The effects of warnings on harmful products: literature review and potential applications to the automotive sector Open Access

Some of the most commercially successful everyday products and services, such as tobacco, alcohol, legal drugs and automobiles, pose significant risks to human health, society and the environment (World Health Organization [WHO], 2018). Government responses to these risks include taxes, bans and information campaigns and warnings, aiming to reduce consumption (Dossou et al., 2017; Government of Canada, 2019; WHO, 2003, 2020). Warnings are increasingly widespread, notably on tobacco, alcohol and legal drug packaging, and more recently in automobile advertising.

The transport sector is a major contributor to climate change, with fossil fuel emissions directly linked to rising global temperatures (Intergovernmental Panel on Climate Change [IPCC], 2023). According to the WHO, 99% of the world's population breathes air exceeding safe quality limits, causing around 7 million deaths annually from pollution-related diseases (WHO, 2022). Therefore, recent initiatives intend to enforce such warnings in gasoline and diesel vehicle advertising. France, for example, issued a directive to amend Article D.328-3 of its Highway Code requiring car advertising as of March 1, 2022, to include messages encouraging the choice of more environmentally friendly modes of transport (Légifrance, 2021). Specifically, automobile advertisements must include the following information: “Prioritise walking or cycling for short trips”, “Consider ride-sharing” and “Make public transport a part of daily life”. Moreover, advertisements should include the hashtags #SeDéplacerMoinsPolluer and #MovePolluteLess to emphasise the risks of pollution. French regulations mandate that warnings on reducing car use must also appear on television, online, in print and other media in a format that is “easily readable or audible” and “clearly distinguishable from advertising messages and other mandatory information”. Non-compliance may result in fines of up to €50,000. Notably, Ireland is also considering introducing warnings in automobile advertisements targeting high-emission vehicles such as Sports Utility Vehicles (NewsTalk, 2022).

This study proposes a new research direction for ‘warnings to reduce car use’. Given the lack of prior studies in this field, it is crucial to establish a solid theoretical foundation. This study integrates findings from health warning research (in particular from Spain and Latin America) with insights from environmental communication, social marketing, green advertising and car consumption. The focus on these regions is based on considerations which underscore the importance of health and environmental warning research. Firstly, Spain and much of Latin America still face major health challenges linked to these harmful products (Alonso-Ayuso, 2020; Diario, 2025; Lidón-Moyano et al., 2017; Organización Panamericana de la Salud [OPS], 2021). Additionally, the transportation sector accounts for approximately 30% of total greenhouse gas emissions in Spain (Statista, 2024) and 40% in Latin America (Calatayud et al., 2023), while both regions are vulnerable to climate change yet lack robust environmental strategies (IPCC, 2023; United Nations, 2021; World Wide Fund for Nature, 2004). Ultimately, this study aims to not only contribute to further research on health warnings in the region but to also aid in the development of future research related to global vehicle control policies and their implementation.

This study employs an integrative literature review (ILR) method, a systematic approach designed to synthesize and critique diverse literature to generate new theoretical frameworks or research directions. This methodology aligns with the frameworks proposed by Snyder (2019), Torraco (2005) and Whittemore and Knafl (2005), which emphasizes problem formulation, data collection, evaluation, analysis and synthesis to address interdisciplinary research gaps, particularly expanding health warning label research to environmental domains.

The review was driven by the question: “How can insights from health warning label research inform the design and evaluation of environmental warning messages?” This aligns with ILR's purpose of integrating evidence across fields to reconceptualize topics.

The data collection phase was structured in two iterative stages. Searches were conducted in Scopus, ScienceDirect, Web of Science and Taylor & Francis Online using Boolean operators and predefined keywords to capture the most relevant literature. The search fields were titles, abstracts and keywords.

• Stage 1: Health warning literature

The initial search targeted the health sector (tobacco, alcohol, cannabis). Using the core search string (“warning” OR “warning label” OR “health warning” OR “health message” OR “labelling”) AND (“tobacco” OR “cigarette” OR “alcohol” OR “drug” OR “cannabis” OR “marijuana”), studies were identified that examined behavioural outcomes (e.g. risk perception, cessation attitudes and intentions, attention, recall, purchasing behaviour, etc.). From the reviewed literature, we identified key psychological mechanisms in health warnings, such as cognitive dissonance theory, risk perception theory, self-reference effect and fear appeals. Additionally, we identified the academic research fields related to health warnings, such as health communication, social marketing, advertising, packaging and consumption.

• Stage 2: Transition to environmental warning

Based on the psychological mechanisms and research fields identified in Stage 1, we expanded to environmental literature using the following strategies:

1. Theoretical validation search: using combination terms such as “cognitive dissonance theory” AND “environmental warning”; “risk perception theory” AND “environmental warning”; “fear appeal” AND “environmental warning” to examine the application of these theories in the environmental field.

2. Interdisciplinary searches: Combine environmental communication, social marketing, advertising and consumer research with environmental warnings/automotive keywords, such as “environmental communication” AND “warning”; “social marketing” AND “environmental warning”; (“ad” OR “advertisement” OR “advertise”) AND “environmental warning”; “environmental communication” AND (“car” OR “automobile”); “social marketing” AND (“car” OR “automobile”); (“ad” OR “advertisement” OR “advertise”) AND (“car” OR “automobile”).

Through the above process, the five dimensions of the theoretical framework for this paper were constructed: existing literature on health warnings and behavioural psychological theories originate from the first stage; insights from environmental communication, social marketing, advertising and consumption emerge in the second stage. As a result of the searches, a total of 560 documents were initially retrieved.

This study employed a phased screening method to rigorously screen the 560 documents obtained from the initial search, ultimately including 103 studies. First, an initial screening was conducted based on titles and keywords to exclude documents clearly unrelated to the research topic (e.g. pure medical treatments, environmental science, biological and chemical analysis studies), as well as non-empirical studies (e.g. reviews, editorials) and non-peer-reviewed literature (e.g. conference abstracts, book chapters). A full-text review was then conducted, focusing on two core criteria: (1) the study had to examine the impact of warning labels on consumer behaviour, cognition, attitudes or attention; and (2) the publication type had to be either a peer-reviewed journal article or a technical report/policy document issued by a governmental agency.

The 103 documents show distinct industry and regional distribution characteristics: In terms of industry distribution, tobacco-related studies account for the highest proportion (39.8%), followed by alcohol (24.3%), cannabis (18.4%) and other domains such as climate, environmental, car and social marketing research (17.5%); From a geographical perspective, research from North America and Europe dominated (70.9%), while research from Spain and Latin America was relatively scarce (10.7%), with other regions (such as Asia, Africa and Australia) accounting for 18.4%. This distribution pattern provides an important data foundation for this study to focus on Spain and Latin America and propose new research directions and recommendations.

This study draws insights from the field of health warning research that can be applied to environmental warning research. By systematically integrating existing literature, the following key findings have been identified. Current warning research primarily focuses on health-related areas such as cigarettes, alcohol and cannabis. These studies primarily concentrate on two main directions. First, a substantial body of research examines the effectiveness of warnings, assessing outcomes across three dimensions: (1) cognitive outcomes, such as risk perception, attitude change and information recall; (2) behavioural outcomes, such as cessation intent, purchase intent and actual consumption changes and (3) attention mechanisms, including visual attention measured with methods such as eye-tracking. Second, many studies investigate the design of warnings, with particular emphasis on the comparative effectiveness of design elements such as text versus images, gain versus loss framing, message size and location, colour and wear-out effects. This multi-dimensional evaluation framework provides a referenceable methodological system for environmental warning research.

The structure and content of the subsequent sections of this paper are directly derived from the two-stage integrative literature review process described in Section 2.2. First, the behavioural and psychological theories identified in the first stage form the theoretical basis for comparing health warnings and environmental warnings, which is elaborated in Section 3.1. This stage also identifies empirical insights from research across various fields of health warnings, which are discussed in Section 3.2 and combined with the analysis in Section 2.4 to identify research gaps and potential areas for further study. The conceptual framework proposed in Section 3.3 is built around five analytical dimensions, which are derived from the integration of the research findings from the two stages: (1) theoretical and empirical results from health warning research, and (2) perspectives identified in the second stage from environmental communication, social marketing and advertising. This approach ensures that the transition from health warnings to environmental warnings is theoretically grounded.

The analogy between car warnings and health warnings can be based on their links and similarities in theory and practice (see Tables 1 and 2). It is also based on previous research on gambling and environmental communication using analogical methods similar to this article.

Cognitive dissonance theory (Festinger, 1957) suggests that both types of warnings create a conflict between current behaviour and new information, prompting a change in attitude or behaviour. According to the risk perception theory (Slovic, 1987) and the self-reference effect (Rogers et al., 1977), individuals tend to react more strongly to threats that are personally relevant. Health warnings highlight direct and internal risks (harm to consumers themselves), leading to higher risk perception. In contrast, environmental warnings emphasise indirect and external consequences (harm to the environment and others), which generally results in lower risk perception. Given the indirect nature of environmental risks, car-related warnings may evoke less psychological discomfort than health warnings – unless they are framed in a way that emphasises personal health consequences.

Furthermore, Newall et al. (2023) provide a useful precedent by comparing gambling warnings with health warnings. Even though the potential negative consequences of gambling are externalised to a greater extent, health warnings nevertheless provide a valuable framework for addressing them. Similarly, Peattie and Peattie (2009) drew on health-oriented social marketing campaigns (anti-smoking) to explore environmentally-focused social marketing, and they also noted relevant similarities and differences in terms of influencing changes in health and environmental behaviour. Additionally, Jacobs et al. (2018) argue that there may be consistency between consumer responses to health and environmental risk, suggesting that the two should be integrated and therefore conducive to changing behavioural intentions in order to improve public health and the environment. Their study not only establishes a foundation for drawing analogies between the two but also offers valuable insights for designing future car warnings.

By demonstrating the hazards of a product, warnings influence consumer attitudes and behaviour. This section reviews research on health-related warnings, in particular in Latin America and Spain, in relation to existing theories as well as cultural and economic context.

While empirical results on the impact of warnings on smoking cessation remain inconclusive, in Spain, visual health warnings on tobacco packages have been found to significantly increase university students’ awareness of smoking-related risks (Portillo and Antoñanzas, 2002). Qualitative studies indicate that European Commission-designed visual warnings have a greater impact on behaviour than text-only warnings (Rey-Pino et al., 2010). The Spanish national smoking bans reduced its prevalence from 34.5% to 26.1% and decreased second-hand smoke exposure, although the use of hand-rolled cigarettes has risen amongst young people (Lidón-Moyano et al., 2017). Part of this policy failure may be related to a “wear-out effect” (Woelbert and D'Hombres, 2019), wherein prolonged anti-smoking campaigns may actually lead to reduced sensitivity to health warnings among some groups, particularly among Spanish adolescents with a strong social smoking culture.

In Latin America, similar tobacco control policies have shown significant effects, with graphic warnings potentially preventing an estimated 69,369 deaths and saving $5.3 billion in healthcare costs over ten years, while plain packaging could avert approximately 155,857 deaths and save $13.6 billion (Alcaraz et al., 2020). In Brazil research has suggested that picture warnings contributed to a reduction in smoking (Nascimento et al., 2008).

Moreover, research in other countries has yielded similar findings. Studies indicate that warning images can prevent tobacco initiation (Hammond, 2011; Hammond et al., 2003; WHO Regional Office for Europe, 2014) while multi-national studies suggest that health warnings contribute to reducing or quitting smoking. An estimated one in five smokers in the European Union reported that health warnings helped them reduce or attempt to quit (European Commission, 2008). In Canada, picture warnings have been found to significantly impact smokers' intentions to quit as well as the number of cessation attempts and likelihood of starting to smoke (Azagba and Sharaf, 2013). Additionally, smokers who engage with health warnings are more likely to reduce consumption in the following six months (Hammond et al., 2003).

Conversely, some studies have concluded that picture warnings have little dissuasive effect. The implementation of smoke-free laws and health warnings in Hong Kong, for one, was not found to have led to a decline in “hardcore smokers” from 2005 to 2008 (Leung et al., 2016). A study in Canada found that graphic warnings did not significantly impact sales (Irvine and Nguyen, 2021). Similarly, Wilson et al. (2012) could not obtain sufficient evidence to quantify the direct effect of health warning labels. While different studies offer different conclusions, the degree of addiction is relevant to consider, as a strong enough addiction can outweigh the dissuasive effects of warnings (Dewi et al., 2021; Leung et al., 2016).

Another factor which affects smoking prevalence is the tobacco industry's vast economic power. According to Aktan (2018), the profits of the top six multinational tobacco companies in 2013 far exceeded those of Coca-Cola, Walt Disney, FedEx, Google, McDonald's and Starbucks combined. These substantial profits, along with political lobbying, complicate the implementation of control policies. However, while short-term effects may be unclear, some authors suggest that global health warnings will strengthen anti-smoking culture for decades, indicating long-term positive effects (Irvine and Nguyen, 2021).

Studies from Spain, Latin America and other regions show that cigarette warnings are generally effective in reducing smoking but their impact is limited by socio-cultural contexts, wear-out effects from prolonged messaging and the tobacco industry's economic and political power.

The WHO recognises the importance of labelling bottles in its Global Strategy to Reduce Harmful Alcohol Use (WHO, 2010). While research on alcohol labelling is less common than that of tobacco (Staub and Siegrist, 2022), studies show the effects of alcohol labelling reduces consumption. Zhao et al. (2020) found alcohol labelling decreased retail sales, and Hobin et al. (2020) showed that awareness of cancer warnings reduced consumption. Evidence also indicates that cancer warnings can affect high-risk drinkers (Pettigrew et al., 2016) and can lower consumption (Clarke et al., 2021a; Winstock et al., 2020). Other studies suggest different alcohol warnings influence risk perception and consumption (Clarke et al., 2021b; Dimova and Mitchell, 2021; Llance et al., 2020; Maynard et al., 2018; Neufeld et al., 2020; Stafford and Salmon, 2017; Winstock et al., 2020).

Despite indications that health warnings may raise alcohol risk awareness and reduce abuse, findings are not unanimous. A literature review by Stockley (2001) concluded that warnings are generally ineffective in changing consumer behaviour and do not affect alcohol consumption. Other studies suggest these labels have no impact (Brown et al., 2015; Dossou et al., 2017; Krischler and Glock, 2015; Scholes-Balog et al., 2012). Their effect on pregnant women is also limited, as there is no evidence that heavy drinkers reduce consumption during pregnancy (Thomas et al., 2014).

Alcohol consumption remains a major public health problem in Latin America and Spain (Diario, 2025; OPS, 2021) but there is limited relevant research in the region. Fundamentally, policy interventions on the issue in Latin America have been insufficient (Maristela, 2013). Although Spain has taken measures to limit alcohol marketing, according to Sillero-Rejon et al. (2020), current alcohol labelling is insufficient to attract consumers’ attention and larger, more appealing pictorial warnings should be used. Alcohol consumption is influenced by different complex economic and cultural factors which highlights the potential for research in the region.

Research on cannabis warnings is limited compared to tobacco and alcohol (Goodman and Hammond, 2021; Hammond et al., 2020; Mutti-Packer et al., 2018; Winstock et al., 2021) and largely focused on the impact of these warnings on risk perception and product appeal. In 2018, the Canadian government enacted a Cannabis Act requiring health warning labels on packaging (Government of Canada, 2019). A study suggested that mandatory cannabis warning labels increase consumer awareness of its hazards and may enhance exposure to risk information, especially among regular and legal purchasers (Goodman and Hammond, 2021). The rapid growth of this industry is accompanied by lower perceptions of its dangers among youth (Frieden et al., 2016), a problem that could be mitigated by more warning labels (Pepper et al., 2020) and risk information (Goodman et al., 2022). Furthermore, products with health warnings are perceived as less attractive than those without (Goodman et al., 2019, 2021; Leos-Toro et al., 2019, 2021; Mutti-Packer et al., 2018).

Despite strong warnings, cannabis use remains high in Spain, especially among adolescents (Alonso-Ayuso, 2020). This suggests some deficiencies in cannabis risk communication. Adolescents who feel well-informed and whose friends who smoke are more likely to consume cannabis themselves (Belzunegui-Eraso et al., 2020). Cannabis use in Spain is closely linked to its cultural, social and legal environment. There is strong social acceptance, particularly among youth, where cannabis is seen as a social activity or a way to relax (Belackova et al., 2016), with peer pressure playing a key role (Gómez-Fraguela et al., 2008). Moreover, Spain's policy of decriminalisation of cannabis and the legal existence of cannabis clubs provides a relatively liberal legal environment for cannabis consumption (Barriuso, 2011). Therefore, it is crucial to provide Spanish adolescents with effective warnings, strengthen regulatory frameworks and reduce the influence of the social environment on potentially harmful cannabis use.

In Latin America, attitudes toward drugs, including cannabis, are polarised. Some countries are gradually legalising and regulating the market (e.g. Uruguay, Chile) (Hudak et al., 2018), while others maintain strict prohibitions (e.g. Bolivia, Brazil) (Mendiburo-Seguel et al., 2017). Additionally, efforts to destigmatise cannabis, such as a study in Sinaloa, Mexico, have shown that education can shape perceptions, leading to more positive or neutral attitudes, potentially reducing stigma (Camberos-Barraza et al., 2023). This contrasts with the limited practice of health warnings on cannabis products in Europe. In general, cannabis health warnings in Latin America are not widespread, research on their effectiveness is minimal, and policy primarily focuses on combating illicit trafficking and reducing social harms rather than information about health risks.

Varied or conflicting results can potentially be explained in terms of several factors or theories which affect decision-making and human behaviour.

Most studies have concluded that warnings have a positive effect on improving adverse behaviours, results which can be explained in terms of protection motivation theory. When faced with a health threat, individuals decide whether to adopt protective behaviours based on their perceived severity, susceptibility, coping efficacy and self-efficacy (Rogers, 1975). Warnings enhance risk perception and can, when paired with supportive measures like quitlines, improve coping efficacy and perceived feasibility of quitting. Social norms theory (Cialdini, 2003) also supports these findings, suggesting that collective behaviour patterns and social consensus may enhance message effectiveness. These insights confirm the potential of strengthening health warnings to reduce harmful behaviours.

For studies which have found health warnings to be ineffective, cognitive dissonance theory (Festinger, 1957) may offer a partial explanation. When there is a discrepancy between an individual's cognition and behaviours, they feel psychological pressure to reconcile the two. In addition to changing smoking behaviour, smokers may ease dissonance by rationalising their behaviour, such as denying health risks or downplaying the harms of moderate smoking.

Overall the consensus of various studies on the effectiveness of visual warnings can be expressed in the old adage “A picture is worth a thousand words”. Visuals and images evoke stronger fear responses in the human brain and better motivate behavioural change.

While such fear-based behaviour avoidance aligns with fear appeal theory (Witte, 1992), improved perceptions and intentions do not always translate into behavioural change. This phenomenon can be explained through motivated reasoning (Kunda, 1990), the health belief model (Janz et al., 1984) and the attitude-behaviour gap (Ajzen and Fishbein, 1977). Firstly, motivated reasoning suggests individuals accept beliefs that justify their behaviour while ignoring contradictory evidence. In cultures with entrenched drinking habits, such as parts of Europe, people may disregard warnings in favour of their routines (Staub and Siegrist, 2022). Secondly, according to the health belief model, whether an individual will change their behaviour depends on the perception of disease severity and susceptibility. Where moderate alcohol consumption is seen as beneficial (Annunziata et al., 2016), perceived harm is lower (Bocquier et al., 2017), limiting warning impact. Additionally, even when aware of risks, individuals may not adjust their behaviour – a phenomenon explained by the attitude–behaviour gap. Ultimately, socio-cultural norms and the pursuit of instant gratification (Staub et al., 2022) can reduce the effectiveness of health risk messages.

A conceptual framework for the application of warnings in the automotive sector is proposed, based on five analytical dimensions.

Table 3 summarises how findings from prior research on health warnings can be applied to car warnings, highlighting similarities between the two in terms of mechanism of action and regulation, as well as key areas for future research. Health warnings primarily target risk perception related to personal health threats, while car-related warnings may need to emphasise both environmental risks and health-related risks to enhance their perceived relevance. Regarding attitudes, cognition and behavioural intentions, prior research has explored how health warnings influence risk perception, awareness, product appeal and behavioural intentions. Similarly, car warnings warrant investigation into not only their effects on environmental and health risk perception but also in relation to consumer intentions to reduce car purchases or usage. In terms of attention, numerous studies have analysed how different design features of health warnings influence visual engagement. Car warnings could benefit from similar approaches, employing eye-tracking methods to explore the effects of format, framing, size and other relevant factors. While health warnings have been shown to influence actual behaviours, whether car warnings can lead to behavioural shifts in car consumption (e.g. usage frequency) remains an open question. Finally, unlike health warnings, which are supported by decades of implementation and robust empirical evidence, car warnings are still in their infancy. Therefore, future research findings will also contribute to the development of regulatory frameworks for car warning policies.

According to cognitive dissonance theory (Festinger, 1957), when people see car warnings, they may experience cognitive dissonance if their behaviour (e.g. frequent use of a private car) is inconsistent with their environmental values. To reduce this discomfort, they may adjust their behaviour (drive less) or attitudes (reassess the warning message). Research has shown that fear appeals (Witte, 1992) can improve message persuasion; cognitive load theory (Sweller, 1988) suggest that graphic messages require fewer cognitive resources than textual messages and can be understood and remembered more quickly, thus facilitating more effective information processing. Recent research has further investigated the role of affective visual cues in enhancing the persuasiveness of warnings (Peters et al., 2019), highlighting the idea that high emotion pictorial warnings are most effective in conveying messages. It can therefore be predicted that negative picture-based car warnings may be more effective than positive text-based warnings. Research in environmental psychology has shown that there is a significant gap between people's environmental attitudes and their behaviour (Ajzen and Fishbein, 1977; Kaiser et al., 1999). Therefore, while car warnings may increase consumers’ environmental awareness or intention to reduce consumption, they may not lead to a reduction in car ownership levels or usage.

Environmental communication plays a crucial role in shaping public awareness and behavioural change. Previous studies on environmental communication have laid the groundwork for the exploration of “car warnings”, as climate change-related images and texts have been found to increase perceived efficacy, which, along with risk perception, is positively correlated with pro-environmental behaviours (Hart and Feldman, 2016; Maartensson and Loi, 2021). Research on environmental advertising or environmental messages suggests that message framing has a significant impact on consumer responses. Gifford and Comeau (2011) noted that perceived climate change competence, engagement and behavioural intentions were much greater after exposure to motivationally-framed messaging than after messaging involving sacrifice. Nevertheless, some studies have also pointed out that positive environmental pictures, while increasing pro-environmental attitudes, are not sufficient to influence their behaviour and that a gap exists between attitudes and behaviour (Soutter and Boag, 2019). In the context of tourism, Song et al. (2023) investigated the effectiveness of message framing in different scenarios, noting that prescriptive-framed appeals (appeals that offer advice or directions) were more effective than proscriptive-framed appeals (appeals that focus on restrictions) with proscriptive-framed appeals being more effective in daily contexts.

From a social marketing perspective, Peattie and Peattie (2009) emphasise the potential of social marketing to reduce consumption and promote sustainable development. Environmental warnings in car advertisements are similarly designed to influence consumers' purchasing decisions through the promotion of environmental messages. However, Corner and Randall (2011) reveal the limitations of social marketing in addressing complex environmental issues. These limitations may also apply to “car warnings”. Specifically, environmental warning messages may face challenges due to simplification, over-reliance on individual behavioural change, inappropriate message dissemination and neglect of consumer values and identity. Inconsistent findings on the effectiveness of social marketing to improve environmental issues has further stimulated research on “car warnings”. Exploring the effectiveness of car-related warnings and how to effectively convey environmental messages in car advertisements is therefore a worthy topic of research.

While an increasing number of automobile consumers are now considering “positive environmental impact” as a factor in purchasing a vehicle (AlixPartners, 2022), research on green or eco-friendly advertising in the automotive industry suggests complex consumer responses. Consumers tend to exhibit brand bias when evaluating eco-sustainability factors in car purchases (Russo et al., 2015), and the effectiveness of green advertising is influenced by the environmental credibility of the advertiser and the green attributes of the product (Kim and Han, 2015). Eco-friendly marketing and environmental knowledge influence purchase intention of eco-friendly automotive products (Kathuria and Nigam, 2024), and the quality of green advertising messages – especially their usefulness, attractiveness and truthfulness – significantly affects consumers’ responses to these advertisements as well as their willingness to buy the related products (Wang and Li, 2022). These findings highlight the complexity of environmental communication in the automotive industry, and factors such as branding, environmental knowledge and message quality should be considered in future research on automotive warnings.

Given these conceptual frameworks, future research could explore how “car warnings” affect consumer attitudes, cognition, behavioural intentions, visual attention and actual behaviour, as well as the effects of different design elements and quality of warning. Moreover, the implementation of warnings in automotive advertising may encounter regulatory challenges and industry resistance. As with tobacco, economic interests, the symbolic value of cars as status and freedom and consumer lifestyles may hinder such measures (Aktan, 2018; Steg, 2005). Altogether these various factors and challenges merit further attention from policymakers and researchers.

This study extends the application of warning messages to the automotive domain, introducing a new research direction. Drawing on existing literature and theoretical perspectives, this study establishes a conceptual framework to guide future research.

The findings contribute to a better understanding of how warning messages work in non-health domains, thus bridging the gap between health and environmental warning research. By integrating perspectives from health warnings, environmental communication, social marketing and advertising, this study lays the foundation for further interdisciplinary theory development. In addition, it extends behavioural and psychological theories – including theories of cognitive dissonance, risk perceptions, fear appeals and self-reference effect – into the unexplored area of environmental warnings in the automotive industry, which is closely linked to society and the environment.

From a regulatory perspective, the findings provide initial guidance on the design of environmental warnings in car adverts. Consistent with evidence from health communication (e.g. Hammond, 2011), combining text and images and experimenting with the gains and losses framing may increase the salience and persuasiveness of messages. However, empirical research is needed to determine whether these strategies can be effectively applied in the automotive sector. For companies, such warnings could be incorporated into advertising and corporate social responsibility strategies to encourage more sustainable consumption. Their implementation may also help strengthen corporate image as responsible actors in the face of the climate crisis, enhance brand reputation and generate a competitive advantage in markets increasingly sensitive to environmental issues. At the same time, the introduction of these messages opens up a debate on the role of counter-marketing as a tool for promoting sustainability, since it involves using advertising spaces, traditionally intended to stimulate consumption, to convey messages that aim to reduce car use. Importantly, the analysis highlights the lack of empirical evidence in Spain and Latin America, where the introduction of such warnings may have important implications for both public policy and marketing practices.

Caution should be exercised in interpreting the results of the study. Comparisons between health warnings and environmental warnings were conceptual rather than empirical, and the two types of warnings differed considerably in terms of the medium in which the information was provided (packaging/advertising), as well as in terms of personal relevance and perceived risk, which limits the generalisability of the findings. Future research should empirically test car warnings and their design elements to assess their actual impact on attitudes, risk perceptions and behaviour.