The JD-R model offers a comprehensive and empirically validated framework for understanding how job demands and resources interact to influence employee well-being and performance, particularly in remote work environments. According to the model, job demands such as excessive screen time, extended working hours, constant connectivity, and high email volume can lead to stress and burnout, especially when not balanced by sufficient resources like sleep, physical activity, autonomy, or recovery time (Bakker and Demerouti, 2007; Bakker et al, 2023). Empirical research has consistently demonstrated that the presence of job resources can mitigate the adverse effects of high demands and enhance productivity (Bakker et al, 2004; Crawford et al, 2010). The model is particularly relevant in digital work contexts where the intensity of online interactions may blur work-life boundaries and increase cognitive load (Bakker and Demerouti, 2017).

Schaufeli and Taris (2014) critically reviewed the JD-R model, identifying two core mechanisms: the health impairment process, where excessive demands deteriorate well-being, and the motivational process, where adequate resources foster engagement. They emphasized the model’s adaptability across diverse occupational settings and discussed the integration of personal resources—such as resilience and optimism—as buffers against job strain. Their review also highlighted the need for clearer conceptual boundaries and the investigation of reciprocal relationships between demands and resources.

Recent research has extended the model’s scope. For example, personal resources like physical activity and sleep have been shown to buffer the negative effects of digital intensity in remote work (Lesener et al, 2019). Leadership also plays a vital role in ensuring access to resources that support well-being (Tummers and Bakker, 2021). Moreover, contextual stressors such as job insecurity during COVID-19 can amplify demand-resource imbalances and reduce employee resilience (Cao et al, 2024). In this study, the JD-R model guides the investigation of how specific digital work patterns and health behaviors interact to shape productivity among remote employees. Together, these contributions underscore the model’s relevance for designing sustainable and human-centered remote work strategies. While the JD-R model offers a valuable lens for analyzing how digital work patterns influence employee well-being and productivity, it does not fully capture the spatial and psychological boundary challenges that emerge in remote work environments. As digital demands increasingly permeate personal spaces, understanding how employees manage and negotiate the boundaries between their professional and private lives becomes essential. To address this dimension, the present study also draws on Work-Life Border Theory, which provides a complementary perspective by focusing on how individuals navigate and construct the borders between work and home domains. This theoretical approach is particularly relevant in the context of remote work, where digital connectivity often disrupts conventional boundaries and reshapes the rhythm of everyday life.

Work-Life Border Theory offers a useful framework for understanding how remote employees manage the boundaries between work and personal life in digitally intensive environments. As subjective well-being is increasingly shaped by the integration—or lack thereof—of work and personal domains, understanding the boundaries that separate or blend these roles becomes critical (Diener, 1984). As Clark (2000) originally posited, individuals actively construct, maintain, and negotiate the borders between work and family domains based on role salience and boundary permeability. In remote work settings, these borders are often blurred, increasing the potential for conflict (Olson-Buchanan and Boswell, 2006). Daily transitions between roles—such as logging off from a meeting and immediately engaging in household tasks—require cognitive and emotional adjustment (Delanoeije et al, 2019). Kreiner (2006) emphasizes that the degree of fit between personal boundary preferences and organizational expectations significantly influences well-being. He identifies three strategies—segmenting, integrating, and cycling—that individuals use to manage work-home boundaries, and notes that misalignment between preferred and actual boundary management can lead to emotional exhaustion, dissatisfaction, and role confusion. Moreover, work-related technology use at home can hinder psychological detachment, a key process for recovery and sustained productivity. This study applies Work-Life Border Theory to explain how digital work conditions reshape boundary management and affect remote employee outcomes.

Work-Life Border Theory continues to evolve as scholars explore how remote work reshapes the psychological and spatial boundaries between professional and personal domains. Adisa and Gbadamosi (2021) propose the Work-Life Border Control Model, emphasizing individual agency in constructing flexible borders that resist institutional overreach. This is especially relevant in digital work environments, where constant connectivity challenges separation (Kossek et al, 2012). Karassvidou and Glaveli (2015) highlight that organizational culture plays a critical role in either facilitating or hindering effective boundary management, especially in fast-paced, performance-driven contexts. Felstead and Henseke (2017) demonstrate that the expansion of remote work intensifies the risk of blurred boundaries, reducing work-life balance and overall well-being. Complementing these findings, Allen et al (2014) argue that sustainable boundary management requires both structural support and personal alignment. Notably, Sarker et al (2018) found that asynchronous communication and cultural distance in globally distributed teams exacerbate work-life conflict, making proactive boundary negotiation essential in digital work contexts. Complementing this, Cowan and Hoffman (2007) show that employees in flexible organizations actively reconstruct the work-life border through communication and relational practices rather than rigid schedules or policies. Karwa (2025) reinforces this view by framing work-life balance as a “modern psychological dilemma”, emphasizing that the cognitive strain of navigating digital boundaries contributes significantly to burnout and disengagement in contemporary workplaces.

Recent scholarship has provided crucial insights into remote work dynamics that directly inform this study’s theoretical framework. Li et al (2025) conducted a longitudinal analysis of 3500 remote workers, demonstrating that work autonomy serves as a significant buffer against privacy violations and job dissatisfaction through stress reduction mechanisms. Their findings challenge traditional applications of the Job Demands-Resources model by revealing that personal resources may operate differently in digital environments compared to traditional workplace settings. However, their study was limited to technology sector employees, raising questions about generalizability across diverse occupational contexts that this research addresses through its multi-industry approach.

Jacobs et al (2025) investigated work-from-home arrangements across 2000 participants, identifying that flexible work hours significantly impact life satisfaction, particularly when combined with regular break-taking behaviors. Their research provides empirical support for Work-Life Border Theory’s emphasis on boundary management flexibility, yet their findings were restricted to North American samples with specific cultural expectations regarding work-life integration. This limitation highlights the need for research that examines these relationships across different cultural contexts, particularly in Central and Eastern European settings where work-life boundary expectations may differ substantially.

Karwa (2025) conceptualized work-life balance as a “modern psychological dilemma”, arguing that digital boundary navigation contributes significantly to contemporary workplace burnout. While this theoretical contribution advances understanding of psychological mechanisms underlying remote work challenges, the analysis remained primarily conceptual without empirical validation. The present study addresses this gap by providing quantitative evidence for the psychological processes Karwa identified.

These recent contributions collectively suggest that traditional theoretical models require adaptation for digital work contexts, yet significant empirical gaps remain regarding the simultaneous interaction of multiple work pattern variables, the cultural specificity of boundary management strategies, and the identification of distinct remote worker profiles requiring differentiated support approaches.

This study draws on these perspectives to explain how digital intensity and personal health behaviors jointly influence remote worker outcomes. Building upon these theoretical foundations, the present study adopts a quantitative research design to empirically examine how digital work patterns and well-being indicators interact to shape productivity outcomes among remote employees.

The inclusion of non-Western theoretical perspectives enhances the global relevance of this research framework, particularly given the international scope of remote work arrangements. Adisa and Gbadamosi’s (2021) Work-Life Border Control Model, developed through African organizational contexts, emphasizes individual agency in constructing flexible borders that resist institutional overreach. This perspective proves particularly relevant for understanding remote work dynamics in Central and Eastern European contexts, where employees navigate tensions between traditional hierarchical organizational cultures and emerging flexible work expectations.

The global nature of this study’s dataset, encompassing remote workers across multiple continents and cultural contexts, necessitates theoretical frameworks that transcend Western-centric assumptions about work-life integration. Non-Western perspectives on collective versus individual boundary management strategies provide essential theoretical grounding for interpreting findings across diverse cultural settings. These perspectives inform the study’s analytical approach by recognizing that work-life balance satisfaction may operate differently across cultural contexts, requiring nuanced interpretation of mediation and moderation effects.

Based on the theoretical integration of the Job Demands-Resources model and Work-Life Border Theory, enhanced by recent empirical developments, this study proposes a comprehensive conceptual model of remote work dynamics. The model positions digital work patterns (daily working hours, screen time, meeting frequency, email volume) as job demands that directly influence well-being outcomes (stress levels, sleep duration, work-life balance satisfaction). Physical health behaviors (physical activity, sleep duration) function as personal resources that moderate the relationship between job demands and well-being while directly predicting productivity outcomes.

Work-life balance satisfaction serves as a critical mediating variable, transmitting the effects of work demands through boundary management processes to influence productivity. The model incorporates work intensity as a key moderating variable that amplifies the importance of well-being factors under high-demand conditions. This framework acknowledges the heterogeneity of remote worker experiences by proposing that distinct employee profiles emerge from different combinations of work patterns, well-being indicators, and productivity outcomes.

Drawing from the integrated theoretical framework and recent empirical evidence, this study tests the following specific hypotheses (Fig. 1):

Hypothesis 1a: Digital work intensity variables (daily working hours, screen time, meeting frequency, email volume) will positively predict stress levels among remote workers.

Hypothesis 1b: Digital work intensity variables will negatively predict sleep duration and work-life balance satisfaction among remote workers.

Hypothesis 2: Physical health behaviors (physical activity and sleep duration) will positively predict productivity scores after controlling for digital work pattern variables.

Hypothesis 3: Work-life balance satisfaction will partially mediate the negative relationship between work demands and productivity outcomes, such that higher work demands will reduce work-life balance satisfaction, which in turn will decrease productivity.

Hypothesis 4a: Work-life balance satisfaction will emerge as the strongest predictor of productivity among all well-being variables examined.

Hypothesis 4b: The relationships between well-being factors and productivity will be stronger under high work intensity conditions compared to low work intensity conditions.

Hypothesis 5: Cluster analysis will reveal distinct remote worker profiles characterized by different combinations of work patterns, well-being indicators, and productivity outcomes, with significant differences in mean productivity scores across profiles.

These hypotheses provide testable predictions that directly address the identified literature gaps while building upon established theoretical foundations and recent empirical developments in remote work research.

The sample consists of 10,000 remote employees (N = 10,000) identified by unique employee identifier (ID). While demographic variables such as age, gender, industry, and geographical location were anonymized to protect participant privacy, the dataset provides rich behavioral and self-reported metrics covering work patterns, productivity, and well-being indicators. The large sample size enhances statistical power and enables robust analysis of complex relationships between multiple variables.

All measurement instruments employed in this study demonstrated acceptable to excellent psychometric properties, as detailed below (Appendix A).

This study relies exclusively on self-reported measures for all primary variables, creating substantial risk for common method bias and measurement error. The use of self-reported productivity assessments represents a particularly significant limitation, as participants may overestimate their performance due to social desirability bias or lack objective benchmarks for evaluation. Several factors compound these measurement concerns in the remote work context. Participants may experience difficulty accurately estimating screen time and working hours due to the blurred boundaries between work and personal digital device usage. Sleep duration and physical activity assessments depend on participant recall accuracy, which may vary based on individual monitoring habits and awareness levels. The single-source nature of all measurements increases the likelihood that observed correlations reflect shared method variance rather than true substantive relationships. This concern is particularly relevant for relationships between well-being variables and productivity outcomes, where participants with generally positive or negative work experiences may systematically bias their responses across multiple domains. To partially address these concerns, the research design incorporated several bias reduction strategies. Response anonymity was ensured to minimize social desirability effects, and scale anchoring was provided to enhance response accuracy. However, these measures cannot eliminate the fundamental limitations inherent in self-report methodology.

Response quality was assessed through multiple indicators including completion time analysis, attention check performance, and pattern detection for systematic response bias. Participants completing the survey in less than fifteen minutes or demonstrating straight-line response patterns were excluded from analysis. Geographic and industry distribution verification confirmed representative sampling across target demographics. Missing data analysis revealed less than 3% missing values across all primary variables, with Little’s test for Missing Completely At Random (MCAR) indicating data were missing completely at random ($\chi$² = 24.67, p = 0.18). List-wise deletion was employed given the large sample size and minimal missing data impact on statistical power. Data efficiency was enhanced through careful instrument selection prioritizing validated brief measures over lengthy comprehensive assessments to minimize participant burden while maintaining measurement quality. The final survey required approximately twenty-five minutes for completion, balancing thorough assessment with practical completion rates.

Descriptive statistics for all study variables are presented in Table 1. The average daily working hours among remote employees was 7.95 hours (SD = 1.87), with considerable variation across the sample (range: 4–12 hours). Mean screen time was 6.14 hours (SD = 1.23), representing approximately 77% of the total working day. Participants reported moderate levels of stress (M = 5.53, SD = 2.41) and work-life balance satisfaction (M = 5.48, SD = 2.36). The average productivity score was 5.54 (SD = 2.73) on a 10-point scale.

Bivariate correlations between all study variables are presented in Table 2. Notably, daily working hours showed a moderate positive correlation with stress levels (r = 0.38, p 0.001) and a moderate negative correlation with work-life balance satisfaction (r = –0.42, p 0.001). Sleep duration was negatively correlated with screen time (r = –0.31, p 0.001) and stress levels (r = –0.34, p 0.001). Productivity score showed significant, albeit weak, correlations with physical activity (r = 0.12, p 0.001) and work-life balance satisfaction (r = 0.14, p 0.001).

Multiple regression analyses examining digital work habits as predictors of well-being outcomes yielded significant results with moderate effect sizes. The regression model predicting stress levels achieved substantial explained variance (R² = 0.27, F(4, 9995) = 922.56, p 0.001, f² = 0.37, indicating a large effect size according to Cohen’s conventions). Daily working hours emerged as the strongest predictor ($\beta$ = 0.32, p 0.001, sr² = 0.08, representing 8% unique variance), followed by email volume ($\beta$ = 0.17, p 0.001, sr² = 0.02) and meeting frequency ($\beta$ = 0.13, p 0.001, sr² = 0.01).

The model predicting work-life balance satisfaction demonstrated large practical significance (R² = 0.31, F(4, 9995) = 1123.04, p 0.001, f² = 0.45). Daily working hours served as the primary predictor ($\beta$ = –0.36, p 0.001, sr² = 0.11), with screen time contributing additional unique variance ($\beta$ = –0.15, p 0.001, sr² = 0.02). The sleep duration prediction model achieved moderate effect sizes (R² = 0.19, F(4, 9995) = 585.76, p 0.001, f² = 0.24), with screen time as the dominant predictor ($\beta$ = –0.28, p 0.001, sr² = 0.06).

Mediation analysis employed Hayes PROCESS Model 4 (simple mediation) to examine work-life balance satisfaction as a mediator between work demands and productivity. The composite work demands variable was created through standardized sum scoring of working hours, screen time, meeting frequency, and email volume ($\alpha$ = 0.73). Model fit assessment through bootstrap procedures (5000 samples) confirmed adequate estimation stability.

The mediation model demonstrated significant pathway effects with meaningful practical importance. The total effect of work demands on productivity (c = –0.26, Standard Error (SE) = 0.03, t = –8.67, p 0.001, 95% CI [–0.32, –0.20]) decomposed into a direct effect (c = –0.17, SE = 0.03, t = –5.67, p 0.001, 95% CI [–0.23, –0.11]) and an indirect effect through work-life balance satisfaction (ab = –0.09, SE = 0.01, 95% CI [–0.11, –0.07]). The proportion of mediation calculated as 35% (Proportion Mediated (PM) = 0.35, 95% CI [0.28, 0.43]), indicating that approximately one-third of the work demands effect operates through boundary management satisfaction. The mediation model explained 12% of productivity variance (R² = 0.12, F(2, 9997) = 681.45, p 0.001).

Moderation analysis utilized Hayes PROCESS Model 1 (simple moderation) to examine work intensity as a moderator of well-being-productivity relationships. Work intensity was operationalized as a composite variable combining standardized working hours and digital communication volume (r = 0.67, p 0.001).

The interaction between work-life balance satisfaction and work intensity significantly predicted productivity ($\beta$ = 0.15, SE = 0.02, t = 7.83, p 0.001, 95% CI [0.11, 0.19], $\mathrm{\Delta }$R² = 0.02, f² = 0.02). Simple slopes analysis revealed stronger relationships under high work intensity conditions ($\beta$ = 0.23, SE = 0.03, t = 7.67, p 0.001) compared to low intensity conditions ($\beta$ = 0.09, SE = 0.03, t = 3.00, p 0.01). The moderation effect size, while statistically significant, represented a small practical impact according to Cohen’s guidelines.

Similarly, the physical activity by work intensity interaction achieved significance ($\beta$ = 0.12, SE = 0.02, t = 6.00, p 0.001, 95% CI [0.08, 0.16], $\mathrm{\Delta }$R² = 0.01, f² = 0.01), with stronger effects under high intensity conditions ($\beta$ = 0.21, SE = 0.03, t = 7.00, p 0.001) versus low intensity conditions ($\beta$ = 0.12, SE = 0.03, t = 4.00, p 0.001).

Cluster analysis employed k-means clustering with multiple validation procedures to ensure solution stability and interpretability. The optimal three-cluster solution was determined through multiple criteria including elbow method analysis, silhouette coefficient assessment, and theoretical interpretability.

Multiple regression analyses revealed significant relationships between digital work habits and employee well-being outcomes (Table 3). Together, the four work pattern variables explained 27% of the variance in stress levels (R² = 0.27, F(4, 9995) = 922.56, p 0.001), 19% of the variance in sleep duration (R² = 0.19, F(4, 9995) = 585.76, p 0.001), and 31% of the variance in work-life balance satisfaction (R² = 0.31, F(4, 9995) = 1123.04, p 0.001).

Daily working hours emerged as the strongest predictor of stress levels ($\beta$ = 0.32, p 0.001) and work-life balance satisfaction ($\beta$ = –0.36, p 0.001). Screen time was the strongest predictor of reduced sleep duration ($\beta$ = –0.28, p 0.001). Meetings attended showed significant but weaker associations with stress ($\beta$ = 0.13, p 0.001) and work-life balance satisfaction ($\beta$ = –0.09, p 0.001). Email volume demonstrated significant associations with all three well-being outcomes, with standardized coefficients ranging from $\beta$ = –0.11 to $\beta$ = 0.17 (all p 0.001).

Hierarchical regression analysis examining the relationship between physical health behaviors and productivity is presented in Table 4. After controlling for work pattern variables in Step 1 ($\mathrm{\Delta }$R² = 0.07, p 0.001), the addition of physical health behaviors in Step 2 explained an additional 5% of variance in productivity scores ($\mathrm{\Delta }$R² = 0.05, p 0.001).

Physical activity demonstrated a significant positive association with productivity ($\beta$ = 0.16, p 0.001), with each additional 1000 steps corresponding to a 0.14-point increase in productivity (b = 0.14, SE = 0.02). Sleep duration also showed a significant positive association with productivity ($\beta$ = 0.13, p 0.001), with each additional hour of sleep corresponding to a 0.25-point increase in productivity (b = 0.25, SE = 0.04). The final model explained 12% of the variance in productivity scores (R² = 0.12, F(6, 9993) = 226.58, p 0.001).

Mediation analysis revealed that work-life balance satisfaction partially mediated the relationship between work demands and productivity. The direct effect of work demands on productivity was significant (c = –0.17, SE = 0.03, p 0.001), as was the indirect effect through work-life balance satisfaction (ab = –0.09, SE = 0.01, 95% CI [–0.11, –0.07]). The total effect of work demands on productivity was c = –0.26 (SE = 0.03, p 0.001). The ratio of indirect to total effect was 0.35, indicating that approximately 35% of the negative impact of work demands on productivity was mediated through reduced work-life balance satisfaction. Fig. 2 presents the mediation model with standardized path coefficients.

Relative weight analysis identified the strongest predictors of productivity among the seven predictor variables (four work pattern indicators and three well-being metrics). Work-life balance satisfaction emerged as the strongest predictor, accounting for 28.3% of the explained variance in productivity. Physical activity was the second strongest predictor (22.7%), followed by sleep duration (19.4%), daily working hours (12.8%), stress level (9.5%), screen time (4.2%), and emails sent (3.1%).

Moderation analysis revealed significant interaction effects between work intensity and well-being predictors. The relationship between work-life balance satisfaction and productivity was stronger under high work intensity conditions ($\beta$ = 0.23, p 0.001) compared to low work intensity conditions ($\beta$ = 0.09, p 0.01), suggesting that work-life balance becomes more critical for productivity as work demands increase. Similarly, the relationship between physical activity and productivity was stronger under high work intensity conditions ($\beta$ = 0.21, p 0.001) compared to low work intensity conditions ($\beta$ = 0.12, p 0.001). Fig. 3 illustrates these interaction effects graphically.

Contrary to expectations, the relationship between sleep duration and productivity did not significantly differ across work intensity levels (interaction term: $\beta$ = 0.04, p = 0.08), suggesting that adequate sleep is consistently important for productivity regardless of work intensity.

Cluster analysis identified three distinct employee profiles: “Balanced Performers” (32.4% of sample), characterized by moderate working hours, high physical activity, adequate sleep, and high productivity; “Digital Immersers” (41.7%), characterized by long working hours, high screen time, low physical activity, and moderate productivity; and “Struggling Remotes” (25.9%), characterized by moderate working hours, high stress, poor sleep, and low productivity. These profiles differed significantly in mean productivity scores (F(2, 9997) = 743.26, p 0.001), with Balanced Performers showing the highest productivity (M = 7.21, SD = 1.94), followed by Digital Immersers (M = 5.18, SD = 2.42) and Struggling Remotes (M = 3.82, SD = 2.31). Post-hoc comparisons (Tukey’s HSD) confirmed that all pairwise differences were statistically significant (all p 0.001).

Fig. 4 presents a visualization of these employee profiles across key study variables, highlighting distinctive patterns in work habits, well-being indicators, and productivity outcomes.

The observed associations between digital work habits and employee well-being outcomes support existing theoretical frameworks while raising important questions about underlying mechanisms. The strong association between daily working hours and stress levels aligns with work intensification theories, yet several alternative explanations warrant consideration.

Alternative Explanation 1: Employees experiencing higher baseline stress levels may gravitate toward longer working hours as a coping mechanism or due to reduced efficiency requiring extended work periods. The cross-sectional design prevents determination of whether extended working hours precede stress increases or whether stressed individuals subsequently work longer hours to maintain productivity standards.

Alternative Explanation 2: Unmeasured organizational factors such as management style, job security concerns, or performance pressure may simultaneously drive both extended working hours and elevated stress levels. Similarly, individual personality traits including perfectionism or achievement orientation could explain both the tendency toward longer work sessions and higher stress reports.

Alternative Explanation 3: The observed relationships may partially reflect measurement overlap, where participants reporting high work demands unconsciously align their stress and work-life balance assessments to maintain cognitive consistency rather than reflecting genuine independent relationships.

The negative association between screen time and sleep duration demonstrates consistency with existing research on digital device impacts, yet alternative mechanisms require consideration. Sleep difficulties may drive increased evening screen time rather than screen exposure causing sleep disruption. Additionally, underlying circadian rhythm differences could influence both technology usage patterns and sleep quality independently.

The positive associations between physical health behaviors and productivity outcomes suggest multiple potential explanatory mechanisms beyond direct causal pathways. While these relationships align with established health psychology theories, alternative explanations merit examination.

Self-Selection Effects Individuals with naturally higher energy levels or superior self-regulation capabilities may demonstrate both increased physical activity and enhanced productivity independent of any causal relationship between these variables. Similarly, employees in less demanding roles may have greater opportunities for both physical activity and optimal sleep, creating spurious associations with productivity measures.

Workplace cultures that support employee wellness may simultaneously encourage physical activity and implement productivity-enhancing practices, creating correlational relationships that reflect organizational factors rather than individual health behaviors. The remote work context may particularly influence these relationships, as employees with supportive home environments may demonstrate advantages across multiple domains.

Our mediation analysis revealed that work-life balance satisfaction partially mediates the relationship between work demands and productivity, accounting for approximately 35% of the negative impact of work demands on productivity. This finding supports Wosny et al (2023), who emphasized that improving the experience of using health information technologies can notably impact clinician well-being and mitigate burnout. The partial mediation observed in our study suggests that while work-life balance is a crucial pathway through which work demands affect productivity, other mechanisms, such as psychological detachment and recovery experiences, likely play important complementary roles.

The mediating role of work-life balance suggests that interventions designed to enhance work-life boundaries may buffer against the negative effects of high work demands on productivity. This aligns with Jacobs et al (2025), who found that work flexibility and work-life balance are crucial elements in promoting positive well-being outcomes for those working from home. Their findings that having less flexible work hours adversely impacts the probability of self-reporting life satisfaction particularly for those who take more frequent breaks resonates with our results on the importance of work-life balance for productivity.

Our moderation analysis revealed that the relationships between well-being factors and productivity are significantly impacted by work intensity. Specifically, the associations between work-life balance satisfaction and productivity ($\beta$ = 0.23 vs. $\beta$ = 0.09) and between physical activity and productivity ($\beta$ = 0.21 vs. $\beta$ = 0.12) were stronger under high work intensity conditions compared to low intensity conditions. These findings suggest that well-being factors become more critical for sustaining productivity as work demands increase. The stronger relationship between well-being factors and productivity under high work intensity aligns with Li et al (2025), who found that work autonomy weakens the negative impact of privacy violations on job satisfaction through reduced stress. Similar to our findings regarding work intensity, their results indicate that protective factors (in their case, autonomy; in our case, well-being practices) become more important under challenging work conditions. The observed interaction effects also connect to Pendse et al (2024)’s findings that both organizational and individual supports are necessary to maintain remote worker productivity and wellbeing. Our results extend this understanding by demonstrating that the relative importance of different support mechanisms (e.g., promoting work-life balance vs. physical activity) may vary depending on work intensity levels, suggesting the need for adaptive intervention strategies.

Our cluster analysis identified three distinct employee profiles with differing patterns of work habits, well-being indicators, and productivity outcomes. “Balanced Performers” (32.4%) demonstrated moderate working hours, high physical activity, adequate sleep, and high productivity. “Digital Immersers” (41.7%) showed long working hours, high screen time, low physical activity, and moderate productivity. “Struggling Remotes” (25.9%) exhibited moderate working hours, high stress, poor sleep, and low productivity. The identification of these distinct profiles aligns with Rodríguez-Modroño and López-Igual (2021), who found that different types of telework by workplace and Communication Technology (ICT) use intensity affect working conditions and job quality differently. Our “Balanced Performers” profile parallels their finding that occasional teleworkers demonstrate the best job quality, while our “Digital Immersers” and “Struggling Remotes” profiles reflect the challenges faced by highly mobile teleworkers and home-based teleworkers, respectively. The significant differences in productivity across these profiles (F(2, 9997) = 743.26, p 0.001) highlight the importance of tailored approaches to supporting remote workers. This finding resonates with Jacobs et al (2025), who emphasized that tailored approaches may be necessary to support different subgroups of remote workers, as demographics, break behaviors, and workspace configurations all influence well-being outcomes differently.

The emphasis on Central and Eastern European contexts requires important methodological qualifications regarding sample representativeness. While the dataset includes participants from the specified geographic regions, the sampling methodology does not ensure statistical representativeness of these populations.

Recruitment through professional networks and remote work organizations likely overrepresents urban, educated, and technologically proficient workers while potentially underrepresenting rural employees, traditional industry workers, and those with limited digital access. This bias particularly affects generalizability within Central and Eastern European contexts, where significant urban-rural digital divides and varying remote work adoption rates exist across countries.

The study design did not include cultural measurement instruments or country-specific validation procedures that would support claims about region-specific applicability. The absence of comparative analysis across different national contexts within the dataset prevents confirmation that observed relationships operate consistently across diverse Central and Eastern European organizational cultures.

Future research should employ stratified sampling procedures that ensure proportional representation across countries, urban-rural divisions, and industry sectors within target geographic regions. Cross-cultural validation of measurement instruments and explicit testing of relationship invariance across national contexts would strengthen claims about regional applicability.

This research advances theoretical understanding through several specific mechanisms that extend existing frameworks rather than replacing established models.

The study contributes new knowledge by demonstrating how personal health resources operate differently in remote work environments compared to traditional workplace settings. Previous applications of the Job Demands-Resources model focused primarily on workplace-based resources, while this research shows that individual health behaviors assume enhanced importance when organizational resources become less accessible in remote settings. The identification of work intensity as a moderator reveals that resource importance varies dynamically based on demand levels, extending theoretical understanding beyond static demand-resource relationships.

The research provides empirical validation for boundary management mechanisms in digital work environments, specifically demonstrating how work-life balance satisfaction functions as a mediating pathway between work demands and performance outcomes. This finding advances theoretical knowledge by quantifying the proportion of demand effects transmitted through boundary management processes, providing empirical support for previously conceptual propositions about border permeability impacts.

The study contributes methodological innovation by demonstrating how Job Demands-Resources and Work-Life Border theories operate simultaneously rather than independently. The identification of distinct employee profiles represents a novel application of person-centered analysis to remote work research, showing how theoretical predictions manifest differently across population subgroups rather than uniformly across individuals.

The research advances applied understanding by quantifying specific relationships between work pattern modifications and well-being outcomes, providing evidence-based foundation for intervention design. The identification of work intensity as a moderating factor contributes new knowledge about when well-being interventions become most critical, informing adaptive management strategies that respond to varying demand conditions.

These theoretical contributions extend existing knowledge through empirical validation, quantification of previously unmeasured relationships, and demonstration of theoretical integration possibilities rather than introducing entirely new conceptual frameworks. The advances support evidence-based applications while maintaining grounding in established organizational psychology theories.