Introduction

Globally, road traffic accidents have been a significant concern, resulting in numerous fatalities. Around 1.19 million people died in 2021 due to road accidents. Most of these accidents happen in Low- and middle-income countries, with 90% of deaths occurring there. In the Southeast Asia Region, things are even worse, with 28% of fatalities happening there. The number of deaths of people using motorcycles and similar vehicles has gone up by 30%. Although efforts are being made to make roads safer, they are still not meeting the goal set by the UN to halve road deaths by 2030 [1]. Helmet use represents the single most effective intervention for preventing motorcycle-related head injuries, reducing the risk of death by 42% and head injury by 69% (World Health Organization, 2023). Among various risky riding behaviors, non-helmet use accounts for the highest proportion of preventable fatalities in low- and middle-income countries. The selection of helmet use as our focal behavior is based on: (1) its proven efficacy as a protective measure, (2) its modifiability through behavioral interventions, and (3) the significant gap between knowledge and practice documented in prior Pakistani studies [2].

The use of motorcycles is rapidly increasing alongside a global rise in motorisation, especially in low- and middle-income countries (LMICS) where two-wheelers serve as an affordable mode of transport for low socioeconomic groups. Motorcycles provide no protection to the rider’s body, and a crash can lead to partial disability, permanent damage to vital organs, and fatalities in severe cases. In India and Pakistan, 73% of all motor vehicles are motorcycles and motorised three-wheelers [3].

Pakistan is a developing country that suffers tremendous economic and life losses from road crashes. The fatality rate (road deaths per 10 thousand registered vehicles) in Pakistan is among the highest in the world, with more than 27 thousand people dying of crashes every year while the number of crashes is continuously increasing. The death rate in crashes in Pakistan is 14.2 per 100,000 population annually, among which more than 50% of crashes involve motorcyclists [4].

In Pakistan, although the legal minimum age for obtaining a motorcycle license is 18 years, enforcement is weak, and motorcycle use among adolescents is prevalent. Many families rely on motorcycles as the primary mode of transport, and adolescents often begin riding before reaching legal age. This reality makes it crucial to understand safety behaviors and knowledge among this age group to develop appropriate interventions [5].

Karachi, Pakistan’s largest city and economic hub with over 20 million residents, presents a critical context for this research. Pakistan ranks among countries with the highest road traffic fatality rates globally (14.2 per 100,000 population), with motorcyclists comprising over 50% of these deaths [6]. Karachi has experienced a 150% increase in motorcycle registrations since 2010, yet helmet compliance remains below 15% [7]. Unlike other major Pakistani cities, Karachi’s unique combination of dense traffic, weak enforcement infrastructure, and socioeconomic diversity makes it an ideal site for examining barriers to helmet adoption across different population segments [7].

Secondary school students (ages 15–19) represent a critical intervention window for several reasons: (1) This age group experiences the highest rate of motorcycle-related injuries in Pakistan, with 35% of motorcycle crash victims being 15–19 years old; (2) Behavioral patterns and safety attitudes formed during adolescence tend to persist into adulthood, making this a crucial period for intervention; (3) School-based settings provide systematic access to this population for educational programs; (4) In Pakistan’s context, many adolescents begin riding before the legal age of 18 due to economic necessity and weak enforcement, making early intervention essential; (5) Prior research demonstrates that interventions targeting younger riders show greater behavioral plasticity compared to adult populations. Unlike college students or working youth, secondary school students can be reached through established educational infrastructure, allowing for scalable interventions [8].

The school environment and children’s learning abilities are often seen as promising opportunities for public health improvement efforts. It is well known that knowledge and behaviours adopted early in life tend to last into adulthood. Several health initiatives have been linked to positive changes in students’ health and well-being, such as increased physical activity and fitness [9], and efforts to prevent smoking and drug abuse, among others. However, recognising that injuries are the leading cause of death among the 10–19 age group [10], there is a lack of significant published data on programmes that address risk-taking behaviours or promote safety measures, and their effectiveness.

This study integrates the Health Belief Model (HBM) and Theory of Planned Behavior (TPB) to provide a comprehensive understanding of helmet use behaviors. While both frameworks address health behaviors, they offer complementary perspectives that together provide richer insights than either model alone. The HBM focuses on individual threat perception and cost-benefit analysis, examining how perceived susceptibility to injury, perceived severity of consequences, perceived benefits of helmet use, and perceived barriers influence behaviour [11]. However, HBM does not adequately account for social influences or volitional control—critical factors in adolescent decision-making. TPB addresses these limitations by incorporating subjective norms (peer and family influences), attitudes toward the behavior, and perceived behavioral control (self-efficacy and practical constraints). In the context of adolescent helmet use, where peer pressure and parental modeling significantly influence behavior, TPB’s social components are essential. The integration allows us to examine: (1) individual risk perceptions (HBM) alongside social influences (TPB); (2) practical barriers like cost and comfort (HBM barriers) in conjunction with perceived control over helmet access (TPB); and (3) both cognitive threat appraisal (HBM) and attitudinal factors (TPB). This combination has been successfully used in previous adolescent safety research and aligns with the socio-ecological model of health behavior, which recognizes that adolescent decisions result from both individual cognitions and social contexts.

Materials and methods

Results

Socio-demographic characteristics

A total of 502 students aged 15–19 years were interviewed. The mean age of the participants was 16.80 ± 1.20 years. The majority of the students belonged to the lower-middle class group, 63% (n = 319), followed by 30% (n = 151) to the lower class group, and only 6% (n = 32) belonged to the middle class (Table 1). Among the students, 25% (n = 125) reported owning a personal motorbike and riding it, representing the largest single category of motorcycle users. Regarding past experiences with accidents, 46% (n = 231) of students reported never having had an accident while riding, while 27% (n = 137) reported having an accident while not wearing a helmet (Table 1).

Download:

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

Table 1. Demographic Characteristics and Use of Motorbikes.

https://doi.org/10.1371/journal.pone.0340608.t001

Distribution of demographic and transport characteristics according to knowledge, HBM scores and TPB scores of children

Table 2 displays demographic and transport characteristics distribution based on participant knowledge, HBM scores, and TPB scores regarding road safety. The majority of individuals with lower HBM scores were from the lower-middle class (63%, n = 283) and the lower class (32%, n = 141), showing a strong association between these groups and their HBM scores (p value = 0.002). The study found no significant association between the participants’ knowledge and TPB scores with demographic variables. Regarding the different types of motorbike users, individuals who rode their bikes as passengers (31%, n = 139) and those who were owners or self-riders (28%, n = 123) had lower HBM scores than those in other groups. The type of motorcycle riders and HBM predictors were strongly associated (p-value <0.001). Furthermore, compared to those using 70cc motorbikes or who were unsure about the type of motorbike, 50% (n = 225) of those using 100cc motorbikes had lower HBM ratings, indicating a significant association between the range of motorcycles and HBM indicators. Similarly, a significant association was recorded between the range of motorcycles and TPB scores (p value = 0.016). Participants using 100cc motorbikes (n = 219, 44%) recorded the lowest TPB scores.

Download:

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

Table 2. Distribution of demographic and transport characteristics according to knowledge, HBM scores and TPB scores of children.

https://doi.org/10.1371/journal.pone.0340608.t002

Table 3 presents data on various characteristics and behaviours related to helmet usage among individuals, divided into groups with inadequate knowledge and adequate knowledge based on the Health Belief Model (HBM) and Theory of Planned Behaviour (TPB) scores. The participants who rarely wore helmets (48%, n=214), or never wore helmets (41%, n=182), while riding motorbikes, were more with lower HBM scores compared to participants with high HBM scores. However, a strong association was found between helmet use while riding a motorbike and HBM scores (p-value = 0.013). A significant association was found between passengers who never wore helmets and their HBM scores (p-value = 0.043). The participants who never wore helmets as passengers were more (47%, n=210), with low HBM scores compared to participants with high HBM scores. Regarding participants’ siblings’ helmet use for road safety, participants reported that their brothers rarely wore helmets (51%, n=192), and this was strongly associated with inadequate knowledge about road safety measures (p-value = 0.019). Participants’ friends’ helmet use for road safety was significantly associated with low HBM scores. Specifically, (51%, n=229) of participants reported that their friends rarely wore helmets, and (n=179, 40%) reported that their friends never wore helmets. (41%, n=208) of the participants with lower TPB scores reported that people rarely wore helmets with whom they sat for a ride on motorbikes, and this association was significant (p-value = 0.032). Interestingly, participants who scored inadequate knowledge and low HBM scores about road safety indicated a willingness to wear helmets while riding motorbikes next time, (37%) always or (42%) sometimes. This association was statistically significant (p-value<0.001)

Download:

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

Table 3. Distribution of type of helmet user according to Knowledge, HBM scores and TPB scores of children.

https://doi.org/10.1371/journal.pone.0340608.t003

Discussion

The study findings revealed a significant lack of awareness among the surveyed students regarding road safety rules and regulations, particularly regarding motorcycle helmet use. The strong association between lower socioeconomic status and reduced HBM scores reflects multiple interconnected mechanisms. First, economic constraints create tangible barriers. Helmet costs (PKR 500–2000, equivalent to 3–14 days of minimum wage income) represent substantial expenses for lower-income families, who must prioritise basic necessities. This aligns with previous research in Bangladesh and India, which shows cost as the primary barrier to helmet adoption among lower-SES populations [18,19]. Second, lower SES correlates with reduced access to safety education, as schools in lower-income areas typically lack road safety programs. Third, enforcement disparities exist, with traffic police focusing enforcement in wealthier neighborhoods, creating differential normative pressures. These findings underscore that helmet promotion cannot rely solely on education but must address structural economic barriers through subsidy programs or free helmet distribution [20,21].

Our finding that ‘siblings’ and ‘friends’ helmet behaviours significantly predict participants’ own practices aligns with Social Learning Theory and TPB’s subjective norms construct. Adolescents in collectivist cultures like Pakistan particularly rely on peer referents for behavioral guidance. The strong peer influence suggests that interventions targeting individual riders will have limited effectiveness without addressing social networks. Successful approaches might include peer education programs, visible role modeling by influential figures, and whole-school campaigns that shift normative perceptions.

The complementary insights from HBM and TPB validate our integrated approach. HBM effectively identified cognitive barriers (cost, discomfort) and threat perceptions, while TPB revealed the powerful role of social norms and perceived control. Interestingly, participants demonstrated awareness of helmet benefits (HBM perceived benefits) yet reported intentions not to wear helmets, with TPB variables (particularly subjective norms and perceived control) explaining this attitude-behavior gap. This suggests that threat awareness alone is insufficient; interventions must enhance social support and self-efficacy while reducing practical barriers.

Our findings parallel studies from Vietnam and Thailand showing socioeconomic gradients in helmet use [22] but contrast with Western contexts where enforcement is the primary driver. This suggests that interventions must be context-specific, addressing local economic realities and social structures rather than simply importing Western enforcement models.

Additionally, this study identified a strong association between the type of motorbike users and their HBM scores those who rode as passengers or were self-riders tended to have lower HBM scores compared to other groups. Furthermore, participants using 100cc motorbikes showed lower HBM (p < 0.001). and TPB scores (p = 0.016), suggesting a potential link between the type of motorcycle used and safety perceptions. This implies that riders of lower-powered motorcycles might perceive risks and safety differently compared to those using more powerful bikes. Different studies discussed the possible interactions between the type of driven motorcycle (different classes and/or engine capacities) and helmet usage [23]. In Skalkidou and colleagues’ study [24], there was a significant relationship between the engine size category and helmet usage (14.3% for engines less than 50 cc, 12.6% for 51–200 cc, 31.8% for 201–400 cc, and 32.5% for greater than 400 cc engines), in which an increase in the category of the engine had an OR of 1.52 for helmet usage (95% CI = 1.30–1.75, P = 0.0001).

Furthermore, students with inadequate knowledge and low HBM scores expressed a willingness to wear helmets in the future, suggesting the potential impact of educational campaigns on attitudes and safer practices. A similar study conducted in Iran concluded that, using the TPB, drivers with more positive attitudes were more likely to have intentions and perceptions of safe driving. The study also recommended that programs to improve attitudes should be designed to reduce road traffic injuries [25].

However, barriers such as perceived costs and comfort issues were reported, emphasising the importance of addressing these practical concerns to facilitate the widespread adoption of helmet-wearing behaviours. In Iribhogbe and Odai’s study, they found that many motorcyclists complain of the cost of helmets, while many had helmets but refused to wear them due to “inconveniencies” [26]. Despite overwhelming evidence that helmet use prevented head trauma [27]. Drawing comparisons with international studies focusing on youth helmet usage and road safety can provide valuable insights into best practices and effective strategies that have been implemented in other global contexts. Research from countries such as Australia, the United States, and European nations has highlighted successful interventions and policies that have increased helmet use among adolescents and young adults [28].

These findings endorse multi-level interventions: (1) Economic strategies like helmet subsidies or free distribution programmes targeting low-SES populations; (2) Educational initiatives: school-based programmes that fill knowledge gaps, build self-efficacy, and challenge peer norms; (3) Environmental strategies: better enforcement with education-first approaches rather than punitive measures that may unfairly burden low-SES riders; (4) Social marketing: campaigns featuring peer role models and tackling comfort and style concerns identified as barriers.

This study’s strengths include its theory-driven approach, large sample size, validated instruments, and inclusion of diverse socioeconomic groups. However, several limitations warrant consideration. First, the cross-sectional design precludes causal inference—we cannot determine whether low HBM scores lead to non-helmet use or whether non-use shapes cognitive appraisals. Longitudinal studies tracking behavior change are needed. Second, self-reported behaviors may be subject to social desirability bias, though we attempted to minimize this through trained interviewers and assured anonymity. Observational studies would strengthen findings. Third, interviewer-administered data collection, while necessary for ensuring comprehension across varying literacy levels, may have introduced interviewer effects despite training. Fourth, our sample from public schools in Karachi may not represent private school students or rural populations, limiting generalizability. Fifth, we did not assess helmet quality or proper usage, which affect actual protection. Sixth, unmeasured factors like traffic density exposure, previous near-misses, or family motorcycle crash history may confound observed associations. The inclusion of 42 non-riders (8.4%), while providing comparative insights, may have diluted effect sizes. Finally, recall bias may affect accident history reporting, though restricting recall to 12 months should minimize this issue.

Future studies should: (1) Conduct longitudinal research tracking behavior change over time; (2) Implement and evaluate theory-based interventions combining education, economic support, and social marketing; (3) Examine differential intervention effectiveness across socioeconomic groups; (4) Assess quality and proper use of helmets, not just ownership; (5) Explore family dynamics and parental modeling effects; (6) Investigate gender differences through studies including female riders and passengers.

Conclusion

This study of 502 adolescent students in Karachi revealed critical gaps between awareness and practice of helmet usage. While most students recognised the importance of helmets, only a minority consistently wore them. Lower socioeconomic status and specific motorcycle user types (passengers and self-riders) were associated with lower HBM scores, indicating disparities in safety behaviours. Social influences (siblings and peers) significantly affected helmet use decisions. Targeted interventions addressing socioeconomic barriers, peer norms, and practical concerns (cost, comfort) are needed. School-based education programmes combined with improved enforcement and community engagement can effectively promote helmet usage among adolescent motorcyclists in urban Pakistan.

Supporting information

Acknowledgments

We would like to thank all the participants who cooperated with us, without whom this study would not have been possible.