In this large observational follow-up study of individuals with T1D, our main finding was that excessive occupational sitting is associated with an increased risk of cardiovascular events, independent of leisure-time physical activity. Notably, no such association was observed for all-cause mortality. However, in a subanalysis of current smokers, excessive occupational sitting was linked to an increased risk of all-cause mortality. Additionally, our cross-sectional analysis identified variables independently associated with excessive occupational sitting. Age and higher occupational category were positively associated, while current smoking, moderate albuminuria, and high levels of leisure-time physical activity were negatively associated. Overall, our findings indicate that current physical activity guidelines need updating to better address sedentary behavior and improve outcomes for individuals with T1D. Occupational sitting, as a specific domain of sedentary behavior, should be considered a key area for interventions aimed at reducing overall sedentary time.

This is the first study, to our knowledge, assessing long-term health outcomes of occupational sitting or sedentary behavior in general in T1D.

Our findings in individuals with T1D align with a positive association between sedentary behavior and cardiovascular events in the general population [32]. Despite a significant increase in research over the past decade, evidence regarding occupational sitting as a specific domain of sedentary behavior remains inconclusive [33, 34]. Several factors may contribute to this uncertainty. One key issue is the potential differences between the correlates of occupational sitting and those of leisure-time or total sitting time, which underpin existing conclusive evidence linking sedentary behavior to adverse health outcomes. For instance, while socioeconomic status is positively associated with occupational sitting, the association is inverse for leisure-time sitting [35]. Similarly, current smoking, a fundamental cardiovascular risk factor, is negatively associated with occupational sitting in contrast to leisure-time sedentary and total sedentary time [35, 36]. Consequently, when considering the impact of occupational sitting on adverse health outcomes, socioeconomic status and its related multitude of health factors might offset a positive association without thorough adjustments in the analyses [37]. Our findings among individuals with T1D highlight this complexity. We observed that those with excessive occupational sitting were more likely to be in higher occupational categories—a proxy for socioeconomic status—and less likely to be current smokers, both of which act as protective factors despite the risks associated with excessive occupational sitting. However, not all beneficial lifestyle habits were positively associated with excessive occupational sitting, as these individuals were less frequently highly physically active during leisure time. Another reason for the inconclusive evidence on occupational sitting is the substantial heterogeneity in measurement methods, study designs, and outcome variables. This variability makes it difficult to draw definitive conclusions [33, 34, 38]. In contrast, assessments of overall sedentary behavior are generally more consistent and uniform [34]. Additionally, many studies rely on categorical measures (e.g., ‘sitting most of the time’ versus ‘hardly ever’), and the absence of more precise, quantified measures may help explain the absence of significant associations between occupational sitting and health outcomes [33].

Despite the inconclusive evidence regarding the impact of occupational sitting on cardiovascular outcomes in the general population, we managed to demonstrate a positive association in our cohort of individuals with T1D. One plausible explanation emerges as we explore the independent mechanisms by which sedentary behavior is hypothesized to impair vascular health. In addition to potential sitting-induced alterations in traditional cardiovascular risk factors (e.g. lipid profile), the hypothesized vascular function impairing mechanisms include metabolic- (e.g., skeletal muscle insulin resistance), inflammatory- (e.g., promotion of low-grade inflammation), and hemodynamic (e.g., reduced shear stress) processes [39, 40]. Simultaneously, individuals with T1D have a pronounced risk of cardiovascular disease due to the inherent nature of the disease, where atherosclerosis occurs at younger age, and the progression is more aggressive [41, 42]. Therefore, one might hypothesize that the cardiovascular risk associated with sedentary behavior may be more pronounced in individuals with T1D, potentially enabling the demonstration of a positive correlation between occupational sitting and cardiovascular events, compared to studies among populations without diabetes. Furthermore, we managed to adjust for key confounders including occupational category, current smoking, and leisure-time physical activity. These adjustments are critical as they have the potential to offset a positive association between occupational sitting and cardiovascular events, as previously discussed and supported by our unadjusted analyses. Finally, using a more quantified measure of sitting time, as recommended by the seminal systematic review on occupational sitting and adverse health outcomes by van Uffelen et al. [33], instead of broad categorical variables like ‘sitting most of the time’ versus ‘hardly ever’, might have helped our study demonstrate a positive association between excessive occupational sitting and cardiovascular events.

In the general population, sedentary behavior is consistently associated with all-cause mortality [6, 7]. This association has recently been replicated in individuals with type 2 diabetes [43]. However, the evidence regarding occupational sitting and all-cause mortality, akin to cardiovascular events, remains inconclusive [33, 34]. While we identified an association between excessive occupational sitting and increased risk of cardiovascular events, we did not observe a corresponding association with all-cause mortality in our cohort of individuals with T1D. There are several plausible explanations for this discrepancy. First, it should be highlighted that our specific focus on occupational sitting led to the exclusion of a significant subset of our initial cohort. Specifically, participants lacking occupational sitting data were excluded, as outlined in the methods section. The predominant reason was individuals on disability pension. Considering the older age and worse health condition among those excluded, especially regarding diabetic complications, our final cohort may have been affected by survival bias. Consequently, this bias might have diminished a potential positive association between excessive occupational sitting and all-cause mortality. If we acknowledge the possibility of a survival bias, it prompts us to question why a positive association between excessive occupational sitting and cardiovascular events persisted. It is conceivable that the deleterious effects of sedentary behavior primarily target the development of cardiovascular events. In essence, although cardiovascular deaths are the leading cause of premature mortality among individuals with T1D [44], the impact of sedentary behavior may have relatively less significance for other causes of premature mortality. Another simplistic rationale is that cardiovascular events typically precede mortality, as not all cardiovascular events result in death, thus mitigating the impact of potential survival bias.

However, among currently smoking individuals with T1D, we managed to demonstrate that excessive occupational sitting increased the risk of all-cause mortality, even after adjusting for leisure-time physical activity, duration of diabetes, occupational category, and diabetic kidney disease. Although a positive trend was observed, the association between occupational sitting and cardiovascular events did not remain statistically significant after final adjustments. In both the general population and individuals with T1D, smoking is associated with cardiovascular events and all-cause mortality [45]. Furthermore, in the general population, smoking has also been shown to exacerbate the harmful effects of sitting [13]. Both behaviors contribute for example to dyslipidemia by inhibiting lipoprotein lipase activity, leading to higher triglyceride levels and lower HDL-cholesterol [39, 46], thus increasing the risk of cardiovascular events and mortality. Consequently, based on biological plausibility and literature, smoking may have accelerated the harmful impacts of occupational sitting in our study cohort, revealing a significant association with all-cause mortality. The lack of statistical significance for cardiovascular events may stem from insufficient statistical power in the analysis, even though a positive trend was observed. In contrast, the significant association with all-cause mortality, observed with comparable statistical power, suggests that smoking may exacerbate the harmful effects of occupational sitting beyond cardiovascular risks, potentially including links to cancer and depression [47,48,49]. Based on these findings, there might be certain high-risk populations among individuals with T1D, such as smokers, where occupational sitting is particularly hazardous, and targeted interventions needed. In the interpretation of this subanalysis among current smokers, however, it is important to acknowledge the limited number of events.

Finally, in addition to the prospective results, our cross-sectional results are novel, as there is no previous literature on variables independently associated with occupational sitting in individuals with T1D. Perhaps the most intriguing cross-sectional finding in our study was the inverse association between excessive occupational sitting and moderately increased albuminuria. This contrasts with prior, albeit limited, research from the general population reporting a positive association between occupational sitting and proteinuria, and declined kidney health [50, 51]. Notably, we found no significant relationship between more severe kidney conditions and excessive occupational sitting, only with moderate increases in albuminuria.

Several factors may explain our discrepant findings with existing literature. Firstly, our cohort may have experienced survival bias, as a significant number of individuals were excluded due to missing occupational data, particularly those on disability pensions. In contrast, only 9% of participants were excluded for this reason in the study by Tsai et al. [50]. Importantly, our higher exclusion rate reflects reliable measures of participants’ current occupational sitting status, as well as the reduced likelihood of labor force participation among individuals with T1D [52]. Additionally, since other studies were conducted in Asian populations, there could be residual confounding from dietary factors. Diets high in red meat, animal fat, or those with a high acid load—more common in countries like Finland—may adversely affect kidney function, particularly in conditions associated with chronic kidney disease such as T1D [53].

An alternative explanation is that occupational sitting may not significantly influence kidney disease in individuals with T1D. Instead, the observed association between moderate increases in albuminuria and reduced sitting time during work could be attributed transient albuminuria due to various benign causes, including posture, physical activity, and hydration status [54, 55]. For instance, postural albuminuria may be possible in less sedentary, physically active jobs performed in an upright position. Additionally, both strenuous physical activity and dehydration can cause transient albuminuria, both possible in less sedentary jobs. It is unlikely that any single factor can fully explain our observed association. Postural proteinuria, for example, is rare in individuals over 30 years old, at least in the general population [54]. However, individuals with T1D may experience transient proteinuria more easily due to a heightened sensitivity to urinary protein fluctuations. For instance, an exercise-induced increase in urinary albumin excretion rate can occur even with lighter physical activity in individuals with diabetes [55], in contrast to healthy peers, where such effects are typically seen only with more intense physical activity [56].

Major strengths of this study include the nationwide cohort of individuals with T1D and a long follow-up period. Furthermore, the physical activity questionnaire is detailed and has been previously validated in Finnish conditions. We were also able to adjust for important confounders such as occupational category, current smoking, and leisure-time physical activity. In this study, however, several limitations must also be acknowledged. Firstly, the use of a self-report questionnaire introduces the potential for both report and recall bias. While objective tools like accelerometers could mitigate these biases, their use in a large nationwide cohort would have been impractical and poses its own challenges. Another limitation is the inability to examine sedentary behavior patterns in detail. Emerging evidence suggests that not only the total sedentary time but also the pattern of sedentary behavior, such as shorter bouts with frequent breaks, correlates with cardiometabolic markers like postprandial glycemia, insulin responses, and endothelial function [57,58,59]. In T1D, a recent interventional study demonstrated that interrupting prolonged sitting with light activity every 30 min improved postprandial glucose levels, with benefits lasting up to 48 h [16]. However, while these studies used accelerometers, large-scale studies like ours rely on self-reported data, which significantly limits the ability to reliably capture sedentary patterns, such as the duration of continuous sedentary bouts and the frequency of breaks. Moreover, our study did not account for occupational physical activity (OPA). Additionally, the exclusion of a significant subpopulation from our initial cohort limits the generalizability of our findings to individuals with T1D who are actively employed. Lastly, despite adjustments for important confounders, there remains a risk of residual confounding due to unaccounted lifestyle factors such as diet and alcohol consumption.