Working against the clock: does night shift work drive multisystem disease?

This editorial refers to ‘Night shift work and its interaction with sleep duration and chronotype, and risk of cardiovascular-kidney-metabolic diseases in the UK Biobank’, by D.T. Tran et al., https://doi.org/10.1093/eurjpc/zwaf774. Modern societies depend heavily on a workforce that operates around the 24-hour clock. Night shift work has therefore become an integral feature of healthcare, manufacturing, transportation, and service industries worldwide, affecting more than one in five workers. However, accumulating epidemiological and mechanistic evidence suggests that working against one’s endogenous circadian rhythm may have significant long-term health consequences. Night shift work is known to disrupt circadian alignment; alter sleep timing and duration; adversely affect neuroendocrine, metabolic, and inflammatory pathways; and potentially underlie the development of chronic disease.1 Most epidemiological research on night shift work to date has focused on isolated disease outcomes, particularly cardiovascular disease2 and type 2 diabetes,3 reinforcing the notion that circadian disruption is metabolically harmful. While these studies have yielded important insights, this approach may underestimate the broader systemic burden of circadian disruption by overlooking the interconnected nature of chronic disease. Importantly, the relationship between night shift work and chronic kidney disease has been relatively understudied, despite the rising global prevalence of kidney disease and strong mechanistic plausibility, sharing multiple underlying pathways with cardiometabolic disease. Experimental and clinical evidence suggests that renal circadian dysregulation, metabolic stress, inflammation, and sleep loss all converge on pathways known to drive kidney disease.4 The relative paucity of epidemiological evidence for kidney outcomes likely reflects slow disease progression, under-ascertainment of early disease, and the exclusion of kidney endpoints from occupational health studies, rather than an absence of biological susceptibility. There is growing recognition that cardiovascular, kidney, and metabolic disorders frequently cluster and share common pathophysiological mechanisms. Recognizing this interconnectedness, the American Heart Association has recently emphasized cardiovascular-kidney-metabolic (CKM) health as a clinically relevant and integrated framework for risk stratification and prevention.5 Examining occupation exposures through this lens may offer a more realistic and clinically meaningful assessment of disease burden than analyses restricted to single endpoints. However, few large-scale population studies have operationalized this approach in the context of night shift work, and evidence linking night work to kidney disease remains especially limited. The study by Tran et al.6 addresses several of these gaps by leveraging the depth and scale of the UK Biobank. Drawing on more than 100 000 individuals with detailed occupational data, the authors examined associations between history of night shift work and incidence of a composite CKM outcome, ascertained through linkage to hospital inpatient records, death registries, and self-reported diagnoses. This approach allows for longitudinal assessment while minimizing outcome misclassification. Importantly, the analysis focused on incident disease, thereby reducing the likelihood that associations were driven by pre-existing morbidity or reverse causation. A key strength of the study is its detailed characterization of night shift work. By examining multiple dimensions of night shift exposure, including duration (years of night shift work), intensity (night shifts per month), and cumulative lifetime exposure (number of lifetime night shifts), the analysis moves beyond simplistic exposure definitions. Across exposures, the authors observed graded associations between night shift work and CKM risk. Participants with long-term or high-intensity night shift work exhibited 32–36% higher risk of CKM outcomes compared with day workers. Interestingly, short sleep duration (≤ 6 h) was found to substantially amplify the association between night shift work and CKM risk, with an HR 1.51 (95% CI 1.27,1.79) observed among short sleepers with long-term night shift exposure (vs. day workers), compared with HR 1.28 (95% CI 1.13, 1.44) among those with normal sleep duration and HR 0.95 (95% CI 0.67, 1.35) among longer sleepers. These findings were consistent across both self-reported and accelerometry-derived sleep measures and were evident on both additive and multiplicative interaction scales, strengthening confidence in the findings. These findings are biologically plausible since short sleep may exacerbate circadian misalignment by increasing sympathetic tone, dysregulating the hypothalamic–pituitary–adrenal axis, impairing metabolic control, and limiting the capacity to adapt sleep schedules around night work. In this context, sleep duration may represent a marker of shift work vulnerability and a potential point of intervention. In contrast, chronotype did not appear to modify the association between night shift work and CKM outcomes. Although evening chronotypes are overrepresented among night workers and are often hypothesized to adapt better to nocturnal schedules, any adaptive advantage may be offset by adverse behavioural or metabolic profiles. Alternatively, the null finding may reflect limitations of the chronotype measurement, which was assessed using a relatively coarse self-report measure, and misalignment between endogenous circadian timing and imposed work schedules may not be fully captured by chronotype categories alone. These findings contrast with some earlier studies,7 but are broadly consistent with prior UK Biobank analyses suggesting limited effect modification by chronotype in the relationship between night shift work and diabetes risk.3 Improved phenotyping of circadian timing may be required to resolve these inconsistencies. However, findings from Tran et al. suggest that behavioural sleep loss may be more critical than intrinsic circadian preference in influencing CKM risk among night workers. From a methodological perspective, the study undertook extensive efforts to address confounding and bias. Models were adjusted for demographic and socioeconomic factors, with partial attenuation of effect estimates after socioeconomic adjustment, highlighting the role of social patterning in occupational health risks. Notably, body mass index (BMI) was not included as a primary confounder, justified on the basis of its role as a potential mediator on the causal pathway between night work and CKM disease. This decision is defensible, but it highlights a broader challenge in interpreting the observational evidence. Descriptive analyses demonstrated a substantially higher prevalence of obesity among long-term night shift workers compared with day workers (26.6% in those working >20 years vs. 16.7% in day workers). Plausibly, BMI could act as both a mediator and a source of confounding through shared upstream determinants of shift work participation and CKM risk. Importantly, genetic evidence from the same cohort suggests that higher BMI increases the likelihood of selection into shift work,8 further complicating causal inference. While including BMI in the model could be deemed to be an overadjustment, incomplete control for adiposity-related pathways may also contribute to residual confounding. Further work to disentangle these relationships will be essential for determining how much of the observed risk is attributable to night work, and how much reflects social, economic and behavioural drivers of nighttime occupations. Other limitations merit consideration. Competing risk analyses and comparisons of mortality among current and former night workers provided limited evidence for survival bias, but healthy worker effects, selection into night work, and selection bias related to participation in the occupational history questionnaire cannot be excluded. The study could not distinguish between permanent and rotating night shift schedules, follow up was relatively short (median follow-up of 7.2 years), and several relevant pathways, including sleep apnoea, light-at-night exposure, diet timing, and sleep quality and timing, were not assessed. Finally, multiple testing considerations should be acknowledged when interpreting ‘statistical significance’ as defined in the study. Overall, this study represents an important advance in occupational cardiometabolic epidemiology. By combining detailed lifetime measures of night shift work with an integrated cardiovascular-kidney-metabolic outcome and rigorous sensitivity analyses, the findings suggest that chronic night work may confer systemic health risks that extend beyond isolated disease endpoints. While causal inference remains limited by potential residual confounding, particularly related to adiposity and behavioural factors, the results highlight night shift work and short sleep duration as potentially important risk factors for cardiovascular-kidney-metabolic disease (Central Illustration). These findings underscore the importance of incorporating work schedules into risk stratification and prevention strategies for cardiometabolic and renal disease, particularly as societies continue to rely on an expanding night-time workforce. Rebecca Richmond (Conceptualization [lead]; Funding acquisition [equal]; Writing—original draft [lead]; Writing—review & editing [equal]), and David W. Ray (Funding acquisition [equal]; Writing—review & editing [equal]) This project is supported by the National Institute for Health and Care Research (NIHR) Programme Grant for Applied Research programme (NIHR203667) and the Oxford Health Biomedical Research Centre funded by the National Institute for Health and Care Research (NIHR) (NIHR203316). The views expressed are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care. R.C.R. and D.W.R are supported by the Medical Research Council (UKRI518). R.C.R. works in the Integrative Epidemiology Unit which is funded by the Medical Research Council (grant number MC_UU_00032/1). No new data were generated or analysed in support of this editorial.