Introduction
Globally, more than half of the population lives in urban environments (UN, 2022). These are characterised by a complex and diverse set of interacting environmental exposures that shape and emerge from urban life. Interest in how urban life, namely, urban environmental exposures, shapes the health of urban inhabitants has a longstanding history (Galea and Vlahov, Reference Galea and Vlahov2005). Yet, in the past 30 years, this research avenue has garnered greater attention (Ribeiro et al., Reference Ribeiro, Gullón and Murray2022; Ribeiro, Reference Ribeiro and Liamputtong2025). With urban population growth, it becomes increasingly important to understand and improve the impact of the urban habitat on health outcomes (Stimson, Reference Stimson2013; Caprotti, Reference Caprotti2018). Also, new urban challenges, such as pollution, heat, and chronic and mental health burdens, require an integrated understanding of the health effects of urban exposures.
Exposome research addresses the health impacts of the totality of environmental exposures across the life course, including both the internal exposome – such as omics-based biomarkers – and the external exposome shaped by social and physical environments (Vineis et al., Reference Vineis, Robinson, Chadeau-Hyam, Dehghan, Mudway and Dagnino2020). The present manuscript focuses specifically on the general external urban exposome, the level at which integrating qualitative and quantitative methods is particularly meaningful and methodologically fruitful. The notion of the ‘urban exposome’ has been proposed to describe the totality of environmental exposures external to the individual present in an urban environment, from pollutants in air (both indoor and outdoor) and water to built features such as housing, parks, and walkability (Robinson et al., Reference Robinson, Tamayo, de Castro, Valentin, Giorgis-Allemand, Hjertager Krog, Marit Aasvang, Ambros, Ballester, Bird, Chatzi, Cirach, Dėdelė, Donaire-Gonzalez, Gražuleviciene, Iakovidis, Ibarluzea, Kampouri, Lepeule, Maitre, McEachan, Oftedal, Siroux, Slama, Stephanou Euripides, Sunyer, Urquiza, Vegard Weyde, Wright, Vrijheid, Nieuwenhuijsen and Basagaña2018). The field of urban exposome research is both ambitious and promising as it comprehensively incorporates all lifelong cumulative environmental exposures. These exposures are subsequently linked to diverse health outcomes across multiple stages of the life course, enabling the evaluation of potential causal relationships. This approach has the potential to minimise residual confounding by assessing a large amount of simultaneous and interacting environmental exposures and strengthen causal inference by exploring the pathways linking urban exposures to health outcomes (Vermeulen et al., Reference Vermeulen, Schymanski, Barabási and Miller2020; Fang et al., Reference Fang, Hu, Chen, Guo, Liu, Lan, Gong and Wang2021). Past and ongoing exposome projects have been fruitful in overcoming challenges associated with exposome assessment, via the development of standardised exposure assessment, sophisticated and multifactorial statistical regressions, and data harmonisation procedures, which limit selective reporting and deal with false positives (Valérie et al., Reference Valérie, Lydiane and Rémy2016).
Many urban exposome-related projects rely heavily on information derived from birth cohort studies, in which the urban exposome is often characterised using Geographic Information Systems (GIS) and related geotechnologies. These tools link participants’ residential locations and movements with geospatial datasets, enabling the quantification of environmental exposures throughout their life course and their association with health outcomes. GIS is a potent powerhouse that, with the abundance of open-access datasets obtained from earth observation systems, allows the assessment of millions of environmental exposures, which are analysed using advanced statistical methods and machine learning algorithms (Lakerveld et al., Reference Lakerveld, Wagtendonk, Vaartjes, Karssenberg, Lakerveld, Penninx, Beulens, Timmermans, Huisman, Wagtendonk, Kramer, van Wier, Boomsma, Willemsen, Schuengel, Oosterman, Stronks, Karssenberg, Vermeulen, Vaartjes, Koster, Stehouwer, van den Hurk, Koomen, de Mutsert, Mt, Verschuren, Picavet, Beenackers, van Lenthe, Ikram, Jaddoe, Oldehinkel, de Jong, Mulder, Dotinga and Consortium2020). Some urban exposome studies also incorporate direct exposure assessments using mobile and personal sensing technologies, along with self-reported questionnaires. This approach captures short-term and spatial variations in participants’ environmental exposures, including indoor environments (Turner et al., Reference Turner, Nieuwenhuijsen, Anderson, Balshaw, Cui, Dunton, Hoppin, Koutrakis and Jerrett2017).
Despite notable methodological progress, urban exposome research is mostly data-driven, relying heavily on large datasets and computational methods to identify patterns and associations without predefined hypotheses, and there has been a comparatively limited focus on contextualising the observed effects (Patel, Reference Patel2019). This paper is grounded in the premise that integrating quantitative and qualitative methods in epidemiological research within a biosocial framework provides a robust approach to examining the intricate relationships between urban environments and health outcomes. The initial section discusses this integration, followed by a discussion on the challenges of implementation and the necessity of embracing interdisciplinarity. Finally, birth cohorts are presented as a promising platform for employing mixed methods.
Overcoming urban exposome research limits with a biosocial perspective
Few projects seeking to operationalise the exposome have addressed the broader and more complex social, economic, cultural, and political forces – conceptualised as distal, outer-layer exposures – that connect individuals to their environments and perpetuate exposure to environmental hazards (Senier et al., Reference Senier, Brown, Shostak and Hanna2017; Prior et al., Reference Prior, Manley and Sabel2018).
Social-exposome research has developed conceptual and methodological tools that may help contextualise exposures and reconstruct their fabric. However, interdisciplinary approaches that draw on knowledge from both the biological and social sciences remain underutilised in the literature, often resulting in overly simplistic approaches to defining neighbourhoods and subsequently exposures.
Despite its many virtues, the quantitatively-oriented urban exposome research tends to define the neighbourhood as a geographic and space-oriented component, ignoring the more social and people-oriented component tied to the concept of ‘place’ – which some describe as a unique set of resources/exposures and social relationships interwoven within geographical boundaries (Bernard et al., Reference Bernard, Charafeddine, Frohlich, Daniel, Kestens and Potvin2007). In a relational approach, places are viewed as multi-scalar and dynamic nodes in networks (instead of fixed bounded places); separated by socio-relational distance (instead of physical/metric distance); composed of highly mobile populations (instead of immovable local populations); and are contingent on social power relations and hold cultural meaning (instead of being neutral) (Cummins et al., Reference Cummins, Curtis, Diez-Roux and Macintyre2007). For instance, consider a hypothetical study on the effects of food outlet availability and food insecurity among older urban residents. A quantitatively oriented approach might draw a buffer around a person’s home and quantify the number and diversity of food outlets, followed by a statistical regression analysis to conclude that no associations exist. In contrast, a relational approach might involve qualitative methods to inquire, with depth, flexibility, and nuance, about residents’ activity spaces (locations visited on a daily basis) and levels of access to food and finding that they no longer shop in the local neighbourhood because gentrification has led to rising prices and the transformation of local businesses into trendy health food stores, as suggested by a Portuguese cohort study conducted by us (Santos et al., Reference Santos, Silva, Astell-Burt, Barros, Torres and Ribeiro2024). This would help explain the lack of statistical association between food outlet availability and food insecurity observed in the quantitative analysis, which can eventually lead to the construction of unforeseen hypotheses.
This leads us to the biosocial perspective, which conceptualises the biological and the social as entangled domains (Harris and McDade, Reference Harris and McDade2018). This is fundamental for causal epidemiology, as single environmental factors, taken in isolation and decontextualised, only play a partial role in the aetiology of multifactorial diseases (Giroux, Reference Giroux2023). For instance, establishing a statistical association between exposure to a specific air pollutant and a given disease without considering the contextual history in which this exposure occurs (e.g., the establishment of a chemical factory in a disadvantaged urban area) provides an incomplete understanding of the disease’s aetiology in that community. Besides, urban life is typically not experienced through isolated factors; rather, these elements are intricately interconnected and cannot be meaningfully separated (Söderström et al., Reference Söderström, Empson, Codeluppi, Söderström, Baumann and Conus2016). Failure to account for these aspects would impede the ability to intervene effectively.
Qualitative methods can help address these challenges – including gaps in contextualisation, biosocial integration, and neighbourhood definitions – and complement quantitative data to advance urban exposome research. Employing a combination of multiple research methods – multi-methods studies (D’Angelo and Her, Reference D.’Angelo and Her2019) and integrating both qualitative and quantitative approaches (mixed-methods studies (Creswell and Clark, Reference Creswell and Clark2017) – allows tapping into the strengths of both quantitative, extensive measurements and qualitative, in-depth, and contextualised accounts. This combination can help to obtain a deeper understanding of the situated phenomena by revealing and contextualising locally acting mechanisms (and their wider connections) relating urban environmental factors and health (Cummins et al., Reference Cummins, Curtis, Diez-Roux and Macintyre2007), which is needed for both intervention and hypotheses development. With its open-endedness, flexibility, and focus on situated meaning-making processes and practices, the integration of a qualitative perspective may reveal relevant information that could otherwise remain hidden.
However, combining qualitative approaches with the quantitative assessment of exposures also poses specific challenges. Despite the emergence and achievements of the mixed-methods movement, research combining qualitative and quantitative methods remains underutilised (Knappertsbusch, Reference Knappertsbusch2023). Furthermore, in specific multi- and mixed-methods research, qualitative methods are often fitted under a dominant quantitative framework. While mixed methods proponents have advanced a philosophical underpinning for their methodological project based on pragmatism (Morgan, Reference Morgan2014), and while the idea that qualitative and quantitative research are tied to different, cohesive and incompatible ontological and epistemological ‘paradigms’ has been thoroughly criticised (Kelle and Reith, Reference Kelle and Reith2023), one can still argue that researchers with quantitative or qualitative training, and, perhaps more so, from different disciplines, will often share different ontological and epistemological viewpoints. However, these tensions may be productively worked, as demonstrated, for instance, by Roberts and colleagues’ brand of bioethnography (2021).
Promising qualitative methods and the unique role of bioethnography
Ethnographic approaches (Hammersley and Atkinson, Reference Hammersley and Atkinson2019) allow researchers to build a detailed and in-depth understanding of a given phenomenon while accounting for its context. Diving into the field for long time periods, the ethnographer may employ all five senses and use several methods, such as participant observation, several types of interviews, and audio and video recordings, among others, to collect data. This typically invites an open-ended approach and results in a large volume of situated, rich, qualitative data. Bioethnography (Roberts and Sanz, Reference Roberts, Sanz, Meloni, Cromby, Fitzgerald and Lloyd2018; Roberts, Reference Roberts2021) goes beyond merely combining ethnographic and quantitative approaches. It critically interrogates the assumptions underlying both methodologies, fostering innovative ways of understanding health phenomena. Rather than subordinating qualitative insights to dominant quantitative frameworks, bioethnography seeks to synthesise these perspectives, emphasising the contextual, complex, and multidimensional nature of health. Roberts (Reference Roberts2021) describes bioethnographic endeavours as making better numbers together with epidemiologists and statisticians since the beginning of a given research project. This requires mutual adjustments, such as including an open-ended ethnographic stage to inform hypotheses generation instead of beginning with fixed hypotheses, or struggling to convert vast, untidy ethnographic data into numerical variables.
Qualitative methods, described in the next paragraphs, can complement quantitative approaches in bioethnographic research by providing contextual and interpretative insights, helping to generate stronger hypotheses and deepen understanding in urban health and exposome studies (Silva and Ribeiro, Reference Silva and Ribeiro2024).
Open-ended participant observation is one such method. It is a discovery-based ethnographic method that enables researchers to closely observe people in their natural, real-life contexts, uncovering the complex social, cultural, and environmental factors that shape their experiences over time, generating fine-grained data about human–environment relations (Hardon et al., Reference Hardon, Téllez-Rojo, Anastario, Tan, Alcala, Echague, Kuritzky, Gordon, Boudart, Sandoval and Roberts2025). For instance, open-ended participant observation added value in the Mexico City ELEMENT study by uncovering why families continued using lead-glazed trastes de barro despite years of health education – because of their deep cultural and social significance – providing information that surveys could not, and enabling more culturally attuned intervention strategies (Hardon et al., Reference Hardon, Téllez-Rojo, Anastario, Tan, Alcala, Echague, Kuritzky, Gordon, Boudart, Sandoval and Roberts2025).
Interviews (including go-along interviews) coupled with environmental and biosensing represent a valuable duo. The go-along interviews involve a researcher walking or travelling alongside a participant in a familiar environment and asking questions along the way (Clark, Reference Clark2017). This enables the researcher to integrate observational data on how the interviewee engages with a specific socio-material context, anchoring their accounts within that very environment. Simultaneously, the participant may carry a Global Positioning System (GPS) to track locations, environmental sensors to measure biophysical exposures, and biosensors to assess physiological responses. For instance, to assess what are the physiological and emotional responses of individuals with psychosis when exposed to different sensory and built environments in urban settings, authors have used a mixed-method approach, which includes tracking physiological responses (like skin conductance), geolocating participants via GPS, and conducting go-along interviews to capture qualitative data (Winz and Söderström, Reference Winz and Söderström2021). Alternatively, this journey can be made without the presence of the researcher, and an interview is conducted after the walk to identify and describe significant moments (Paiva et al., Reference Paiva, Gonçalves, Ferreira, Pedro and Portugal2023). In this scenario, the researcher cannot directly observe how the interviewee interacts with the chosen route and navigates the physical and social environment. However, the participant retains the ability to observe and report their experiences. Additionally, the researcher can utilise data collected by sensors to refine the interview guide, ensuring a comprehensive understanding of the participant’s interactions within their environment.
This multiple approach enriches the understanding of individual experiences in their activity spaces and reveals short-term or immediate links between how individuals experience urban environments (exposures) and their health responses (Prior et al., Reference Prior, Manley and Sabel2018). By tracking an individual’s activity spaces through walk-along interviews, one may also obtain a more truthful and justified delimitation of each person’s neighbourhood, contributing to a more accurate assessment of a person’s urban exposome. This is of considerable importance since the neighbourhood limits often used by public health researchers may resonate little with the lived experience of space (Petrović et al., Reference Petrović, Manley and van Ham2019). By providing insightful information on how people navigate space, this approach could also provide valuable data on how diverse people are differently exposed to various risk factors, which affect them differently, as their relationship with places is also different. This type of misalignment can bring valuable insights, revealing gaps in knowledge or areas for further study.
Images can also be particularly advantageous. Visual methods utilise visual data, ranging from photographs to videos (Emmison et al., Reference Emmison, Smith and Mayall2012). Among the different visual methods, photovoice has become widely popular in urban health research (Gravina et al., Reference Gravina, Jauregi, Estebanez, Fernández-Aedo, Guenaga, Ballesteros-Peña, Díez and Franco2020; Hatala et al., Reference Hatala, Njeze, Morton, Pearl and Bird-Naytowhow2020; Silva et al., Reference Silva, Santos, Torres, Martínez-Manrique, Barros and Ribeiro2023). It uses photographs to reveal meaning (Sutton-Brown, Reference Sutton-Brown2014), but it is also concerned with promoting collective reflection about community issues and fostering social change (Wang and Burris, Reference Wang and Burris1997). Besides its capacity to empower communities and promote collective action, photovoice yields a diverse range of spatialised information about the urban micro-environments not easily assessed using the conventional geospatial technologies commonly employed in urban exposome research, such as food quality and diversity (Gravina et al., Reference Gravina, Jauregi, Estebanez, Fernández-Aedo, Guenaga, Ballesteros-Peña, Díez and Franco2020); spatialised emotions (Meenar and Mandarano, Reference Meenar and Mandarano2021); trash, disorder, and drug and alcohol consumption (D’Angelo and Her Reference D.’Angelo and Her2019).
Group interaction methods are highly valuable as well. This is the case of focus groups (Kitzinger, Reference Kitzinger1995) and the world café (Löhr et al., Reference Löhr, Weinhardt and Sieber2020). These methods involve a collective discussion focused on a certain topic, moderated by a research team member. They can be particularly useful to study group dynamics, brainstorm ideas, capture collective opinions, and eventually reach a solution related (in this case) to the creation of healthier urban environments. For instance, a group of participants could use sensors to collect data on exposures and physiological processes in a given setting (e.g., home, school, work) and then collectively discuss the results of the analysis of that data, including the discussion of measures and solutions to reduce those exposures. Within the Generation XXI birth cohort (Portugal), to deepen the understanding of the interconnections between green space and health – previously studied through quantitative space-based approaches (Cavaleiro Rufo et al., Reference Cavaleiro Rufo, Paciência, Hoffimann, Moreira, Barros and Ribeiro2021; Paciência et al., Reference Paciência, Moreira, Moreira, Cavaleiro Rufo, Sokhatska, Rama, Hoffimann, Santos, Barros and Ribeiro2021; Queiroz Almeida et al., Reference Queiroz Almeida, Paciência, Moreira, Cavaleiro Rufo, Moreira, Santos, Barros and Ribeiro2022) – a multidisciplinary team conducted focus groups with over 60 participants. Through this study, they were able to better comprehend how adolescents engage with green spaces, identify perceived health benefits, and explore the barriers and facilitators influencing their utilisation of these spaces (Sampaio Costa et al., Reference Sampaio Costa, Queiroz Almeida, Silva, Barros, Ribeiro and Leão2024). Importantly, this study also enabled us to uncover new potential pathways underlying previous associations found between green space exposure and intelligence (Queiroz Almeida et al., Reference Almeida, Barros and Ribeiro2022) – one of them being the use of green spaces for creative activities and for relieving stress related to schoolwork – as well as links to respiratory health (Queiroz Almeida et al., Reference Queiroz Almeida, Paciência, Moreira, Cavaleiro Rufo, Moreira, Santos, Barros and Ribeiro2022), particularly the role of cleaner air, especially beneficial for adolescents already experiencing respiratory problems.
Regarding integration, and drawing on established mixed-methods principles (Creswell and Clark, Reference Creswell and Clark2017), upstream, simultaneous, and downstream combinations of qualitative and quantitative components represent appropriate and complementary pathways for strengthening urban exposome research. Upstream, qualitative methods, such as ethnography, go-along interviews, and participatory mapping, can help refine environmental exposure definitions prior to GIS assessments and quantitative modelling in urban exposome research. The simultaneous collection of qualitative and quantitative data, for instance, through the combination of GPS tracking, environmental sensing, interviews, and visual data, allows for the contextualisation of real-time environmental fluctuations and enhances the identification of underlying mechanisms. Downstream, qualitative insights can be used to establish comparisons, identify similarities, complementarities, and divergences (O’Cathain et al., Reference O’Cathain, Murphy and Nicholl2010), and explain unexpected quantitative findings. In doing so, they enhance causal inference and support the development of more robust hypotheses and theories.
Addressing challenges and embracing interdisciplinary tensions
Despite the many opportunities, integrating quantitative methods and measurements of exposures with qualitative descriptions in exposome research may raise several challenges. Some of these challenges are described by Roberts and Sanz (Roberts, Reference Roberts, Sanz, Meloni, Cromby, Fitzgerald and Lloyd2018). First, ethnography (and much of qualitative research) and quantitative epidemiological and public health research have very different practices, cultures, and rhythms. While ethnographic and other qualitative approaches evolve iteratively, epidemiological designs – despite collecting extensive, multi-domain data in exposome birth cohorts – still require predefined hypotheses and standardised measures through time, which limit flexibility to introduce new variables over time. Despite the advances in the mixed methods field, effectively translating ‘thick descriptions’ (Geertz, Reference Geertz1973) of phenomena from qualitative studies into operational hypotheses for public health and epidemiological research remains challenging. Qualitative data often provide rich insights into situated phenomena that are not easily broken into isolated and quantifiable variables, making it difficult to integrate them into the structured frameworks typically used in epidemiological studies. Conversely, the numerical data organised by epidemiologists to meet their research objectives may lose sight of context and how it shapes the phenomena that the former intend to represent, leading to potential misalignment between the two approaches.
Second, research teams working in exposome research need diverse skills, including expertise in GIS, sensing technologies, statistical analysis, and qualitative research methodologies, among others. Additionally, each researcher on the team may need a basic grasp of several different skills. Adequate instruction is essential to ensure a harmonised and effective research process, but training experiences so far have focused on quantitative methodologies (Niedzwiecki Megan and Miller Gary, Reference Niedzwiecki Megan and Miller Gary2017).
Third, integrating qualitative approaches into epidemiological or exposome research introduces new interdisciplinary tensions regarding ethics and ethical oversight. Indeed, the increasing influence of ethical oversight procedures and principles imported from the health sciences into the social sciences has been considered a challenge for ethnography (Cefaï and Costey, Reference Cefaï and Costey2009; Hammersley, Reference Hammersley2018). The open-ended and flexible nature of such approaches fits poorly with the predictability of procedures and rigid plans expected by ethics committees in the health sciences. Moreover, participatory methods, such as photovoice, raise specific ethical challenges related to power issues, the rights of groups or communities, ownership and dissemination of data, anonymity and confidentiality (and tensions between these and the former), and the blurring of the distinction between researchers and research participants. These may also be difficult to reconcile with the ethical oversight of medical research and its focus on pre-defined protocols, clear distinction between professional researchers and research participants (Durham Community Research Team, 2011), and reliance on individualistic notions such as personal autonomy and privacy and the procedure of informed consent (Williams, Reference Williams2005).
Nonetheless, tackling the challenges of interdisciplinary and mixed-methods research holds the potential to generate ground-breaking knowledge, practices, and relationships that surpass traditional disciplinary boundaries and conventional metrics of accountability and innovation, paving the way for transformative and inventive outcomes (Barry and Born, Reference Barry and Born2013).
Birth cohort studies: a promising platform for biosocial research
Cohort studies offer a robust foundation for bioethnographic research, owing to their longitudinal nature and capacity to integrate diverse disciplinary approaches. By following individuals across life stages, cohort studies capture rich, multidimensional data on social, biological, and environmental factors that shape health outcomes. Moreover, cohort study teams are typically interdisciplinary, comprising epidemiologists and experts from various backgrounds, such as medicine, biology, sociology, and geography, all of whom are proficient in multiple research methodologies. As highlighted in the Biosocial Birth Cohort Research Network’s work, combining cohort studies with bioethnographic insights allows researchers to address complex health disparities by integrating nuanced social narratives with robust biological data, enriching the overall scope of exposome research (Gibbon et al., Reference Gibbon, Roberts, Hardy, Béhague, Téllez Rojo, Goncalves-Soares, Mathers, Penkler, Fraga, Wooyoung Kim, Pentecost, Tabor, Jeffries-Hein, Lappé, Borra, Rossmann, Lloyd, Filipe and Silva2024).
However, birth cohorts rarely employ mixed methods systematically. Hence, it is crucial to develop hybrid methodologies that adapt ethnographic techniques for systematic application within longitudinal frameworks. Such enhancements could include, for instance, periodic ethnographic ‘snapshots’ or participatory approaches, including participatory mapping (Hasanzadeh, Reference Hasanzadeh2022), that engage cohort participants as co-researchers to capture evolving experiences over the space-time continuum. Such snapshots would also be valuable for generating new hypotheses and guiding quantitative data collection in subsequent waves of the cohort.
Conclusion
Integrating qualitative and quantitative approaches within the study of the urban exposome offers a powerful pathway to reveal the contextualised, relational, and biosocial processes that shape how the urban outdoor exposome influences health. By bridging methodological traditions, researchers can generate richer hypotheses, refine exposure assessments, and illuminate mechanisms that remain hidden in purely data-driven frameworks. Advancing this interdisciplinary agenda, particularly through longitudinal cohort infrastructures, will be essential for developing more nuanced and actionable evidence about the complex interplay between cities and health.
Funding statement
This work was supported by FEDER through the Operational Programme Competitiveness and Internationalisation and national funding from the Foundation for Science and Technology – FCT (Portuguese Ministry of Science, Technology and Higher Education) under the Unidade de Investigação em Epidemiologia – Instituto de Saúde Pública da Universidade do Porto (EPIUnit) (UIDB/04750/2020), Instituto de Sociologia da Universidade do Porto (UID/00727/2025), and Laboratório para a Investigação Integrativa e Translacional em Saúde Populacional (ITR) (LA/P/0064/2020). Ana Isabel Ribeiro was supported by National Funds through FCT, under the ‘Stimulus of Scientific Employment – Individual Support’ programme within the contract CEECIND/02386/2018. This investigation also received funding from the European Union’s Horizon 2020 research and innovation programme (ATHLETE, grant agreement number 874583). This research also received support from the Centre of Studies in Geography and Spatial Planning (CEGOT), funded by national funds through the Foundation for Science and Technology (FCT) under the reference UIDB/04084/2025.
Competing interests
The authors declare no conflicts of interest. Funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Ethical standard
Not applicable.
Open access
