Skip to main content
NCBI home page
Health Promotion International logoLink to Health Promotion International
. 2026 Sep 28;41(5):daag144. doi: 10.1093/heapro/daag144

Beach safety as a family affair: exploring the role of parents on young adolescents’ knowledge, bathing risk perceptions, and behaviours

Jeoffrey Dehez 1,2,✉, Robert W Brander 3,4, Amy E Peden 5,6, Sandrine Lyser 7, Bruno Castelle 8,9
PMCID: PMC13616521  PMID: 42802102

Abstract

Beaches are popular destinations among young adolescents despite the safety risks they may pose. This study aimed to explore how young adolescents are influenced by parents at the beach and how parental presence impacts both their risk perception and their self-reported risk-taking behaviour. We conducted a cross-sectional survey among pupils aged 11–15 years (n = 416) in a public high school in south-western France and assessed their sociodemographic characteristics, water-based recreational use, beach safety knowledge, risk perceptions, and risky behaviours. Pupils were also asked about their likelihood of entering the water outside of lifeguard-supervised areas, as well as their perception of risk, both with and without parental supervision. To analyse pupils’ responses, we estimate two cumulative link mixed models with random intercepts for individuals. Two-thirds of pupils reported bathing outside supervised areas within the last 12 months. Just over half of the pupils were able to correctly define rip currents or supervised bathing areas. Our findings highlight the crucial but ambivalent influence of parents’ behaviours as parental presence decreased the odds of declaring higher perceived risks and increased the odds of declaring higher willingness to enter the water. Older pupils declared lower perceived risks compared to younger ones, and males were more likely to report entering the water outside of supervised areas than females. Parents constitute the primary information source on water safety for young adolescents. Therefore, beach safety prevention messages should attempt to reach the entire family unit.

Keywords: adolescence, beach safety, parental monitoring, drowning, risky behaviour

Contribution to Health Promotion.

  • Parents should encourage young adolescents to bathe in supervised areas and act as role models by doing so themselves.

  • Young adolescents should be provided with specific swimming and water safety skills relevant to natural water environments.

  • Parents must have access to relevant information about ocean water safety so that they can pass it to their children.

  • Prevention efforts must prioritize boys, particularly the older ones.

Introduction

Beaches are among the world’s most popular destinations for tourism and leisure (Ghermandi and Nunes 2013), offering a wide range of benefits for visitors (Britton et al. 2020). However, beaches come with inherent physical hazards associated with dynamic environmental conditions including large waves, strong currents, and variable water depths, and it is not uncommon for drowning and injury to occur (Castelle et al. 2018, Koon et al. 2021, 2023b). Among beachgoers, adolescents aged 10–19 years are one of the most at risk groups. Indeed, drowning is a leading cause of adolescent death worldwide (Peden et al. 2023). In high-income countries, drownings often occur during leisure activities in open waters, with beaches accounting for a significant proportion (Willcox-Pidgeon et al. 2017, Sakamoto et al. 2020). Within this context, there are various risk profiles depending on gender and age-specific factors. Males, in particular, consistently account for a disproportionate share of incidents and fatalities at the beach (Quan et al. 2011), a phenomenon generally attributed to factors such as lower aversion to risk and greater self-confidence in dealing with hazardous situations (Moran 2011, Willcox-Pidgeon et al. 2018). Age is another important risk factor as adolescence is a period during which adolescents’ perceptions of risk and one’s own vulnerability begin to change (Millstein and Halpern-Felsher 2001). It is therefore very common to observe an increase in drowning deaths among older adolescents (Sakamoto et al. 2020, Abihasira et al. 2023). In an effort to reduce the number of aquatic injuries and drowning deaths, supervision is often regarded as the primary preventive action (Leavy et al. 2016, WHO 2024). Such supervision may be provided by professional lifeguards or trained volunteers (Surf Life Saving Australia 2025).

Parents’ influence on preventing water-related risks among adolescents is largely overlooked, however. Typically, research on parental control has focused on the supervisory conditions for young children (Morrongiello et al. 2013, Cenderadewi et al. 2024). When it comes to adolescents, the emphasis generally shifts to studying individual knowledge, attitudes, and behaviours, especially among those aged 15 years and older (Moran 2009, Willcox-Pidgeon et al. 2018, Peden et al. 2024). In particular, no studies have examined the relationship between parental supervision and adolescent risk-taking behaviour in the context of ocean surf beaches. Parental supervision can reduce the risk among adolescents by encouraging them to voluntarily refrain from risky behaviour in a variety of situations (Li et al. 2000, Dittus et al. 2023, Ghanbari et al. 2025), but this relationship can become complicated when adolescents reach an age when they seek greater independence (Moran 2009). Moreover, very little is known about the beach safety knowledge and behaviour of young adolescents aged 11–15 years (Koon et al. 2023a). We believe that this period is important to study because it marks the beginning of an increase in the desire for autonomy and in risky behaviour (Millstein and Halpern-Felsher 2001, Daddis 2011).

In this study, we aim to address this important gap for drowning prevention by analysing parent’s influence on young adolescents’ beach safety knowledge, risk perception, and their self-reported risk-taking behaviour. We also examine the impact of gender and age on parental influence.

Method

Study area

Our analysis is based on data collected through a cross-sectional survey conducted in a public high school in south-western France. The study took place at a public high school in the municipality of Merignac located in the Bordeaux metropolitan area in south-western France (Fig. 1). The high school is located about 50 km (≈1 hour drive) from the nearest ocean beaches, which are among the most hazardous in France, if not Europe, particularly due to the presence of strong offshore-flowing rip currents (Castelle et al. 2016, 2018). According to Castelle et al. (2018), ∼40% of injuries in the surf zone within this area involve individuals aged between 10 and 20 years.

Figure 1.

For image description, please refer to the figure legend and surrounding text.

Open in a new tab

Location of the high school where the survey was conducted in regard to Western Europe (a) and the French coast (b).

The nearby ocean beaches are patrolled by lifeguards, mainly during the summer. Since 2022, lifeguard-supervised bathing areas are marked with yellow and red rectangular flags on the beach, in line with international guidelines that denote a safer and supervised bathing zone. Additionally, another coloured flag is hoisted at the lifeguard post to indicate whether bathing is dangerous or permitted. This flag comes in three colours: (i) green flag means that bathing is supervised with no particular danger; (ii) yellow-orange flag means that bathing is dangerous, but supervised; and (iii) red flag means that bathing is forbidden.

Questionnaire

The questionnaire included 25–29 questions, depending on pupils’ responses and took less than 10 min to complete. The questionnaire collected data on pupil demographics, self-reported swimming skills, beach recreational activities, beach safety knowledge, hazard identification, and risk perception. The full questionnaire is available in the Supplementary Table S1. The questionnaire was administrated in French and translated in English for publication. Data were numerical, ordinal, or categorical.

Pupils who reported visiting ocean beaches during the past 12 months were asked whether they had entered the water or not and whether they did so outside of lifeguard-supervised areas. Questions about beach safety knowledge, hazard identification, information sources, and hypothetical scenarios were presented to all pupils. Beach safety knowledge focused on three main topics: identifying supervised bathing areas, significance of the coloured flag on the flagpole at the lifeguard post, and the definition of a rip current. For all questions, pupils were asked to select the correct response from a range of options. Identification of main local hazards and information sources about local hazards were assessed using multiple-choice questions based on predesigned lists.

In order to analyse pupils’ risk perception and self-reported risk-taking behaviour in a variety of situations, two scenarios were presented in the questionnaire, each consisting of a set of two questions. The first asked, ‘Suppose you are at the beach by the ocean on a fine day and the waves are chest high. You are accompanied by friends as old as you who want to go for a swim. There is no one around and no lifeguards on duty. Your parents are not there either. What are the chances that you will go bathing?’. Five options were offered along a 5-point rating scale: ‘extremely unlikely’, ‘somewhat unlikely’,’ likely’, ‘somewhat likely’, and ‘extremely likely’. The second question asked the pupils to assess their perception of the risk of bathing in the given scenario. Again, answers were presented on a 5-point rating scale: ‘not at all risky’, ‘somewhat risky’, ‘moderately risky’, ‘very risky’, and ‘extremely risky’. The second scenario was similar with the only difference being that the pupil’s parents were nearby. The assessment of perceived bathing risk using a 5-point scale has previously been validated by Dehez et al. (2024) in an adult population and by Willcox-Pidgeon et al. (2018) in adolescents.

One teacher and a group of pupils were involved in designing the questionnaire and testing it. Involving pupils ensured that the questions were correctly interpreted by adolescents aged 11–15 years (Koon et al. 2023a). A first draft was presented to the pupil group who were invited to provide feedback. The final version was approved by the pupil group, who were then excluded from the survey to avoid biased responses. At every stage of the research, the teacher acted as a facilitator between the pupils and the researcher. The teacher was also responsible for organizing on-site administration of the survey.

Data collection

The survey collected responses between 5 May and 12 May 2025 in a dedicated classroom during school time. Pupils from every grade level from Year 6 (i.e. aged 11–12 years old) to Year 9 (i.e. aged 14–15 years old) were surveyed. Groups of about 15 pupils were formed to avoid pupils talking to each other. As it was not possible to survey all the pupils at the high school due to scheduling constraints, we used a convenience sampling method, aiming to interview as many pupils as possible. We sought to survey approximately the same number of pupils from each school year however. The surveys were administered using ODK Collect, a free and open-source field data collection software (Hartung et al. 2010). Pupils filled out questionnaires on laptops, which synchronized responses to a central database. All data were anonymous and securely stored on a password-protected network. Prior to undertaking the questionnaire, the teachers and some of the pupils involved in the survey design delivered a brief statement describing participant information including the rights of the participant.

The survey was not introduced from the perspective of safety or bathing risks so as not to create an overly anxiety-inducing atmosphere which might skew responses. Instead, the survey was introduced from the perspective of leisure activities and visiting beaches. Thereafter, completion of the questionnaire was considered as implied consent. No pupils had previously received any beach safety education within the school curriculum. No questions pertaining to personal information were included in the questionnaire to ensure it was General Data Protection Regulation (GDPR) compliant regarding the collection, processing, and storage of E.U. residents’ personal data. The project received ethical approval from the INRAE Research Ethics Committee (IRB00013805/IORG0011649-CEPR39a), and the high school staff validated that these conditions were in conformity with French regulations on adolescents’ consent.

Data analysis

To enable comparisons with other international studies, we used age rather than school level. We found two missing values for gender. As the number of cases was insufficient to consider it as a separate category, we excluded these observations whenever necessary in statistical tests. The pupils’ responses regarding the number of laps they declared they would be able to swim in the pool were converted into distance groupings. Responses regarding the three beach safety knowledge questions were recoded into a dichotomous variable (incorrect answer/correct answer). Responses regarding hazard identification and hazard information sources were also recoded in a way that each modality became a dichotomous variable (no/yes). Preliminary statistical analyses consisted of frequency calculations. χ² tests and Fisher’s exact tests were used to examine the relationship between two categorical variables, Mann–Whitney U tests were used to examine the relationships between an ordinal variable (e.g. distance swum in a pool), and categorical variables (e.g. gender) and univariate regression were used to examine the relationship between a categorical variable and a continuous variable (i.e. age).

In order to investigate the relationship between adolescents’ perceived risk, self-reported risk-taking behaviour, and parental supervision, we estimated two cumulative link mixed models (logit link) with random intercepts for individuals, considering perceived risk and self-reported risk-taking behaviour (i.e. propensity to enter the water) as dependent variables, respectively. The two responses regarding risk perception (i.e. without and with parental supervision) were treated as two observations for a repeated single question. We proceeded in the same way with the two responses regarding the likelihood of entering the water. To distinguish between the two scenarios presented to the pupil, we created a specific dichotomous variable (without/with parental supervision). Gender, age, parental supervision, and the interaction effects between parental supervision and both age and gender were considered as potential predictors in both of the models. All predictors have been introduced simultaneously in the models. Age was introduced as a normalized numerical variable. All statistical analyses were conducted in the R statistical programming environment (R Core Team 2025). The two cumulative link mixed models with random intercepts for individuals were estimated with the ordinal package (Christensen 2023). We only consider statistical relationships that are statistically significant at the 0.05 level or higher.

Results

Sample

Swimming ability

A total of 416 pupils completed the questionnaire with an even distribution of males (50.7%) and females (49.3%). Sample mean age was 12.5 years (median = 13, SD = 1.24). Table 1 summarizes the results of responses related to swimming ability and gender.

Table 1.

Pupils’ responses to questions relating to swimming ability.

Female Male Total P-value Age
n (%) n (%) n (%) AOR [95% CI]
Has attended swimming lessons 136 (66.7) 138 (65.7) 274 (66.2) 0.919 1.10 [0.90–1.36]
Current or former member of sports swimming club 59 (28.9) 58 (27.6) 117 (28.3) 0.919 1.08 [0.87–1.34]
Self-assessed maximum distance swum in a pool – – – <0.0001*** 1.09 [0.92–1.30]
 <50 m 41 (20.1) 20 (9.5) 61 (14.7) – –
 50–100 m 57 (27.9) 49 (23.3) 106 (26.5) – –
 100–200 m 50 (24.5) 35 (16.7) 85 (20.5) – –
 200–400 m 27 (13.2) 51 (24.3) 78 (18.8) – –
 >400 m 29 (14.2) 55 (26.2) 84 (20.3) – –
Self-assessed capability to swim the same distance in the ocean 76 (37.3) 104 (49.5) 180 (43.5) 0.015** 0.93 [0.76–1.13]
Open in a new tab

*** P < 0.001; **P < 0.01; *P < 0.05.

About two-thirds of pupils (66.2%) declared having had swimming lessons outside of school time, but only 28.3% are, or used to be, regular members of swimming sports clubs (Table 1). Slightly over half (59.6%) reported being able to swim >100 m in a swimming pool without touching the bottom or having rest. Statistically significant gender differences exist with regard to self-assessed distance swum in a pool (W = 16060, P < 0.0001) and self-assessed capability to swim the same distance in the ocean (χ(1) = 5.848, P = 0.015). Compared to males (9.5% Table 1), more females (20.1%) declared they were able to swim less than 50 m without stopping or touching the bottom. Conversely, more males (24.3%) declared that they were able to swim between 200 and 400 m without stopping compared to females (13.2%) and more males (26.2%) declared they were able to swim more than 400 m without stopping compared to females (14.2%). Compared to females (37.3%), more men (49.5%) declared being able to swim the same distance in a swimming pool and in the ocean. No statistically significant influence of age was found over each of the four aforementioned variables.

Beach-related recreational activities and beach safety knowledge

Table 2 describes results for pupils’ declared beach-related recreational activities and declared beach safety knowledge. Almost all pupils (92.3%) stated they had visited local ocean beaches at least once in the past 12 months. One in five (19.1%) declared they engaged in a recreational activity related to ocean beaches (e.g. surfing, bodyboarding, sailing). Almost all pupils who declared having been to the ocean in last 12 months said they bathed (97.6%), and 68.1% declared that they had entered the water outside of supervised areas at least once. A statistically significant influence of age was observed on the likelihood of reporting entering the water outside supervised areas at least once during the past 12 months (AOR = 1.27, 95% CI [1.02–1.59]), suggesting older pupils have higher odds of having bathed outside supervised areas during the last 12 months than younger pupils.

Table 2.

Pupils’ declared beach-related recreational activities and declared beach safety knowledge.

Females Males Total P-value Age
n (%) n (%) n (%) AOR [95% CI]
Declared beach-related recreational activities
 Has been to ocean local beaches during past 12 months 190 (93.1) 192 (91.4) 382 (92.3) 0.640 1.42 [0.98–2.09]
 Bathe once at the ocean 183 (96.3) 190 (99.0) 373 (97.6) 0.103 0.64 [1.31–1.26]
 Has bathed outside of supervised areas during past 12 months 120 (63.2) 140 (72.9) 260 (68.1) 0.053 1.27 [1.02–1.59]*
 Practice ocean-based recreational activity 41 (20.1) 38 (18.1) 79 (19.1) 0.694 1.01 [0.79–1.30]
Declared beach safety knowledge
 Correct identification of supervised bathing areas 105 (51.5) 110 (52.4) 215 (51.9) 0.930 1.07 [0.89–1.31]
 Correct definition of the yellow-orange flag 116 (56.9) 96 (45.7) 212 (51.2) 0.029* 1.01 [0.84–1.23]
 Correct definition of rip current 113 (55.4) 129 (61.4) 242 (58.5) 0.251 1.49 [1.22–1.82]***
Open in a new tab

*** P < 0.001; **P < 0.01; *P < 0.05.

Approximately half of the pupils (51.9% and 51.2%, respectively) were correct in identifying supervised areas (i.e. ‘with red and yellow flags on the beach’) and interpreting the meaning of the yellow-orange flag (i.e. ‘swimming is dangerous but supervised’). Slightly more (58.5%) provided a correct answer for defining a rip current. Regarding gender, a statistically significant relationship was only found for the meaning of the yellow/orange flag (χ(1) = 4.710, P = 0.029) with more females (56.9%) providing the correct definition compared with males (45.7%). A statistical influence of age on the probability of providing the correct answer on the definition of a rip current was also found (AOR = 1.49, 95% CI [1.22–1.82]) with older pupils having higher odds of selecting the correct answer for rip current than younger ones.

Ocean beaches hazards identification and sources of information

Table 3 presents descriptive statistics relating to the pupils’ responses for hazard identification and source of information about ocean hazards. In response to the various options proposed to them in the questionnaire, pupils selected ‘rip currents’ as the main hazard (63.8%) followed by ‘waves’ (36.2%) and ‘tide’ (17.1%). A statistically significant difference was observed between gender and hazard (χ(1) = 8.68, P = 0.002) with rip currents chosen by 56.7% of males and 71.1% of females, respectively, and ‘waves’ (χ(1) = 7.72, P = 0.005) chosen by 29.5% of males and 43.1% of females, respectively. No statistically significant influence was observed between age and hazard selection.

Table 3.

Pupils’ selected ocean beaches hazards and sources of information about ocean hazards.

Females Males Total P-value Age
n (%) n (%) n (%) AOR [95% CI]
Selected ocean beach hazards
 Rip currents 145 (71.1) 119 (56.7) 264 (63.8) 0.003** 1.14 [0.93–1.40]
 Waves 88 (43.1) 62 (29.5) 150 (36.2) 0.005** 0.82 [0.67–1.01]
 Tide 38 (18.6) 33 (15.7) 71 (17.1) 0.512 0.91 [0.70–1.18]
 Jellyfishes 18 (8.8) 32 (15.2) 50 (12.1) 0.064 1.14 [0.85–1.54]
 Rocks 19 (9.3) 21 (10.0) 40 (9.7) 0.944 0.93 [0.69–1.33]
 Sharks 12 (5.9) 4 (1.9) 16 (3.9) 0.065 0.63 [0.36–1.06]
 Sun 1 (0.5) 2 (1.0) 3 (0.7) >0.9 1.08 [0.33–3.57]
Sources of information
 Parents or family 120 (58.8) 115 (54.8) 235 (56.8) 0.462 0.88 [0.72–1.06]
 TV, radio 40 (19.6) 45 (21.4) 85 (20.5) 0.736 1.30 [1.02–1.65]*
 Internet and social network 45 (22.1) 40 (19.0) 85 (20.5) 0.524 1.28 [1.01–1.64]*
 School 37 (18.1) 34 (16.2) 71 (17.1) 0.692 1.32 [1.02–1.72]*
 Newspapers 15 (7.4) 17 (8.1) 32 (7.7) 0.921 1.04 [0.72–1.48]
 Friends 10 (4.9) 22 (10.5) 32 (7.7) 0.052 0.88 [0.60–1.26]
 Sports club 12 (5.9) 16 (7.6) 28 (6.8) 0.611 0.91 [0.61–1.33]
Open in a new tab

***P < 0.001; **P < 0.01; *P < 0.05.

Pupils were then asked where they had received previous information on ocean hazards with the main source being from ‘parents or family’ (56.8%), followed by ‘TV, radio’, or ‘internet and social networks’ (20.5% each; Table 4). No statistically significant differences were observed by gender. Age has a statistically significant influence on the likelihood of choosing the following three sources of information, i.e. ‘TV, radio’ (AOR = 1.30, 95% CI [1.02–1.65]), ‘internet and social network’ (AOR = 1.28, 95% CI [1.01–1.64]), and ‘school’ (AOR = 1.32, 95% CI [1.02–1.72]). In each case, the older the pupil, the higher their odds of choosing each of these three sources of information.

Table 4.

Adjusted odds ratio and 95% confidence interval (CI) for each parameter of the models.

Parameters AOR 95% CI lower bound 95% CI upper bound Signif.
Perceived risks equation models
 Parents supervising (reference: parent not supervising) 0.19 0.13 0.29 ***
 Gender (reference: female) 0.92 0.56 1.52 –
 Age 0.75 0.58 0.97 *
 Parents supervising × gender 0.79 0.46 1.34 –
 Parents supervising × age 1.37 1.05 1.79 *
LogLik −1112.38 – – –
AIC 2244.76 – – –
Self-reported risk-taking behaviour equation model
 Parents supervising (reference: parent not supervising) 8.43 5.47 12.99 ***
 Gender (reference: female) 2.06 1.17 3.60 *
 Age 1.82 1.37 2.42 ***
 Parents supervising × gender 1.01 0.58 1.74
 Parents supervising × age 0.63 0.48 0.83 **
LogLik −1198.06 – – –
AIC 2416.11 – – –
Open in a new tab

*** P < 0.001; **P < 0.01; *P < 0.05.

Perceived risk and risk-taking behaviour

Finally, the parameters of the model designed to analyse pupils’ responses regarding their perceived risk and their self-reported risk-taking behaviour are summarized in Table 4. In the perceived risk models, the fixed effect of parental supervision was large and clearly negative (AOR = 0.19, 95% CI [0.13–0.29]). Consequently, parental presence decreased the odds of young adolescents declaring higher perceived risks by a factor of five. Similar conclusions hold for age as the effect was negative and moderate (AOR = 0.75, 95% CI [0.58–0.97]). The older the pupils were, the lower they perceived the risks. There were no meaningful differences in risk perception between males and females. Interaction terms between scenario (i.e. without and with parental supervision) and age had a positive effect (AOR = 1.37, 95% CI [1.05–1.79]) suggesting that the influence of parental supervision becomes weaker for older pupils. In other words, the decrease in risk perception associated with parental supervision is lower for older pupils. In contrast, younger pupils were more sensitive to parental supervision at the beach. Notably, the variance of the individual random intercept is large (σ2 = 2.95) indicating substantial unobserved heterogeneity across individuals.

Regarding the components of the model for pupil self-reported risk-taking behaviour, fixed effect associated with parental supervision was also large, but clearly positive (AOR = 8.43, 95% CI [5.47–12.99]) suggesting that pupils’ odds of declaring higher willingness to bathe increased by a factor greater than eight if pupils were informed that their parents would be at the beach with them. Both parameters associated with gender (AOR = 2.06, 95% CI [1.17–3.60]) and age (AOR = 1.82, 95% CI [1.37–2.42]) were clearly positive indicating that males and older pupils reported higher propensity to enter the water. Similar to what we observed in the previous model, only the interaction term parameter between parental supervision and age had a significantly statistical influence (AOR = 0.63, 95% CI [0.48–0.83]) suggesting that older pupils were less sensitive to the presence of their parents when it comes to entering the water as compared with younger ones. Once again, we did not identify any significant statistical influence of interaction term between parental supervision and gender. As in the previous model, the variance of the individual random intercept is large (σ2 = 4.30) indicating substantial unobserved heterogeneity across individuals.

Discussion

Parents have a vital role to play in reducing young adolescents’ vulnerability to injury. This is certainly the case in the area of beach safety, where such supervision has proven to be an effective strategy in most situations (Leavy et al. 2016). However, no research had previously been conducted into how young adolescents (i.e. those aged 11–15 years) are influenced by parents and how parental presence impacts both their risk perception and their self-reported risk-taking behaviour. To address this gap, we conducted a survey among pupils in a public school in south-western France located about an hour’s drive from high energy ocean beaches characterized by hazards such as rip currents. We found that these young adolescents were highly exposed to ocean hazards with 92.3% having visited ocean beaches at least once in the last 12 months. Of these, 97.6% reported having bathed and 68.1% reported having bathed outside lifeguard-supervised areas. We believe our findings have several important implications in beach safety, but also more generally in the areas of drowning prevention.

First, we confirm that young adolescents have only a fragmented understanding of beach safety. In our sample, pupils felt that rip currents are something to pay attention to at the beach, as they ranked them first among all potential hazards, but only 58.5% were able to correctly define them. It is worth noting that public awareness of the rip current hazard is widespread among the local adult population, largely due to the long-standing public information campaigns (Dehez and Lyser 2024) and most reported feeling ‘anxious’ or ‘very anxious’ about the hazard (Dehez et al. 2024). Given that parents and family remain the primary source of information about beach hazards for the pupils in our study, it is likely that parents have shared their concerns about rip currents to their children. A similar scenario likely applies to waves, which pupils cited as the second-most common hazard, and which adults surveyed in the same region also often identified as a major hazard (Dehez et al. 2024, 2025). As Koon et al. (2023a) pointed out in a survey carried out in New South Wales, Australia, there is a good chance that parents decide which beach and location on the beach relative to the supervised areas to go to once they arrive. By doing this, parents do not give their children the chance to question the meaning of the flags for themselves. This might help explain the incorrect identification of supervised bathing areas by a large proportion of young adolescents in our study.

Second, we observed that parental supervision had an influence on both young adolescents’ perception of risk and their intention to enter the water outside lifeguard-supervised areas. Not only does parental supervision lower the level of perceived risk among young adolescents, it also tends to increase their likelihood of entering the water in an unsupervised area. Unlike what has been observed with other adolescent behaviours (Li et al. 2000, Ghanbari et al. 2025), parental supervision might not decrease the propensity of engaging in risky behaviours at the beach. We believe that this relationship highlights a distinctive challenge for beach safety—that water activities, regardless of lifeguard presence, are perceived to be inherently enjoyable rather than risky, by both adolescents and parents (Gstaettner et al. 2017, Oliver et al. 2023). This sense of complacency may be reinforced if adolescents have previously observed their parents entering the water outside supervised areas. Indeed, such behaviour is very widespread among adults in south-western France (Dehez and Lyser 2024). By doing so, parents may unknowingly transmit risky behaviour to their children.

Third, we noted that the above relationships are influenced by the respondents’ gender and age. Overall, our findings regarding gender-related effects are consistent with those in the literature, which identify that young males often declare engaging in risky behaviour at the beach more than young females due in part to an excess of self-confidence or sensation seeking (Moran 2011, Sakamoto et al. 2020, Lawes et al. 2021). Our results also emphasized the critical influence of age on pupils’ safety at beaches as older pupils perceiving risks to be lower and being more likely to report entering the water at unsupervised beaches than younger ones. These findings are consistent with documented observations that adolescents’ risk perceptions gradually decrease and converge with those of adults as they get older (Millstein and Halpern-Felsher 2001, Blankenstein et al. 2024). The results of our regression models allow us to identify a two-fold influence of age. On the one hand, age directly affects risk perception and self-reported risk-taking behaviour, and on the other hand, age lowers the influence of parental supervision on both of these outcomes. In the models, the estimated value for the interaction terms between parental supervision and age is consistent with the literature-documented phenomenon of adolescents’ ‘desire for autonomy’ (Daddis 2011). In the present case, this induced additional effect of age may reduce the risk among older pupils who, being less sensitive to parents’ influence, would be less likely to underestimate the risk and less likely to enter the water. However, based on odds ratio values, such an impact appears to be of limited magnitude in the final decision-making process compared with other variables (e.g. fixed effect of parental presence or age). Further research is needed on this finding. Statistical differences regarding information sources were slightly pronounced with the older adolescents tending to receive beach safety information more often on TV and radio, internet and social networks, but also at school.

We are aware of the limitations of this study and how they may be addressed in future research. We acknowledge that the results can only be extended beyond this particular cohort of pupils with caution as they are not necessarily representative of the entire age group population in France or elsewhere. For this reason, it would be useful to replicate the study in other geographical areas including those further away from the ocean and in less affluent locations, given known impacts of social determinants of health and cultural norms on drowning risk (Leavy et al. 2024, Willcox-Pidgeon et al. 2025). Increasing the sample size might make it possible to identify effects that are smaller in magnitude that our current sample was too small to detect. We are also aware of traditional limitations of declarative responses that can be influenced by ‘social desirability bias’ among pupils who may have provided answers they perceived to be socially acceptable rather than fully reflecting their actual attitudes or behaviours (Zaal et al. 2026). A further limitation stems from the use of hypothetical scenarios, which allow for more refined control of the situations to which participants are exposed, but can skip the potential affective component of risk assessment that could be grasped in ‘real-world’ situations (Wolff et al. 2019). Nevertheless, involving a group of pupils in the design of the questionnaire helped to make the material more relevant to young adolescents (Koon et al. 2023a). We are also limited by the use of cross-sectional data, particularly when studying the impact of age. Longitudinal data would help to address this issue.

By highlighting the various ways in which parents influence the knowledge and behaviour of young adolescents regarding ocean-bathing hazards, we believe that our findings could have important implications for beach safety and health promotion in general. Parents’ influence comes first through their own behaviour. We have shown that the parents’ presence can act as an unintentional incentive for young adolescents to bathe outside supervised areas. This tendency can be further accentuated if parents themselves bathe at unsupervised beaches. Parents must be aware that their own supervision can hardly replace the supervision provided by professional lifeguards. Indeed, bystander rescuers, often parents, can drown whilst trying to rescue a child in the ocean (Lawes et al. 2020). Parental supervision may also be incomplete. In a study based on a sample of surfing and non-surfing beaches in New Zealand, Moran (2010) found that 29% of people observed did not simultaneously meet the supervisory criteria of attention, proximity, and continuity (Saluja et al. 2004). Among those who had failed to simultaneously meet the criteria, one-third lay on the beach sunbathing, one-quarter talked to others, and one-quarter used cell phones. Consequently, parents should encourage their young adolescents to bathe in supervised areas and act as role models by doing so themselves. Lack of vigilance may be fostered by the belief among parents that the need to supervise children decreases as they learn to swim (Morrongiello et al. 2013, 2014). In our study, it appears that parents prioritized having their children learn basic swimming skills as two-thirds of the pupils reported having taken swimming course outside of school. This misconception undermines the important difference between swimming skills and water safety knowledge (Petrass and Blitvich 2014). Therefore, young adolescents should be provided with specific swimming and water safety skills relevant to natural water environments. We also have shown that parents serve as a key source of information for young adolescents regarding ocean-bathing hazards. In this context, there is a clear need to treat adolescent beach safety as a family affair, involving both parents and adolescents equally. Parents must have access to relevant information about ocean water safety so that they can pass it to their children, particularly young males.

Conclusion

Surf beaches are popular destinations for young adolescents despite the safety risks they may pose. It is therefore crucial to ensure young adolescents can enjoy them under the safest possible conditions. Whilst parental supervision remains a critical component of drowning prevention, we show that the influence of parents must be viewed in a broader perspective. Indeed, we show that parents may unwittingly influence young adolescents to take risks through their own behaviour and that they represent an essential source of information on ocean hazards and beaches safety for them. Young adolescents’ knowledge and specific ocean-related skills still need to be improved. Measures should not only target young adolescents but should be designed to involve the whole family.

Supplementary Material

daag144_Supplementary_Data
daag144_supplementary_data.docx (21.8KB, docx)

Acknowledgements

We wish to sincerely thank the study participants and the teachers for their valuable time and insights provided for this study.

Contributor Information

Jeoffrey Dehez, ETTIS, INRAE Nouvelle Aquitaine Bordeaux, 50 avenue de Verdun Gazinet, Cestas F-33612, France; UNSW Beach Safety Research Group, UNSW Sydney, Kensington, NSW 2052, Australia.

Robert W Brander, UNSW Beach Safety Research Group, UNSW Sydney, Kensington, NSW 2052, Australia; School of Biological, Earth and Environmental Sciences, Kensington, UNSW Sydney, NSW 2052, Australia.

Amy E Peden, UNSW Beach Safety Research Group, UNSW Sydney, Kensington, NSW 2052, Australia; School of Population Health, UNSW Sydney, Kensington, NSW 2052, Australia.

Sandrine Lyser, ETTIS, INRAE Nouvelle Aquitaine Bordeaux, 50 avenue de Verdun Gazinet, Cestas F-33612, France.

Bruno Castelle, UNSW Beach Safety Research Group, UNSW Sydney, Kensington, NSW 2052, Australia; CNRS, University of Bordeaux, Bordeaux INP, EPOC, UMR 5805, Allée Geoffroy Saint-Hilaire, Pessac F-33600, France.

Author contributions

Jeoffrey Dehez (Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Supervision, Writing—original draft), Robert W. Brander (Conceptualization, Formal analysis, Writing—original draft, Writing—review & editing), Amy E. Peden (Conceptualization, Formal analysis, Writing—original draft, Writing—review & editing), Sandrine Lyser (Methodology, Writing—original draft, Writing—review & editing), and Bruno Castelle (Funding acquisition, Writing—original draft, Writing—review & editing)

Supplementary material

Supplementary material is available at Health Promotion International online.

Conflicts of interest

None declared.

Funding

This work was supported by Region Nouvelle Aquitaine Council (PSGAR CORALI) and Agence Nationale de la Recherche (ANR-22-EXIR-0004-Project IRICOT/PEPR IRIMA).

Data availability

The anonymized raw survey dataset is available from the corresponding author upon reasonable request.

References

  1. Abihasira  S, Moran  DS, Orr  D  et al.  Drowning rates among children and adolescents (aged 7–17) in Israel during year 2008–2018. BMC Public Health  2023;23:1695. 10.1186/s12889-023-16671-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Blankenstein  NE, van Hoorn  J, Dekkers  TJ  et al.  Adolescent risk-taking likelihood, risk perceptions, and benefit perceptions across domains. Pers Individ Dif  2024;231:112806. 10.1016/j.paid.2024.112806 [DOI] [Google Scholar]
  3. Britton  E, Kindermann  G, Domegan  C  et al.  Blue care: a systematic review of blue spaces interventions for health and wellbeing. Health Promot Int  2020;35:50–69. 10.1093/heapro/day103 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Castelle  B, Brander  R, Tellier  E  et al.  Surf zone hazards and injuries on beaches in SW France. Nat Hazards  2018;93:1317–35. 10.1007/s11069-018-3354-4 [DOI] [Google Scholar]
  5. Castelle  B, Scott  T, Brander  R  et al.  Rip current types, circulation and hazard. Earth-Sci Rev  2016;163:1–21. 10.1016/j.earscirev.2016.09.008 [DOI] [Google Scholar]
  6. Cenderadewi  M, Franklin  RC, Fathana  PB  et al.  Child drowning in Indonesia: insights from parental and community perspectives and practices. Health Promot Int  2024;39:daae113. 10.1093/heapro/daae113 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Christensen  R. H. B. 2023. Ordinal regression models for ordinal data (R package version 2023.12-4.1). https://CRAN.R-project.org/package=ordinal
  8. Daddis  C. Desire for increased autonomy and adolescents’ perceptions of peer autonomy: “Everyone Else Can; Why Can’t I?”. Child Dev  2011;82:1310–26. 10.1111/j.1467-8624.2011.01587.x [DOI] [PubMed] [Google Scholar]
  9. Dehez  J, Lyser  S. How ocean beach recreational quality fits with safety issues? An analysis of risky behaviours in France. J Outdoor Recreat Tour  2024;45:100711. 10.1016/j.jort.2023.100711 [DOI] [Google Scholar]
  10. Dehez  J, Lyser  S, Castelle  B  et al.  Investigating beachgoer's perception of coastal bathing risks in southwest France. Nat Hazards  2024;120:13209–30. 10.1007/s11069-024-06715-w [DOI] [Google Scholar]
  11. Dehez  J, Lyser  S, Castelle  B. Predicting individual's decision to enter the water at a high-energy recreational surf beach in France. Inj Prev  2025. 10.1136/ip-2024-045574. Epub ahead of print. [DOI] [PubMed] [Google Scholar]
  12. Dittus  PJ, Jingjing  L, Verlenden  JV  et al.  Parental monitoring and risk behaviors and experiences among high school students—youth risk behavior survey, United States, 2021. MMWR Suppl  2023;72:37–44. 10.15585/mmwr.su7201a5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Ghanbari  A, Howard  M, Cavanaugh  J  et al.  Teen perceptions of parental monitoring and its impact on their risky road behavior: an analysis of the national youth risk behavior survey. J Safety Res  2025;93:342–7. 10.1016/j.jsr.2025.03.007 [DOI] [PubMed] [Google Scholar]
  14. Ghermandi  A, Nunes  PALD. A global map of coastal recreation values: results from a spatially explicit meta-analysis. Ecol Econ  2013;86:1–15. 10.1016/j.ecolecon.2012.11.006 [DOI] [Google Scholar]
  15. Gstaettner  AM, Rodger  K, Lee  D. Visitors perspectives of risk management in natural tourism setting: an application of the theory of planned behaviour. J Outdoor Recreat Tour  2017;19:1–10. 10.1016/j.jort.2017.04.001 [DOI] [Google Scholar]
  16. Hartung  C, Lerer  A, Anokwa  Y  et al.  Open data kit: tools to build information services for developing regions. Paper presented at the ICTD ‘10: Proceedings of the 4th ACM/IEEE International Conference on Information and Communication Technologies and Development, London, UK. 2010, December 13–16. 2010.
  17. Koon  W, Brander  R, Alonzo  D  et al.  Lessons learned from co-designing a high school beach safety education program with lifeguards and students. Health Promot J Austr  2023a;34:222–31. 10.1002/hpja.664 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Koon  W, Brander  RW, Dusek  G  et al.  Relationships between the tide and fatal drowning at surf beaches in New South Wales, Australia: implications for coastal safety management and practice. Ocean Coast Manag  2023b;238:106584. 10.1016/j.ocecoaman.2023.106584 [DOI] [Google Scholar]
  19. Koon  W, Peden  A, Lawes  JC  et al.  Coastal drowning: a scoping review of burden, risk factors, and prevention strategies. PLoS One  2021;16:e0246034. 10.1371/journal.pone.0246034 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Lawes  J, Ellis  A, Daw  S  et al.  Risky business: a 15-year analysis of fatal coastal drowning of young male adults in Australia. Inj Prev  2021;27:442–9. 10.1136/injuryprev-2020-043969 [DOI] [PubMed] [Google Scholar]
  21. Lawes  JC, Rijksen  EJT, Brander  RW  et al.  Dying to help: fatal bystander rescues in Australian coastal environments. PLoS One  2020;15:e0238317. 10.1371/journal.pone.0238317 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Leavy  JE, Crawford  G, Leaversuch  F  et al.  A review of drowning prevention interventions for children and young people in high, low and middle income countries. J Community Health  2016;41:424–41. 10.1007/s10900-015-0105-2 [DOI] [PubMed] [Google Scholar]
  23. Leavy  JE, Crawford  G, Scarr  J-P  et al.  Drowning prevention: a global health promotion imperative, now more than ever. Health Promot J Austr  2024;35:860–3. 10.1002/hpja.830 [DOI] [PubMed] [Google Scholar]
  24. Li  X, Stanton  B, Feigelman  S. Impact of perceived parental monitoring on adolescent risk behavior over 4 years. J Adolesc Health  2000;27:49–56. 10.1016/S1054-139X(00)00092-6 [DOI] [PubMed] [Google Scholar]
  25. Millstein  SG, Halpern-Felsher  BL. Perceptions of risk and vulnerability. In: Fischhoff  B, Nightingale  EO, Innotta  JG (eds.), Adolescent Risk and Vulnerability. Concept and Measurement. Washington, DC: National Academy Press, 2001, 15–49. [Google Scholar]
  26. Moran  K. Parents, pals, or pedagogues? How youth learn about water safety. Int J Aquat Res Educ  2009;3:article 4. 10.25035/ijare.03.02.04 [DOI] [Google Scholar]
  27. Moran  K. Watching parents, watching kids: water safety supervision of young children at the beach. Int J Aquat Res Educ  2010;4:269–77. 10.25035/ijare.04.03.06 [DOI] [Google Scholar]
  28. Moran  K. (Young) men behaving badly: dangerous masculinities and the risk of drowning in aquatic leisure activities. J Leis Res  2011;14:260–72. 10.1080/11745398.2011.615719 [DOI] [Google Scholar]
  29. Morrongiello  BA, Sandomierski  M, Schwebel  DC  et al.  Are parents just treading water? The impact of participation in swim lessons on parents’ judgments of children's drowning risk, swimming ability, and supervision needs. Accid Anal Prev  2013;50:1169–75. 10.1016/j.aap.2012.09.008 [DOI] [PubMed] [Google Scholar]
  30. Morrongiello  BA, Sandomierski  M, Spence  JR. Changes over the swim lessons in parents’ perceptions of children's supervision needs in drowning risk situations: “His swimming has improved so now he can keep himself safe”. Health Psychol  2014;33:608–15. 10.1037/a0033881 [DOI] [PubMed] [Google Scholar]
  31. Oliver  DM, McDougall  CW, Robertson  T  et al.  Self-reported benefits and risks of open water swimming to health, wellbeing and the environment: cross-sectional evidence from a survey to Scottish swimmers. PLoS One  2023;18:e0290834. 10.1371/journal.pone.0290834 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Peden  AE, Minniti  S, Isin  A  et al.  Unintentional injuries in adolescents: a neglected issue in global health. Lancet Child Adolesc Health  2023;7:447–9. 10.1016/S2352-4642(23)00134-7 [DOI] [PubMed] [Google Scholar]
  33. Peden  AE, Williamson  S, Fozard  F  et al.  The effect of ‘Swim Repear’ program on water safety awareness, drowing mortality and morbidity among males aged 15–34 years in Aotearoa, New Zealand. J Safety Res  2024;88:190–8. 10.1016/j.jsr.2023.11.006 [DOI] [PubMed] [Google Scholar]
  34. Petrass  LA, Blitvich  JD. Preventing adolescent drowning: understanding water safety knowledge, attitudes and swimming ability. The effect of a short water safety intervention. Accid Anal Prev  2014;70:188–94. 10.1016/j.aap.2014.04.006 [DOI] [PubMed] [Google Scholar]
  35. Quan  L, Pilkey  D, Gomez  A  et al.  Analysis of paediactric drowning deaths in Washington state using deat review (CDR) for surveillance: what CDR does and does not tell us about lethal drowning injury. Inj Prev  2011;17:i28–33. 10.1136/ip.2010.026849 [DOI] [PubMed] [Google Scholar]
  36. R Core Team . R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing, 2025. [Google Scholar]
  37. Sakamoto  I, Stempski  S, Srinivasan  V  et al.  Adolescent water safety behaviors, skills, training and their association with risk-taking behaviors and risk and protective factors. Children  2020;7:301. 10.3390/children7120301 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Saluja  R, Brenner  B, Morrongiello  BA  et al.  The role of supervision in child injury risk: definition, conceptual and measurement issues. Inj Control Saf Promot  2004;11:17–22. 10.1076/icsp.11.1.17.26310 [DOI] [PubMed] [Google Scholar]
  39. Surf Life Saving Australia . National coastal safety report 2025. SLSA: Sydney, Australia. 2025.
  40. WHO . Global Status Report on Drowning Prevention. Geneva, Switzerland: WHO, 2024. [Google Scholar]
  41. Willcox-Pidgeon  S, Devine  SG, Franklin  RC. Adult migrants urgent need for drowning prevention in Australia: water safety perceptions, attitudes, and behaviours. Health Promot Int  2025;40:daaf109. 10.1093/heapro/daaf109 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Willcox-Pidgeon  S, Kool  B, Moran  K. Knowledge, attitudes, and behaviours of New Zealand youth in Surf Beach environments. Int J Aquat Res Educ  2017;10:6. 10.25035/ijare.10.02.06 [DOI] [Google Scholar]
  43. Willcox-Pidgeon  S, Kool  B, Moran  K. Perceptions of the risk of drowning at surf beaches among New Zealand youth. Int J Inj Contr Saf Promot  2018;25:365–71. 10.1080/17457300.2018.1431939 [DOI] [PubMed] [Google Scholar]
  44. Wolff  K, Larsen  S, Ogaard  T. How to define and measure risk perceptions. Ann Tour Res  2019;79:102759. 10.1016/j.annals.2019.102759 [DOI] [Google Scholar]
  45. Zaal  E, Ongena  Y, van der Velden  N  et al.  Unraveling honest responding: a systematic review on the effectiveness of social desirability bias reduction methods in survey research. Qual Quant  2026;60:10359–91. 10.1007/s11135-026-02664-7 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

daag144_Supplementary_Data
daag144_supplementary_data.docx (21.8KB, docx)

Data Availability Statement

The anonymized raw survey dataset is available from the corresponding author upon reasonable request.

Articles from Health Promotion International are provided here courtesy of Oxford University Press

RESOURCES