The Epidemiology of Working Life: Reorganizing Occupational Health Science Around Work‐Structured Exposure Systems

ABSTRACT

Occupational epidemiology has played a central role in identifying workplace hazards and informing prevention. However, its focus on discrete exposures, bounded worksites, and linear models is insufficient to capture how contemporary work shapes health. Many risks arise not from isolated hazards, but from cumulative, interacting exposures generated by work organization that extend beyond the workplace and accumulate across working life. This paper proposes the Epidemiology of Working Life (EWL) as a framework for reorganizing occupational health science. Anchored in the working‐life‐exposome, EWL centers analysis on work‐structured exposure systems—structures that generate recurring configurations of co‐occurring exposures under specific employment conditions. EWL changes what counts as exposure, how risk is analyzed, and where prevention can act. Rather than treating work as a setting of discrete hazards, EWL treats it as a system that generates and distributes risk across working life. EWL shifts the unit of inference from individual hazards to recurring exposure configurations and the systems that generate them. This shift enables earlier detection of risk and identifies intervention points that reduce multiple harms simultaneously—less visible in hazard‐by‐hazard approaches. EWL yields three prevention‐oriented outputs: (1) recurrent exposure configurations associated with elevated risk, (2) upstream drivers that generate and sustain them, and (3) early‐warning indicators for anticipatory intervention. Proof‐of‐concept scenarios illustrate how EWL reveals prevention opportunities obscured when exposures are examined in isolation. By centering occupational health science on work‐structured exposure systems, EWL strengthens alignment between epidemiological evidence and prevention in contexts where risks are cumulative, interacting, and structured by employment conditions.

1. Introduction: Why Occupational Epidemiology Must Evolve

Occupational epidemiology has been one of the most influential sciences in public health, identifying harmful workplace exposures and informing regulations that have prevented disease and injury [1]. Its strength lies in linking specific hazards to specific outcomes and translating that evidence into prevention.

However, many contemporary work‐related risks no longer arise from isolated hazards alone. Instead, they emerge from combinations of conditions shaped by how working life is organized—for example, unstable scheduling, income volatility, extended or irregular hours, algorithmic management, and limited access to protections [2, 3]. These conditions interact and extend beyond the worksite, affecting sleep, stress, housing, and access to care [4, 5, 6]. As a result, health risks are increasingly produced through interacting exposure systems. When this occurs, analyzing hazards one at a time may fail to capture how risk is produced, reinforced, and distributed over time [7, 8].

This creates a mismatch between how risk is produced and how it is typically analyzed. Hazard‐based approaches remain essential, but they do not always capture how risks arise from recurring combinations of conditions or where interventions can reduce multiple risks simultaneously [8]. In such contexts, risk may be under‐specified or treated as residual confounding rather than modeled as a central causal process [8]. Where exposures are strongly interacting and structured by employment conditions, the exposure system becomes the more appropriate unit of inference because the causal process lies in the configuration of exposures rather than in their individual effects [8, 9].

These challenges do not reflect entirely new hazards. Many work‐related strains—such as job insecurity, hazardous environments, and social inequalities—have long histories [10]. What has changed is how they are combined, intensified, and coordinated under contemporary labor arrangements, including platform‐based work, fragmented employment trajectories, and climate‐related stressors [2, 3, 4, 7, 8, 9].

At the same time, the preventable burden of work‐related disease and injury remains substantial. WHO/ILO e stimates attribute approximately 1.9 million deaths globally to occupational risks [11]. Despite advances in identifying hazards, current approaches are often better suited to documenting harm than to anticipating system‐level risk under rapidly changing working conditions and identifying earlier opportunities for prevention.

The following vignette illustrates this limitation. A truck driver does not experience “fatigue,” “sleep loss,” “income insecurity,” and “regulatory pressure” as separate exposures [12, 13, 14, 15, 16]. These conditions are structurally linked: payment structures encourage long hours, scheduling limits recovery, and economic constraints restrict access to care [12, 13, 14, 15, 16]. Together they form a recurring pattern of exposure that is only partially captured when analyzed separately [17, 18]. What matters epidemiologically in such a case is not only whether each condition is harmful but how conditions co‐occur, reinforce one another, and generate a sustained pattern of elevated risk.

Vignette: Why hazard‐based epidemiology may miss contemporary work.

A long‐haul truck driver does not experience “fatigue,” “sleep loss,” “income insecurity,” and “regulatory pressure” as separate exposures. He experiences a work‐organized system of exposure. His pay structure rewards longer driving hours. Scheduling practices erode recovery time. Regulatory compliance is unevenly enforced. Sleep occurs in irregular environments. Economic precarity limits access to health care. Each element reinforces the others, producing recurrent exposure configurations that accumulate across weeks, years, and working lives. A hazard‐based study isolates sleep duration, adjusts for income and work hours, and estimates an association with crash risk. The system that produces sleep loss remains largely invisible. By the time elevated crash rates are detected, harm is already embedded—biologically, organizationally, and economically. The Epidemiology of Working Life (EWL) begins from a different analytical starting point. It asks how employment relations generate recurring exposure configurations, how those configurations evolve over time, and where intervention could disrupt multiple downstream risks at once. What appears complex is not analytical excess, but the ordinary structure of risk in contemporary working life. EWL does not add complexity; it restores coherence by aligning analysis with how work is actually lived.

This paper addresses that gap by proposing the Epidemiology of Working Life (EWL) as a framework for shifting analytical focus toward work‐structured exposure systems. This shift redefines what counts as exposure, what constitutes the unit of inference, and where prevention is directed. Rather than treating work as a setting of discrete hazards, EWL treats it as a system that generates and distributes interacting exposures across the working lifecourse.

Where risk is generated through work‐structured exposure systems, EWL redirects inference toward three questions: (1) what combinations of exposures are associated with elevated risk, (2) what upstream mechanisms generate these patterns, and (3) where intervention could reduce multiple downstream harms simultaneously. This reframes analysis from isolated hazards to the exposure configurations through which risk is produced, and prevention can be improved.

EWL is anchored in the working‐life‐exposome, which captures interaction, accumulation, and socially patterned vulnerability among exposures and their biological consequences [8, 17, 18, 19]. In this framework, the focus is on how exposure systems are produced, translated into health trajectories, and modified through identifiable organizational and policy levers in contexts where hazard‐by‐hazard analysis alone does not adequately represent the structure of risk.

The remainder of the paper proceeds as follows. Section 2 reviews the contributions and limitations of occupational epidemiology. Section 3 introduces EWL. Section 4 situates it in relation to existing approaches. Section 5 outlines methodological considerations. Section 6 presents illustrative scenarios. Section 7 examines implications for research, practice, and policy. Section 8 concludes.

2. Occupational Epidemiology: Contributions and Constraints

Occupational epidemiology has produced some of the most important advances in public health. Through cohort studies, case–control designs, and job–exposure matrices, it has identified causal links between workplace hazards and disease, contributing to prevention and regulation [20, 21, 22, 23]. Evidence on asbestos, lead, benzene, and ergonomic risks has informed standards that have reduced exposure and prevented illness and injury [24, 25, 26, 27]. These achievements reflect the field's core strength: generating robust causal evidence linking discrete exposures to specific outcomes and translating that evidence into prevention.

This hazard‐centered focus is less well suited to risks produced under contemporary working conditions, where exposures often co‐occur and are shaped by work organization [8]. For example, scheduling practices, compensation systems, and job insecurity can influence sleep, stress, and health‐related behavior in interacting and reinforcing ways [28, 29]. These processes are not fully captured when exposures are analyzed independently, creating a misalignment between the dominant unit of analysis and how risk is generated. In these contexts, risk is an emergent property of interacting exposures rather than the sum of individual effects, and estimating independent effects may not adequately represent the underlying causal structure [30, 31].

A related limitation concerns how exposures beyond the immediate worksite are treated. Occupational epidemiology has traditionally focused on hazards encountered during work tasks, while conditions such as income instability, housing insecurity, or disrupted sleep are often treated as external. Yet these are largely shaped by employment conditions and represent key pathways through which risks accumulate and affect health [8, 32, 33, 34]. When not explicitly modeled, they may be underestimated or treated as background context rather than as components of the exposure system through which work‐related risk is generated.

Methodologically, many studies rely on models emphasizing relatively simple exposure–outcome relationships [7, 8, 17, 18]. These approaches are effective for bounded hazards, but less well suited to time‐varying, interacting exposures within changing employment trajectories. While exposomic measurement, wearable technologies, network analysis, and machine learning offer ways to capture this complexity [19, 35, 36, 37, 38], they are often incorporated as extensions to existing research designs rather than used to redefine the primary unit of analysis around exposure systems.

There are also challenges in translating evidence into prevention. Occupational epidemiology has been effective in identifying hazards, but in many settings, evidence is often generated after harm has occurred. Structural conditions shaping exposure—such as labor arrangements, regulatory environments, and access to protections—are not always explicitly incorporated into epidemiological models, even when central to how risks are produced [8, 39, 40]. Equity dimensions are frequently treated as confounders or effect modifiers rather than as structural determinants that organize exposure and vulnerability [41]. As a result, opportunities for earlier or more upstream intervention may be less visible because processes generating exposure are not themselves the primary objects of analysis.

These limitations do not diminish the importance of hazard‐based occupational epidemiology, which remains essential for identifying and controlling discrete risks. Rather, they indicate that its primary analytical focus aligns best with bounded hazards, while many contemporary risks are generated through interacting, work‐organization‐driven exposure systems. In such contexts, adding more individual exposures to existing models may still not capture how risk is produced. Where exposure arises from interconnected employment conditions unfolding over time, the limitation is structural: the unit of analysis may not match the causal process that organizes exposures in practice.

As summarized in Table 1, prevailing approaches prioritize discrete hazards, bounded worksites, and relative stability, whereas emerging conditions require frameworks capable of representing cumulative, interacting, and socially patterned exposures across the lifecourse. This mismatch underscores the need to reorganize how exposure and inference are specified when risk is systemically produced.

Table 1.

Prevailing occupational epidemiology versus epidemiology of working life.

Analytical dimension	Prevailing occupational epidemiology	Epidemiology of working life
Unit of analysis	Job, hazard, or discrete exposure	Work‐structured exposure system (cumulative, interacting, lifecourse‐structured)
Exposure focus	Single, isolated, often chemical or ergonomic hazards	Cumulative, interacting, and work‐structured exposures spanning behavioral, psychosocial, occupational, environmental, and structural domains
Time frame	Cross‐sectional or short‐term longitudinal	Lifecourse, time‐accumulating, and intergenerational trajectories
Causality	Linear: one exposure → one outcome	Systemic and nonlinear: interacting exposures, feedback loops, and time‐dependent processes
Methods	Job–exposure matrices, regression, cohort studies	Integrated longitudinal designs using exposomic, systems, network, and machine‐learning approaches, combined with causal inference for time‐varying exposures
Equity	Often treated as confounding, stratification, or effect modification	Modeled as causal structure that shapes exposure generation, distribution, and embodiment
Translation	Reactive, hazard‐by‐hazard regulation	Anticipatory, system‐level prevention and intervention on upstream drivers of exposure

Taken together, these considerations indicate the need for approaches that make these patterns explicit—approaches that can represent interacting exposures, identify how they are generated, and link them to prevention‐relevant intervention points. EWL is therefore proposed as a system‐level approach to risk centered on work‐structured exposure systems. It does not replace hazard‐based analysis, but repositions it within a framework that captures how exposures are generated, combined, and sustained over time—particularly in contexts where hazard‐by‐hazard analysis does not adequately specify risk.

3. The EWL: A Reorganizing Framework

This paper proposes EWL as a system‐level approach to understanding how working life structures health risk under contemporary conditions. In practical terms, EWL shifts analysis from single exposures to recurring patterns of co‐occurring conditions across working life. In many settings, the dominant unit of analysis—discrete hazards—does not fully capture how risks are produced. Instead, risks often arise from recurring, interacting exposures structured by work organization.

EWL addresses this by shifting the primary unit of analysis from isolated hazards to work‐structured exposure systems in contexts where exposures are interacting, time‐varying, and organizationally produced [8, 17, 18]. In these settings, modeling exposures independently can yield incomplete or unstable estimates of risk. EWL therefore treats recurring configurations of co‐occurring exposures as the primary object of explanation, thereby redefining exposure, the unit of inference, and targets of prevention.

Risk emerges from interacting exposure systems rather than from the additive effects of independently modeled exposures. Accordingly, the central causal question shifts from estimating the effect of single exposures to identifying how exposure configurations are generated, recur, and produce risk over time.

3.1. Conceptual Foundation

EWL is anchored in the working‐life‐exposome, which captures the network of occupational, psychosocial, behavioral, and structural exposures that accumulate and interact across the working lifecourse and their biological consequences [8, 17, 18]. It draws on lifecourse epidemiology, systems science, exposomics, and causal inference to represent interaction, feedback, and socially patterned vulnerability [19, 31, 42, 43].

Within this framework, equity is not treated only as a characteristic of individuals, but as a feature of how exposure is organized [41]. Conditions such as precarious employment, migration status, and labor market segmentation shape how exposures are distributed, combined, and experienced, and are therefore part of the causal structure shaping health [44, 45, 46]. In this view, inequities arise through the organization of exposure systems rather than solely through individual‐level differences. Here, precarious employment refers to work characterized by instability, limited protections, and reduced worker control over wages, schedules, and working conditions [44].

3.2. What Counts as Exposure in EWL

EWL expands the scope of exposure while maintaining clear analytical boundaries. An exposure is included when it is produced, shaped, or intensified by employment conditions, and is plausibly modifiable through workplace practices, labor regulation, or social protection. This includes exposures encountered during work (e.g., chemical agents), exposures generated by work organization (e.g., income volatility), work‐mediated exposures extending beyond the workplace (e.g., housing instability linked to income insecurity), and macrosocial conditions shaping employment relations (e.g., social protection regimes) that condition exposure and vulnerability [8].

Exposures not meaningfully connected to employment conditions are excluded from the EWL analytical frame. In this formulation, exposure is defined relationally—as part of a system generated through working life—rather than as an isolated attribute, so that measurement aligns with how exposure is produced and therefore how it can be prevented. Figure 1 illustrates this structure as interconnected layers spanning the working lifecourse, with exposures accumulating, interacting, and feeding back across levels—from macrosocial conditions to biological responses—producing prevention‐relevant patterns over time [8, 17, 18].

Figure 1.

Layers, feedback, and nonlinearity in the Epidemiology of Working Life (EWL).

3.3. From Hazards to Exposure Systems

In many contemporary settings, work‐related exposures rarely operate independently—they cluster and interact. For example, unstable scheduling can lead to sleep disruption, increasing fatigue and injury risk, and elevating long‐term cardiometabolic strain [47, 48]. These linked processes are difficult to capture when exposures are modeled in isolation. EWL treats these linked conditions as manifestations of exposure systems—structures that generate recurring configurations of co‐occurring exposures over time. In these contexts, the exposure system becomes the central analytical focus, with individual hazards interpreted as components of a broader causal structure that organizes their co‐occurrence, timing, and combined effects.

This shift changes the inference target. Rather than asking whether a single exposure increases risk, EWL asks which exposure configurations recur, how they develop and persist, and how they are generated by modifiable organizational and structural conditions. Risk is thus specified at the level of the system that organizes these exposures, rather than at the level of isolated components—allowing inference to better reflect how exposures are experienced in practice.

For clarity, “exposure systems” refer to structures that generate exposures, whereas “exposure configurations” refer to recurring observable patterns within those systems. For example, hospital staffing and shift‐rotation practices form an exposure system; repeated short rest intervals, sleep disruption, and high workload constitute an exposure configuration.

3.4. What EWL Produces

EWL generates three prevention‐relevant outputs:

• 1.Exposure‐system phenotypes—that is, recurrent exposure configurations associated with elevated risk (e.g., long hours + unstable schedules + sleep restriction);
• 2.Leverage‐node maps (i.e., upstream drivers of exposure)—upstream features of work organization (e.g., scheduling policies, payment structures) that generate these configurations and represent intervention points;
• 3.Early‐warning indicators—surveillance‐ready signals that harmful exposure patterns are increasing before disease or injury becomes evident (e.g., rising short rest intervals alongside increasing fatigue).

These outputs shift epidemiological inference from estimating isolated exposure effects to identifying structured exposure configurations, their generators, and their trajectories. The emphasis is on where and how intervention can disrupt systems that produce risk. Figure 2 summarizes these outputs and their corresponding inference targets, highlighting how EWL links pattern detection, causal explanation, and anticipatory intervention. The accompanying Box 1 specifies EWL's core commitments—including scope, inference targets, and criteria for prevention‐relevant evidence.

Figure 2.

Three prevention‐oriented outputs of the Epidemiology of Working Life (EWL) and their inference targets.

Box 1. The Epidemiology of Working Life (EWL): Core commitments.

1.

1. Work is a system, not a hazard list.	Health effects of working life arise from cumulative, interacting, and recurrent exposure configurations generated through employment relations, not from isolated hazards abstracted from their organizational and governance contexts.
2. Exposure must be employment‐structured and policy‐modifiable.	An exposure belongs in the working‐life‐exposome only if it is generated, patterned, or amplified through identifiable employment‐based mechanisms and is plausibly modifiable through occupational standards, labor regulation, organizational governance, or social protection.
3. The primary unit of inference is the exposure system.	Inference targets recurrent exposure configurations, their transitions over time, and the upstream mechanisms that generate them—not single variables treated as independent causes.
4. Equity is a causal structure, not a modifier.	Stratification processes (e.g., precarity, racialized labor segmentation, migration regimes) are modeled as exposure‐generating mechanisms shaping exposure assignment, intensity, and recovery, not merely covariates for adjustment.
5. Prevention relevance is a scientific criterion.	EWL is judged by whether it yields surveillance‐ready indicators and policy‐actionable leverage points that enable earlier, structural intervention—not by isolated exposure–outcome coefficients alone.

3.5. Relationship to Existing Approaches

EWL addresses limitations of hazard‐centered analysis as the primary organizing framework in contexts where risk is systemically produced. Its role is not additive but reorganizing: it reorients how hazard‐based evidence is interpreted when exposures are interacting and temporally structured. This shift is context‐dependent and does not displace hazard‐based approaches where exposures are discrete and well‐bounded. However, where exposures accumulate, interact, and are generated by work organization, analyzing hazards individually can understate total risk and obscure causal mechanisms. EWL therefore reorients analysis toward systems through which exposures are generated and combined—that is, toward the causal organization of work‐related risk in settings where those structures drive exposure.

3.6. Linking Analysis to Action

A central aim of EWL is to strengthen the connection between epidemiological analysis and prevention, particularly because many exposure‐generating conditions—such as scheduling practices, payment systems, or staffing policies—are shaped by labor markets, employers, and regulatory environments [28, 49]. EWL does not assume that identifying upstream drivers is sufficient to produce change, nor does it substitute for political or collective action. Importantly, EWL does not replace hazard‐based evidence in driving change; rather, it complements it by identifying where intervention can act earlier and on multiple exposures at once. Its contribution is to make exposure‐generating processes empirically visible and to identify points at which interventions could simultaneously alter multiple exposures. In many cases, these conditions are controlled by employers or shaped by market incentives—meaning that prevention depends not only on evidence but also on regulatory, economic, and collective action. EWL, therefore, aims to make actionable structures of risk empirically visible while recognizing that changing them depends on forces beyond epidemiology itself.

3.7. Evaluating EWL

EWL is intended to be empirically testable. Its value can be assessed by whether it identifies high‐risk exposure patterns earlier than hazard‐based approaches, reveals intervention points that affect multiple downstream risks, and improves the detection and explanation of inequities in exposure and health outcomes. These criteria can be evaluated through comparative analyses of exposure‐system‐based and hazard‐specific models. Evaluation focuses on explanatory power and prevention yield in settings where risk is systemically produced. In this way, EWL specifies a distinct, system‐oriented logic of inference that is testable, falsifiable, and explicitly linked to prevention. EWL is therefore characterized not only by expanded scope, but by a clearly specified, system‐oriented inference logic centered on prevention.

4. Position Within and Extension Beyond Key Frameworks

EWL builds on established frameworks in population and occupational health, but reorganizes how they are applied when risk is produced through interacting, time‐varying exposures structured by working life. Its contribution is not to introduce new determinants, but to specify risk differently by redefining the unit of analysis—from individual exposures or levels to work‐structured exposure systems as the primary object of explanation and intervention.

Lifecourse epidemiology has demonstrated how exposures accumulate and shape long‐term health trajectories [42]. However, work is often operationalized narrowly—for example, as occupation at a single time point or as isolated exposures. EWL extends this perspective by treating working life as a continuous exposure system, specifying how employment conditions combine, interact, and persist over time—thereby clarifying mechanisms of cumulative risk and the structured sequencing of exposure over time.

Research on the social determinants of health emphasizes income, education, housing, and inequality [50], with work typically treated as one determinant among many. EWL instead positions work as an organizing system through which these determinants are generated, mediated, and combined—via mechanisms such as compensation, scheduling, and access to protections—linking structural conditions directly to exposure patterns and their health consequences.

Syndemic theory highlights the interaction of co‐occurring conditions and their combined effects on health [51]. EWL locates these interactions within work‐structured exposure systems, shifting from describing co‐occurrence to specifying how such interactions are generated, sustained, and made amenable to intervention through identifiable features of work organization.

Socioecological models link individual, workplace, community, and policy levels [34], but are often applied as static frameworks. EWL treats these relationships as dynamic, emphasizing feedback, accumulation, and time‐dependent interaction, and modeling how work organization connects and structures exposures across levels over time, rather than treating levels as separate domains of influence.

The Total Worker Health® (TWH) approach integrates occupational safety and health promotion [52]. EWL provides an epidemiological specification of this integration, defining interacting exposure systems as explicit objects of analysis and enabling more precise identification of intervention points across domains through the joint production of work and nonwork exposures.

Across these frameworks, EWL introduces a common shift: from identifying exposures to specifying how they are structured, combined, and produced through working life. This constitutes a shift in analytical focus rather than a synthesis of perspectives, particularly in contexts where risk is systemically generated and cannot be adequately represented by independent exposures or levels alone. EWL does not sit alongside existing frameworks as an additional perspective; instead, it reorients how they are operationalized when exposure is organized as a system. Table 2 summarizes these relationships, aligning these approaches around a shared exposure‐system‐based unit of analysis and a common focus on how risk is generated, rather than only where it is observed.

Table 2.

Positioning the Epidemiology of Working Life (EWL) relative to established frameworks.

Framework	Core contribution	Limitations	EWL extension
Lifecourse epidemiology	Demonstrates how early exposures, cumulative risks, and critical periods shape long‐term trajectories.	Work is often reduced to midlife occupational status or single exposures; limited integration of employment dynamics.	Re‐specifies work as a continuous, evolving exposure system across the lifecourse, linking employment insecurity, shiftwork, and psychosocial strain to intergenerational effects.
Social determinants of health	Highlights structural influences of education, housing, income, and inequality.	Work is typically treated as background or outcome, not as a central determinant.	Elevates work as a structural exposure—a generator, mediator, and amplifier of inequities within the broader system of social determinants.
Syndemics	Reveals how co‐occurring conditions interact to create burdens greater than the sum of their parts.	Often descriptive; limited operationalization in occupational contexts.	Specifies syndemic processes as work‐structured exposure systems, mapping how occupational fatigue, psychosocial stress, toxic exposures, and cardiometabolic strain converge, and identifies upstream intervention points.
Socioecological models	Provides multilevel perspectives linking individual, workplace, community, and policy contexts.	Typically static snapshots, underplaying feedback, adaptation, and time.	Re‐specifies these relationships as dynamic, time‐dependent exposure systems with feedback and adaptation, enabling prediction and intervention across levels.
Total worker health (TWH)	Integrates occupational safety with health promotion, recognizing that workplace and nonwork factors jointly shape well‐being.	Programmatic rather than theoretical; often intervention‐focused without a unifying epidemiological framework.	Provides an epidemiological specification for TWH by embedding multidomain exposures within exposure systems that can be modeled longitudinally to inform structural, equity‐oriented interventions.

5. Methodological and Infrastructural Considerations

For EWL to be operational, its conceptual framework must be supported by feasible study designs, analytical approaches, and data infrastructures. The aim is not to replace established methods, but to redirect their application toward work‐structured exposure systems when these define risk. Where exposures are interacting and time‐varying, alignment between analytical focus and causal structure is necessary for valid, prevention‐oriented inference, so that models reflect how risk is actually produced.

A key feature of EWL is that it can be implemented incrementally. Studies need not capture all exposures at once. Researchers can begin with a limited set of high‐relevance domains—such as scheduling, income stability, and sleep—and expand as measurement improves. This modular approach allows implementation within existing study designs and is intended to make EWL feasible in practice, rather than making it dependent on entirely new data infrastructures or prohibitively complex measurement strategies.

5.1. Expanding the Unit of Analysis

EWL shifts the unit of analysis from isolated hazards or job titles to work‐structured exposure systems—combinations of exposures that interact and evolve over time. In practice, this requires collecting information not only on workplace hazards, but also on how work is organized and how it shapes recovery, behavior, and living conditions. The goal is not exhaustive measurement, but sufficient representation of key interacting domains and their temporal dynamics—including co‐occurrence, sequencing, and reinforcement.

Equity‐related factors—such as precarious employment, migration status, and labor market segmentation—are treated as components of exposure systems rather than adjustment variables. They are mechanisms through which exposure patterns and differential risk are generated, allowing inequities to be analyzed as features of system structure rather than residual variation or unexplained heterogeneity.

5.2. Study Design and Data Collection

EWL can be implemented using extensions of established epidemiological designs. Prospective cohort studies remain central, particularly those incorporating repeated measures over time. Existing cohorts can be adapted by adding targeted measures of work organization (e.g., scheduling patterns, income variability) and related outcomes (e.g., sleep, stress).

Data sources may include surveys, administrative or platform data, wearable devices, and biological measures, where feasible. Not all components are required in every study. Lower‐cost approaches—such as structured surveys and repeated self‐reports—can capture key exposure‐system features, particularly in resource‐constrained settings.

The goal is not maximal measurement, but adequate representation of interacting exposure domains and their evolution over time. This flexibility supports scalability across diverse labor contexts and allows studies to prioritize the most relevant components of the exposure system in a given setting.

5.3. Analytical Approaches

Analyzing exposure systems requires methods that can accommodate time‐varying, interacting exposures.

These may include longitudinal models, network approaches, machine learning, and other methods that identify co‐occurrence patterns, temporal sequences, and dependencies among exposures [53, 54]. Causal inference methods remain essential for evaluating interventions [43]. Pattern detection and causal analysis are distinct but complementary: pattern identification informs hypotheses, while causal models evaluate intervention‐relevant mechanisms linked to exposure‐system drivers.

EWL does not prescribe specific methods; it redefines the inference problem. The task is to identify and evaluate structured co‐occurrence, temporal evolution, and upstream generation of exposures, rather than to estimate isolated effects. Methods are selected to match the structure of the exposure system, not the reverse, thereby reducing the risk of oversimplifying complex exposure processes.

5.4. Linking Data to Prevention

A central purpose of EWL is to support earlier and more effective intervention. This requires linking analytical outputs to modifiable features of work organization—including scheduling policies, compensation systems, staffing levels, and access to protections.

By identifying recurrent exposure patterns and their drivers, studies can indicate where changes in work organization may simultaneously reduce multiple risks. The focus shifts from estimating individual effects to identifying leverage points within exposure systems, thereby linking evidence directly to interventions that act on multiple exposures at once.

5.5. Ethical and Practical Considerations

The use of detailed and repeated data on working people raises important ethical concerns, particularly regarding privacy and potential misuse [55, 56]. These risks are especially salient where workers have limited control over data collection and use.

Core safeguards include informed consent, data protection, and governance structures that prevent misuse for surveillance or managerial control [56]. Participatory approaches involving workers and their representatives can help align research with worker interests. These safeguards are essential in contexts involving platform data or employer‐linked monitoring systems, where data may otherwise be used to intensify control rather than improve conditions [56].

5.6. Scalability and Global Applicability

EWL is intended to be applicable across different labor contexts, including settings with limited resources or large informal sectors. Lower‐cost approaches—such as surveys, participatory mapping, and mobile health assessments—can capture key exposure‐system features without complex infrastructure.

At the same time, coordination across studies—through shared measures and comparable designs—can support a broader understanding of exposure systems across contexts. Scalability depends on consistent representation of core exposure‐system dimensions, rather than technological intensity, enabling comparability while preserving flexibility in data collection.

6. Illustrative Proof‐of‐Concept Scenarios

The following scenarios illustrate how EWL can be applied in practice. They are schematic and designed to show how exposure systems, their upstream drivers, and early‐warning indicators can be identified under realistic conditions. They demonstrate how shifting the unit of analysis—from isolated hazards to exposure systems—changes both explanation and intervention targets, and therefore what counts as actionable evidence.

While hazard‐based occupational epidemiology has identified important hazards and outcomes, when exposures are cumulative, interacting, and shaped by employment conditions, analyzing hazards in isolation can underrepresent total risk and obscure intervention opportunities. The scenarios below show how EWL makes these systems explicit, linking recurring exposure patterns, their generators, and early signals of harm into a single analytical framework.

6.1. Platform‐Based Gig Work

Platform‐based delivery workers experience elevated risks of traffic injury and musculoskeletal strain [57, 58]. Hazard‐based analyses capture these risks but often treat working conditions—such as income instability or algorithmic management—as background factors rather than as drivers of exposure patterns. A hazard‐based analysis may identify elevated injury risk, whereas EWL identifies the system producing recurrent fatigue‐related risk.

EWL instead conceptualizes gig work as a work‐structured exposure system in which multiple conditions co‐occur and reinforce one another. Algorithmic dispatch, pay variability, extended hours, and environmental stressors interact through pathways such as sleep disruption and fatigue [59]. For example, incentives tied to delivery speed and volume can increase work intensity, reduce rest opportunities, and elevate fatigue‐related risk. In this context, risk arises from the recurring configuration of exposures rather than from any single component.

EWL identifies three prevention‐relevant outputs: (a) exposure‐system phenotype (i.e., a recurrent exposure configuration): recurring patterns of long hours, unstable schedules, fatigue, and environmental stressors associated with elevated injury risk, (b) leverage‐node map (upstream driver of exposure): platform governance rules, including dispatch algorithms and pay structures that shape work intensity and recovery opportunities, and (c) early‐warning indicator: concurrent increases in fatigue, shortened rest intervals, and near‐miss events. Intervening at the level of platform design can therefore modify multiple exposures simultaneously, while early signals create opportunities for anticipatory intervention rather than waiting for injury rates to rise.

6.2. Healthcare Shiftwork

Healthcare workers face risks related to shiftwork, including sleep disruption, fatigue, and psychosocial strain [60, 61]. These exposures are often studied separately, yielding fragmented evidence that does not fully capture how they accumulate across repeated work cycles. A hazard‐based approach may isolate sleep disruption or workload, whereas EWL identifies the recurring system that produces their combined and cumulative effects.

EWL treats these conditions as a temporally‐structured exposure system: rotating shifts disrupt sleep and metabolic regulation; staffing shortages increase workload; and limited recovery amplifies fatigue and injury risk. Risk reflects recurring, interacting exposure patterns rather than isolated effects.

EWL again yields three distinct outputs: (a) exposure‐system phenotype (recurrent exposure configuration): repeated cycles of short rest intervals, sleep disruption, and high workload, (b) leverage‐node map (upstream driver of exposure): scheduling practices, staffing levels, and recovery policies, and (c) early‐warning indicator: increasing frequency of short rest intervals, rising fatigue reports, and error or near‐miss events. Interventions targeting scheduling systems (e.g., limiting rapid shift rotation or enforcing minimum rest periods) can simultaneously modify several exposure pathways. Monitoring early indicators enables system‐level surveillance, rather than relying solely on downstream outcomes such as injury or burnout.

6.3. Informal and Small‐Scale Mining

Artisanal and small‐scale miners face risks from silica dust, mercury exposure, and related health outcomes [62, 63, 64]. Hazard‐focused studies identify these risks but often isolate them from broader conditions shaping exposure. A hazard‐based approach may focus on silica or mercury exposure, whereas EWL identifies the broader system that intensifies and sustains these risks.

EWL situates these hazards within a work‐structured exposure system shaped by economic and organizational constraints. Informal labor arrangements, income insecurity, heat stress, and limited access to protective equipment interact to influence both exposure intensity and vulnerability. For example, income instability may increase time spent in high‐exposure tasks while limiting access to protective resources.

Within this system, three prevention‐relevant outputs are identifiable: (a) exposure‐system phenotype (recurrent exposure configuration): co‐occurring silica exposure, heat stress, and prolonged high‐risk work under constrained conditions, (b) leverage‐node map (upstream driver of exposure): informal labor structures, unstable income, limited regulation, and lack of protective infrastructure, and (c) early‐warning indicator: concurrent increases in respiratory symptoms, heat strain, and workload intensity. Interventions that stabilize working conditions or improve access to protective resources can therefore reduce both exposure intensity and vulnerability simultaneously. Early signals reflect system‐level escalation rather than isolated symptom trends.

6.4. Synthesis

Across these scenarios, the same outputs recur: high‐risk exposure configurations (exposure‐system phenotypes), organizational drivers that generate them (leverage‐node maps), and early‐warning indicators that can support earlier intervention. EWL does not replace hazard identification; it repositions it within the systems that shape exposure intensity, distribution, and recurrence. Its contribution is inferential: identifying system‐level leverage points and anticipatory signals that are not visible when exposures are analyzed independently. This system‐level approach moves prevention from reacting to isolated hazards toward intervening on the systems that produce and sustain them—enabling earlier, more integrated, and more effective intervention in contexts where risk is generated through interacting, recurrent exposure patterns.

7. Implications for Research, Practice, and Policy

EWL has implications for how occupational health research is conducted, how risks are addressed in practice, and how evidence informs policy. Its central contribution is to redefine how risk is specified, investigated, and targeted for intervention in contexts where risk is generated through interacting, cumulative exposures shaped by employment conditions.

7.1. Implications for Research

For researchers, EWL shifts attention from isolated exposures to how exposures co‐occur, interact, and evolve over time. This does not require abandoning established methods, but reorganizing their application around exposure systems as the primary unit of analysis when these reflect the underlying causal structure of risk. This shift also changes how key variables are treated. Conditions such as precarious employment, migration status, and labor segmentation are no longer treated primarily as covariates for adjustment, but as components of the causal structure shaping exposure patterns. This allows inequities to be analyzed as features of exposure systems rather than as residual variation.

In practice, this implies targeted extensions to existing designs—such as measures of scheduling, income stability, or recovery time—rather than redesigning epidemiological research from the ground up. The key change is not adding variables, but specifying exposure systems and their drivers as primary inference targets. More broadly, EWL reframes what constitutes an adequate epidemiological explanation in these contexts: not simply estimating the effect of individual exposures, but identifying recurring exposure configurations, mechanisms that generate them, and their implications for intervention.

7.2. Implications for Practice

For occupational health practice, EWL supports a shift from responding to individual hazards toward identifying and managing recurring exposure patterns. When risks arise from combinations of conditions, interventions targeting a single hazard may have limited impact if the underlying exposure‐generating conditions remain unchanged. By contrast, targeting features of work organization—such as scheduling, workload, or access to rest—can modify multiple exposures simultaneously. Intervention, therefore, shifts from controlling individual hazards to altering the systems that generate them.

EWL also supports earlier detection of risk by emphasizing leading indicators. Patterns such as repeated short rest intervals, rising fatigue, or clusters of near‐miss incidents can signal instability within exposure systems before more severe outcomes occur. This enables a transition from reactive approaches based on injury or disease rates to anticipatory, system‐level risk management.

7.3. Implications for Policy

At the policy level, EWL highlights the importance of linking epidemiological evidence to conditions that generate exposures. By focusing on exposure systems, it identifies policy‐relevant leverage points—such as scheduling standards, compensation structures, and workplace protections—that influence multiple outcomes simultaneously through shared underlying mechanisms. This shifts the role of epidemiological evidence from documenting harm to identifying where and how intervention can alter exposure generation. It also supports the development of surveillance systems that track interacting exposure patterns rather than single hazards, enabling earlier detection of system‐level risk at the population level.

Importantly, EWL reframes regulatory targets. While hazard‐specific regulation remains essential, EWL is situated within a broader strategy that addresses upstream determinants of exposure, including employment arrangements and labor market structures. Policy can therefore act not only on hazards themselves but also on the systems that produce and distribute them. EWL clarifies that identifying exposure‐generating conditions does not, by itself, produce change; rather, it provides a stronger empirical basis for regulatory, institutional, and collective interventions.

7.4. Summary

Across research, practice, and policy, EWL makes visible work‐structured exposure systems that are not captured when hazards are analyzed independently. By specifying how exposures are structured, combined, and generated, it supports a more prevention‐oriented use of epidemiological evidence. In these contexts, EWL does not simply extend existing approaches—it reorganizes how risk is defined, how evidence is produced, and how intervention targets are identified. This shift aligns epidemiological inference with the processes through which risk is generated, strengthening the capacity of occupational health science to anticipate and reduce harm under contemporary working conditions.

8. Conclusion

Occupational epidemiology has made essential contributions to identifying harmful workplace exposures and informing prevention [65]. Its strength in linking specific hazards to specific outcomes remains central to protecting the health of working people. At the same time, many contemporary risks arise from recurring, interacting exposures structured by how work is organized. In these settings, risks accumulate over time and extend beyond the worksite, making them less visible to hazard‐by‐hazard approaches alone. This creates a growing mismatch between how risk is produced and how it is analyzed and prevented.

EWL is proposed as a system‐level approach in contexts where risk is generated through work‐structured exposure systems. By focusing on exposure systems, their upstream drivers, and early‐warning indicators, it provides a more prevention‐relevant representation of how work‐related health risks are produced and where intervention may be most effective. It aligns what is studied, how inference is made, and where intervention is directed with how risk is generated in practice.

EWL does not replace hazard‐based approaches, but repositions them within a broader exposure‐system framework. Hazard identification remains essential, while EWL adds the capacity to identify how exposures are combined, sustained, and patterned over time—and where those patterns can be disrupted. The primary object of explanation thus shifts from individual exposures to the systems that organize their co‐occurrence, timing, and cumulative effects.

The continued evolution of work—through changes in labor organization, technology, protections, and environmental conditions—requires approaches that can capture these dynamics while remaining empirically grounded. EWL provides such a framework. It enables earlier detection of risk, identifies upstream leverage points, and strengthens the basis for more effective prevention.

Where risk is generated through work‐structured exposure systems, the central task of occupational health science is not only to identify hazards, but to specify, monitor, and intervene on the systems that generate and sustain them. Engaging with this shift is increasingly necessary if epidemiological evidence is to remain aligned with effective prevention. How the field responds will shape its capacity to anticipate and reduce avoidable harm in contemporary working life.

Author Contributions

Yorghos Apostolopoulos conceived and developed the Epidemiology of Working Life (EWL) framework; formulated the conceptual and theoretical structure of the manuscript; synthesized and integrated relevant literature and interdisciplinary perspectives; developed the analytical and prevention‐oriented framework; wrote, revised, and finalized the manuscript; and approved the final version for publication.

Funding

The author has nothing to report.

Ethics Statement

This study uses publicly available, anonymized secondary data from various sources. No primary data collection involving human participants was conducted, and therefore, no additional ethical approval was required.

Conflicts of Interest

The author declares no conflicts of interest.

Data Availability Statement

This study uses publicly available data and sources.