Sugar-sweetened beverages: consumption, purchasing behaviour, and opinions about policy measures in a nationally representative panel of Dutch secondary school adolescents

Abstract

Despite substantial public health efforts, reducing adolescents’ sugar-sweetened beverage (SSB) consumption remains challenging. While much is known about determinants, the underlying motivations for sustained popularity remain unclear. This study aims to (1) gain insights into the prevalence of perceptions and contextual influences surrounding SSB purchase and consumption in a representative sample of Dutch adolescents; and (2) to investigate adolescents’ opinions on presumed effects of policies to reduce SSB consumption. A cross-sectional, nationally representative survey in June 2021 among 839 Dutch students aged 12–16 was conducted, using an online questionnaire to assess dietary behaviours, including weekly consumption of SSBs (glasses/week). The results showed that over two-thirds of adolescents (72.6%) agreed that ‘Drinking soft drinks is normal.’ Across all adolescents, average consumption was 9.5 glasses per week; over half (51.7%) were defined as high consumers (≥ 7 glasses/week). Adolescents with constant home availability of SSBs had 14 times higher odds of being high consumers (95% CI 12.32–14.48), a significant association; they also tended to purchase SSBs more frequently and spent significantly more money on them, compared to those without home availability. Nearly half (49.7%) agreed that price increases would reduce their consumption. Overall, this study confirms that SSBs are widely consumed by adolescents and perceived as a normal, everyday choice. This normalization is reinforced by their availability and low cost. Fiscal measures and targeted interventions in settings such as supermarkets hold promise for reducing SSB consumption. Future research should explore alternatives to SSBs and assess the effectiveness of relevant policies to reduce consumption.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-026-26557-4.

Introduction

Frequent sugar-sweetened beverage (SSB) consumption contributes significantly to the development of overweight and obesity, type 2 diabetes, cardiovascular diseases, some cancers, and tooth decay [1, 2]. Although SSB intake has declined among Western European adolescents [3], socioeconomic disparities remain a significant factor [4, 5]. Research indicates that adolescents from lower socioeconomic groups still consistently consume more SSBs, and changes in income later in life have minimal impact on these rigid consumption patterns [6]. Moreover, adolescence is a formative period in which lifelong behaviours, including dietary habits, are shaped [7]. Their stage of life [8]—characterized by a desire for autonomy, weaker impulse control, and a growing reliance on peer approval [9]—might make adolescents more likely to consume SSBs. In addition, recent research found that drinking SSBs came with a sense of ‘fitting in’ among peers [10]. Therefore, adolescence represents a critical period for targeted interventions to reduce SSB consumption and address related inequalities in beverage consumption behaviours.

Among adolescents specifically, SSB consumption remains generally high [3]. In the Netherlands, adolescents continue to be a key consumer group of SSBs, with a large proportion in 2021 exceeding the WHO’s recommendation to reduce free sugar intake to less than 10% of total energy by consuming nearly half a litre of SSBs per day [11, 12]. Such behaviour is shaped by both personal and local environmental factors, as well as the growing influence of social media. Over decades of public health efforts to reduce SSB consumption, many policies and interventions have been developed across different levels of governance aiming to alter social norms and limit SSB intake in specific populations [3, 12]. In this context, the Netherlands recently introduced a volume-based tax on non-alcoholic beverages—including sugary and non-sugary drinks but excluding water—to encourage healthier choices, reflecting a policy approach distinct from sugar-focused taxes in other countries [13]. Within schools, efforts to promote healthier options, such as through school canteens and vending machines, have been inconsistent. There are no mandatory requirements, and varying availability of sugar-sweetened beverages varies across schools [14]. However, with some exceptions, such efforts have yet to achieve their intended goals or creating meaningful, sustained impact. This is despite the well-established link between SSBs and adverse health outcomes, as well as the widespread awareness and ongoing public health efforts. Therefore, given the rapidly changing context, it is essential to continuously understand adolescents’ perspectives and adapt interventions accordingly to create actions that may be more effective [15]. Yet, most research on SSB purchasing and consumption decisions focuses primarily on adults [16, 17].

Adolescent-focused research is often qualitative and due to its limited generalizability insufficient to inform large-scale policy development. To address this gap, we need more representative quantitative research to tailor interventions to adolescents’ behaviours and preferences and to learn their opinions about potential policy measures that can resonate more effectively. The aim of this study is therefore twofold: (1) to gain insights into the prevalence of perceptions and contextual influences surrounding SSB purchase and consumption among a representative sample of Dutch secondary school students aged 12 to 16; and (2) to investigate adolescents’ opinions on presumed effects of policies to reduce SSB consumption.

Methods

Study population and design

To examine adolescents’ SSB consumption and purchase behaviour, and their views on potential policy measures to reduce SSB consumption, we conducted a questionnaire study in a nationally representative sample of Dutch secondary school students aged 12 to 16. Recruitment and sampling of participants were conducted by Ipsos I&O Research [18], a Dutch research agency certified for three internationally recognized quality standards; ISO 9001, ISO 20,252, and ISO 27,001. Participants were randomly selected from two existing Dutch research panels: the I&O Research Panel (over 31,000 active members) and the Dynata Panel (over 500,000 active members), both recruited through random sampling. Respondents did not receive financial compensation for joining the panels. No specific dietary, physiological, or nutritional characteristics were used as inclusion or exclusion criteria; all adolescents aged 12–16 within the panels were eligible. To achieve a net sample of 1,000 respondents, email invitations were sent to approximately 1,550 adolescents in the I&O Research Panel and 900 adolescents in the Dynata Panel, with reminders send as necessary. Our sample was drawn in accordance with the Gold Standard, using the latest Statistics Netherlands (CBS) data to ensure representativeness in terms of age, gender, education and region among Dutch adolescents aged 12 to 16 [18]. Parents of recruitees younger than 16 provided active informed consent for their child to participate; recruitees aged 16 were contacted directly and provided informed consent themselves. Ten €10 gift vouchers were randomly allocated to respondents. Data collection took place in June 2021. The Research Ethics Review Committee (BETCHIE) of the Faculty of Science, Vrije Universiteit Amsterdam, determined in advance that a full ethical review would not be required under the faculty’s Code of Ethics.

To capture a combination of relevant behavioural and policy-related constructs, which to our knowledge, are not jointly covered in existing instruments, we developed a questionnaire with Ipsos I&O Research [18] that was appropriate for the adolescent target group based on “The developmental science–based guide” by Omrani and colleagues [19], (see Appendices 1 and 2). Items were, where possible, adapted from established surveys, such as national [20] and school-level monitoring tools [21], comparable studies [22–24], and informed by sources such as the Family Finance Information Institute (Nibud) [25]. In designing the questionnaire, we took factors into account including questionnaire length, wording, response categories, and question sequence. With two exceptions, we opted for closed-ended questions, which are deemed appropriate for the age group [19]. The two open-ended questions were: (1) “How can you tell whether or not a soft drink is healthy?” and (2) “About how much money do you spend per week on soft drinks?” The first question generated qualitative data that were not included in the final analyses, as it fell outside the scope of this paper. The questionnaire was conducted online, accessible from any device with an internet connection, and took approximately 10 min to complete. Parents were instructed that, whilst they might assist if needed, the child was to complete the questionnaire independently.

Outcome variables

SSB consumption

Prior to starting their questionnaire, respondents were provided with a definition of SSBs: ordinary sugar-sweetened soft drinks (excluding light and no-cal versions), fruit juices, fruit syrup drinks, sports drinks and energy drinks. For ease of interpretation, SSBs were denoted as ‘soft drinks’ (frisdrank) throughout the questionnaire. Participants were asked to recall ‘On how many days in the past week did you consume soft drinks?’ and ‘How many beverage glasses [250 ml] on average would you drink on one such day? One half-litre bottle would count as 2 glasses.’ Weekly SSB consumption was calculated by multiplying responses to these two questions. The responses were then categorized as ‘low to moderate’ (0–6 glasses weekly) and ‘high’ (≥ 7 glasses weekly) to approximate consumption levels as commonly reported in dietary interviews and other studies on SSB consumption when they differentiate between daily and non-daily consumption patterns [20, 22].

Contextual influences: SSB availability at home and social norms about SSBs

Participants were asked about SSB availability at home: ‘Are soft drinks present in your home?’ (rarely to never, sometimes, most of the time or always). They were also shown the statement ‘Drinking soft drinks is normal’ to explore the social norm applying to SSBs (5-point Likert-type scale: 1 = strongly disagree, 5 = strongly agree).

Contextual influences: purchase of SSBs and expenditure

Participants were asked whether they purchase SSBs themselves and, if so, where (e.g., school canteen or supermarket, with multiple answers permitted). Purchase frequency was grouped into four categories (daily, weekly, monthly, and less than monthly). They were also asked how much they spend on SSBs weekly. To determine the price point at which participants would stop purchasing SSBs, we asked them to indicate a price per can (330 ml) beyond which they would no longer buy them, in six categories: ‘I would stop buying the soft drink if the price would be €1, €2, €3, €4, €5, other.’ This method, commonly used in tobacco research to assess price sensitivity [26], estimates the potential impact of price increases on SSB consumption. At data collection, the regular supermarket price range for 330 ml cans of ordinary cola was €0.39 for store brands and €0.89 for premium brands [27]. Participants who reported rarely to never buying SSBs were automatically directed to subsequent questions, bypassing those on purchasing considerations, purchase frequency and expenditures. As a result, analyses regarding purchasing behaviour and expenditure were based on a reduced subsample of participants who reported buying SSBs.

Perceptions of proposed policy interventions to reduce SSB consumption

Questions about potential policy interventions were based on previous qualitative research where teenagers reflected on measures that might effectively reduce adolescent SSB consumption [10]. Our respondents were presented with four statements on a 5-point Likert scale (1 = strongly disagree, 5 = strongly agree). These statements did not represent a single underlying construct; instead, they captured different policy-related perceptions. The specific construct addressed by each item depended on the extent to which the corresponding policy measure is currently implemented in the Dutch context. For policies that are already (partly) in place, items primarily assessed perceived effectiveness, whereas for policies that are not yet implemented, items focused on support or preferences. (1) If I can’t buy soft drinks at school, I drink them less. (2) An age limit should be put on the sale of soft drinks. By this we mean a minimum age to buy soft drinks, similar to that for tobacco and alcohol. (3) If the price of soft drinks were to be increased, fewer teenagers would buy them. (4) It would be better to ban soft drinks entirely rather than increasing the price. For statement 2, agreeing respondents were asked to choose a minimum age (12, 13, 14, 15, 16, 17, 18, other, don’t know).

Sociodemographic variables

Demographic characteristics included gender, age, school type, and year in secondary school; no other socioeconomic variables were available. The research panels consisted of adolescents aged 12 to 16. School type was categorized in three groups: (1) pre-vocational education (PVSE, Dutch vmbo) (2), higher general secondary education (HGSE, havo), and (3) pre-university education (PUE, vwo). Adolescents enrolled in practical training (praktijkonderwijs, a vocational track in special education) were grouped under PVSE. Year in secondary school ranged from year 1 to 4. In this study, the few students in HGSE year 5 were grouped into HGSE 4 and those in PUE 5 into PUE 4.

Statistical analyses

The dataset was weighted for gender, age and school type according to selection probabilities to match the characteristics of the overall Dutch population of adolescents aged 12 to 16 [18]. Within gender, 1.0% (n = 8) identified as ‘other’; due to that low number and statistical purposes, we randomly reassigned these to either ‘female’ or ‘male’. To assess the impact of sample design on precision, we calculated the design effect using an assumed margin of error of 3.1% at 50%; the result was approximately 1.0, indicating practically non-existent inflation of variance. We adopted complex sampling procedures, including stratification and weighting [28]. In the results, demographic characteristics of the participants are presented both unweighted and weighted to reflect the sample distribution and show sampling differences compared to the actual Dutch population (Table 1).

Table 1.

Demographic characteristics of participants, unweighted and weighted

Total study population	Unweighted N	Unweighted %	Weighted %
	839	100	100
Gender
Female	428	51.0	48.1
Male	411	49.0	51.9
Age in years
12	71	8.5	12.3
13	194	23.1	22.2
14	200	23.8	22.2
15	185	22.1	22.6
16	189	22.5	20.7
School type
Pre-vocational secondary	326	38.9	41.9
Higher general secondary	238	28.4	30.1
Pre-university	275	32.8	28.0
Year in secondary school (with mean ages in years, 95% CI)
1 (12.60, 12.60–12.60)	178	21.2	24.2
2 (13.69, 13.67–13.69)	204	24.3	23.0
3 (14.61, 14.60–14.61)	214	25.5	27.2
4 (15.63, 15.62–15.63)	233	27.8	25.6

Chi-square tests were conducted to analyse whether consumption level and the presence of soft drinks at home differed by gender, age or school type. To examine these factors in association with SSB purchase location and frequency, we conducted additional analyses. Since participants could select multiple purchase locations, each location type was analysed separately. However, because purchase locations may be influenced by school-specific factors and policies, we used school year rather than age in those models. Price thresholds were also analysed by chi-square test. As the continuous variable ‘expenditure’ was not normally distributed, we performed a log transformation to obtain a normal distribution and enable linear regression to examine associations with all outcomes of interest. Outliers were identified using statistical methods in SPSS (including stem-and-leaf plots, descriptive extremes and 95% confidence intervals) and were carefully evaluated [29]. For SSB consumption, values greater than 48 glasses (> 12 L) per week, identified in 7 of 839 respondents, were removed prior to further analysis. Expenditure responses were screened beforehand using data from the Family Finance Information Institute (Nibud) as a reference [25] and were excluded if deemed implausible or extreme. Specifically, weekly values exceeding €28 (n = 26) were omitted from further analysis.

For the outcomes ‘social norm’ and ‘statements regarding policy interventions’, responses were expressed on 5-point Likert scales. For ease of interpretation and facilitate analysis, responses were dichotomized as ‘agree’ (combining agreeing and strongly agreeing responses) and ‘don’t agree’ (combining strongly disagreeing, disagreeing and neutral responses). Logistic regression analyses were then performed to examine associations between sociodemographic characteristics (gender, age, school type, and home availability of SSBs) and (1) agreement with the statement that drinking soft drinks is normal and (2), support for various policy initiatives. For the school-related statement, school year was analysed rather than age. In addition, age was analysed as a categorical variable, as a linear relationship between age and SSB consumption among 12–16-year-olds was not expected, based on contextual factors such as the home environment (e.g., parental rules on soft drinks) and the school environment (e.g., policies restricting younger students from leaving school premises). These factors vary with age, making a categorical approach more appropriate. In addition, logistic regression analyses were conducted to ascertain whether SSB consumption levels were associated with those sociodemographic characteristics. For all analyses, both unadjusted and adjusted estimates are presented in the results. For age, 12 years was the reference category; for school type, PUE; and for home availability ‘rarely to never’.

Odds ratios (OR) are presented with 95% confidence intervals (CI). In analyses of purchase locations, multiple tests may have occurred, increasing the risk of finding significance due to chance. We therefore focused primarily on strongly significant results (p < 0.01). All analyses were performed using IBM SPSS Statistics version 26.

Results

A total of 982 adolescents completed the questionnaire, recruited from the I&O Research Panel (n = 612) and the Dynata panel (n = 374), for an average response rate of 40% over the two panels. Given our inclusion criterium of participants having to be attending secondary school, we left out all others who e.g. still were in primary school or who already finished secondary school. This is reflected in the number of 12-year-olds, which is relatively small, as a portion of this age group was still in primary school at the time of data collection. In total, 839 participants were ultimately included in data analyses. For the selection and distribution across the panels, see Fig. 1. Of respondents, 51.9% were categorized as male and 48.1% as female. The number of participants was evenly distributed across age groups (mean 14.1 years, SD = 0.91) and school types. See Table 1 for further demographic characteristics of the participants, presented both unweighted as weighted.

Fig. 1.

Flowchart of participant recruitment and inclusion. Overview of the number of adolescents recruited via the I&O Research Panel and the Dynata panel, and the final sample included in the analyses based on the inclusion criterion of attending secondary school

SSB consumption

Among Dutch secondary school adolescents, the average consumption was 9.5 glasses of 250 millilitres per week (95% CI 8.93–10.17), while the median consumption was 7 glasses per week. Over half of the adolescents (51.7%) were high consumers, drinking seven or more glasses per week. Among high consumers, the average consumption was 16.4 glasses per week (95% CI 15.70-17.11), while low to moderate consumers drank 2.2 glasses on average (95% CI 2.02-2.38). The majority (88%) reported having consumed SSBs in the past week. Only 10.0% had consumed SSBs but not in the past week, and 2.0% reported never consuming them. Among consumers, the mean intake was 2.38 glasses per day (95% CI 2.28–2.48) in the past week. For 68.6% of adolescents, SSBs were usually or always available at home, whilst only 7.9% reported that SSBs were rarely or never available.

Significant associations were found in SSB consumption with age, gender, school type, and SSB availability at home, see Table 2. Among 12-year-olds, 39.5% were high consumers. Higher percentages were observed for 13-year-olds (61%), 14-year-olds (50.3%), 15-year-olds (54.8%) and 16-year-olds (47.8%). Yet only the difference between 13- and 12-year-olds showed to be statistically significant. Boys were more likely to be high consumers than girls (OR 1.41, 95% CI 1.04–1.92). Students in pre-vocational education (PVSE) had approximately double the odds of being high consumers as compared with pre-university (PUE) students (OR 2.01, 95% CI 1.39–2.89). Students in higher general secondary education (HGSE) likewise had greater odds of being high consumers than PUE students (OR 1.67, 95% CI 1.08–2.33). No significant difference was shown between HGSE and PVSE students.

Table 2.

Odds ratios for high Sugar-Sweetened Beverage consumption by socio-demographics and home availability (N = 839)

Variable	% indicator	Adjusted OR	(95% CI) lower	(95% CI) upper	p-value
Age
12	39.5	-	-	-	-
13 vs. 12	60.2	2.25*	1.23	3.84	< 0.001
14 vs. 12	50.2	1.43	0.85	2.43	0.18
15 vs. 12	53.9	1.68	0.99	2.84	0.06
16 vs. 12	47.0	1.56	0.91	2.67	0.10
Gender
Girls	46.0	-	-	-	-
Boys	52.9	1.41*	1.04	1.92	0.03
School type
PUE	58.0	-	-	-	-
PVSE vs. PUE	52.6	2.01*	1.39	2.89	< 0.001
HGSE vs. PUE	39.8	1.67*	1.08	2.33	0.01
Availability of SSBs at home
(almost) Never	11.1	-	-	-	-
Usually vs. (almost) never	29.8	3.00*	1.21	7.45	0.01
Sometimes vs. (almost) never	50.2	7.36*	3.15	7.22	< 0.001
Always vs. (almost) never	65.4	13.65*	5.89	31.65	< 0.001

Abbreviations: SSB Sugar-sweetened beverage, PVSE Pre-vocational secondary education, HGSE Higher general secondary education, PUE Pre-university education

Reference categories: Age = 12 years, School type = PUE; Home availability = (almost) never

*P < 0.01 indicate statistical significance

Availability of SSBs at home – sometimes (OR 3.00, 95% CI 1.21–7.45), usually (OR 7.36, 95% CI 3.15–17.22) or always (OR 13.65, 95% CI 5.89–31.65) – was significantly associated with greater odds of being a high consumer, in comparison with adolescents who rarely to never had SSBs at home. Adolescents who always had SSBs at home had approximately 14 times greater odds of being high consumers, drinking an average of 13.4 glasses per week (95% CI 12.32–14.48), compared to 2.5 glasses (95% CI 1.65–3.35) for those who rarely to never had SSBs at home.

Social norm with respect to SSBs

More than two thirds (73.1%) of adolescents agreed or strongly agreed with the statement ‘Drinking soft drinks is normal.’ Gender, age, high consumption and SSB availability at home were significantly associated with that perception, see Table 3. Boys had higher odds of agreement (OR 1.50, 95% CI 1.07–2.10). Some 58.8% of twelve-year-olds agreed, and higher percentages were observed among 13-year-olds (76.3%), 14-year-olds (77.9%), 15-year-olds (73.7%) and 16-year-olds (72.5%). All older age groups except 15-year-olds showed statistically significant differences compared to 12-year-olds. High consumers had greater odds of considering SSBs normal compared to low-to-moderate consumers (OR 1.80, 95% CI 1.25–2.59). Respondents with SSBs available at home were also more likely to view them as normal. Those with always SSBs available at home had over 3 times greater odds for that perception than those who rarely to never had SSBs at home (OR 3.17, 95% CI 1.78–5.65); similarly, usual availability also increased the odds (OR 2.02, 95% CI 1.12–3.62), but those who sometimes had SSBs at home were no more likely to consider them normal. School type was not significantly associated with this perception.

Table 3.

Social norms on SSBs: odds ratios for agreement with ‘Drinking soft drinks is normal’ (N = 839)

Variable	% indicator	Adjusted OR	(95% CI) lower	(95% CI) upper	p-value
SSB consumption
Low-to-moderate	63.6	-	-	-	-
High	81.0	1.80*	1.25	2.59	0.01
Age
12	58.8	-	-	-	-
13 vs. 12	75.2	1.96*	1.07	3.57	0.01
14 vs. 12	77.6	2.56*	1.40	4.67	< 0.001
15 vs. 12	73.3	1.75	0.95	3.22	0.43
16 vs. 12	72.5	2.01*	1.09	3.70	0.01
Gender
Girls	67.8	-	-	-	-
Boys	77.2	1.50*	1.07	2.10	0.01
School type
PUE	75.7	-	-	-	-
PVSE vs. PUE	73.3	1.19	0.81	1.76	0.37
HGSE vs. PUE	67.7	1.19	0.79	1.81	0.40
Availability of SSBs at home
(almost) Never	48.6	-	-	-	-
Usually vs. (almost) never	63.0	1.57	1.05	2.35	0.03
Sometimes vs. (almost) never	71.3	2.02*	1.24	3.32	0.01
Always vs. (almost) never	80.8	3.17*	1.78	5.65	< 0.001

Abbreviations: SSB Sugar-sweetened beverage, PVSE Pre-vocational secondary education, HGSE Higher general secondary education, PUE Pre-university education

Reference categories: Age = 12 years, School type = PUE; Home availability = (almost) never

*P < 0.01 indicate statistical significance

Purchase locations and purchase frequency

Supermarkets were reportedly the most common SSB purchase location among adolescents (Table 4), followed by schools (including canteens and vending machines). No gender differences were apparent, but both secondary school year and school type were significantly associated with purchase location. Supermarket purchases were lower for first-year students (44.5%), gradually increasing across the second (55.0%), third (58.1%) and fourth year (60.5%). Buying SSBs at school was significantly less common for PUE students (9.9%) than for those in HGSE or PVSE.

Table 4.

SSB purchase location by sociodemographic characteristics (N = 839)

Reported purchase location	School %	Supermarket %	Sports %	On the go (e.g. public transport, petrol stations) %	Delis, fast-food outlets %	Never buy SSBs %
Gender
Female	18.7	53.0	6.3	6.8	9.0	36.8
Male	17.1	55.2	7.2	5.2	6.7	35.4
Year in secondary school
First	15.5	44.5*	2.1*	3.6	3.3	44.1*
Second	20.1	55.0	7.4	6.6	5.3	35.4
Third	22.2	58.1	11.0*	8.6	7.2	35.7
Fourth	14.0	60.5	6.4	5.8	15.4*	27.6*
School type
Pre-vocational secondary	22.0	60.2*	6.7	7.4*	6.6	30.9*
Higher general secondary	19.5	50.8	7.3	5.8	8.5	34.7
Pre-university	9.9*	48.9	6.2	4.6	8.9	45.4*

Percentages do not sum to 100%, as participants could select multiple responses

SSB  Sugar sweetened beverage

*P < 0.01 indicate statistical significance

Non-purchasers were almost evenly distributed between boys and girls, with no significant gender differences. Significant associations for non-purchase emerged for both school year and school type, with first-year students and PUE students being, respectively, more likely to be non-purchasers than students in higher years and students in HGSE or PVSE.

For purchase frequency, sample size differs due to the preceding selection question (‘Where do you usually buy soft drinks?’), which included the option ‘I never buy soft drinks’; n = 538 adolescents reported purchasing SSBs themselves. School type was found significantly associated with purchase frequency (Table 5). Students in PVSE were more likely to be frequent purchasers, with 11.9% buying SSBs daily, 40.8% weekly, and 26.5% monthly. These rates were higher than those for HGSE and PUE students. Conversely, only one in five PVSE students (20.7%) purchased SSBs less frequently than once a month, a proportion increasing in HGSE (32.6%) and PUE (41.8%). Adolescents with constant home availability tended to purchase SSBs more frequently themselves, P = 0.05. Yet, with our chosen threshold of P < 0.01 this finding was not significant. No gender differences were observed.

Table 5.

SSB purchase frequency by sociodemographic characteristics (N = 538)

Reported purchase frequency	Daily %	Weekly %	Monthly %	Less than monthly %
Gender
Female	7.9	29.5	29.2	33.4
Male	11.7	36.7	25.8	25.8
Age*
12	3.0	20.7	24.8	51.5
13	18.5	37.9	23.0	20.6
14	12.9	33.1	28.0	26.1
15	6.9	41.4	29.0	22.6
16	5.2	26.9	30.6	37.3
School type*
Pre-vocational secondary	11.9	40.8	26.5	20.7
Senior general secondary	8.8	30.1	28.6	32.6
Pre-university secondary	7.5	23.2	27.6	41.8
SSB availability at home
Rarely to never	5.3	23.0	26.5	45.2
Sometimes	7.5	31.9	32.4	28.2
Usually	6.3	32.5	33.1	28.2
Always	13.5	36.3	22.1	28.3

SSB  Sugar sweetened beverage

* P<.01

Average weekly SSB expenditure and price threshold

On average, adolescents (N = 538) spent nearly four euros (€3.90, 95% CI 3.50–4.31) per week on SSBs. After adjusting for age, gender, school type and SSB availability at home, high consumers indicated to spend over twice as much as lower consumers, averaging €4.81 (95% CI 4.23–5.39) versus €2.14 (95% CI 1.78–2.51). Adolescents from households where soft drinks are always available spent more (€1.96, 95% CI €1.45–€2.65) than those who usually (€1.69, 95% CI €1.27–€2.25), sometimes (€1.63, 95% CI €1.24–€2.13), or rarely to never (€1.59, 95% CI €1.17–€2.15) have soft drinks at home after adjusting for age, gender, school type and consumption level. The average price at which adolescents would cease purchasing SSBs was €2.57 per can (330 ml) with 84.5% claiming to stop should the price reach €3. The price threshold was associated with age and with consumption level, but not with gender, school type or SSB availability at home.

Adolescents’ perceptions of proposed policy interventions

The statement ‘If the price of soft drinks were to be increased, fewer teenagers would buy them’ received the highest level of agreement, with nearly half of the adolescents agreeing (49.7%; Table 6). No significant differences between agreers and non-agreers were observed for this premise in relation to gender, age, school type, or consumption level. The statement ‘If I couldn’t buy soft drinks at school, I would drink them less’ received only 28.8% agreement, but a significant difference emerged for PUE students, who showed higher odds of agreement compared to PSVE students (OR 1.83 95% CI 1.21–2.76) or HGSE students OR 1.63, 95% CI: 1.05–2.52). Also, after adjusting for gender, school year, and school type, high consumers were, less likely than low-to-moderate consumers to be non-agreers (OR 0.58, 95% CI 0.41–0.81), insinuating that when being a high consumer, even a school restriction on SSB purchasing would not deter consumption. Over three-quarters of adolescents disagreed or strongly disagreed with a complete ban. Remarkably, high consumers had significantly greater odds of favouring a complete ban than more occasional consumers, with percentages of non-agreement being much lower (5.2% versus 17.6%, unadjusted OR 95% CI OR 0.28, 95% CI 0.17–0.47). HGSE students were significantly more likely than PVSE students to indicate non-agreement with a complete ban, estimated in an additional model with PVSE as the reference (OR 0.52, 95% CI 0.30–0.90).

Table 6.

Adolescents’ perceptions of proposed policy interventions to reduce SSB consumption (N = 839)

Policy measure	Strongly agree %	Agree %	Neutral %	Disagree %	Strongly disagree %
If I can’t buy soft drinks at school, I drink them less.	5.9	22.9	19.0	30.4	21.8
Age limit should be put on sale of soft drinks.	5.4	10.3	17.9	35.7	30.7
Increasing the price results in fewer teenagers buying soft drinks.	12.1	37.6	18.8	19.5	12.1
Soft drinks should be banned rather than increasing the price.	3.9	7.7	11.9	36.2	40.3

No further significant differences were found between school types. Although 84.3% of participants did not assent to age restrictions, those endorsing a minimum age for buying SSBs were evenly distributed across age groups, with nearly all proposing their own age as an appropriate threshold. No significant differences were found for sociodemographic background characteristics.

Discussion

The aim of this study was twofold: (1) to gain insights into the prevalence of perceptions and contextual influences surrounding SSB purchase and consumption among a representative sample of Dutch secondary school students aged 12 to 16; and (2) to investigate adolescents’ opinions on presumed effects of policies to reduce SSB consumption. The findings revealed that the vast majority (88%) of adolescents had consumed sugar-sweetened beverages (SSBs) within the past week. Among all adolescents, over half were classified as high consumers. Consistent with previous research [12, 30, 31], boys reported higher consumption and were more likely to be high consumers. Among school type, a disproportionate percentage of the PVSE students were identified as high consumers. Although PVSE (pre-vocational education) is not inherently indicative of low SEP [32], previous research has shown that high consumption is associated with vocational education and with lower socioeconomic position (SEP) [33]. As SEP was not collected in our study, this association should be interpreted with caution. Availability of SSBs at home – whether sometimes, usually or always – was associated with significantly greater odds of high consumption compared with no home availability of SSBs. In particular, adolescents with constant availability of SSBs at home exhibited 14-fold greater odds of high consumption. Moreover, high consumption was significantly associated with the perception that consuming SSBs is normal. High consumption was also associated with spending more on SSBs and indicating to be willing to spend more on SSBs in future than moderate and low consumers said they would be. With respect to what adolescents in our study viewed to be potentially impactful policy interventions to reduce SSB consumption, increasing the price of SSBs received the most support, with a complete ban facing the most opposition. Over four fifths of respondents did not favour age restrictions, while those in favour often suggested their own age as an appropriate threshold for purchasing SSBs.

Starting with adolescents’ perceptions of SSB consumption, the vast majority of teenagers in our study perceive drinking soft drinks as normal: 90% of them indicated to consume SSBs. Furthermore, the widespread acceptance of this norm – reflected by our findings that nearly three-quarters agreed with the statement ‘Drinking soft drinks is normal’ – was strongly associated with higher SSB consumption. This is in line with previous research that also found that perceiving sugary drinks as unhealthy and agreeing with social norms discouraging their consumption were linked to lower intake, demonstrating both sides of the influence social norms can have on SSB consumption [34]. Furthermore, our finding aligns with existing evidence that such social norms may lead people to overestimate the actual prevalence of SSB consumption in the population, thereby strengthening the perception that SSB consumption is acceptable [35]. Moreover, since SSBs remain readily accessible and heavily marketed in locations where teenagers live, go to school, work and socialize – both offline and online – this further reinforces the norm that drinking SSBs is normal. This, in turn, fuels preferences for ultra-processed beverages and foods [35, 36]. Shared eating and drinking occasions play a key role in building and bolstering friendships [37]. Given that peer norms exert an even greater impact during adolescence [9], SSBs may be viewed as an essential aspect of adolescents’ social identity. As a result, the strong social norm encouraging SSB consumption, combined with the perceived role of SSBs in adolescent identity, may help explain the low preference found in our study for a complete ban on soft drinks, with only 11.6% preferring such a ban over increasing the price. Building on such factors, the food industry cleverly capitalizes on the vital role of SSBs in adolescents’ lives by marketing products that tap into their identity and evoke feelings of togetherness – thus further fuelling the perception that SSBs are a normal and essential part of their social life [38]. Notably, high consumers had significantly greater odds of favouring a complete ban over increasing the price than more occasional consumers. While this may seem counterintuitive, it could reflect greater awareness among high consumers of their own excessive intake and the perceived ineffectiveness of subtler measures such as price increases. Alternatively, it may reflect a preference for clear, unambiguous regulations that are easier to follow, or social desirability in reporting attitudes toward stricter policies. Taken together, these insights further underscore the need for restrictions on marketing unhealthy products like SSBs to children – as repeatedly advocated by WHO and UNICEF [36, 39]. Comprehensive bans have also been proposed given children’s exposure to a wide range of marketing mediums in multiple settings [40]. Additionally, it may be crucial to raise the age threshold in such restrictions, in order to protect adolescents during vulnerable developmental stages [41, 42].

As further evidence of contextual influences on perceptions and consumption behaviour, we found that the availability of SSBs at home was associated with adolescents being high consumers. This concurs with literature showing a link between the mere presence of SSBs in the household and higher consumption [31, 43, 44]. It highlights the strong influence of family habits and parental modelling. Adolescents whose parents are high SSB consumers, who maintain high availability at home and who hold positive attitudes towards SSBs have been found to consume higher amounts themselves [43]. Such parental behaviour and household routines may help foster a stronger social norm, as supported by the association we found between ready home availability and the perception that drinking SSBs is normal. Moreover, SSBs being readily available at home showed to be associated with purchasing SSBs more frequently (borderline significant in our findings) and with indicating to spend more money on them despite already having easy access at home. This could be explained by the perception of SSB consumption as normal and by a certain habituation to regular intake – potentially reinforced through family habits and household routines. A recent review has demonstrated similar insights, highlighting that social norms, along with habit and intention, are key determinants of both the amount and frequency of SSB consumption [45]. Overall, these findings emphasize the crucial role of parental habits concerning the availability and purchasing of SSBs in shaping adolescents’ habits and consumption patterns. Interventions that target parents’ purchasing behaviour could therefore potentially help to reduce adolescent SSB consumption.

In terms of local environmental influences on purchasing behaviour, the supermarket emerged as adolescents’ prime location for purchasing SSBs, with over half of respondents obtaining them there. Other points of sale, such as schools and sports venues, were far less common, and that aligns with other studies on purchasing behaviour [24]. Despite this, many public health interventions focus on removing soft drinks from sports canteens and schools [46–50]. While this may help reduce the broad availability, visibility and accessibility of SSBs, its impact on SSB purchases by adolescents is likely to be minimal. An additional observation in our study involved the increasing percentages of students in later school years who purchased SSBs in supermarkets compared to their younger counterparts. Besides the greater spending power of older students, this might also reflect common policies in Dutch schools whereby first- and second-year students (typically aged 12–13) may not leave school premises during school hours. Yet if we truly want to reduce adolescents’ SSB purchases, and thereby their consumption, supermarkets may offer serious potential for impact parallel to creating healthier school settings. Our findings support the need for future interventions to prioritize supermarkets in combination with schools and other frequently visited locations, with the goal of reducing availability and modifying product offerings for SSBs. In addition, measures such as restricting in-store marketing and promotions for unhealthy beverages, and promoting healthier alternatives through targeted displays or discounts, may be considered [51]. A recent paper by Middel et al. [52], revealed that promotion of unhealthy products, such as SSBs, is deeply ingrained in supermarket business models and that simple interventions such as nudges or food labels would not be effective. Instead, the study recommends policy actions such as promoting sustainable corporate governance models, adjusting financial incentives through taxes and subsidies, and ensuring a ‘level playing field’ to encourage healthier business practices.

Finally, we would like to examine contextual influences on spending and purchasing behaviour in conjunction with potential policy measures targeting SSB prices. Adolescents reported spending an average of €4 per week on SSBs – at prices ranging from €0.39 for store brands to €0.89 for premium brands for 330-millilitre cans of cola [27]. Only at prices exceeding €3 per can would over 80% of participants reportedly refrain from buying SSBs. Although such intentions do not necessarily translate directly into behaviour [53], and teenagers might overestimate their willingness to spend, this nonetheless suggests that adolescents consider SSBs important enough and are willing to allocate significant portions of their weekly budgets to SSBs. Yet we also saw that the price increase policy to reduce SSB consumption garnered the most support, which aligns with previous findings that adolescents favour population-level interventions, like sugar taxes [54]. Evidence supporting the health benefits of sugar taxes continues to accumulate, and taxing already has broad support among adults as a measure to reduce SSB consumption [16, 17]. This study extends such findings by demonstrating adolescent support for such policies, with half of our respondents indicating that a price hike would reduce their consumption. Even so, our findings also suggest that small price increases may have limited impact on adolescent purchasing, consistent with other research that indicates that taxes must be sufficiently high, at least 20%, to affect affordability if they are to help reduce SSB consumption [16].

Assessing the impact of beverage taxes on adolescents’ consumption is essential, yet evidence remains mixed. In the UK, the Soft Drinks Industry Levy led to significant reductions in soft drinks consumption [54]. Similarly, a Spanish study reported a decline in sugar-sweetened beverage purchases among young people living in poorer neighbourhoods following the introduction of an excise tax [55]. In contrast, a recent study across six European countries with sugar-based beverage taxes found no greater post-tax reductions in adolescent consumption than in countries without a such taxes [56]. Moreover, concerns have been raised that consumers may shift toward artificially sweetened beverages (ASBs) as substitutes [57]. While ASBs were considered an intermediate step toward ultimately replacing SSBs with water or other healthier beverages [58], taxes may inadvertently encourage their consumption. This is concerning since the health benefits and long-term safety of ASBs are increasingly debated [59]. Yet, research examining such substitution effects is largely absent, and studies focusing specifically on adolescents are, to our knowledge, lacking.

Recently, the Netherlands increased the volume-based excise tax on non-alcoholic beverages, including regular, light, no-calorie soft drinks, and fruit juices - excluding bottled water - aiming to promote healthier beverage choices. Unlike sugar-based taxes, this tax targets volume [13]. To date, no formal evaluation on its impact has been published. Examining its potential effects on adolescents’ consumption behaviours is needed.

Strengths and limitations

While much of the existing research currently focuses on adults, this study employed a nationally representative sample of Dutch adolescents (aged 12–16) to explore their SSB consumption, purchasing habits, and policy measures they would consider effective to reduce adolescent consumption. To our knowledge, our study is the first to do so. Besides its strengths, this study also has some limitations. While the sample is nationally representative of Dutch adolescents aged 12–16, these finding may not be fully generalizable to other countries, age groups, or time periods, due to differences in cultural, environmental, and policy contexts. It was conducted in June 2021, shortly after the lifting of COVID-19 restrictions, which may have had some unpredictable effects given the uniqueness of the situation. In addition, the questionnaire was developed solely by experts and was not tested beforehand within the target population, meaning we cannot be certain that respondents interpreted all questions correctly. Seasonal effects, such as higher beverage consumption during warmer months, may have influenced the reported intake levels. Furthermore, the study’s cross-sectional design prevents any causal inference regarding the relationships observed. We relied on self-reported measures of SSB consumption and purchasing habits, which may be subject to recall bias or social desirability bias. Adolescents’ own educational level was used as a proxy for socioeconomic status (SES), which may not fully capture family-level SES or other dimensions such as household income or parental occupation. Finally, the study was not pre-registered, which limits transparency regarding the a priori hypotheses and analytical approach.

Conclusion

This study confirms that SSBs are extensively consumed by adolescents and perceived as a routine, normalized part of daily life. This normalization is reinforced by the widespread availability and low cost of SSBs, contributing to elevated free sugar intake and excess caloric consumption, which are important nutritional concerns for adolescent health. Notably, adolescents perceive price increases as a more acceptable measure compared to bans, and believe that such increases could be effective in reducing the number of teenagers buying SSBs, highlighting the potential of fiscal measures to reduce consumption. The results emphasize the need for interventions that effectively target appropriate audiences – such as parents – and can strategically intervene in key settings, particularly supermarkets. While taxes could hold promise, the recently introduced sugar tax in the Netherlands has not yet been formally evaluated. Assessing its potential impact on adolescents’ consumption behaviours would provide valuable insights for policymakers. Interventions, such as adjustments in product availability and accessibility, taxation, and marketing restrictions, could play pivotal roles in challenging entrenched social norms and ultimately driving change. Additional research is also needed to explore alternatives to SSBs and to help design policies and systems that can facilitate favourable shifts in consumption patterns.

Abbreviations

SSBs

Sugar sweetened beverages

WHO

World Health Organisation

ISO

International Organisation for Standardization

PVSE

Pre-vocational secondary education

HGSE

Higher general secondary education

PUE

Pre-university education

OR

Odds Ratios

95% CI

95% Confidence Interval

SEP

Socioeconomic position

Authors’ contributions

RP co-conceptualized the study, analysed and interpreted the data and led the writing of this manuscript. WW contributed to the analysis and interpretation of the data and critically reviewed and edited the manuscript. BG critically reviewed the manuscript. CV interpreted the data and critically reviewed the manuscript. AV and JS contributed to the data analysis, interpretation and critically reviewed and edited the manuscript. VB co-conceptualized the study, analysed and interpreted data, and critically reviewed and edited the manuscript. All authors have approved the final manuscript.

Funding

This work was supported by the Dutch Diabetes Research Foundation, project number 2022.15.001. The funder had no role in the design, analysis or writing of this article.

Data availability

The datasets generated and/or analysed during the current study are available from the Public Health Service Amsterdam (GGD Amsterdam), Department of Healthy Living, via email on reasonable request. Contact: gezondleven@amsterdam.nl.

Declarations

Ethics approval and consent to participate

In accordance with the Declaration of Helsinki, this study was reviewed and declared not subject to the law on research involving human subjects. The protocol was assessed and considered compliant with scientific due diligence by the BETHCIE, the Research Ethics Review committee of the Faculty of Science, Vrije Universiteit Amsterdam, The Netherlands.

Parents of recruitees younger than 16 provided active informed consent for their child to participate; recruitees aged 16 were contacted directly and provided informed consent themselves.

Competing interests

The authors declare no competing interests.