Introduction. The serious problem of child suicide in Spain

Child suicide is a critical public health issue in Spain. Between 2020 and 2021, the rate of child suicide increased by approximately 57%, with 22 children aged 15 or younger taking their own lives in 2021, compared to 14 in 2020 (Instituto Nacional de Estadística n.d.). These may be extraordinary cases resulting from the effects of the Covid-19 pandemic in the country —which recorded one of the highest suicide rates globally (Tandon, 2021; Lantos and Nyári, 2024)—, and its impact on children’s mental health, as explained in UNICEF (2024). In 2022, however, the number of child suicides decreased to 12 cases (Observatorio del Suicidio en España, 2022), aligning with the average rate observed from 2012 to 2022 (Ministerio de Sanidad, 2024).

Child suicide is a complex phenomenon, given the presence of psychological, social, and economic factors that can influence an individual to end their life. Suicidal ideation is a major risk indicator within this etiology (Sedgwick et al., 2019), referring to those negative thoughts related to the desire to take one’s own life (Nobile et al., 2021). Among the triggers of child suicide, discrimination and exclusion stand out as primary predictive factors (Ballesteros, 2022), as well as mental disorders (depression, anxiety), lack of social support (Castro et al., 2014; George, 2019; Navarro Gómez, 2020; Garcés et al., 2020; Moreno et al., 2022; Garcés et al., 2023a; 2023b), intrafamily violence (Vianchá et al., 2013), bullying (Azúa Fuentes et al., 2020; Kwan et al., 2022), and online risks such as cyberbullying, sexting, or sextortion (Navarro-Gómez, 2020), among others.

Although these are the most significant triggers, recent years have seen the emergence of a line of research linking media content consumption and suicidal ideation (Herrera et al., 2015; Acosta et al., 2017; Lois et al., 2018; Durán and Fernández, 2020; Niederkrotenthaler et al., 2020; Garcés et al., 2023a). Regarding online content, Spain’s Ministry of Health, Social Policy, and Equality (2020) indicates that most available content on the internet “could be classified as pro-suicide content” (p. 75), noting that youth groups are particularly vulnerable to it. The impact on mental health among youth is partly explained by their tendency to favor social platforms (Newman et al., 2023) and, more broadly, internet platforms (Ministry of Health, Social Policy, and Equality, 2020) as sources of information. These platforms often provide guidance and content not aligned with World Health Organization (2017), which may contribute to—but do not solely determine—the development or worsening of mental health concerns.

There are also other factors, such as the impact of audience attraction and retention strategies on social media or certain websites (Rodrigues et al., 2022; Caro-Castaño, 2022) which may exacerbate psychological disorders and lead to an increase in suicidal ideation. However, previous literature also suggests that social media can serve as beneficial spaces by fostering the development of support communities (Rodrigues and Araújo, 2022). From this perspective, some network-led initiatives—such as keyword filtering or flagging “inappropriate content,” as YouTube does (p. 205)—, may help mitigate the impact of certain content.

In media consumption, it is also worth highlighting the responsibility of the media themselves, especially in covering topics like celebrity suicides. A poorly constructed media narrative can encourage suicidal ideation, particularly among children, who are more vulnerable (Martínez and Robles, 2016). As demonstrated by Garcés et al. (2023a), certain media products can either increase or decrease suicidal ideation. This is explained by the fact that poor media coverage can create a contagion effect, whereas inspiring coverage can provide guidelines to counteract destructive behavior. Academic literature points to a strong association between media consumption and suicidal ideation, which in some cases may contribute to suicidal actions (Sedgwick et al., 2019; Durán and Fernández, 2020; Niederkrotenthaler et al., 2020; Garcés-Prettel et al., 2023a). This relationship is not necessarily causal, as media use is intertwined with multiple individual and contextual factors. Media exposure can foster both an imitation-contagion effect (the Werther or Copycat effect) (Carmichael and Whitley, 2019) and a protective Papageno effect that may reduce suicidal ideation (Sinyor et al., 2018), both of which are amplified by the widespread use of social media.

The technical complexity of addressing this topic has led some conventional media outlets to adopt self-censorship when reporting on suicide (Barredo, 2011; Lois et al., 2018). However, numerous countries have developed guidelines to support safe and responsible reporting, emphasizing constructive coverage rather than avoidance. The World Health Organization (2017) provides global recommendations, and countries such as the United States (Action Alliance, 2019) and the United Kingdom (National Institute for Health and Care Excellence, 2019) have national guidelines. Australia stands out for the strength of its implementation programs and demonstrated success in encouraging safe media coverage, including specific guidelines for young people communicating about suicide online or via social media (Robinson et al., 2018; La Sala et al., 2021). In this article, our primary objective is to compare differences in suicidal ideation based on whether Spanish children are exposed to media or social media content. Specifically, we aim to a) identify which media spaces or formats can be considered protective or risky for suicide among the selected population; and b) identify the factors associated with suicidal ideation in relation to media and social media consumption in this demographic. Unlike previous research on adolescents and young adults, this work provides empirical data, and a detailed analysis of how different media consumption formats affect children, a highly vulnerable group.

It is important to distinguish between suicidal ideation, suicidal behaviour, and suicide mortality. The present study focuses exclusively on suicidal ideation, which refers to thoughts about engaging in behaviour intended to end one’s life. While ideation can be an indicator of psychological distress and a risk factor for later behaviours, it does not necessarily lead to attempts or suicide deaths. In children, suicide deaths are statistically rare in absolute terms, yet they can still represent one of the leading causes of mortality within this age group due to the overall low incidence of death in childhood.

Literature review: media consumption and suicidal ideation in children

Research on the impact of social media as a potential risk factor for the mental health of children and adolescents has grown over the past decade in response to the high penetration and intensity of use of these platforms within this age group (Prinstein, Nesi, and Telzer, 2020; Newman et al., 2023). In this sense, the systematic review by Sedgwick et al. (2019) on suicidal behavior among adolescents and their use of the internet and social media found contradictory results. The authors indicated that the existence of a causal relationship was unclear, and that studies demonstrating such a link could not establish the direction of causation between the two variables. In a similar way, Giancola, Bentley, and Torous (2024) concluded in their systematic review that, while a relationship between suicidal behavior and social media use among young people can indeed be affirmed, the nature of this relationship remains unclear, with significant variability observed among individuals. Moreover, George (2019) suggested that adolescents who are already vulnerable, such as those suffering from depression, are more likely to have stronger reactions to exposure to messages about self-harm or suicide. Moreno et al. (2022) also noted that digital technologies tend to amplify the mental state that adolescents bring into the online environment, whether positive or negative.

Other research has focused on analyzing the attitudes and specific uses that young people make of social media and the impact on their mental health. Thus, the study by Hamilton et al. (2021) found that infrequent social media use could be symptomatic of experiencing social anhedonia, a condition that may increase severe suicidal ideation. In this regard, the systematic review by Keles, McCrae, and Grealish (2020) suggested the importance of analyzing specific behaviors such as social comparison, active/passive social media use, or usage motivations as factors that could have a greater influence than the “time of use” variable on the development of mental health issues like depression, anxiety, or stress in adolescents. In this vein, research on suicidal ideation and social media focuses on exploring the types of emotional experiences adolescents report on these platforms. Studies have identified not only risk factors but also protective factors against suicidal ideation (Boyd et al., 2024; Hamilton et al., 2024; Massing-Schaffer et al., 2020) and against depression (Moreno et al., 2022; Nessi et al., 2022).

Among the potentially harmful effects of social media use, there have been indications of the normalization and contagion of self-harm practices and suicidal behavior (Marchant et al., 2017), their glamorization (Graham, 2024), and the facilitation of learning (Kline et al., 2023), as well as negative social comparison (Hamilton et al., 2024; Weinstein et al., 2021) fueled by social media metrics (Kline et al., 2023). Potentially protective effects include social connection (Boyd et al., 2024; Hamilton et al., 2024; Kline et al., 2023; Marchant et al., 2017), particularly through the ability to escape the social isolation experienced in face-to-face settings by creating and maintaining online friendships (Massing-Schaffer et al., 2020); the evasion of negative emotions (Kline et al., 2023); and the ability to seek support and a sense of belonging from individuals who have faced similar mental health challenges (Zhang et al., 2024).

Social media as sensory environments of everyday life

Social media have become central spaces for young people’s self-presentation and socialization today (Charteris et al., 2018). In fact, adolescents rarely choose to abandon these platforms for fear of severely affecting their connection with their peer groups (Jorge et al., 2023). Social media thus establish a framework of conditions and rules for content consumption, social interaction, and identity exploration. These platforms constitute online infrastructures that mediate everyday life experiences (Van Dijck, 2016), prioritizing certain forms of knowledge and reality experimentation over others. Following McLuhan’s classic proposal (1964/1980), every technological innovation introduces changes in the senses, favoring certain modes of perception over others. When the way of experiencing the world changes, thinking about that reality, one’s identity, and relationships with others also changes. Thus, while printed text brought with it a linear, sequential, and rational mode of thinking linked to individuality, audiovisual media present an immersive quality that promotes a more superficial and fragmentary type of knowledge, within a logic that prioritizes the emotional over the rational.

On the other hand, the predominance of audiovisual content in social media has transformed images into an essential element for individual self-presentation and establishing connections in these spaces (Lobinger et al., 2021; Olivio et al., 2022), to such an extent that the lack of self-images on these media tends to be interpreted among young people as a problematic and suspicious indication (Aziz, 2014; Caro-Castaño, 2015). Likewise, the full integration of these platforms into everyday life experiences has normalized users sharing visual and audiovisual content from their private lives as a useful strategy to enhance the sense of copresence (Zhao and Elesh, 2008), a form of phatic communication that allows maintaining and strengthening social bonds (Lobinger et al., 2021) and convey authenticity and proximity in mediated interaction (Caro-Castaño, 2023).

In this sense, various studies have confirmed that social media can affect self-esteem through social comparison mediation (Jiang and Ngien, 2020) —especially as a result of others’ tendency to select positive social information (Yang and Robinson, 2018)—, potentially producing intrusion phenomena (Rivas-Herrero and Igartua, 2021), and suicidal ideation in cases of high levels of negative social comparison (Hamilton et al., 2024; Spitzer et al., 2023).

Moreover, since these are platforms that integrate techno-commercial infrastructure, users, and socio-cultural practices (Poell, Nieborg, and Van Dijck, 2019), it is the commercial vector that governs these companies’ interests in designing the affordances of these spaces. Thus, these platforms follow a business model based on content personalization: they collect user data to build micro-segmented profiles, determine which commercial, informational, or entertainment content to display next, and maximize attention. This often involves offering increasingly extreme material and normalizing content harmful to mental health (Graham, 2024). In this sense, in recent leaks of TikTok’s internal documentation published by NPR (Allyn, Goodman, and Kerr, October 11, 2024), the company acknowledged that the navigation experience personalization offered by their algorithms produces filter bubbles (Pariser, 2011) that “can serve potentially harmful content expeditiously” (Allyn et al., October 11, 2024).

Methodology

The methodological design of this study was developed in accordance with regulations concerning children’s rights. The Charter of Fundamental Rights of the European Union (2010/C83/02) and the Treaty on European Union establish that “the best interests of the child” must prevail in all public or private actions. This regulatory framework guided the study’s approach and the ethical procedures implemented.

The research was structured as a non-experimental study, with a transversal extraction of data, utilizing a survey questionnaire as the data collection technique. The methodology and design of the study was evaluated and approved by the Ethics Committee of the University of Málaga. The endorsement is recorded under number 79, with CEUMA registration number 129-2023-H. To select the sample, it was applied a sampling method based in a multistage cluster with stratification of the first-stage units of children aged 10 to 15 years. A total of 309 schools and Parent-Teacher Associations (AMPA) were contacted across ten autonomous communities between February and June, 2024. Of these, 17 educative centers agreed to participate, which facilitated the coordination of ethical aspects and the obtaining of informed consent from parents or guardians. The locations and times for administering the questionnaire were scheduled, ensuring the assent of the participants. Data were collected anonymously and confidentially by Sigma Dos, a well-regarded organization in Spain known for its public perception and opinion research. The consent document included information on data usage and allowed participants to skip any uncomfortable questions. In addition, all participants and their families received information about available mental health resources and helplines, including contact information for specialized support services, to ensure that any child or guardian could seek assistance if needed. Due to the anonymous nature of the study, it was not possible to directly identify or refer individual cases with higher levels of suicidal ideation. This approach also sought to minimize potential stigma or unintended social consequences for the children involved. While suicidal ideation is a recognized marker of psychological distress and a potential risk factor for later suicidal behaviour, it does not necessarily lead to attempts or suicide deaths, which remain comparatively rare in this age group.

The paper questionnaire began with the question: “Do you voluntarily choose to participate in this study?” Participants who agreed completed the questionnaire within a maximum time frame of 30 min.

A total of 804 valid questionnaires were completed, distributed across 67 classrooms. Participation varied by grade and age: 9% were 10 years old (n = 72), 24.1% were 11 years old (n = 194), 25.4% were 12 years old (n = 204), 26.4% were 13 years old (n = 212), 12.9% were 14 years old (n = 104), and 1.6% were 15 years old (n = 13). Regarding gender distribution, 50.6% identified as female (n = 407), 47.4% as male (n = 381), 1.2% identified as “other”, and 0.7% did not respond. The lower proportion of participants aged 14 and 15 is attributable to the sampling process and the availability of these grades in the schools that agreed to participate. Several schools did not have complete upper-grade cohorts or were unable to schedule sessions for these students due to academic calendar constraints, which limited their representation in the final sample.

The questionnaire included the Paykel Suicidal Scale (Paykel et al., 1974), proposed and validated in minors by Fonseca and Pérez de Albéniz (2020). This scale consists of five questions that evaluate thoughts of death (items 1 and 2), ideation (items 3 and 4), and suicide attempts (item 5). It demonstrated high reliability (Cronbach’s alpha of 0.811 and 0.828 in its standardized version), explaining 51.2% of the total variance in suicidal behavior. Additionally, the Media Use and Attitudes Scale validated by Rosen et al. (2013) was simplified, according to the population, from 14 to 11 items. The 11 items used presented a Cronbach’s alpha of 0.836 (see Table 1), explaining 64.1% of media content consumption in the surveyed child population. The correlations between the items of both scales are satisfactory, with values ranging from 0.352 to 0.680, indicating good coherence.

Table 1 Media use and attitudes consumption recognized by the surveyed children in Spain (2024).
Full size table

Data analysis was conducted using a three-step strategy. First, the frequency of suicidal ideation was determined using the Paykel Scale, classifying risk levels as low (percentiles 5 to 25), medium (percentiles 50 to 75), and high (percentiles 90 to 95). Second, correlations between the items of both scales were calculated using Kendall’s Tau-b correlation to identify significant relationships. Finally, a binary logistic regression model was estimated to assess the independent associations of the interrelated variables with medium to high levels of suicidal ideation.

Results. From media content consumption to digital social media content consumption

First, we assessed the media content consumption recognized by the surveyed children, as these symbolic products have been linked to suicidal ideation (Sinyor et al., 2018; Carmichael and Whitley, 2019; Garcés et al., 2023a). Among violent and/or sensationalist content, the most frequently indicated consumptions were reality shows or competitions (10.2%) and action series or movies with violent content (9.2%). Regarding cultural and/or educational content, the most frequently consumed materials were programs that promote social support (13.3%) and general news to stay well-informed (12.7%) (Table 1).

The least consumed media products among the surveyed children were related to the information or videos searched on the internet (88,7%) or on social media (85,1%) about cases of suicide. The other items with highest scores marked as “never” include exposure to series or movies depicting acts of suicide (63.3%), watching programs that promote spiritual life (45.1%), and viewing news about violent deaths (41%).

However, we also explored the content consumption and attitudes on social media and the internet recognized by the surveyed children in Spain. In this regard, we observed that the most frequent activities included reading posts (34.6%), liking posts (25.9%), and browsing friends’ or family members’ profiles from their mobile devices (21.3%) (Table 2).

Table 2 Content consumption and attitudes on digital social media among the surveyed children in Spain (2024).
Full size table

In contrast, the primary online activities marked as “never” by the surveyed children include posting pictures of their private lives on social media (68.9%), interacting with people online whom they have never met in person (63.1%), connecting with such individuals (57.6%), and posting pictures of their public lives on social media (56.7%).

The relationship between suicidal ideation and the use and reception of media and digital social media content

To determine the risk levels of suicidal ideation among the surveyed children, the scores from the five items of the Paykel Suicide Scale were summed, and a percentile ranking process was applied using the Waverage method, which considers the distribution of the data. Three risk levels were established based on the frequency of suicidal ideation, a proximal indicator of suicide risk (Nobile et al., 2021): low risk (percentiles 5 to 25), medium risk (percentiles 50 to 75), and high risk (percentiles 90 to 95). According to our findings, 62.7% of the surveyed Spanish children exhibited low risk of suicide, 26.1% were at medium risk, and 11.2% were at high risk (Table 3).

Table 3 Waverage percentile ranking of suicidal ideation among the surveyed children in Spain (2024).
Full size table

Once the previous ranking was completed, we tested the relationship between the use and reception of media and digital social media content. At this point, we found two important results: on one hand, we observed a positive correlation between suicidal ideation and searching for information about suicide on the internet (r = 0.336; p < 0.01), particularly when the content is not aligned with WHO technical guidance (2017). On the other hand, multiple forms of engagement on social media were statistically associated with higher levels of suicidal ideation among children aged 10–15 in Spain. This pattern was observed for both active uses, such as commenting on posts or status updates (r = 0.248; p < 0.01), and more passive uses, such as clicking “Like” on a post or photograph (r = 0.213; p < 0.01) (Table 4).

Table 4 Relationship between suicidal ideation and the consumption of media content among the surveyed children (2024).
Full size table

In the case of media content disseminated by mass media (Table 4), the evidence is not as conclusive as it is for digital social media. Only three media products showed positive correlations with higher levels of suicidal ideation within the selected population profile: exposure to action series or movies with violent content (r = 0.175; p < 0.01); viewing contests or reality shows (r = 0.077; p < 0.05); and exposure to news about violent deaths (r = 0.163; p < 0.01).

Binary logistic regression analysis of the association between media and social media exposure and suicidal ideation

In the regression models estimated with and without the inclusion of age as a moderating variable, the analysis identified consistent associations with medium–high levels of suicidal ideation among Spanish children aged 10–15 years. In line with previous literature noting the absence of definitive evidence for a causal link between social media use and suicidal behaviour (Sedgwick et al., 2019; Giancola et al., 2024), the present findings indicate that frequent online searches for information about suicide (β = 0.525 in Model 1; β = 0.522 in Model 2, both p < 0.01), regular posting of private-life photographs on social media (β = 0.435 in Model 1; β = 0.436 in Model 2, both p < 0.01), and frequent viewing of contests or reality shows (β = 0.243 in Model 1; β = 0.236 in Model 2, both p < 0.05) are the variables most strongly associated with elevated suicidal ideation in this population (Table 5).

Table 5 Binary logistic regression of suicidal ideation and the reception of media and social media content among the surveyed children in Spain (2024).
Full size table

The findings reaffirm that these behaviors are associated with a higher risk of suicide among Spanish children. The model that includes age (Model 2) shows a slight increase in predictive power compared to Model 1, with Cox and Snell R² ranging from 10.9% to 15.6% versus 10.6% to 15.1%, respectively. Both models share the same significant predictor variables, indicating their robustness concerning suicide risk.

Discussion and conclusions

This study advances understanding of the relationship between media exposure and children’s mental health, focusing on the associations between suicidal ideation and engagement with various types of media and social media content among Spanish children aged 10–15, and provides evidence relevant for academics and policymakers seeking effective interventions in children’s mental health in the digital environment.

According to our data, three variables showed statistically significant and consistent associations with higher levels of suicidal ideation among the surveyed minors in Spain. In the first case (frequently seeking information about suicide), the primary explanation may be linked to existing individual conditions. Children who are already experiencing mental disorders may engage in this active and intentional search as part of an acute risk profile, in which pre-existing vulnerabilities intersect with algorithmic recommendations and platform design, a phenomenon also noted in previous studies (George, 2019; Moreno et al., 2022).

In the second case, regularly posting private photos on social media, this association is consistent with dynamics of visibility-seeking, social comparison, and exposure to potentially harmful online interactions, with the individual pressure to gain attention from both the community and the platform’s algorithms (Rodrigues et al., 2022; Rodrigues and Araújo, 2022; Caro-Castaño, 2022). This pressure contributes to a need for social prestige in an opaque, algorithm-governed visual regime (Miyazaki, 2019), where users attempt to decode platform logic through the visibility and interactions their content generates (Cotter, 2019). This process constructs an “algorithmic imaginary” (Bucher, 2017) that defines which communicative practices are privileged and rewarded through social metrics (Georgakopoulou et al., 2020). In this complex scenario, posting private photos could be also linked to toxic interactions or cyberbullying (Coyne et al., 2021), which involve violent interactions in the social environment that children inhabit. Additionally, sharing private photos on social media may exacerbate negative social comparisons with peers, as highlighted in previous literature (Hamilton et al., 2024; Kline, 2023; Jiang and Ngien, 2020; Rivas-Herrero and Igartua, 2021; Weinstein et al., 2021), and/or contribute to social isolation (Massing-Schaffer et al., 2020). The interaction design of these platforms demands continuous audiovisual self-presentation from users’ private lives, serving as a form of phatic communication (Lobinger et al., 2021; Olivio et al., 2022) and fostering a sense of authenticity and intimacy (Caro-Castaño, 2023). This compulsory exposure from private spaces creates a paradoxical situation for young users: they feel vulnerable due to constant social validation and comparison yet simultaneously pressured to participate to maintain peer relationships. Given these challenges, it is crucial for platforms to implement more effective content detection systems. Importantly, this content moderation should be conducted by human moderators rather than algorithms, as they can better distinguish between harmful content that may promote risky behavior and content that facilitates genuine mutual support (Zhang et al., 2024).

In the third case, frequent viewing of competitions or reality shows was associated with higher levels of suicidal ideation, as such media content often fails to adhere to the technical guidelines and recommendations proposed by the WHO (2017). These programs construct emotional narratives that emphasize conflict among participants. In contrast to media products that promote prevention through social companionship (Garcés et al., 2023a), competitions and reality shows tend to exacerbate the contagion effect by amplifying negative emotions.

The results must be interpreted in the context of the study’s focus on suicidal ideation rather than on suicidal behaviour or mortality. Although ideation can co-occur with other risk factors, most children who report such thoughts will never attempt suicide, and deaths by suicide in this age group are uncommon. This distinction is essential to avoid overestimating the lethality implied by the findings, particularly given the way suicide is often portrayed in the media.

Despite the interest, this study has several limitations. One relates to the generalisability of the findings due to the age distribution of the sample. The proportion of participants aged 14 and 15 years was comparatively small, largely because some participating schools did not have complete upper-grade cohorts or were unable to schedule sessions for these students due to academic calendar constraints. This underrepresentation may limit the extent to which the results capture the media use patterns characteristic of older adolescents, who typically engage more intensively with social media than their younger peers. Therefore, caution is warranted when extrapolating the findings to the upper end of the surveyed age range.

Future research should aim for a more balanced representation across all ages, ideally incorporating longitudinal designs or in-depth qualitative approaches to better understand the mediating and moderating factors behind the associations observed. It would also be valuable for future analyses to differentiate between children with and without a history of suicide attempts, as this could help identify whether certain patterns of media and social media use are more prevalent in specific risk groups. In the present study, this distinction was not possible due to the anonymous nature of the data collection and the ethical safeguards implemented.

Similarly, this study raises the urgency of addressing the ethical implications of media consumption among minors, as exposure to sensitive content can affect their mental health. The algorithmic nature of social media raises important questions about platform responsibility in protecting younger users. The lack of oversight and the promotion of harmful content highlight the tension between commercial interests and the well-being of minors, reinforcing the need for policies that regulate access to such content. This also calls for schools to develop guidelines for parents and educators on managing media exposure and to encourage the active involvement of platforms in regulating suicide-related content, in line with WHO (2017) recommendations.