Introduction

Rapid changes in society, family structure, and socioeconomic factors are contributing to severe adversities and stress in young people [1]. Self-harm and suicide rates have been going up in young populations prompting an examination of factors that trigger this behaviour [2]. Globally, it is reported that each year, more than 700,000 people die due to suicide and suicide is the fourth leading cause of death in 15–19 year old [3]. Additionally, suicide contributes to 1.4 percent of the global burden of disease and also emerged as the second cause of death among young people aged 15–29 years [4]. Hence, making suicide and concomitant issues of loneliness, isolation and poor support important public health concerns [5]. The African annual prevalence of suicide was found to be 34,000 with incidence of 3.2 per 100,000 [6]. A recent systematic review found that the median point prevalence of suicidal ideation in adolescents aged 10–19 from several countries in sub-Saharan Africa (SSA) was 11.6% [7]. It is noteworthy suicide does not occur sporadically; rather, it operates through a process which usually constitutes suicidal ideation and suicidal attempt [8]. Prior exposure to suicide and prior exposure to suicide attempts in the general population are associated with increased odds of subsequent suicidal behaviour [9]. Therefore, understanding suicidal ideation provides an enabling framework to tackle suicides in the general population.

In its generic form, suicidal ideation simply refers to when an individual thinks, conceives, and conceptualizes the processes to kill themselves for reasons known to them. Suicidal ideation can also be defined as “any self-reported passive thought about wanting to be dead or active thoughts about killing oneself not accompanied by preparatory behaviour” [10, p. 1]. It is estimated that the global lifetime prevalence of suicidal ideation ranges between 3.1 and 56 percent [11, 12]. Nonetheless, this prevalence might be understated as suicide is a sensitive topic, and is therefore rarely discussed openly in the society [13]. Among European countries, the lifetime prevalence of suicidal ideation among adolescents aged 15–16 years was 31.5 percent in Hungary and 15 percent in Armenia [14]. In a study conducted in Tanzania, Dunlavy et al. [15] found that among adolescents aged 11–16 years, seven percent of them reported suicidal ideation. However, in Ethiopia, a reported 23.7 percent of in-school adolescents reported suicidal ideation [16].

Articulating the pathway through which suicidal ideation operates is a challenging endeavor given the fact that it is affected by a multiplicity of factors which can broadly be categorized in biological (sex and age), psychological (affective disorder, major depression, and anxiety disorders) and environmental factors [17, 18]. Furthermore, literature shows that psychosocial triggers are one of the strongest predictors of suicidal ideation [11]. Among these psychosocial triggers are the experiences of loneliness and poor social support.

Loneliness is defined as the “cognitive discrepancy between the social relations an individual wishes to have and those one perceives to have and the affective reactions of sadness and emptiness resulting from this discrepancy” [19, p.133]. Evidence available shows that loneliness has adverse effects on the health of adolescents [20, 21]. Protracted or intense loneliness can easily become a trigger for suicidal ideation and other self-harming behaviours. For example, in a study among HIV/AIDS patients in Africa, loneliness and social support were significant predictors of suicidal ideation and attempt [22, 23]. Although Yadegarfard et al.’s [14] study was conducted among HIV/AIDS patients, which included adolescents, their findings suggest the potential association between loneliness, social support, and suicidal ideation among adolescents in the general population.

There is an urgent need to study mental health outcomes in low-resource settings like SSA because of the high unmet mental health needs due to unavailable human and financial resources [24]. Notwithstanding, there is a paucity of empirical evidence in SSA to substantiate the hypothesis that loneliness, and low social support increase the risk of suicidal ideation in adolescents. This presents a significant gap in the existing scientific discourse and evidence on correlates of adolescent suicidal ideation. The present study, therefore, seeks to test this hypothesis by investigating the association between loneliness, social support, and suicidal ideation among adolescents in SSA. Our findings provide important evidence which can facilitate targeted support and services to the most vulnerable sub-populations, as well as engineer countries in SSA towards the attainment of the Sustainable Development Goal (SDGs) 3, target 3.4 that points to self-harm and suicide prevention as key adolescent health strategies to be scaled at the population level.

We premise our study on the interpersonal-psychological theory of suicide (IPTS). This theory was developed by Joiner with the goal of providing some theoretical model of suicide behaviour [25]. According to the IPTS, suicidal desires and intentions stem from two main factors: perceived burdensomeness and a sense of low belongingness [26]. Joiner [27] describes the perceived burdensomeness as a situation whereby an individual feels or thinks that their existence is a burden for their friends and family as well as the larger society. Such sense of burdensomeness leads to a situation of self-hate and feelings of being a liability rather than an asset to everyone [25]. The second tenet of the IPTS is the low sense of belongingness [27]. This is the crust of our study. Low sense of belongingness denotes the feelings of being alienated from one’s family or social group, and thus, comprises loneliness and absence of reciprocal care [25, 27]. Our study perfectly fits into this IPTS as it shows how loneliness, and the absence of reciprocal care (i.e., social support) can lead to low sense of belongingness and consequently increase the risk of suicidal desires and ideation among adolescents. This relationship has been tested in previous studies that have found a statistically significant association between perceived burdensomeness and suicidal ideation, as well as between low sense of belongingness and suicidal ideation [28, 29].

Methods and methods

Sampling method and sample size

The respondents were selected using a two-stage cluster sampling technique. The first stage involved the selection of study schools with probability proportional to the school’s enrolment size. Later, classes with the study schools were randomly selected and within those schools, all the eligible students aged 10–19 years were included in the study. Among the selected students, only those who assented and their parents consented were given questionnaires to complete. In the GSHS, numerical weights were applied to each respondent to enable the generalization of the findings. A total of 19,119 adolescents from the eight countries in SSA were included in the study (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Distribution of the study sample.

https://doi.org/10.1371/journal.pone.0275660.t001

Statistical analyses

We analysed the data using Stata software version 16.0 (Stata Corporation, College Station, TX, USA). Percentages were used to present the results of the prevalence of suicidal ideation, loneliness, and social support (Fig 1). We computed the distribution of the study sample per country across the variables used. Next, we examined the association between loneliness and social support and suicidal ideation using a mixed-effect multivariable binary logistic regression. Three models (Model O—Model II) were built to examine this association. Model O was fitted to contain only the outcome variable with no explanatory variables or covariates. Model I had only the key explanatory variables (loneliness and social support). Model II on the other hand contained the key explanatory variables and the covariates. The results were presented using adjusted odds ratios (aORs) with their respective 95% confidence interval (CIs). Statistical significance was set at p<0.05. We used the Stata command “melogit” in fitting the three models. We used the Akaike’s Information Criterion (AIC) tests to compare the fitness of the model with the last model being the best fitted model because of the least value of the AIC. Later, we examined the country-level association between the key explanatory variables and suicidal ideation using the Stata command “bysort” (Table 4). All the analyses were weighted. We applied a complex sample analysis (svy) and the inherent sample weight in all analyses we performed to reduce bias from non-response and improve generalisability to all in-school adolescents in sub-Saharan Africa. This manuscript was drafted by relying on the “Strengthening the Reporting of Observational Studies in Epidemiology” (STROBE) statement writing the manuscript [35].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Prevalence of loneliness, peer support, close friends, and suicidal ideation among in-school adolescents.

https://doi.org/10.1371/journal.pone.0275660.g001

Results

Distribution of the covariates across the individual countries in sub-Saharan Africa

Table 2 shows the covariates across the individual countries in sub-Saharan Africa.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Distribution of the covariates in each individual country.

https://doi.org/10.1371/journal.pone.0275660.t002

Overall, there were many older in-school adolescents in the countries considered in this study. The distribution of sex varied from one country to the other. Most of the in-school adolescents in the studied countries did not feel hungry, were not anxious, did not use alcohol, cigarette, marijuana, and tobacco. Apart from in-school adolescents in Tanzania, the rest of the in-school adolescents did not receive parent or guardian supervision. In general, there was no parental or guardian connectedness and bonding. However, most of the in-school adolescents had respect for privacy from their parents or guardians.

Association between loneliness, peer support, close friends, and suicidal ideation

Table 3 shows the results of the association between loneliness, peer support, close friends, and suicidal ideation among in-school adolescents. The study found that in-school adolescents who felt lonely (aOR = 1.88, 95% CI = 1.69, 2.09) were more likely to experience suicidal ideation. However, in-school adolescents who had peer support (aOR = 0.89, 95% CI = 0.82, 0.97) and those who had one or more close friends (aOR = 0.77, 95% CI = 0.68, 0.86) had lower odds of having suicidal ideation.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Mixed effect analysis of the association between loneliness, peer support, close friends and suicidal ideation among the in-school adolescents.

https://doi.org/10.1371/journal.pone.0275660.t003

Association between loneliness, peer support, close friends, and suicidal ideation by country

Table 4 shows the results of the association between loneliness, social support, close friends and suicidal ideation segregated by country. The study found that in-school adolescents who felt lonely and lived in Benin (aOR = 1.97, 95% CI = 1.42, 2.74); Ghana (aOR = 1.53, 95% CI = 1.19, 1.98); Mauritius (aOR = 2.68, 95% CI = 1.91, 3.75); Namibia (aOR = 1.57, 95% CI = 1.25, 1.96); Seychelles (aOR = 4.09, 95% CI = 2.97, 5.64); and Tanzania (aOR = 1.94, 95% CI = 1.32, 2.85) were more likely to experience suicidal ideation. Conversely, in-school adolescents who had close friends and were from Mozambique (aOR = 0.50, 95% CI = 0.29, 0.85) and Tanzania (aOR = 0.42, 95% CI = 0.30, 0.58) were less likely to experience suicidal ideation.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Association between loneliness, social support and suicidal ideation segregated by country.

https://doi.org/10.1371/journal.pone.0275660.t004

Discussion

Our study examined the associations between loneliness, social support and suicidal ideation among adolescents in SSA. We found that the prevalence of loneliness, peer support, one or more close friends, and suicidal ideation were 10%, 33.4%, 90.1%, and 14.5%, respectively. Namibia recorded the highest (14.9%) prevalence of loneliness while Tanzania had the least (6.4%). A possible reason for this finding of Namibian adolescents having the highest prevalence of loneliness in SSA could be as a result of the increased consumption of alcohol and other illicit drugs (e.g., tobacco) in the country [36]. It could also be that the substance use prevents formation of more meaningful social and peer relationships [36]. It has also been widely reported that negative peer influences can drive substance use and other mental health problems. While Liberia recorded the highest (20.7%) prevalence of suicidal ideation by the in-school adolescents, Tanzania recorded the least (12.0%). A possible explanation for this finding from Liberia could be the recent history of massive ethnic conflicts and exposure to traumatic violence in the country that have been widely discussed [37]. It has also been well documented that in conflict affected, humanitarian settings, intersectional violence especially sexual violence and displacement of families, communities can trigger suicidal ideation and both acute and long term psychological distress [38–40]. Evidence also suggests that mental health policies and programs are deficient in planning and poorly implemented in SSA countries. This is even more glaring in countries going through massive conflicts and humanitarian disasters [41, 42] and these could have accounted for the observed finding in our study.

The study found that in-school adolescents who felt lonely were more likely to experience suicidal ideation, a finding that was supported by other previous studies [15, 32, 43–45]. This finding could be attributed to the aggravation of the negative effects of psychological problems that are associated with having no one to share the problems one encounters [32]. This study provides additional empirical evidence to the second tenet of the IPTS which espoused that loneliness can trigger a reduction in belongingness which may result in some self-harming thoughts and behaviours [27]. This finding underscores the need to design and implement programmes that eliminate loneliness among in-school adolescents to help alleviate having thoughts of suicide. Essentially, making counselling services more accessible to all in-school adolescents without any form of restrictions would help in alleviating suicidal ideations.

However, in-school adolescents who had peer support had lower odds of having suicidal ideation. Some previous studies [15, 31, 32, 46] also reported similar findings. A possible reason for this observation could be a reflection of the indirect relationship between peer support and suicidal ideation [15, 31, 32]. It could also be that, since in-school adolescents have greater peer support or school or parental surveillance, they can share the problems they are facing with them for immediate solution, hence, reducing their likelihood of considering suicide [47]. This certainly provides credence for the second tenet of IPTS on which this study was premised [27]. This finding suggests the need for school authorities to create clubs or social groups in schools such as debate clubs, brigade and drama clubs, among others which will provide meaningful platforms for in-school adolescents to share and discuss issues of mental health to help alleviate having suicidal ideations whenever they encounter any problem.

In-school adolescents who had one or more close friends had lower odds of having suicidal ideation, suggesting the protective effect of having one or more close friends against suicidal ideation. Other previous studies in Mozambique[31], Nigeria [48] and Canada [49] have also confirmed this finding. It could be that those who had close friends are more likely to discuss their problems with their friends, reducing their likelihood of having suicidal ideation [50]. Our finding situates well with the IPTS framework that points to sense of belonging as a protective factor.

The study also found that in-school adolescents who felt lonely and lived in Benin, Ghana, Mauritius, Namibia, Seychelles and Tanzania were more likely to experience suicidal ideation. A possible reason could be that counselling services may be absent in many schools [51] in Benin, Ghana, Mauritius, Namibia, Seychelles and Tanzania, causing an increase in suicidal ideations among in-school adolescents. It could also be that mental health and counselling services in schools have been neglected or given less attention, increasing the risk of suicide behaviours as a result of bullying and other aggressive behaviours. This calls for urgent revamp in mental health and counselling services in schools in the identified countries in the SSA region.

Conclusion

The findings point to significant role of loneliness, and lack of social support, in understanding suicidal ideation. Countries in SSA need to improve child and adolescent mental health policies and programmes to respond to these risk factors and mental health challenges. It is also important to develop life-skills and psychosocial competencies in young people to promote social support, self-expression and support for common mental disorders so that suicidal ideation and attempts can be mitigated in time. It is important that countries in SSA advance child and adolescent mental health policies that address treatment, prevention and promotion in this population. Early screening and developing interventions for risk factors for suicidal ideation, including addressing experiences of loneliness and lack of social support needs prioritization to improve adolescent well-being in the region.

Supporting information

Acknowledgments

We acknowledge the World Health Organization for making the Global School-based Student Health Survey freely accessible for our study.