Introduction

Adolescence is considered a crucial stage of human development, during which individuals are subjected to significant physical, psychological, and social changes. This period presents numerous opportunities for personality development that will establish the groundwork for adult behavior. However, it is also a time when individuals may encounter considerable conflicts and risks to their mental health. (^{Caceda, 2014}; ^{Arias, 2021}). One of the risks is suicidal behavior, which is a consequence of the interaction of multiple causes and predisposing factors with a high potential for prevention. (^{Carballo, 2008}; ^{Clayton, 2018}).

According to ^{Andrade (2012)}, the ideation and suicide attempt in adolescence are, in most cases, seen as the expression of a desire to change an existing social and psychological chaos. It is regarded as a form of reaction to the feeling of impotence to change a situation that has become unbearable. While suicidal behavior is understood as a process that is gradually occurring and is manifested in various ways, including the desire or manifestation of dying, the image of death, and the suicidal attempt, up to completed suicide.

In Latin America, high incidences of suicidal behavior in adolescents and young populations have been reported in several recent studies. For example, two ecological studies were conducted by ^{Dávila and Luna in Mexico in two different years; the first in 2018} using data from the National Health and Nutrition Survey (Ensanut), where 21,509 adolescents were included, and the second in 2019 using data from the national drug use survey where 26,503 Mexican students were included. In the first study, the incidence of attempted suicide was found to be 2.74%, with 1.45% occurring in the twelve months prior to the survey and 1.29% taking place before the period of study (Davila y Luna, 2018). It was reported that men had 4.3% ideation, 3.6% planning, and 3.3% intent, while women had 7.9% ideation, 8% planning, and 10.3% attempted suicide. In Colombia, according to the latest figures from the National Institute of Health, attempted suicide is increasing, with a rate that went from 52.41 in 2017 to 62.17 per 100,000 inhabitants in 2019 (Dávila, 2019).

Given that suicide attempts and suicide mortality are considered a Public Health problem, it is contended by this research that suicidal behavior is etiologically determined by a cluster of biological, psychological, social, and environmental conditions that are manifested at the individual level as risk factors. Therefore, it is deemed pertinent to be aware of the associated factors of suicide ideation and attempts that are considered as a foundation for the implementation of interventions from the perspective of modern pidemiology founded on scientific evidence. (^{Alvarez, 2008}). The reviews conducted on the subject were found to be non-systematic (^{Serrano, 2017}; Azua, 2020; Londoño, 2020 and ^{Arias, 2021}), lacking in thorough research in databases (2-5), not in accordance with the PRISMA statement guidelines, and with no assessment of the studies' methodological quality. The aforementioned characteristics are met by only one review (Hernandez, 2020); however, it is focused solely on the adolescent population and lacks specificity to Latin America. Therefore, the objective of this systematic review was to identify the risk and protective factors associated with ideation and suicide attempts in adolescents and young people in Latin American studies, and as a secondary objective, to evaluate the methodological quality of the included studies.

Method

A systematic review of the literature, in accordance with the PRISMA guidelines declaration (^{Urrutia, 2010}) was conducted in November 2021 and updated in December 2023 on risk and protective factors for suicidal ideation and attempt in adolescents and young people. The databases CUIDEN, LILACS, Pubmed, Scielo, Science Direct, SCOPUS, EBSCO, and Medline were consulted, with the results being restricted to articles in the last 12 years in order to obtain the most recent data. The following keywords consulted in the Health Sciences Descriptors library were used to design search equations combined with the Boolean operator AND and OR as follows: suicide AND adolescents OR young people and suicide AND risk factors for the Spanish language and its counterparts in the English and Portuguese languages. The search was conducted by two authors (del Pino, 2014).

Data analysis

Data synthesis was conducted by extracting statistical estimates according to the type of variable previously categorized, through a summary of the findings of the articles included in the review. The minimum and maximum percentages indicating the prevalence of suicide ideation and attempts were identified, along with the statistical estimators of association for risk and protective factors. Furthermore, the vote counting technique and the Sign Test have been utilized. The vote count enables a comparison between the number of studies reporting a positive association and those reporting a negative association. In this instance, studies demonstrating a statistically significant association between a given factor and suicide ideation and/or attempt were assigned a positive vote, while a negative vote was given to studies lacking such an association. The sign test was utilized to determine if the difference in the number of positive studies was significantly greater than the opposite result. A significance value of less than 0.05 was established, although these analysis techniques are limited, they help to guide the results of the review (del Pino, 2014). This information was organized and compiled in a table.

Results

Characteristics of the sample

A total of4,055 documents were obtained, of which 20 studies conducted in Latin America were included. 50% (n=10) were conducted in Colombia, 15% (n=3) in Mexico, 10% (n=2) in Cuba, 15% (n=3) in Chile, Jamaica and Peru 5% (n=1) each (figure 1).

Figure 1 Study general outline (PRISMA) 

A cross-sectional analytical design was utilized in 95% (n=19) of the studies, while 5% (n=1) were characterized by a retrospective longitudinal design. Of the total, 35% (n=7) employed probabilistic sampling methods, with 4 utilizing a stratified type and 3 using a simple random type. The remaining 65% (n=13) of studies employed non-probabilistic sampling methods, such as convenience, incidental, or snowball sampling. With respect to the sample of participants, 70% (n=14) of the studies were conducted in adolescents attending secondary school, of which 28.57% (n=4) examined the age range between 15-20 years, while 71.43% (n=10) included adolescents aged 10-15 years. Additionally, 20% (n=4) of the studies incorporated adolescents from mental health institutions, with only 1 studying adolescents between 10-21 years old. 5% of the sample included university students from the Medicine program aged 16-36 years and adolescents from the LGTBIQ+ community, respectively. The smallest sample, consisting of 46 adolescents, and the largest sample, consisting of2,997 individuals, were included in the study.

Risk of study bias.

The evaluation of the biases of the included studies was conducted using the Ciapponi critical reading guide, which assesses the internal validity and overall quality of the investigations with observational designs. A medium internal validity level was obtained by 55% (n=11) of the studies, while a high level was obtained by 30% (n=6) and a low level by 15% (n=3). In terms of global quality, a medium level was obtained by 60% (n=12), a high level by 30% (n=6), and a low level by 10% (n=2) (appendage 1).

Prevalence of suicidal ideation and attempt.

Suicide ideation and attempt were measured with different self-response instruments in 80% (n=16) of the cases. Among these, in 20% (n=4) of the studies, the authors' own questionnaire was designed, while 15% (n=3) utilized the Plutchik scale, and 10% (n=2) employed the Positive and Negative Suicidal Ideation (PANSI) and Okasha, respectively. Various instruments such as student health survey, youth health survey, suicidal ideation scale, Beck scale, CES-D scale, and Columbia University rating scale were used in the rest of the investigations (n=6). On the other hand, suicidal ideation and/or attempt were not measured in 3 studies because the participants were adolescents with suicide attempts who were in mental health institutions.

The range of prevalence of suicidal ideation in school adolescents was found to be between 22.8% and 31.4% according to the Plutchik scale, while it was reported to be between 8.1% and 9.7% with the questionnaires designed by the authors. Suicidal ideation was measured at 19.5% using the suicidal ideation scale, 14.9% with the Beck scale, 43% with PANSI, 14.7% with the CES-D Scale, 23.6% with the student health survey, and 14.23% with the survey of youth health.

In relation to attempted suicide, the prevalence was found to be 12.2% according to the Plutchik scale, 17.5% according to the student health survey, 3.55% according to the youth health survey, between 9% and 16.4% according to Okasha, and between 4.9% and 22.4% with the questionnaires designed by the authors. Additionally, it was determined to be 2.7% according to Beck and 13.9% according to the CES-D Scale.

Instruments usedfor associated factor measurement

The review included studies that utilized a variety of instruments to measure the wide range of factors associated with suicide ideation and attempts, with a total of 34 instruments being identified. Of these, 8.8% (n=3) were constructed by the researchers themselves to measure variables such as cigarette consumption, psychoactive substance consumption, bullying, sexual abuse, and family structure (Mendez 2022, Alvarez 2015, ^{Pineda 2019}). Meanwhile, 88.2% (n=30) were comprised of validated instruments, and only two investigations utilized the structured interview to investigate variables such as history of suicidal behavior, family mental disorders, among others (^{Perez 2020}, ^{WD 2012}).

The family APGAR, used to measure family dysfunction, was found to be the most utilized among the validated instruments (^{Alvarez 2013}, ^{Cañon 2018}, Aguirre 2013, ^{Perez 2012}, ^{Valdivia 2015}) and the Rosemberg Self-Esteem Scale (^{Silva 2017}, Cañon 2018, Aguirre 2013, Perez 2012, ^{Sarmiento 2011}), Both were used in 5 investigations. This was followed by the CAGE Scale for alcohol dependence in 4 studies (Alvarez 2012, Cañon 2018, Pinzon 2013, Aguirre 2013). Thirdly, the Beck Hopelessness Scale (Silva 2017, Garza 2019, Valdivia 2015) and the Cisneros Scale for bullying (Cañon 2023, Cañon 2018, Aguirre 2013) were located in three investigations each.

Several instruments were used in relation to the depression variable, including the CES-D depression scale (Pinzon 2013, ^{Secundino 2020}, ^{Sarmiento 2011}), the Beck Depression Inventory (^{Silva 2017}, Valdivia 2015), the Birleson Scale (Alvarez 2014, Aguirre 2013), and the Columbia Depression Scale (^{Perez 2012}). Furthermore, instruments that measure, in addition to depression, anxiety, and stress variables such as the Goldberg Depression and Anxiety Scale (Cañon 2022), the Hospital Depression and Anxiety Scale (Cañon 2018), and the DASS-21 (Mendez 2022), are utilized.

Discussion

In Latin America, research on factors associated with suicidal ideation and attempt is found to be more focused on the study and correlation of risk factors, as evidenced by the results of this review. Conversely, the investigation of protective factors is typically more restricted and less prevalent. Results that are in agreement with the review conducted by Hernandez (2020) were reported, indicating that only 5 out of 23 analyzed investigations had examined protective factors. The study of risk factors is focused on by several objectives such as predicting a phenomenon, determining the causes, making a diagnosis, or eliminating the risks in order to prevent the occurrence of a problem. However, the possibility of promoting positive strategies that reduce the probability of falling into certain risks is left aside by this position, so it is important for the phenomena to be studied also from this positive sense (^{Amar, 2003}; ^{Giner, 2010}; ^{Perez, 2020}; Servicio de Andaluz, 2010).

The most common risk factors reported in the studies are of a familial nature, such as family dysfunction, poor communication with parents, domestic violence, and negative parenting role or style. Unlike the biological and spiritual factors, which were studied to a lesser extent. Similar findings were reported in the review by ^{Arias, Morantes, Montoya (2021)}, where family factors such as dysfunction and violence were found to be the most frequently reported in the studies analyzed. The important role that the family plays in the development of a healthy environment that prevents the development of suicidal thoughts and attempts is shown by these findings (^{Brent, 2005}).

The poorly structured nature of the family and the development of a conflictive and violent atmosphere expose its members to be considered wrong solutions that lead to self-destructive behavior. The means of communication of feelings, demands, or pleas, which are not known or cannot be expressed by the adolescent in any other way, are become by the lack of adequate coping strategies, as well as the necessary family cohesion that allows fears to be expressed, tensions and anxieties to be resolved, problems to be solved, and clear guidelines to be established based on love, trust, respect, and understanding. According to Garza (2019), confusing norms and rules were received by more than half of the adolescents and young people interviewed in their study, along with few limits and excessive punishments; Furthermore, it was indicated by 20% that they do not have anyone in their family to rely on in case of problems. t is believed that the adaptive behaviors of childhood, which will be solidified in adolescence and youth, are impacted by parenting. A significant role is played by these behaviors in the quality of the parent-child relationship, as well as in the perceived family and social support by the young person when they are confronted with life stresses and mental health issues.

In second place, psychological factors such as hopelessness, mental disorders, and the use of psychoactive substances are considered. Thirdly, social factors such as sexual and domestic violence, bullying, love breakups, and lack of social support are also taken into account. Various reviews (^{Serrano, 2017}; Azua, 2020; Londoño, 2020; and ^{Arias, Morantes, Montoya, 2021}) have indicated that mental disorders, particularly depression and substance use, are directly associated with a higher risk of suicidal ideation and attempt. Similarly, the risk is greatly increased by being in a violent and hostile environment and having little or no social support.

Bullying in all its forms, including cyberbullying, has been observed to have gained strength in recent years as a social factor, resulting in significant emotional distress among adolescents and prompting suicide attempts. However, the risk factor in question may be surrounded by contradictory scientific evidence. It was observed by ^{Lardier (2016)} that in adolescents, the probability of suicidal ideation was increased by bullying; however, after conducting multivariate logistic regression analysis, the effect was found to no longer be significant. This was achieved by including other risk variables such as depression, family conflict, and substance use. It has been reported by other investigations that a statistically significant relationship exists(Aguirre, 2013; Pérez, 2012; ^{Cañon, 2018}; Cañon, 2021).

In relation to the protective factors, the most reported ones were relatives such as family functionality and family support. Results align with the review conducted by ^{Arias, Morantes, and Montoya (2021)}, who also identified these factors as the most extensively studied in a positive light and linked to the lack of suicidal ideation and attempts. The family is seen as a system of relationships, where each member contributes from their own individual characteristics to the creation of a secure environment. In cases where the adolescent is situated within a family unit that effectively regulates these connections and establishes explicit boundaries and guidelines, functional interaction at an emotional and social level can be achieved, aiding in the management of emotions during crises and facilitating positive engagement with life's transitions. This, in turn, diminishes the likelihood of the manifestation of suicidal behavior (^{Cid, 2014}; Sánchez, 2015).

The prevalence of suicidal ideation was found to be variable as a result of the heterogeneity in which it was measured in studies, with figures ranging from 8.1% as the lowest to 43% as the highest in adolescents and young people in school. Similarly, attempted suicide also exhibited a variable prevalence, with the lowest figure being 2.7% and the highest being 17.5%Despite these differences, it is shown by these figures that both suicidal ideation and suicide attempts are a public health issue that occurs frequently in the young population, and the urgent implementation of strategies and programs for prevention is required. Through an articulated, comprehensive, and community approach, steps prior to suicide are taken, greatly increasing the probability that a subject will die from this cause (National Institute Of Mental Health, 2020; Ministerio de salud, 2018). According to ^{Echeburúa (2015)}, if a subject has had a suicide attempt, the risk of committing suicide increases in the weeks after the act, making it important to provide the necessary therapeutic help and activate the family and social support network.

Furthermore, a high social value is placed on the young population by societies due to their highly productive stage. Consequently, the occurrence of suicidal behavior in young individuals results in significant direct and indirect costs, including the loss of potential years of life in those who commit suicide and the high health costs associated with hospitalization and treatment for those who attempt suicide (United Nations Children's Fund, 2021); According to PAHO in the Americas, suicide is considered the fifth cause of this indicator, as well as years of life adjusted for disability (^{Pan American Health Organization, 2019}).

In this sense, a synthesis of the risk factors associated with suicide ideation and attempts is provided by this review, which must be taken into consideration when developing programs for the prevention of suicidal behavior and the promotion of mental health from an interdisciplinary approach in health, as the reduction of these risk factors must be achieved through interventions in order to protect adolescents and young people. However, it is deemed necessary for future research to be bolstered in the examination of protective factors, as it is recommended that positive factors be identified that can serve as a foundation for the promotion of mental health in a constructive manner that not only enables young individuals to utilize and amplify their own resources, but also fortify their familial, educational, and societal surroundings.

It should be noted that most of the studies selected their participants using non-probabilistic sampling, often for convenience. This type of sampling is typically utilized by researchers when there is a restricted timeframe for conducting the investigation or when there are financial constraints hindering the execution. However, this type of sampling is accompanied by certain limitations that result in a low external validity of the findings and introduce weaknesses to the associations, as total representation of the population is not guaranteed and objectivity may be lacking (^{Hernandez, 2014}). In turn, the internal validity of 68.41% (n=13) of the studies included in the review, which were classified as having medium and low validity, was affected by this type of sampling. Therefore, it is deemed important that for future studies from the risk approach, primary research be conducted to reduce the barriers for the execution of probabilistic sampling, ensuring greater objectivity and representativeness of the population (^{Araujo, 2011}; ^{Otzen & Manterola, 2017}).

Similarly, the results of this review indicate that there are few observational studies in the scientific evidence that can be classified as having high internal and global validity according to the criteria of the Ciapponi guide through critical reading, due to the characteristics of the methodological designs. Observational research of higher quality was conducted by researchers, who were evaluated, in order to achieve greater control over biases that may be encountered, and which are unrelated to the limitations of various existing observational designs. (Ciapponi, 2014). Despite the contributions of this review, some limitations can be considered, such as the heterogeneity of the results that prevent quantitative analysis of the data (meta-analysis). Although an extensive search was conducted, it is possible that scientific evidence of interest that could contribute to the results of this review has eluded detection.

In conclusion, the high prevalence of suicidal ideation and attempts in the young population of Latin America is frequently associated with risk factors such as cigarette consumption, psychoactive substance use, severe family dysfunction, alcohol dependence, low self-esteem, sexual abuse, bullying, depression, psychological abuse, female gender, verbal insults, academic performance, parental divorce, romantic disappointments, recent or past suicidal ideation, , anxiety, family abuse, family history of suicide attempt and/or suicide, parental absence and hopelessness.

The evidence concerning protective factors is limited, and the ones that have been most extensively examined are family functionality and family support. The majority of the scientific evidence published on the topic can be categorized as having a medium level of methodological quality.