We aimed to identify short and long-term trajectories of self-rated health (SRH) and physical working capacity during the retirement transition, and investigate whether work-related factors and social class predict belonging to these trajectories. We used the representative, biennial Swedish Longitudinal Occupational Survey of Health (SLOSH) 2006–2018. We applied group-based trajectory modeling with B-spline smoothers to model trajectories of SRH (n = 2,183) and physical working capacity (n = 2,152) during the retirement transition. Multinomial logistic regression analyses were conducted to investigate trajectory belonging by work-related factors and social class. There was a small “honeymoon effect” in SRH for the total sample. We found four trajectories of SRH and five of physical working capacity. The large majority sustained excellent or good SRH and physical working capacity throughout the study period. Almost 6% had Fairly poor SRH and physical working capacity starting from years before retirement, which remained throughout the study period. High job demands, low job control, adverse physical working conditions, and being in manual occupation increased the likelihood of belonging to the trajectory groups Deteriorating or Fairly poor when compared with the Excellent trajectory group for both SRH and physical working capacity. Our findings suggest that for most people health status is already established some years’ preretirement and maintained for years after retirement, except a short improvement in SRH in accordance with a honeymoon effect. In order to improve health and employability, interventions focusing on working environment should be aimed at younger and midlife employees as well as older workers.

Large Language Models (LLMs) have revolutionized natural language processing but present significant technical and legal challenges when confronted with the General Data Protection Regulation (GDPR). This paper examines the complexities involved in reconciling the design and operation of LLMs with GDPR requirements. In particular, we analyze how key GDPR provisions—including the Right to Erasure, Right of Access, Right to Rectification, and restrictions on Automated Decision-Making—are challenged by the opaque and distributed nature of LLMs. We discuss issues such as the transformation of personal data into non-interpretable model parameters, difficulties in ensuring transparency and accountability, and the risks of bias and data over-collection. Moreover, the paper explores potential technical solutions such as machine unlearning, explainable AI (XAI), differential privacy, and federated learning, alongside strategies for embedding privacy-by-design principles and automated compliance tools into LLM development. The analysis is further enriched by considering the implications of emerging regulations like the EU’s Artificial Intelligence Act. In addition, we propose a four-layer governance framework that addresses data governance, technical privacy enhancements, continuous compliance monitoring, and explainability and oversight, thereby offering a practical roadmap for GDPR alignment in LLM systems. Through this comprehensive examination, we aim to bridge the gap between the technical capabilities of LLMs and the stringent data protection standards mandated by GDPR, ultimately contributing to more responsible and ethical AI practices.

Introduction: The objective of this study was to investigate the bi-directional associations between experienced and witnessed gender-based harassment (GBH) on the one hand, and depressive symptoms and psychological treatment on the other, in an occupational setting. GBH are behaviors that derogate, demean, or humiliate an individual based on his or her gender.

Methods: The analyses were based on data from the Swedish Longitudinal Occupational Survey of Health at 2018 (T1) and 2020 (T2), including 6,679 working participants (60.3% women) with a majority in the age range of 45–64. Using cross-lagged structural equational models, we analyzed experienced and witnessed GBH in relation to depressive symptoms and having received psychological treatment (talked to a counselor or psychological professional) over time.

Results: Our results showed that neither experienced nor witnessed GBH was prospectively associated with depressive symptoms or psychological treatment over two years. Both higher levels of depressive symptoms (β = 0.002, p ≤ 0.001) and having received psychological treatment (β = 0.013, p = 0.027) weakly predicted experiences of GBH over time. Having received psychological treatment was furthermore weakly associated with witnessed GBH (β = 0.019, p = 0.012).

Discussion: In conclusion, the hypothesized associations between exposure to GBH and mental health outcomes were not statistically significant, while a weak reverse association was noted. More research addressing bidirectional associations between GBH and mental health outcomes are needed.

Organizational injustice is known to negatively affect employees' health and to increase the risk for sickness absence. The negative health effects are also known to be more pronounced in uncontrollable, strain increasing, situations at the workplace. This study tests whether locked-in status, i.e., being stuck in a non-preferred workplace, modifies the associations between injustice perceptions and frequent (>= 2 times/yr) and long (>= 8 days/yr) sickness absence. The sample contained 2631 permanent employees from the Swedish Longitudinal Occupational Survey of Health in 2018 and 2020. Multigroup structural equation modelling was used to compare the proposed relationships between employees who are locked-in in their workplace and employees who are not. We found a positive association between higher overall organizational injustice and long sickness absence two years later, with the association being stronger for the locked-in group. Also, higher injustice was associated with more frequent sickness absence, but only for those not being locked-in.

Employees being locked-in seem to have higher risk of long-term sickness absence which might indicate more serious health problems. Employees not being locked-in more often take short sickness absence, which could indicate a coping behaviour to handle high strain. This study adds knowledge to the role of locked-in status as a moderator in the much-studied relationship between organizational justice and health as well as to the multiple reasons underlying sickness absence.

Background: The aim of this paper was to investigate if job demands, decision authority, and workplace violence mediate the association between employment in the health and social care industry and register-based sickness absence. Methods: Participants from the Swedish Longitudinal Occupational Survey of Health who responded to questionnaires in 2006–2016 (n = 3951) were included. Multilevel autoregressive cross-lagged mediation models were fitted to the data. Employment in the health and social care industry at one time point was used as the predictor variable and register-based sickness absence >14 days as the outcome variable. Self-reported levels of job demands, decision authority, and exposure to workplace violence from the first time point were used as mediating variables. Results: The direct path between employment in the health and social care industry and sickness absence >14 days was, while adjusting for the reverse path, 0.032, p = 0.002. The indirect effect mediated by low decision authority was 0.002, p = 0.006 and the one mediated by exposure to workplace violence was 0.008, p = 0.002. High job demands were not found to mediate the association. Conclusion: Workplace violence and low decision authority may, to a small extent, mediate the association between employment in the health and social care industry and sickness absence.

Growing evidence indicates that retiring from paid work is associated, at least in the short-term, with dramatic reductions in sleep difficulties and more restorative sleep. However, much is still not known, in particular how universal these improvements are, how long they last, and whether they relate to the work environment. A methodological challenge concerns how to model time when studying abrupt changes such as retirement. Using data from Swedish Longitudinal Occupational Survey of Health (n = 2,148), we studied difficulties falling asleep, difficulties maintaining sleep, premature awakening, restless sleep, a composite scale of these items, and non-restorative sleep. We compared polynomial and B-spline functions to model time in group-based trajectory modelling. We estimated variations in the individual development of sleep difficulties around retirement, relating these to the pre-retirement work environment. Reductions in sleep difficulties at retirement were sudden for all outcomes and were sustained for up to 11 years for non-restorative sleep, premature awakening, and restless sleep. Average patterns masked distinct patterns of change: groups of retirees experiencing greatest pre-retirement sleep difficulties benefitted most from retiring. Higher job demands, lower work time control, lower job control, and working full-time were work factors that accounted membership in these groups. Compared to polynomials, B-spline models more appropriately estimated time around retirement, providing trajectories that were closer to the observed shapes. The study highlights the need to exercise care in modelling time over a sudden transition because using polynomials can generate artefactual uplifts or omit abrupt changes entirely, findings that would have fallacious implications.

Sequential probit approach is used to model differentials in the effects of women’seducation on parity progression in Ethiopia. Since reasons to have a first childmay differ from those to have, say, a second or third child, we allow the effects ofcovariates on the progression propensities to vary between parities in the same model.Data used for illustration come from the Ethiopian Mini Demographic and HealthSurvey of 2019 in which 8885 women from 11 regions were interviewed. Resultsshow that the sequential model provides more insight than conventional modelswhen exploring the association between education and parity progression. We alsofound similarities and differences in the effects of education on parity progressionamong the regions. A random effect term to account for women’s clustering withinhouseholds was significant in a model for the entire country but disappeared whenregion was included as a covariate in the model.

The COVID-19 pandemic has altered workers' possibilities to combine work and private life. Work and private life could either interfere with each other, that is, when conflicting demands arise, or enrich, that is, when the two roles are beneficial to one another. Analyzing data from the Swedish Longitudinal Occupational Survey of Health through individual growth models, we investigated time trends of interference and enrichment between work and private life from 2016 through March to September 2020, which is during the first wave of the pandemic. The sample included workers who had remained in the same workplace throughout the study period and worked at least 30% of full time, reaching 5,465 individuals. In addition, we examined trends in level of interference and enrichment across gender and industries. Results showed that Life-to-work interference increased over time in the Swedish working population, but neither did work-to-life interference nor enrichment. We observed only marginal differences across gender. Also, in the industries of fine manufacturing and real-estate activities, a decrease in interference, work-to-life interference, and life-to-work interference, respectively, was observed. In the human health and social care industry, an increase in interference and life-to-work interference was seen. Our conclusion is that overall changes to the possibilities to balance work and private life have occurred for workers in Sweden during the first period of the pandemic. Further studies are needed to study development time trends throughout the pandemic and across different occupations.

The chapter aims to contribute to the understanding of increased gender differences in sickness absence and health by applying a structural approach, and by analysing and describing the development of and differences in psychosocial working conditions in differently gendered industries. Some psychosocial work factors are expected to vary by differently gendered industries, and others are expected to vary by differently gendered industries and gender combined. Descriptive and comparative data of psychosocial working conditions for women and men in Sweden from the 1990s and onwards are presented. The results show that job demands and decision authority have developed more negatively in female-gendered industries for both genders since the early 1990s, although more negatively for women. Job demands are furthermore higher and decision authority lower in these industries among both men and women. Gender-based and sexual harassment, on the other hand, show a pattern between gendered industries and gender combined, such that more harassment occurs for men in female-gendered industries and for women in male-gendered industries. The results are discussed in relation to contextual and gender perspectives.