet Child & Adolescent Health, 2019; 3(4): 223-233. 25. VELOSO-BESIO C, et al. The prevalence of suicide attempt and suicidal ideation and its relationship with aggression and bullying in Chilean adolescents. Frontiers in Psychology, 2023; 14: 1133916. 26. WEAVER ND, et al. Global, regional, and national burden of suicide, 1990-2021: a systematic analysis for the Global Burden of Disease Study 2021. The Lancet Public Health, 2025; 10(3): e189-e202. 27. WORLD HEALTH ORGANIZATION. WHO COVID-19 dashboard: cases. Geneva: World Health Organization, 2025a. 28. WORLD HEALTH ORGANIZATION. Suicide worldwide in 2021: global health estimates. Geneva: World Health Organization, 2025b. 29. WORLD HEALTH ORGANIZATION. World report on violence and health. Geneva: World Health Organization, 2002. 30. YAN Y, et al. Suicide before and during the COVID-19 pandemic: a systematic review with meta-analysis. International Journal of Environmental Research and Public Health, 2023; 20(4): 3346. 20 Revista Ibero-Americana de Humanidades, Ciências e Educação. São Paulo, v. 12, n. 9, set. 2026. ISSN: 2675-3375