Social media use, risk behavior engagement, and mental health among middle adolescents in the Caribbean

Abstract

Background

Social media plays a pivotal role in adolescents’ lives. Social media encounters, including online risk behaviors, may influence real-world risk behaviors and mental health. This study explored the magnitude and patterns of social media use, risk behaviors, and mental health among boys and girls and examined associations between these factors.

Methods

A survey was administered to Grade 9 students in New Providence, The Bahamas, in 2023 to assess their social media use, including platforms used, online activities, and experiences of online risk behavior, as well as their risk behavior engagement and mental health. We analyzed data from 1,563 students using generalized linear mixed models.

Results

Two-thirds of the students spent three hours or more on social media daily. Online risk behaviors, such as cyberbullying and sexting (sending, receiving, or forwarding sexually explicit messages), were also prevalent. Approximately 48% had suicidal ideation during their lifetime. Girls reported higher rates of social media usage, active and passive engagement on social media, cyberbullying experiences, and mental health issues, while boys were more likely to engage in sexting and offline risk behaviors. Factors associated with suicidal ideation included being female, experiencing sexting or cyberbullying, using X, and engaging in risk behaviors such as inappropriate touching and weapon carrying.

Conclusions

Future studies should clarify the complex interplay among the content consumed by adolescents, their online activities, and gender-specific effects on mental health. This study highlights the need for programs that can both foster a positive and supportive online environment and provide targeted assistance for victims of online harm.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-025-23646-8.

Social media use and mental health among adolescents

Social media has become a crucial part of adolescents’ daily routines, with over three-quarters of American adolescents engaging with this communication method daily [1]. The COVID-19 pandemic intensified social media use (SMU) among adolescents [2], as a way to establish and maintain connections with their peers and the broader world [3]. Social media can be a powerful influence during adolescence when peer relationships take center stage [4]. Although social media can have positive effects, including the sharing of information and ideas, alleviating loneliness, and fostering creativity [5, 6], it can lead to risk behaviors, including substance use, risky sexual behaviors, and delinquency [7], and can worsen adolescent mental health [8].

Mental health is a growing concern for adolescents. Suicide is a leading cause of adolescent deaths globally [9] and has increased steadily in the Americas regardless of gender since 2010 [10]. Depressive symptoms, self-harm, and suicidal ideation are strong predictors of suicide attempts and death by suicide [11, 12]. Concerns and distress from the COVID-19 pandemic worsened adolescents’ depression and self-harm [13, 14], and aggravated their suicidal ideation and behaviors [15]. The sharp increase in depression, self-harm, and suicide among adolescents, coinciding with the rise in digital media use, suggests that digital technologies and social media may play a role in worsening mental health outcomes among adolescents [16].

Impact of social media use on mental health among adolescents

Online risk behaviors, such as cyberbullying, sexting, and online exposure to pornography, prevail among adolescents from 15 to 59% [17–19]. There is a general consensus that cyberbullying and exposure to risk behavior-related content contribute to worsening mental health. Exposure to content about risk behaviors can also lead adolescents to believe that such behaviors are socially approved and promote risk behaviors in the real world, such as delinquency, substance use, and sexual risk behavior [19–21], contributing to mental health problems [22]. However, findings on sexting (sending, receiving, or forwarding sexually explicit messages via cellphone or smartphone) are mixed. One systematic review reported that sexting and suicidal ideation have a positive association [23], but another study, a meta-analysis, did not find an association [24]. The relationship among SMU, risk behaviors, and mental health status should be further explored.

Gender plays a significant role in how SMU influences risk behavior and mental health. Girls are more likely to actively engage in social media activities for connection. In contrast, boys are more likely to spend more time in non-social media activities, such as playing computer games or watching videos and use social media for information [25, 26]. Further, girls are at a higher risk of experiencing online risks, such as cyberbullying and exposure to sexual or violent images [27], and internalizing symptoms like depression, anxiety [28], low self-esteem, and poor body image [29]. These gender-specific dynamics intersect with additional factors, such as parental monitoring, with girls receiving greater parental oversight [30]. Furthermore, the types of social media platforms and interactions on the platforms, coupled with exposure to peer influence and peers’ posts about risk behaviors [31, 32], complicate the relationship between SMU impact and well-being of boys and girls.

Theoretical framework

Social learning theory serves as the theoretical foundation for this study. It suggests that adolescents’ behaviors are shaped through both direct experience and observation of others. Adolescents are more likely to observe, imitate, and adopt behaviors of their role models whom they perceive as similar, relevant, and influential [33]. The rise of SMU among adolescents has extended their social interaction to online platforms, increasing their exposure to risk content or online risk behaviors, including cyberbullying [27, 34]. Peer interaction on social media can create biased normative perceptions, intensify modeling effects, and increase the likelihood of engaging in risk behaviors, such as substance use, risky sexual behavior, and delinquency [31, 35, 36].

Lack of research on social media use in the Caribbean

Internet is widely used in Latin America and the Caribbean, with 75% of individuals in the region using the Internet. The Bahamas, in particular, reported one of the highest rates at 94% [37]. The prevalence of SMU varies, although SMU studies in the region are limited: a Jamaican study found that the prevalence of problematic SMU is 28% [38], and a study from Trinidad and Tobago revealed that 0.6% of Facebook users and 3.7% of TikTok were problematic users [39]. Online risk behavior is fairly common. In a 2018 study in South and Central America and the Caribbean, 20.4% of 13- to 17-year-old students reported having been a victim of cyberbullying [40]. Moreover, a meta-analysis reported that 14.8% had sent a sext and 27.4% had received one [18]. As one in four to five adolescents in the English-speaking Caribbean experience mild to severe mental health symptoms [41], research on the consequences of SMU on adolescents’ mental health is important. However, no studies have examined the impact of SMU among Caribbean adolescents, particularly in The Bahamas. Therefore, we used student data from a national implementation study of school-based HIV prevention to explore the magnitude and patterns of SMU, risk behaviors, and mental health status of adolescent boys and girls in The Bahamas and examined the association among SMU, risk behaviors, and mental health status.

Methods

Study site and participants

All Garde 6 students within the public school system in The Bahamas received the national HIV prevention program. The students completed the Health Risk and Protective Factors Survey before and after program implementation to evaluate the effectiveness of the program. Grades 7 through 9 students received an annual one-hour booster session and completed a follow-up survey annually. This study used cross-sectional data collected in 2023 from Grade 9 students in public schools in New Providence. New Providence is the island where the capital city, Nassau, is located. It has the highest population density, accounting for approximately 75% of the country’s population [42] and is approximately 21 miles in length and 7 miles in width.

Data collection

The Health Risk and Protective Factors Survey is a culturally validated instrument that assesses HIV/AIDS knowledge, sexual and reproductive health knowledge, perceptions about and intentions for risk behaviors, as well as recent engagement in risk behaviors [43]. It was adapted from the Young Risk Behavior Survey, a standardized survey developed by the CDC [44], and has been used in The Bahamas for over 15 years [45, 46]. The follow-up survey for Grade 9 students included additional questions about SMU and mental health status, adapted from literature [32, 47–49] and validated tools such as Patient Health Questionnaire-9 [PHQ-9] [50] (See Supplementary File 1). All data were self-reported. Students completed the survey in paper-and-pencil format in a classroom supervised by program staff. Teachers were not present during the survey.

Measures

Social media use

The intensity of internet use and SMU was assessed as the number of hours daily on average with four options (< 2, 2 to < 3, 3 to < 6, and 6 h). For social media platforms and devices for daily use. The survey presented a list and asked students to mark all that applied. For online activities, the options included active (commenting and messaging) or passive interaction (posting) with others on social media, watching videos, playing games, or doing homework.

Risk behavior

Students’ offline risk behaviors were evaluated by the extent of their engagement in delinquency, substance use, and sexual risk behaviors in the past six months. For delinquency, four items asked whether students had been truant, carried a knife as a weapon, engaged in robbery, or physically fought with someone. For sexual risk behaviors, the items were whether they did feel up or touched someone in a wrong way and whether they had had unprotected sex. Substance use focused on cigarettes, marijuana, and alcohol. Online risk behaviors included cyberbullying, sexting, or online pornographic content consumption, each with a Yes/No answer. Lastly, students were asked if their parents monitored their Internet use. The Cronbach’s alpha for risk behaviors was 0.73.

Mental health status

Three items evaluated students’ mental health status: depressive mood for two or more consecutive weeks in the past year, thoughts of self-harm in the past two weeks, and lifetime suicidal ideation. Responses were binary (yes/no) for mood and suicidal ideation, and self-harm thoughts were assessed on a four-point Likert scale ranging from “never” to “nearly every day.” The Cronbach’s alpha for risk behaviors was 0.67.

Data analysis

Among 1,610 students who participated in the survey, data from 47 students were removed because of missing values in key variables (such as gender and suicidal ideation). With data from 1,563 students, we first used descriptive statistics (mean, standard deviation, and percentage) for age, SMU patterns, health risk behaviors, and mental health. We examined gender differences using Pearson chi-square tests for categorical variables or Student’s t tests for continuous variables. We also examined bivariate associations between SMU, health risk behaviors, and mental health (depression and suicidal ideation) among boys and girls, respectively. Self-harm thoughts were removed due to the high correlation with suicidal ideation (r = 0.605).

We used generalized linear mixed models to examine the association of depression and suicidal ideation, respectively, with SMU and health risk behaviors, adjusting for clustering effects of classroom and/or school as students are clustered within classrooms in schools. We included SMU and health risk behavior variables that were significant in our bivariate analysis and those that have been shown to be significant factors in the existing literature (e.g [28]). Given high correlations among some of the explanatory variables (especially SMU patterns), stepwise variable selection was used for model selection. Adjusted odds ratios (aORs) and their 95% confidence intervals (CIs) were calculated. All statistical analyses used the SAS 9.4. statistical software package (SAS Institute Inc., Cary, NC).

Results

Prevalence of social media use, risk behavior, and mental health status

Among 1,563 students, 52% were girls. Approximately two-thirds of students used social media for three hours or more per day (Table 1). More than 80% used WhatsApp, while 31% used X (formerly Twitter). 68% of students actively interacted with others on social media. Online risk behaviors were prevalent, with 64% consuming pornographic content and 46% having experienced cyberbullying (Table 2). 76% of students reported engaging in one or more offline risk behaviors (not reported in tables). The percentage of risk behavior varied from 12% carrying a weapon to 47% drinking alcohol (Table 2). Most students (62%) felt depressed, and 48% reported having seriously thought about killing themselves.

Table 1.

Social media use among grade 9 students in the Bahamas

Characteristics	Total	Boys (N = 750)	Girls (N = 813)	2	p
	Column %	Column %
Daily internet use
< 2 h	4.8	5.3	4.4	10.2	< 0.05
>= 2 h & < 3 h	8.2	9.4	7.0
>= 3 h & < 6 h	19.0	21.3	16.9
>= 6 h	68.0	64.1	71.6
Daily social media use
< 2 h	17.9	25.7	10.8	84.76	< 0.001
>= 2 h & < 3 h	14.5	17.7	11.5
>= 3 h & < 6 h	22.3	20.3	24.1
>= 6 h	45.3	36.3	53.6
Parental monitoring on Internet access
Yes	19.7	18.9	20.5	3.39	0.18
No	65.8	68.0	63.7
I don’t know	14.5	13.2	15.8
Types of devices for daily use (yes %)
Smartphone	88.1	86.8	89.3	2.35	0.13
Tablet	37.1	33.1	40.8	10.09	< 0.01
Laptop	36.7	36.5	36.8	0.01	0.92
Desktop	14.7	20.4	9.4	38.11	< 0.001
Types of social media use (yes %)
WhatsApp	82.9	78.4	87.0	20.13	< 0.001
Tik Tok	82.3	72.9	90.9	86.34	< 0.001
Instagram	66.2	56.9	74.8	55.58	< 0.001
Snapchat	61.0	46.4	74.5	129.49	< 0.001
Facebook	60.6	56.8	64.2	8.85	< 0.01
Facebook Messenger	43.0	38.3	47.4	13.15	< 0.001
Discord	38.8	50.0	28.5	75.66	< 0.001
X (formerly Twitter)	30.7	33.1	28.4	3.97	< 0.05
Other	19.6	19.3	19.8	0.06	0.81
Online activities (yes %)
Streaming or watching videos	75.6	70.4	80.3	20.79	< 0.001
Active interaction on social media	67.8	54.0	80.4	124.85	< 0.001
Playing games	62.8	73.9	52.6	75.23	< 0.001
Homework	48.9	42.3	55.0	25.24	< 0.001
General information/research	37.7	34.1	41.0	7.74	< 0.01
Passive interaction on social media	33.7	24.5	42.1	53.71	< 0.001
Other activities	16.4	16.7	16.1	0.09	0.77

The total sample size was 1,563; Participants’ mean age was 14 years old (SD = 1.0), and participant age did not differ by gender

Table 2.

Risk behavior engagement and mental health among grade 9 students in the Bahamas

Characteristics	Total	Boys (N = 750)	Girls (N = 813)	2	p
	Column %	Column %
Online risk behavior experience (yes %)
Cyberbullying	45.9	36.5	54.6	50.27	< 0.001
Pornography consumption	64.2	69.4	59.3	16.85	< 0.001
Sexting	36.6	37.5	35.8	0.49	0.48
Offline risk behavior experience (yes %) a
Delinquency
Was truant	32.4	28.9	35.5	7.76	< 0.01
Carried a weapon	12.0	18.7	5.8	60.9	< 0.001
Involved in robbery	5.9	8.2	3.8	13.21	< 0.001
Engaged in a fight	37.7	45.2	30.9	33.52	< 0.001
Sexual risk behavior
Had unprotected sex	10.2	12.4	8.2	7.47	< 0.01
Touched someone inappropriately	25.9	34.9	17.6	59.94	< 0.001
Substance use
Smoked cigarettes	11.8	13.6	10.1	4.57	< 0.05
Used marijuana	13.0	13.4	12.6	0.22	0.64
Drank alcohol	47.3	41.2	52.8	21.06	< 0.001
Mental health
Depression	61.5	44.6	77.0	171.22	< 0.001
Self-harm thoughts in the past two weeks
Never	47.3	64.8	31.1	183.67	< 0.001
Several days	20.8	16.2	25.0
More than half of the days	13.6	9.3	17.6
Nearly every day	18.3	9.7	26.3
Lifetime suicidal ideation	47.8	32.1	62.2	141.71	< 0.001

The total sample size was 1,563.

^aExperience in the past six months

Gender differences were observed in most SMUs, risk behaviors, and mental health problems. Girls more commonly used social media for three hours or more (78% vs. 57%, p < 0.001), had active or passive online activities (active activities: 80% vs. 54%, p < 0.001; passive activities: 42% vs. 25%, p < 0.001) and experienced cyberbullying (55% vs. 37%, p < 0.001). Their mental health was worse than boys. The suicidal ideation rate of girls almost doubled the rate of boys (62% vs. 32%, p < 0.001). Boys reported higher rates of engaging with pornographic content and all risk behaviors than girls did, except for truancy and drinking alcohol.

Association between social media use, risk behavior, and mental health status

In bivariate analyses, utilization of specific apps like Instagram, active and passive interaction on social media, and online risk events were correlated with depression and suicidal ideation among boys and girls (Tables 3 and 4). Specific offline risk behaviors, such as truancy, physical fighting, unprotected sex, and substance use, were correlated with depression, whereas all offline risk behaviors were associated with suicidal ideation. When analyzed separately in boys and girls, the patterns of the association did not differ by gender: most variables that were significant in girls were also significant in boys.

Table 3.

Social media use, risk behaviors, and depression among grade 9 students in the Bahamas

Characteristics	Boys (N = 750)			Girls (N = 813)
	Depression	No depression	p	Depression	No depression	p
	Column %			Column %
All	32.3	62.4		67.7	37.6
Social Media Use
Daily Internet use
< 2 h	6.1	4.4	0.43	4.5	4.3	0.36
>= 2 h & < 3 h	9.5	9.3		7.3	6.5
>= 3 h & < 6 h	18.0	24.2		15.8	21.1
>= 6 h	66.5	62.1		72.4	68.1
Daily social media use
< 2 h	23.7	27.2	0.12	10.6	11.7	0.004
>= 2 h & < 3 h	17.5	17.9		11.0	13.9
>= 3 h & < 6 h	19.1	21.4		21.6	32.2
>= 6 h	39.7	33.5		56.9	42.2
Parental monitoring on Internet access
Yes	21.7	16.4	0.18	19.4	24.7	0.29
No	65.2	70.5		64.5	60.2
I don’t know	13.1	13.2		16.1	15.1
Types of devices for daily use (yes %)
Smartphone	87.7	85.9	0.48	90.5	85.0	0.03
Tablet	34.3	32.5	0.60	40.2	43.0	0.49
Laptop	37.7	35.7	0.58	37.0	37.1	0.98
Desktop	21.4	19.7	0.56	10.0	7.5	0.32
Types of social media use (yes %)
WhatsApp	78.9	77.9	0.74	87.6	84.4	0.26
Tik Tok	76.8	69.9	0.04	91.6	88.7	0.22
Instagram	63.6	51.2	0.001	77.7	64.5	< 0.001
Snapchat	49.4	43.9	0.14	76.7	68.3	0.02
Facebook	62.1	52.7	0.01	67.6	52.7	< 0.001
Facebook Messenger	41.0	35.9	0.16	52.3	31.2	< 0.001
Discord	51.5	49.0	0.50	29.3	26.9	0.53
X (formerly Twitter)	36.5	30.1	0.07	29.7	24.2	0.14
Other	17.5	20.9	0.24	19.3	22.0	0.42
Online activities (yes %)
Streaming or watching videos	71.1	70.0	0.67	80.6	79.6	0.77
Active interaction on social media	27.7	22.3	0.09	45.3	31.7	0.001
Playing games	72.9	75.0	0.51	50.6	60.2	0.02
Homework	44.6	40.3	0.24	54.8	55.4	0.89
General information/research	34.6	33.7	0.80	41.6	39.3	0.56
Passive interaction on social media	59.6	49.8	0.007	82.3	74.2	0.01
Other activities	18.4	15.5	0.30	18.0	10.2	0.01
Online Risk Behavior Experience (yes %)
Cyberbullying	46.3	28.9	< 0.001	59.4	38.0	< 0.001
Pornography consumption	77.9	62.8	< 0.001	61.7	51.2	0.012
Sexting	45.8	30.9	< 0.001	41.2	18.0	< 0.001
Offline Risk Behavior Experience (yes %)
Delinquency
Was truant	34.5	24.6	0.003	37.5	29.2	0.04
Carried a weapon	24.9	13.9	< 0.001	6.1	4.9	0.52
Involved in robbery	11.9	5.4	0.001	4.0	3.2	0.63
Engaged in a fight	52.9	39.4	< 0.001	33.6	21.7	0.002
Sexual risk behavior
Had unprotected sex	16.5	9.3	0.003	9.4	4.3	0.03
Touched someone inappropriately	42.5	28.6	< 0.001	19.4	11.9	0.02
Substance use
Smoked cigarettes	17.5	10.5	0.005	11.5	5.4	0.02
Used marijuana	15.5	11.8	0.14	14.6	6.5	0.004
Drank alcohol	47.6	36.2	0.002	57.9	36.2	< 0.001

The total sample size was 1,563. Age did not differ between groups that have had and not had depression

Table 4.

Social media use, risk behaviors, and suicidal ideation among grade 9 students in the Bahamas

Characteristics	Boys (n = 750)			Girls (n = 813)
	Suicidal ideation	No suicidal ideation	p	Suicidal ideation	No suicidal ideation	p
	Column %			Column %
All	32.3	62.4		67.7	37.6
Social Media Use
Daily Internet use
< 2 h	5.1	5.3	0.34	3.8	5.5	0.20
>= 2 h & < 3 h	7.6	10.3		7.1	6.9
>= 3 h & < 6 h	21.1	21.3		15.9	18.6
>= 6 h	66.2	63.1		73.2	69.0
Daily social media use
< 2 h	21.9	27.5	0.06	10.1	12.0	0.07
>= 2 h & < 3 h	19.7	16.8		10.7	13.0
>= 3 h & < 6 h	16.3	22.2		23.2	25.5
>= 6 h	42.1	33.5		56.0	49.5
Parental monitoring on Internet access
Yes	19.5	18.6	0.96	19.5	22.2	0.21
No	67.4	68.2		63.1	64.7
I don’t know	13.1	13.2		17.4	13.1
Types of devices for daily use (yes %)
Smartphone	87.9	86.3	0.53	89.3	89.3	0.97
Tablet	36.9	31.2	0.12	41.0	40.7	0.96
Laptop	42.3	33.8	0.02	36.4	37.5	0.75
Desktop	70.8	20.2	0.87	10.5	7.5	0.16
Types of social media use (yes %)
WhatsApp	80.1	77.6	0.44	86.0	88.6	0.28
Tik Tok	78.4	70.3	0.02	91.5	80.0	0.44
Instagram	61.4	54.8	0.09	78.3	69.1	0.003
Snapchat	49.4	45.0	0.26	76.0	72.0	0.20
Facebook	61.8	54.4	0.06	67.3	59.0	0.02
Facebook Messenger	41.1	36.9	0.28	50.8	41.7	0.01
Discord	55.2	47.5	0.05	28.9	28.0	0.80
X (formerly Twitter)	41.5	29.1	0.001	30.6	24.8	0.07
Other	21.2	18.5	0.38	19.2	20.9	0.57
Online activities (yes %)
Streaming or watching videos	70.5	70.3	0.95	80.4	80.1	0.92
Active interaction on social media	30.7	21.6	0.007	45.7	36.2	0.008
Playing games	73.9	73.9	1.0	50.4	56.4	0.1
Homework	46.5	40.3	0.11	53.0	58.3	0.14
General information/research	37.8	32.4	0.15	39.9	42.7	0.44
Passive interaction on social media	57.3	52.5	0.22	82.4	77.2	0.07
Other activities	22.4	14.0	0.004	18.8	11.7	0.008
Online Risk Behavior Experience (yes %)
Cyberbullying	48.9	30.8	< 0.001	61.8	42.7	< 0.001
Pornography consumption	81.6	63.6	< 0.001	64.3	50.7	< 0.001
Sexting	53.5	30.1	< 0.001	42.9	24.2	< 0.001
Offline Risk Behavior Experience (yes %)
Delinquency
Was truant	38.4	24.5	< 0.001	38.7	30.4	0.02
Carried a weapon	27.6	14.4	< 0.001	7.9	2.3	0.001
Involved in robbery	13.9	5.5	< 0.001	4.8	2.3	0.08
Engaged in a fight	54.8	40.6	< 0.001	36.9	21.1	< 0.001
Sexual risk behavior
Had unprotected sex	18.6	9.5	< 0.001	11.0	3.6	< 0.001
Touched someone inappropriately	49.2	28.3	< 0.001	22.0	10.5	< 0.001
Substance use
Smoked cigarettes	19.7	10.8	0.001	13.0	5.5	0.001
Used marijuana	18.5	11.1	0.006	16.6	6.2	< 0.001
Drank alcohol	52.1	36.1	< 0.001	61.0	39.4	< 0.001

The total sample size was 1,563. Age did not differ between groups that have had and not had suicidal ideation

In multivariable modeling, being a girl, sexting, cyberbullying, and alcohol drinking were significantly related to both depression and suicidal ideation (Table 5). Using X, weapon carrying, inappropriate touch to someone, and pornography consumption were associated with suicidal ideation only. The top three predictors of both depression and suicidal ideation were gender, cyberbullying, and sexting. The odds of having mental health issues were four-fold higher among girls than boys (depression: aOR = 4.60, 95% CI: 3.56–5.93; suicidal ideation: aOR = 4.20, 95% CI: 3.26–5.42). The odds were roughly two times higher among those who experienced cyberbullying and sexting than those who did not (aOR for cyberbullying: 2.24 and 2.05; aOR for sexting: 1.83 and 1.73). All the ORs were adjusted (i.e., each takes into account the contributions of the other variables in the model). We included school and classroom as random-effect variables; both were not significant (Covariance parameter estimates for classroom: estimate = 0.035; standard error = 0.041).

Table 5.

Generalized linear mixed model assessing the association of mental health with social media use patterns and risk behaviors among grade 9 students in the Bahamas

	Depression				Suicidal Ideation
	aOR	95% CI	t/2	p	aOR	95% CI	t/2	p
Age	1.06	0.93–1.20	0.9	0.37	1.02	0.94–1.11	0.52	0.61
Gender
Girls	4.60	3.56–5.93	11.74	< 0.001	4.20	3.26–5.42	11.1	< 0.0001
Boys	1.00				1.00
Sexting
Yes	1.83	1.37–2.44	4.12	< 0.001	1.73	1.32–2.26	3.95	< 0.001
No	1.00				1.00
Cyberbullying
Yes	2.24	1.76–2.85	6.60	< 0.001	2.05	1.63–2.58	6.12	< 0.0001
No	1.00				1.00
Using X (formerly Twitter)
Yes	1.11	0.85–1.44	0.77	0.44	1.31	1.02–1.68	2.08	0.04
No	1.00				1.00
Weapon carrying
Yes	1.28	0.87–1.87	1.26	0.21	1.64	1.13–2.38	2.61	0.01
No	1.00				1.00
Drinking
Yes	1.39	1.08–1.80	2.56	0.01	1.38	1.08–1.76	2.57	0.01
No	1.00				1.00
Inappropriate touch to someone
Yes	1.02	0.75–1.40	0.13	0.90	1.35	1.00-1.84	1.94	0.05
No	1.00				1.00
Pornography consumption
Yes	1.21	0.92–1.57	1.37	0.17	1.32	1.01–1.72	2.06	0.04
No	1.00				1.00

aOR Adjusted odd ratios, CI Confidence interval

Discussion

This study explored the status of SMU, risk behaviors, and mental health among adolescents and examined the association among them. We found prevalent SMU and online risk behaviors among adolescents. Moreover, their mental health was poor: at least half of our participants reported depression or suicidal ideation. Girls’ mental health status was worse than boys. Adolescents who used X, experienced online risk behaviors, and engaged in specific risk behaviors also had poorer mental health status than their counterparts.

Our findings are concerning. Among our study participants, the prevalence of both depressive symptoms (61.5%) and suicidal ideation (47.8%) far exceeded the respective global average of 34% and 20% among adolescents [51, 52]; this may be due to the socioeconomic disparities in The Bahamas, which are among the worst in the Caribbean [53]. These disparities were exacerbated by the impact of the COVID-19 pandemic since 2020. The economic deterioration affected many students not only by reducing family income but also by limiting access to necessities, coupled with a heightened level of insecurity about the future. However, minimal guidance and counseling were available to process these challenges, which may have further contributed to the poor mental health. Depression is the strongest predictor of suicidal ideation, as our study and other researchers have found, and it can lead to poorer adult mental health [54] and achievement [55]. Thus, the high rates of depressed mood and suicidal ideation observed in our study require immediate attention and further exploration to understand the contributing factors to these troubling statistics.

We found that boys and girls differ in SMU patterns, online risk behavior experiences, and mental health status. Girls, compared to boys, spent more time on social media, experienced more cyberbullying, and reported poorer mental health, aligning with previous research [25, 27]. Gender differences in online activities could also explain girls’ mental health [28], but our study was unable to identify the relationship between types of online activities and mental health. Perhaps this finding may be due to the content with which adolescents mainly engage, for which we did not collect in this study. Other researchers found that content could be inspirational [56] or could reinforce negative thoughts among adolescents [29]. Therefore, further research should comprehensively explore the interplay between content exposure, types of activities on social media, and mental health by gender.

Interestingly, we found that X was the only social media platform that was associated with poor mental health. A study previously found that text-based platforms like X may be less effective in reducing loneliness than image-based platforms, possibly because of a lack of intimacy [57]. However, X has also been studied as a platform through which users can openly discuss mental health problems [58] as a means of coping and finding a shared community without feeling judged [59]. It is also a feasible platform for spreading information about mental health [59] and could be used as an adjunct to more credible mental health resources [60]. Further, multiple algorithms for monitoring mental health based on X data have been explored [61], an approach that can be supported by our finding of an association between X use and poor mental health.

Consistent with existing studies [62, 63], our study found a significant association between online risk behaviors and adolescents’ mental health. The cyberbullying and sexting rates in our study were higher than those reported in a study that included Caribbean adolescents and a meta-analysis [18, 40]. The COVID-19 pandemic transformed social activities into online activities, increasing the chances of being exposed to online risk behaviors. Research indicates that online risk behaviors can result in various externalized and internalized problems, like delinquency, violence, depression, anxiety, and suicidal ideation and behavior [62, 64], with girls having a higher risk of developing suicidal ideation and behavior [65]. Our study also supports the strong impact of cyberbullying and sexting on mental health, which may exceed the influence of offline risk behaviors. This finding highlights the need for gender-specific programs for adolescents to increase awareness of online risk behaviors, improve coping skills, and support victims of cyberbullying and sexting.

The study findings have several implications for adolescents’ mental health in The Bahamas. First, a nationwide comprehensive program is needed to prevent and address mental health challenges among adolescents. While the Ministry of Health in The Bahamas provides counseling services for adolescents with psychosocial and behavioral challenges and their families [66], its primary focus remains on offline risk behaviors and depression, with limited attention to online risk behaviors. Additionally, the service is contingent on adolescents visiting the Adolescent Health Centre, limiting its reach. To effectively address the growing mental health needs of Bahamian adolescents, a more comprehensive approach that includes collaboration with schools and expands to online risk behaviors is necessary.

Strengths and limitations

To our knowledge, this is the first study in The Bahamas to comprehensively examine SMU, risk behaviors, and mental health with a large sample size. Students in our study attended public schools in New Providence, which makes up 75% of the population in The Bahamas. However, the findings may not be generalizable to students from private schools or the remote (more rural) Family Islands. Additionally, our study did not collect adolescents’ information, such as socioeconomic status and academic performance, that may influence adolescents’ motivations and patterns of using social media [67]. Also, our study evaluated lifetime suicidal ideation as a proxy for mental health. Therefore, we cannot know whether the suicidal ideation occurred only once in their earlier life or persisted. Questions on recent suicidal ideation and behaviors would be helpful in the future. Lastly, the cross-sectional nature of our study does not confirm causal relationships.

Conclusions

SMU and online risk behaviors are prevalent among adolescents and can negatively affect their risk behavior and mental health. Our study confirmed their widespread presence in The Bahamas, revealing significant gender differences. Girls spent more time on social media, engaged more in online interactions, were more vulnerable to cyberbullying, and had poorer mental health. High rates of suicidal ideation and the strong impact of online risk behavior experiences on mental health highlight the urgent need for programs that promote healthy online environments and provide tailored mental health support, especially addressing gender-specific risks. Future research efforts should prioritize developing targeted interventions to mitigate the adverse effects of SMU and online risk behaviors on adolescent mental health.

Abbreviations

SMU

Social media use

HIV

Human immunodeficiency virus infection

aOR

Adjusted odd ratios

CIs

Confidence intervals

Authors' contributions

DK and LD conceived the research idea. DK analyzed and interpreted the data. DK, RM, JK, and AW drafted the manuscript. MP and LD were involved in questionnaire development. RA and LD were fully involved in the data collection, entry, and validation process. All authors were involved in the revision of the paper for intellectual content. All authors read and approved the final manuscript.

Funding

The project was funded by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (R01HD095765).

Data availability

Data are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study was approved by UMass Chan Medical School Institutional Review Board and the Institutional Review Board of the Bahamas’ Princess Margaret Hospital, Public Hospitals Authority. Written informed consent was obtained from parents, and informed verbal assent was obtained from students prior to participation in the study. This study was conducted in compliance with the principles of the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.