Accelerometry-measured prolonged and interrupted sedentary behavior and cancer incidence and mortality: A cohort study of 91,292 UK Biobank participants

Figures
Abstract
Background
Current sedentary behavior (SB) guidelines primarily emphasize total time spent sedentary. We explored differences between interrupted and prolonged SB in relation to a range of cancer outcomes.
Methods and findings
This study included 91,292 UK Biobank participants with valid accelerometer data. Participants were followed for a median of 12.38 years (interquartile range 11.56–13.15 years). A two-step approach based on a random forest model was used to classify SB. Multivariable Cox proportional hazards models were applied to overall incident cancers and cancer deaths, plus obesity-related and type-2 diabetes-related cancers, and 23 site-specific cancers. Models were adjusted for demographic, socioeconomic, lifestyle, dietary, and health-status factors, including age, ethnicity, deprivation, education, smoking, alcohol intake, diet, and morbidity count. Isotemporal substitution models were used to estimate the associated cancer risk when replacing prolonged SB with intermittent SB, or physical activity (PA). After adjusting for sociodemographic and lifestyle factors, each additional hour of prolonged SB was associated with a higher risk of overall cancer mortality (hazard ratio [HR] HR1hour 1.09; 95% confidence interval [CI] [1.06, 1.11]; p < 0.001). Replacing 1 hour per day of prolonged SB with light PA (HRLPA 0.88; 95% CI [0.79, 0.99]; p = 0.033) was associated with lower risk of overall cancer mortality. Similarly, replacing 30 min per day of prolonged SB with moderate PA (HRMPA 0.92; 95% CI [0.86, 0.99]; p = 0.024) was associated with a lower risk of overall cancer mortality. The main methodological limitations were observational design, residual confounding, healthy volunteer bias, and measurement imprecision due to having only 7 days of accelerometer wear.
Author summary
Why was this study done?
- Previous studies have shown that spending more total time on sedentary behavior, such as sitting or reclining while awake, is linked to poorer health outcomes.
- Some evidence suggests that replacing sedentary time with physical activity (PA) may lower the risk of cancer death, but less is known about whether the pattern of sedentary time matters.
- This study was done to examine whether long, uninterrupted periods of sedentary behavior and shorter, interrupted periods of sedentary behavior are differently associated with cancer risk.
What did the researchers do and find?
- We studied 91,292 UK Biobank participants with valid activity monitor data and followed them for a median of 12.38 years.
- Each additional hour per day of prolonged sedentary behavior was associated with a 10% higher hazard of cancer death, whereas each additional hour of interrupted sedentary behavior was associated with a 19% lower hazard of cancer death.
- Replacing prolonged sedentary behavior with PA was associated with a lower risk of cancer death: a 12% lower hazard when 1 hour per day was replaced with light PA, an 8% lower hazard when 30 min per day was replaced with moderate PA, and 22% lower hazard when 5 min per day was replaced with vigorous PA.
What do these findings mean?
- These findings suggest that not only the total amount of sedentary time, but also how sedentary time is accumulated, may be important for cancer risk.
- Reducing long, uninterrupted periods of sedentary behavior and replacing them with PA, even those with light intensity, may be a practical target for future interventions.
- These findings should be interpreted with caution because the study cannot prove causality, UK Biobank volunteers may not represent the wider population, and the activity monitor captured behavior only during a limited period without showing the context of sedentary behavior, such as work, television viewing, or driving.
Citation: Zhou Z, Trost SG, Ryde GC, Parra-Soto S, Fang Z, Xu C, et al. (2026) Accelerometry-measured prolonged and interrupted sedentary behavior and cancer incidence and mortality: A cohort study of 91,292 UK Biobank participants. PLoS Med 23(7): e1004767. https://doi.org/10.1371/journal.pmed.1004767
Academic Editor: Steven C. Moore, National Cancer Institute, UNITED STATES OF AMERICA
Received: September 11, 2025; Accepted: May 29, 2026; Published: July 2, 2026
Copyright: © 2026 Zhou et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: This study was conducted using third-party data from the UK Biobank (www.ukbiobank.ac.uk) under application Number 71392. The authors are not permitted to publicly deposit or redistribute the individual-level data or the minimal dataset underlying this study because access to UK Biobank data is governed by legal and contractual restrictions. Data are available to eligible researchers through the UK Biobank Access Management System for health-related research that is in the public interest. Researchers may apply for access through the UK Biobank website (www.ukbiobank.ac.uk/use-our-data/apply-for-access). Further inquiries about access to UK Biobank data may be directed to UK Biobank at access@ukbiobank.ac.uk.
Funding: This work was supported by the American Cancer Society (grant code MRSG-17-200-01-NEC to MS, www.cancer.org), and the U.S. National Institutes of Health (grant code U01CA261961, R01CA263776, and R01CA285851 to MS, www.nih.gov). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: We have read the journal’s policy, and the authors of this manuscript have the following competing interests: FH is a paid statistical consultant on PLOS Medicine’s statistical board. The other authors have declared that no competing interests exist.
Abbreviations: BMI, body mass index; CI, confidence interval; CVD, cardiovascular disease; HR, hazard ratio; ICD-10, International Classification of Diseases, 10th revision; IGF-I, insulin-like growth factor I; ISM, Isotemporal substitution models; LTCs, long-term conditions; METs, metabolic equivalents of task; MLTCs, multiple long-term conditions; MVPA, moderate-to-vigorous PA; PA, physical activity; RCT, randomized controlled trial; SB, sedentary behavior; SBRN, Sedentary Behavior Research Network; SDs, standard deviations; T2D, type-2 diabetes; WC, waist circumference; WHR, waist-to-hip ratio; WHtR, waist-to-height ratio
Introduction
Sedentary behavior (SB), typically defined as any waking activities involving an energy expenditure of 1.5 metabolic equivalents of task (METs) or less while in a sitting, reclining, or lying posture, accounts for ~55% of waking time in both children and adults based on self-reported data [1]. Existing evidence consistently demonstrates that excessive SB, such as prolonged daily sitting, is a risk factor for cardiovascular disease (CVD), type-2 diabetes (T2D), obesity, certain cancers, and all-cause mortality [2,3], particularly among individuals with low levels of physical activity (PA) [4]. This association may result from SB replacing PA or as an independent effect.
Recent evidence suggests that how SB is accumulated is important; not only the total amount of SB accumulated [1]. Prolonged, uninterrupted bouts of SB have been identified as potentially the most hazardous and interrupting prolonged bouts with activity breaks has been recommended by several health agencies as a viable strategy to mitigate associated health risks [5]. Experimental crossover studies provide a plausible biological rationale for this recommendation: interrupting prolonged sitting with brief bouts of light- or moderate-intensity activity attenuates postprandial glucose and insulin responses in both overweight/obese and healthy normal-weight adults, indicating that the pattern of SB accumulation can acutely influence metabolic regulation [6,7]. However, current evidence regarding optimal frequency and duration of sedentary breaks remains limited; it is often based on self-reported data, and experimental studies often lack clinical endpoints [8]. Importantly, the roles of interrupted and prolonged SB in cancer outcomes have not been examined.
In addition to examining whether types of SB could be related to cancer risks, it is also important to examine how best to change or replace it. It is not possible to reduce SB without increasing the time spent on other activities. Potential harms associated with SB may be mitigated through a replacement with PA [1,9]. According to the 2018 Physical Activity Guidelines Advisory Committee Scientific Report [1], higher amount of moderate to vigorous PA (e.g., 40–100 min per day or over 300 min per week) may offset or eliminate the risks associated with prolonged SB. It is also possible that risk could be lowered simply by interrupting SB with some forms of PA without actually increasing the total PA. However, there is a lack of evidence on these important questions in relation to cancer outcomes, especially based on objective measurements of SB and PA.
In addition, it is also important to examine whether the association of SB with cancer differs by obesity status [10,11]. Obesity is an important risk factor for cancer, and previous studies have shown obesity to modify the associations between SB and PA and health outcomes [12,13]. If the cancer risk associated with SB differs between people with and without obesity, interventions may need to be tailored to optimize health outcomes across different population subgroups.
To address these gaps, we conducted a prospective cohort study to examine whether prolonged and interrupted sedentary behavior are differently associated with cancer risk. Specifically, we evaluated associations of prolonged and interrupted sedentary behavior with overall incident cancer, overall cancer mortality, obesity-related cancers, T2D-related cancers, and 23 site-specific cancers. We also examined whether replacing time spent in sedentary behavior with other daily behaviors, including PA, was associated with cancer risk, and whether associations differed by obesity status. We hypothesized that prolonged sedentary behavior would be associated with higher cancer risk, whereas interrupted sedentary behavior and replacing sedentary time with PA would be associated with lower cancer risk.
Methods
UK Biobank is a prospective, population-based cohort study that recruited 502,493 participants aged 37–73 years at baseline from the general population (with a 5.5% response rate) between 2006 and 2010 [14]. Participants attended one of 22 assessment centers across Scotland, England, and Wales [14]. All participants completed touch-screen questionnaires and underwent physical measurements related to social demography, lifestyle, health, and physical assessments at baseline [15]. More information about the UK Biobank protocol can be found online (http://www.ukbiobank.ac.uk).
Of the 103,068 participants with valid device-measured PA data, 11,776 were excluded because they had a history of cancer at baseline, had insufficient wearing time (<72 hours), or withdrew from UK Biobank. A total of 91,292 UK Biobank participants were included in this study (Fig 1). The median (interquartile range) of follow-up was 12.38 (11.56–13.15) years. This study is reported as per the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline (S1 Checklist).
Accelerometer-measured PA and sedentary behaviors
Accelerometer-measured PA was obtained, using Axivity AX3 wrist-worn triaxial accelerometers, from 103,567 UK Biobank participants between 2013 and 2015. Participants who provided an email address to UK Biobank were invited at random [16]. The accelerometer was worn on the dominant wrist for a period of 7 days, with recording at 100 Hz, as has been described elsewhere [16]. The accelerometer is water resistant up to 1.5 meters and the participants were advised to wear it for 7 days continuously. Only participants in the accelerometer sub-study were included in this analysis. Those with insufficient wear time (n = 8,480; < 96 hours, based on GGIR rule) or poor device calibration (n = 11; based on previous UK Biobank analysis [16]) were excluded. There were 95,076 participants with valid accelerometer data.
A validated two-step classifier based on a random forest model was used to categorize raw acceleration as SB, light, moderate and vigorous PA [17,18]. Previous field evaluations showed positive and negative predictive values for sedentary behavior to be 0.88 and 0.97, respectively [19]. Briefly, the random forest model uses features in the raw acceleration signal to classify raw accelerometer data (100 Hz) into 10-s window of SB (e.g., watching television), utilitarian movements (e.g., ironing a shirt, washing dishes), walking, or high energy activities. Light PA (LPA) included utilitarian movements, and walking at a low speed (<3mph); moderate PA (MPA) included walking activities at a moderate speed (3–4 mph); vigorous PA (VPA) included walking activities at brisk intensity (>4 mph) and high energy activities (e.g., running, actively playing a tennis game). It should be noted that accelerometry provided precise measurements of PA intensity in 10-s intervals as opposed to self-reported measurements, which provide average measurements for given activities. For example, playing badminton leisurely is usually classified as MPA based on self-reported data. However, it actually consists of standing with movement, walking, and running, and the classifier would thus identify a mix of these activity classes over the duration of game play. Mapping these to the traditional intensity-based categories, a combination of LPA, MPA, and VPA would be identified, whether the ‘average intensity’ would depend on the proportion of those PA. This two-step classifier was found to be more accurate than classification purely based on accelerations [17,18], and has been previously used and published [20,21]. Sleep was classified based on a validated method [22] to distinguish nonwear from sleep.
Prolonged SB was defined as any bout of SB that lasted at least 30 min and during which at least 90% of the time was sedentary, while interrupted behaviors to be those that did not meet the threshold above, i.e., those lasted less than 30 min or those broken down by >10% of non-sedentary behaviors. Prolonged and interrupted SB are mutually exclusive. We defined prolonged SB different from PAs to better capture how individuals allocate their time and to coincide with existing evidence. Current randomized controlled trial (RCT) and cohort studies indicate often defined prolonged sedentary bouts as those ≥ 30 min [23–25]. The accelerometer data scoring was done using the publicly available R package actimetric (https://github.com/PhysicalActivityOpenTools/actimetric) with the ‘Adult Wrist RF Trost’ algorithm.
Outcomes
The study outcomes were overall cancer mortality and incidence of overall, obesity-related, T2D-related, and site-specific cancers. After excluding cancers with case numbers fewer than 50, 17 cancer sites were included for both sexes (esophageal, kidney, stomach, liver, pancreas, colorectal, bladder, oral, lung, melanoma, thyroid, multiple myeloma and malignant plasma cell neoplasms, brain, gallbladder, leukemia, non-Hodgkin lymphoma and hepatocellular carcinoma), two specific to men (testicular and prostate cancer), and four specific to women (breast, uterine, cervical, and ovary cancer). Obesity-related and T2D-related cancers were identified based on strong or highly suggestive existing evidence of their relationships with obesity and T2D, respectively [26]. Specifically, obesity-related cancers included esophageal, liver, kidney, myeloma, pancreatic, colorectal, gallbladder, breast, ovarian, and thyroid cancer. T2D-related cancers included thyroid, breast, liver, pancreatic, endometrial, esophageal, colorectal, kidney, gallbladder, ovarian, bladder cancer, and non-Hodgkin lymphoma and leukemia.
Obesity measures
Body mass index (BMI) was calculated as weight (kg) divided by height squared (m²). Standing height (cm) was measured using a Seca 202 device (SECA). Weight was measured using the Tanita BC-418MA body composition analyzer (Tanita Corporation of America). Waist and hip circumferences were measured in centimeters by trained nurses following a standardized protocol. Wessex nonstretchable sprung tape was used to measure waist and hip circumferences [28], and the waist-to-hip ratio (WHR) was calculated by dividing waist circumference by hip circumference [29]. Waist-to-height ratio (WHtR) was calculated by dividing waist circumference by height [29]. General obesity was defined as a BMI ≥ 30 kg/m². Central obesity was classified as follows: WHR > 0.95 for men and >0.85 for women; waist circumference (WC) ≥ 102 cm for men and ≥ 88 cm for women; and WHtR > 0.5 [29].
Covariates
Participants’ age at the time of PA data collection was calculated using their date of birth and the date of data collection. Ethnicity was self-identified and grouped into categories: White, South Asian, Black, Chinese, other, or mixed. Socioeconomic deprivation was assessed using area-based continuous Townsend scores derived from participants’ postcodes at baseline [30]. Educational level was determined from the highest qualification reported by participants. Smoking status was self-reported and categorized as never, former, or current smoker. Alcohol consumption was calculated based on self-reported frequency and amount of alcohol intake. Dietary intake, including fruit and vegetables, red meat, processed meat, and oily fish, was assessed through a baseline food frequency questionnaire completed by participants. Morbidity count was assessed by summing 43 self-reported long-term conditions (LTCs), with participants classified as having multiple long-term conditions (MLTCs) if they reported two or more. The list of the 43 conditions is in Table A of S1 Appendix and has been used before in UK Biobank [31,32].
Statistical analyses
The characteristics of participants stratified by quartiles of total, prolonged, and interrupted SB, were summarized using means (standard deviations [SDs]) and frequencies (percentages) for quantitative and categorical variables, respectively. Cox proportional hazards models were used to analyze associations between total, prolonged, and interrupted SB and cancer outcomes, including overall cancer mortality; overall, obesity-related, and T2D-related incident cancer; as well as 23 sites-specific incident cancers. Note that even though the Cox model technically examines hazard (an instantaneous incidence rate), the associations were interpreted as ‘risk’ for easier reading. This approximation should be valid given that the outcomes are rare and that the direction of association between risk and hazard should be consistent.
Two multivariable models were used for each outcome: Model 1 adjusted for age, sex, and ethnicity; Model 2 additionally adjusted for deprivation, education, smoking status, and intake of alcohol, sugar, processed meat, red meat, fruit and vegetables, and oily fish. These covariates were selected as they were potential confounders. Sensitivity analyses further adjusting for BMI and morbidity count were conducted to examine the associations conservatively, since BMI and morbidity count could be confounders or mediators.
Cox proportional hazards models were also applied to analyze associations between different intensities of PA and composite and site-specific cancers. SB-adiposity interactions were additionally assessed by including a multiplicative interaction term in the model. Penalized cubic spline models were used to assess potential nonlinear relationships between each type of SB, PA and the three composite cancer outcomes.
Isotemporal substitution models (ISM) were used to estimate associated change in risk of replacing a specified amount of type-specific SB with equivalent durations of other activities [33]. Two separate ISMs were fitted. In the first ISM, prolonged SB was treated as the activity being replaced. This model estimated the associated change in risk when replacing prolonged SB with an equal duration of another behavior. For example, the coefficient for interrupted SB indicates the change in risk when 1 hour of prolonged SB was replaced by 1 hour of interrupted SB, holding all other behaviors unchanged. Similarly, the coefficients for LPA, MPA, and VPA indicate the associated risk when 1 hour, 30-min, and 5-min of LPA, MPA, and VPA were used to replace the equal duration of prolonged PA. These durations were selected based on the approximate interquartile range of each activity. In the second ISM, interrupted SB was treated as the activity and prolonged SB being replaced and the specification is similar. The formulae for the two ISMs were as follows:
Model 1 (replacing prolonged SB):
h(t) = h0(t) * exp (β1 sleep + β2 LPA + β3 MPA + β4 VPA + β5 interrupted SB + βk Covariates)
Model 2 (replacing interrupted SB):
h(t) = h0(t) * exp (β1 sleep + β2 LPA + β3 MPA + β4 VPA + β5 prolonged SB + βk Covariates)
where h(t) is the hazard function and h0(t) is the baseline hazard, and all behavioral variables were continuous variables representing the individual’s time spent in those behaviors. Due to the perfect multicollinearity inherent in 24-hour behavioral data, where the sum of time spent on sleep, prolonged SB, interrupted SB, LPA, MPA, and VPA must be 24 hours, one behavior was omitted from each model to act as the reference. In this framework, the coefficients of the included behaviors represent the theoretical effect of increasing that behavior by one unit while simultaneously decreasing the omitted behavior by the same amount. For instance, in Model 1, β2 represents the estimated hazard ratio for replacing one unit of prolonged SB with LPA. In this paper, only β2 – β5 were reported because sleep duration could be confounded by mental health and is outside the scope of this study. The ISM analyses were conducted adjusted for deprivation, education, smoking status, and intake of alcohol, sugar, processed meat, red meat, fruit and vegetables, and oily fish. All analyses were conducted using R version 4.0.3. There was no prospective analysis protocol for this study and there were no changes to the analysis plan. The data was re-analyzed once during the peer review process due to a significant update to the actimetric package used to ascertain SB and PA.
Results
There were 91,292 participants included in the final analysis (Tables 1 and Tables A–D in S1 Appendix). Of these, 51,169 [56.0%] were female and the mean age was 56.0 years (interquartile range: 7.4–7.8 years). The mean age of participants increased progressively across quartiles of total sedentary time from 54.0 years in Q1 to 58.2 years in Q4. Participants with higher total sedentary time were more likely to be deprived: mean = −1.9 [SD 2.7] in Q1 versus mean = −1.4 [SD 3.0] in Q4 and report higher educational levels. Higher total sedentary time was associated with increased weekly intake of red meat (2.0 times in Q1 versus 2.1 times in Q4) and processed meat (1.3 times in Q1 versus 1.5 times in Q4). Compared with total SB, similar findings were found across prolonged SB quartiles (Table C in S1 Appendix). Participants with prolonged SB were also more likely to be deprived - mean = −1.8 [SD 2.8] in Q1 versus mean = −1.5 [SD 3.0] in Q4 and consumed red meat more frequently, from 2.0 to 2.1 per week across the quartiles. On the contrary, compared with both total and prolonged SB, participants with higher amounts of interrupted SB had lower red meat intake, from 2.1 to 2.0 times per week and higher fruits and vegetable intake from 4.0 to 4.3 serving per day for Q1 and Q4, respectively (Table D in S1 Appendix). The proportion of participants experiencing cancer mortality, and overall, obesity-related, and T2D-related incident cancers incidence was higher in Q4 of SB than in Q1. Similar findings were also observed for several site-specific cancers: esophageal, kidney, liver, colorectal, lung, and prostate cancer (Table B in S1 Appendix).
Table 2 shows the associations between a 1-hour increment in overall, prolonged and interrupted SB and composite cancer outcomes. Each overall sedentary time was associated with a higher risk of overall cancer mortality (hazard ratio [HR]1 hour 1.12; 95% confidence interval [CI] [1.08, 1.15]; p < 0.001); overall incident cancer (HR1 hour 1.03; 95% CI [1.02, 1.04]; p < 0.001); obesity-related cancer (HR1 hour 1.06; 95% CI [1.04, 1.08]; p < 0.001); and T2D-related cancer (HR1 hour 1.07; 95% CI [1.05, 1.09]; p < 0.001). However, only prolonged SB was associated with a higher risks of cancer mortality (HR1 hour 1.09; 95% CI [1.06, 1.11]; p < 0.001); overall incident cancer (HR1 hour 1.03; 95% CI [1.02, 1.04]; p < 0.001); obesity-related cancer (HR1 hour 1.05; 95% CI [1.03,1.06]; p < 0.001); and T2D-related cancer (HR1 hour 1.05; 95% CI [1.04, 1.07]; p < 0.001). In contrast, interrupted SB was associated with a lower risk of all outcomes: cancer mortality (HR1 hour 0.82; 95% CI [0.78, 0.86]; p < 0.001); overall incident cancer (HR1 hour 0.94; 95% CI [0.92, 0.96]; p < 0.001); obesity-related cancer (HR1 hour 0.91; 95% CI [0.88, 0.94]; p < 0.001); and T2D-related cancer (HR1 hour 0.90; 95% CI [0.88, 0.93]; p < 0.001). Similar associations were observed for several site-specific cancers: breast, lung, and oral cancers, non-Hodgkin lymphoma and leukemia) (Table E in S1 Appendix). For example, overall and prolonged SB were associated with a higher risk of breast cancer, respectively (overall SB: HR1 hour 1.05; 95% CI [1.02, 1.09]; p = 0.002; prolonged SB: HR1 hour 1.04; 95% CI [1.02, 1.06]; p < 0.001), whereas interrupted SB was associated with a lower risk (HR1 hour 0.92; 95% CI [0.88, 0.96]; p < 0.001). The associations between quartiles of total sedentary time (hours/day) and cancer outcomes are detailed in Table F in S1 Appendix.
The two sensitivity analyses did not alter the findings materially. Additional adjustment for BMI and LTCs produced similar patterns for the associations between total, prolonged, and interrupted SB and composite cancer risk (Table G in S1 Appendix). After excluding the first two years of follow-up, overall and prolonged SB were still associated with increased cancer risk, while interrupted SB was still associated with reduced risk of all examined cancer outcomes (Table K in S1 Appendix).
The nonlinear associations between SB and cancer mortality and T2D-related cancer were shown in Fig 2. The relationships between prolonged SB and cancer outcomes were approximately linear across the range of SB in the study cohort, whereas overall SB showed some evidence of nonlinearity. Interrupted SB was inversely associated with cancer outcomes, with the risk reduction being more pronounced at lower levels and then tending to level off at higher levels of interrupted SB. The nonlinear associations for PA are shown in Fig 1 in S1 Appendix.
All models were adjusted for age, sex, and ethnicity, deprivation, education, smoking, intake of alcohol, sugar, processed meat, red meat, fruit and vegetables, and oily fish. Obesity-related cancer including esophagus cancer, liver cancer, kidney cancer, myeloma, pancreatic cancer, colorectal cancer, gallbladder cancer, breast cancer, ovarian cancer, and thyroid cancer.T2D-related cancers including thyroid cancer, breast cancer, liver cancer, pancreatic cancer, endometrial cancer, esophagus cancer, colorectal cancer, kidney cancer, gallbladder cancer, ovarian cancer, non-Hodgkin lymphoma, leukemia, and bladder cancer. Panels (A), (B), (C), and (D) represent the association between overall cancer mortality, the incidence of overall, obesity-related, and T2D-related cancers, respectively.
The associations between different intensities of PA and composite cancer outcomes and site-specific cancers are shown in Tables H and I in S1 Appendix. VPA, even of relatively short duration, was strongly associated with lower risk of cancer mortality, T2D-related cancers, and cancers of esophagus, uterine, liver, colorectal, bladder, and lung, as well as leukemia.
Table 3 shows the interaction tests between SB and adiposity defined using BMI, WHR, and WHtR in relation to composite cancer outcomes. Overall, there was no significant interaction between total or prolonged sedentary time and general or central obesity for risks of overall cancer, obesity-related cancer, or T2D-related cancer.
Table 4 presents hazard ratios for composite cancer outcomes associated with specified reallocation of time from one type of sedentary behavior to another behavior. Replacing 1 hour per day of prolonged SB with other behaviors was generally associated with lower risks of the composite cancer outcomes. For example, replacing 1 hour of prolonged SB with LPA was associated with 12% lower hazard (HR 0.88; 95% CI [0.79, 0.99]; p = 0.033). Replacing 5 min/day of prolonged SB with 5 min/day VPA was associated with the largest reduction in cancer incidence outcomes (HRT2D-related cancer 0.89; 95% CI [0.86, 0.92; p < 0.001]). Hazard ratios for site-specific cancer outcomes under the same reallocation scenarios are presented in Table J in S1 Appendix.
Discussion
Overall time spent in SB is associated with adverse health outcomes. However, in this large population cohort, we demonstrated substantial differences between prolonged and interrupted SB. More time spent in prolonged SB was associated with higher risk of a wide range of cancer outcomes, whilst more time spent in interrupted SB was associated with a lower risk. Replacing prolonged SB with additional PA was associated with additional reductions in risk, even if it was replaced by light intensity PA. To our knowledge, this is the first study that showed the differential associations of interrupted and prolonged SB with cancer outcomes, using state-of-the-art classification methods and objectively measured time use activities.
Our findings are generally consistent with, but meaningfully extend, existing studies that utilized accelerometer-measured data. For instance, a prospective study [34] involving 8,002 US middle-aged and older participants found that both total sedentary time and sedentary bout duration were associated with significantly higher cancer mortality risks. In the unadjusted model, these sedentary behaviors were associated with 82% (95% CI [1.27, 2.60]; p < 0.001) and 61% (95% CI [1.16, 2.24]; p = 0.005) higher risk of cancer mortality, respectively. Another prospective study [35] involving 7,671 European participants, among whom 516 cancer events were recorded, found that SB accumulated in longer bout durations was associated with higher rates of incident cancer. Specifically, prolonged sedentary bouts were associated with an increased risk of cancer incidence 12% (95% CI [1.02, 1.23]; p = 0.023), which is consistent with the association between prolonged SB and overall cancer incidence observed in our study. Beyond studies examining composite cancer outcomes and those using accelerometer-measured data, one observational study [36] investigated the associations between overall SB and the risks of colon, endometrial, and lung cancers. Total SB was associated with 24%, 32%, and 21% increased risk for these cancers, respectively, which is also broadly consistent with some of our site-specific findings.
The ISM results from this study are also consistent with the accelerometer-based findings reported by Gilchrist and colleagues [34], who found that replacing 30 min of total sedentary time with either LPA or moderate-to-vigorous PA (MVPA) was associated with a lower risk of cancer mortality. Specifically, a 30-min replacing was associated with an 8% reduction for LPA (HR 0.92, 95% CI [0.86, 0.97]) and a 31% reduction for MVPA (HR 0.69, 95% CI [0.48, 0.97]). Additionally, that study reported linear dose-response associations of both total sedentary time and mean sedentary bout length with cancer mortality. Total sedentary time was significantly associated with cancer mortality risk in a linear, dose-response (HR per 1 hour/d increase in total sedentary time 1.16, 95% CI [1.03, 1.31]). However, this study focused solely on the association between SB and cancer mortality, without addressing cancer incidence. They also did not explicitly examine differences between interrupted and prolonged SB.
Our findings suggest that the health effects of SB may depend not only on total sedentary time, but also on whether that time is accumulated in prolonged bouts or interrupted by activity. This pattern is biologically plausible: experimental studies have shown that interrupting prolonged sitting with short bouts of activity can improve metabolic responses compared with uninterrupted sitting [37,38]. Prolonged sedentary behaviors were found to promote chronic low-grade inflammation [39] as well as suppressing efficient immune function [40]. In particular, higher device-measured total SB has been associated with endothelial dysfunction and low-grade inflammation independent of PA levels [41]. However, that study found no significant associations for sedentary breaks or average sedentary bout length, possibly because sedentary interruption was operationalized differently and the sample size was smaller (n = 2,363).
Additional evidence suggests that replacing SB with PA may contribute to cancer prevention through several biological pathways that are not limited to body weight [25]. One possible pathway is ectopic fat accumulation [42] whereby triglyceride are deposited in nonadipose tissues such as the liver, skeletal muscles, heart, and pancreas. This pattern of fat storage was found to be particularly prevalent in people who were sedentary [43] and is metabolically harmful, explaining the links between sedentary behavior and cancer risk [25,44]. Prolonged sedentary time may also contribute to impaired glucose regulation and hyperinsulinemia. Insulin and insulin-like growth factor I (IGF-I) regulate glucose metabolism, cell growth, apoptosis, and angiogenesis, and chronic upregulation of these pathways has been linked to several cancers, including breast, prostate, and colorectal cancer [36]. The precise molecular mechanisms underlying these associations could not be examined in this study and warrant future investigation.
Our study has several strengths. First and foremost, it used device-measured PA, which should accurately reflect the actual SB and PA level of participants and is robust against recall or reporting bias. Second, we used a validated, machine learning-informed approach to derive SB, which is likely to be more accurate than questionnaire-based assessment and more informative than traditional cut-point methods [45]. Finally, to our knowledge, our study is the first comprehensive study that explored the associations of type-specific SB with both composite and individual cancer outcomes. However, our study is subject to the following limitations. Firstly, device-measured PA only captured behaviors in a limited time window, which may not necessarily represent the long-term habitual behavior. Importantly, even though we could characterize SB by their bout length, there may be other contextual factors related to PA that could modulate the association. For example, a previous UK Biobank study has found self-reported television viewing, but not driving or using a computer, to be associated with higher cancer risk [46]. Currently, there are no reliable methods to map objectively measured SB to its context and would warrant future studies. Secondly, UK Biobank is not representative of the UK population, with evidence of healthy volunteer bias. Estimates of effect size were found to be consistent with population-representative cohorts, but PA levels are higher among UK Biobank participants. Thirdly, despite best efforts, there are limitations on how cancer outcomes were defined. For example, the number of events was not sufficient to separate leukemia by cell lineage, and there was no information on the proportion of esophageal adenocarcinoma in esophageal cancers. Last but not least, as with other observational studies, residual and occupation confounding cannot be ruled out and could be, partly, driving the observed associations. Specifically, general frailty or LTCs that could affect mobility could be major confounders, which have been mitigated in this study via a long-term condition variable. However, it should also be worth noting SB could also affect the development of LTCs and the adjustment of that variable could induce overadjustment bias.
The findings of this study showed that replacing prolonged SB as well as replacing SB with PA may play a role in cancer prevention. Individuals who accumulate large amounts of prolonged sedentary behavior may therefore be an important target group for future interventions. There are multiple ways to intervene, including breaking down of prolonged sedentary behavior, and replacing that by PA of different intensity. If a person could replace SB with PA, the potential cancer risk reduction is dependent on the PA intensity. Not surprisingly, replacing even 5 min a day of prolonged sedentary time with VPA is associated with the lowest cancer incidence risk [20], which has also been reported previously in analyses of the same UK Biobank accelerometer dataset using a different exposure definition [20]. However, it should be noted that VPA may not be feasible for most people, especially for those who currently spend substantial time in SB—they are most likely to be older, frail, and have multimorbidity [47]. For these individuals, replacing SB with LPA might be a more achievable target, and the study showed that LPA is already associated with tangibly lower risk. It should be noted that LPA is currently not included in any of the PA guidelines, partly because of the lack of measurement—it is hard to measure LPA using questionnaires—and this study showed that LPA should not be overlooked. At the same time, several hazard ratios were modest in magnitude and should not be overinterpreted at the individual clinical level. However, some replacement could still be potentially meaningful, e.g., replacing prolonged sedentary time with LPA was associated with approximately a sizable reduction of cancer mortality and obesity- and T2D-related cancer incidence. Given the high prevalence of sedentary behavior, even modest associations could still have important public health implications, although formal cost-effectiveness analyses would be needed before recommending resource-intensive interventions.
In summary, prolonged, but not interrupted, SB was associated with higher risks of a wide range of cancer outcomes. Replacing prolonged SB with PA was associated with lower risks of cancer outcomes. These findings should be corroborated in intervention studies.
Supporting information
S1 Appendix. Supplementary Tables A–K and Fig A.
Table A. Long-term morbidity groupings. This table lists the long-term morbidity categories used in the analysis and their corresponding definitions or codes. ICD-10, International Classification of Diseases, 10th Revision. Table B. Site specific cancers incidence of participants by total sedentary behavior quartiles. Values are presented as numbers (%) unless otherwise stated. Q1, first quartile; Q2, second quartile; Q3, third quartile; Q4, fourth quartile. Table C. Baseline characteristics of participants by prolonged sedentary behavior quartiles. Values are presented as number (%) unless otherwise stated.: Q1, first quartile; Q2, second quartile; Q3, third quartile; Q4, fourth quartile; LPA, light physical activity; MPA, moderate physical activity; VPA, vigorous physical activity; BMI, body mass index; WHtR, waist-to-height ratio; WHR, waist-to-hip ratio; WC, waist circumference. Table D. Baseline characteristics of participants by interrupted sedentary behavior quartiles. Values are presented as number (%) unless otherwise stated. Q1, first quartile; Q2, second quartile; Q3, third quartile; Q4, fourth quartile; LPA, light physical activity; MPA, moderate physical activity; VPA, vigorous physical activity; BMI, body mass index; WHtR, waist-to-height ratio; WHR, waist-to-hip ratio; WC, waist circumference. Table E. Association between sedentary behavior and site-specific cancers. All models were adjusted for age, sex, and ethnicity, deprivation, education, smoking, intake of alcohol, sugar, processed meat, red meat, fruit and vegetables, and oily fish. Table F. Association between total sedentary behavior and composite cancer outcomes. Q1, first quartile; Q2, second quartile; Q3, third quartile; Q4, fourth quartile. All models were adjusted for age, sex, and ethnicity, deprivation, education, smoking, intake of alcohol, sugar, processed meat, red meat, fruit and vegetables, and oily fish. Obesity-related cancer including esophagus cancer, liver cancer, kidney cancer, myeloma, pancreatic cancer, colorectal cancer, gallbladder cancer, breast cancer, ovarian cancer, and thyroid cancer.T2D-related cancers including thyroid cancer, breast cancer, liver cancer, pancreatic cancer, endometrial cancer, esophagus cancer, colorectal cancer, kidney cancer, gallbladder cancer, ovarian cancer, non-Hodgkin lymphoma, leukemia, and bladder cancer. Table G. Sensitivity analysis of the association between sedentary behavior and composite cancer risk adjusted for BMI and morbidity count. All models were adjusted for age, sex, and ethnicity, deprivation, education, smoking, intake of alcohol, sugar, processed meat, red meat, fruit and vegetables, and oily fish. Obesity-related cancer including esophagus cancer, liver cancer, kidney cancer, myeloma, pancreatic cancer, colorectal cancer, gallbladder cancer, breast cancer, ovarian cancer, and thyroid cancer.T2D-related cancers including thyroid cancer, breast cancer, liver cancer, pancreatic cancer, endometrial cancer, esophagus cancer, colorectal cancer, kidney cancer, gallbladder cancer, ovarian cancer, non-Hodgkin lymphoma, leukemia, and bladder cancer. Table H. Association between intensity of physical activity and composite cancer outcomes. All models were adjusted for age, sex, and ethnicity, deprivation, education, smoking, intake of alcohol, sugar, processed meat, red meat, fruit and vegetables, and oily fish. Obesity-related cancer including esophagus cancer, liver cancer, kidney cancer, myeloma, pancreatic cancer, colorectal cancer, gallbladder cancer, breast cancer, ovarian cancer, and thyroid cancer.T2D-related cancers including thyroid cancer, breast cancer, liver cancer, pancreatic cancer, endometrial cancer, esophagus cancer, colorectal cancer, kidney cancer, gallbladder cancer, ovarian cancer, non-Hodgkin lymphoma, leukemia, and bladder cancer. Table I. Association between intensity of physical activity and site-specific cancers. All models were adjusted for age, sex, and ethnicity, deprivation, education, smoking, intake of alcohol, sugar, processed meat, red meat, fruit and vegetables, and oily fish. LPA, light physical activity; MPA, moderate physical activity; VPA, vigorous physical activity. Table J. Hazard ratios for site specific cancer outcomes associated with replacing type-specific SB with other SB and physical activity in isotemporal substitution models. All models were adjusted for age, sex, and ethnicity, deprivation, education, smoking, intake of alcohol, sugar, processed meat, red meat, fruit and vegetables, and oily fish. LPA, light physical activity; MPA, moderate physical activity; VPA, vigorous physical activity; ISB, interrupted sedentary behavior. Table K. Association between sedentary behavior and incident of composite cancer, excluding first 2 years of follow-up. All models were adjusted for age, sex, and ethnicity, deprivation, education, smoking, intake of alcohol, sugar, processed meat, red meat, fruit and vegetables, and oily fish. Obesity-related cancer including esophagus cancer, liver cancer, kidney cancer, myeloma, pancreatic cancer, colorectal cancer, gallbladder cancer, breast cancer, ovarian cancer, and thyroid cancer.T2D-related cancers including thyroid cancer, breast cancer, liver cancer, pancreatic cancer, endometrial cancer, esophagus cancer, colorectal cancer, kidney cancer, gallbladder cancer, ovarian cancer, non-Hodgkin lymphoma, leukemia, and bladder cancer. Fig A. Nonlinear association between intensity of physical activity and composite cancer outcomes. All models were adjusted for age, sex, and ethnicity, deprivation, education, smoking, intake of alcohol, sugar, processed meat, red meat, fruit and vegetables, and oily fish. Obesity-related cancer including esophagus cancer, liver cancer, kidney cancer, myeloma, pancreatic cancer, colorectal cancer, gallbladder cancer, breast cancer, ovarian cancer, and thyroid cancer. T2D-related cancers including thyroid cancer, breast cancer, liver cancer, pancreatic cancer, endometrial cancer, esophagus cancer, colorectal cancer, kidney cancer, gallbladder cancer, ovarian cancer, non-Hodgkin lymphoma, leukemia, and bladder cancer. Panels (A), (B), (C), and (D) represent the association between overall cancer mortality, the incidence of overall, obesity-related, and T2D-related cancers, respectively.
https://doi.org/10.1371/journal.pmed.1004767.s001
(DOCX)
S1 Checklist. STROBE Statement.
The checklist of items that should be included in reports of cohort studies. The STROBE checklist is best used in conjunction with this article (freely available on the websites of PLoS Medicine at http://www.plosmedicine.org/, Annals of Internal Medicine at http://www.annals.org/, and Epidemiology at http://www.epidem.com/). Information on the STROBE Initiative is available at http://www.strobe-statement.org.
https://doi.org/10.1371/journal.pmed.1004767.s002
(DOCX)
References
- 1. Katzmarzyk PT, Powell KE, Jakicic JM, Troiano RP, Piercy K, Tennant B, et al. Sedentary behavior and health: update from the 2018 physical activity guidelines advisory committee. Med Sci Sports Exerc. 2019;51(6):1227–41. pmid:31095080; PMCID: PMC6527341.
- 2. Biswas A, Oh PI, Faulkner GE, Bajaj RR, Silver MA, Mitchell MS, et al. Sedentary time and its association with risk for disease incidence, mortality, and hospitalization in adults: a systematic review and meta-analysis. Ann Intern Med. 2015;162(2):123–32. pmid:25599350
- 3. Ekelund U, Tarp J, Steene-Johannessen J, Hansen BH, Jefferis B, Fagerland MW, et al. Dose-response associations between accelerometry measured physical activity and sedentary time and all cause mortality: systematic review and harmonised meta-analysis. BMJ. 2019;366:l4570. pmid:31434697
- 4. Cao C, Friedenreich CM, Yang L. Association of daily sitting time and leisure-time physical activity with survival among US cancer survivors. JAMA Oncol. 2022;8(3):395–403. pmid:34989765
- 5. Duran AT, Friel CP, Serafini MA, Ensari I, Cheung YK, Diaz KM. Breaking up prolonged sitting to improve cardiometabolic risk: dose-response analysis of a randomized crossover trial. Med Sci Sports Exerc. 2023;55(5):847–55. pmid:36728338
- 6. Dunstan DW, Kingwell BA, Larsen R, Healy GN, Cerin E, Hamilton MT, et al. Breaking up prolonged sitting reduces postprandial glucose and insulin responses. Diabetes Care. 2012;35(5):976–83. pmid:22374636
- 7. Peddie MC, Bone JL, Rehrer NJ, Skeaff CM, Gray AR, Perry TL. Breaking prolonged sitting reduces postprandial glycemia in healthy, normal-weight adults: a randomized crossover trial. Am J Clin Nutr. 2013;98(2):358–66. pmid:23803893
- 8. Saunders TJ, Atkinson HF, Burr J, MacEwen B, Skeaff CM, Peddie MC. The acute metabolic and vascular impact of interrupting prolonged sitting: a systematic review and meta-analysis. Sports Med. 2018;48(10):2347–66. pmid:30078066
- 9. Sagelv EH, Hopstock LA, Morseth B, Hansen BH, Steene-Johannessen J, Johansson J, et al. Device-measured physical activity, sedentary time, and risk of all-cause mortality: an individual participant data analysis of four prospective cohort studies. Br J Sports Med. 2023;57(22):1457–63. pmid:37875329
- 10. Drenowatz C, Prasad VK, Hand GA, Shook RP, Blair SN. Effects of moderate and vigorous physical activity on fitness and body composition. J Behav Med. 2016;39(4):624–32. pmid:27055817
- 11. Davis ME, Blake C, Perrotta C, Cunningham C, O’Donoghue G. Impact of training modes on fitness and body composition in women with obesity: a systematic review and meta-analysis. Obesity (Silver Spring). 2022;30(2):300–19. pmid:35088563
- 12. Zhang Y, Liu X. Effects of physical activity and sedentary behaviors on cardiovascular disease and the risk of all-cause mortality in overweight or obese middle-aged and older adults. Front Public Health. 2024;12:1302783. pmid:38410660
- 13. Blyth F, Haycraft E, Peral-Suarez A, Pearson N. Tracking and changes in the clustering of physical activity, sedentary behavior, diet, and sleep across childhood and adolescence: a systematic review. Obes Rev. 2025;26(7):e13909. pmid:39967026
- 14. Sudlow C, Gallacher J, Allen N, Beral V, Burton P, Danesh J, et al. UK biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age. PLoS Med. 2015;12(3):e1001779. pmid:25826379
- 15. Fry A, Littlejohns TJ, Sudlow C, Doherty N, Adamska L, Sprosen T, et al. Comparison of sociodemographic and health-related characteristics of UK biobank participants with those of the general population. Am J Epidemiol. 2017;186(9):1026–34. pmid:28641372
- 16. Doherty A, Jackson D, Hammerla N, Plötz T, Olivier P, Granat MH, et al. Large scale population assessment of physical activity using wrist worn accelerometers: the UK biobank study. PLoS One. 2017;12(2):e0169649. pmid:28146576
- 17. Ahmadi MN, Clare PJ, Katzmarzyk PT, del Pozo Cruz B, Lee IM, Stamatakis E. Vigorous physical activity, incident heart disease, and cancer: how little is enough? Eur Heart J. 2022;43(46):4801–14.
- 18. Pavey TG, Gilson ND, Gomersall SR, Clark B, Trost SG. Field evaluation of a random forest activity classifier for wrist-worn accelerometer data. J Sci Med Sport. 2017;20(1):75–80. pmid:27372275
- 19. Pavey TG, Gilson ND, Gomersall SR, Clark B, Trost SG. Field evaluation of a random forest activity classifier for wrist-worn accelerometer data. J Sci Med Sport. 2017;20(1):75–80. pmid:27372275
- 20. Stamatakis E, Ahmadi MN, Friedenreich CM, Blodgett JM, Koster A, Holtermann A, et al. Vigorous intermittent lifestyle physical activity and cancer incidence among nonexercising adults: the UK biobank accelerometry study. JAMA Oncol. 2023;9(9):1255–9. pmid:37498576
- 21. Stamatakis E, Ahmadi MN, Gill JMR, Thøgersen-Ntoumani C, Gibala MJ, Doherty A, et al. Association of wearable device-measured vigorous intermittent lifestyle physical activity with mortality. Nat Med. 2022;28(12):2521–9. pmid:36482104
- 22. Ahmadi MN, Nathan N, Sutherland R, Wolfenden L, Trost SG. Non-wear or sleep? Evaluation of five non-wear detection algorithms for raw accelerometer data. J Sports Sci. 2020;38(4):399–404. pmid:31826746
- 23. Diaz KM, Howard VJ, Hutto B, Colabianchi N, Vena JE, Safford MM, et al. Patterns of sedentary behavior and mortality in U.S. middle-aged and older adults: a national cohort study. Ann Intern Med. 2017;167(7):465–75. pmid:28892811
- 24. Tremblay MS, Aubert S, Barnes JD, Saunders TJ, Carson V, Latimer-Cheung AE, et al. Sedentary Behavior Research Network (SBRN) - Terminology Consensus Project process and outcome. Int J Behav Nutr Phys Act. 2017;14(1):75. pmid:28599680
- 25. Friedenreich CM, Ryder-Burbidge C, McNeil J. Physical activity, obesity and sedentary behavior in cancer etiology: epidemiologic evidence and biologic mechanisms. Mol Oncol. 2021;15(3):790–800. pmid:32741068
- 26. Hu Y, Zhang X, Ma Y, Yuan C, Wang M, Wu K, et al. Incident type 2 diabetes duration and cancer risk: a prospective study in two US cohorts. J Natl Cancer Inst. 2021;113(4):381–9. pmid:33225344
- 27. Conroy MC, Lacey B, Bešević J, Omiyale W, Feng Q, Effingham M, et al. UK Biobank: a globally important resource for cancer research. Br J Cancer. 2023;128(4):519–27. pmid:36402876
- 28. Zhou Z, Parra-Soto S, Boonpor J, Petermann-Rocha F, Welsh P, Mark PB, et al. Exploring the underlying mechanisms linking adiposity and cardiovascular disease: a prospective cohort study of 404,332 UK biobank participants. Curr Probl Cardiol. 2023;48(8):101715. pmid:37004891
- 29. Zhong P, Tan S, Zhu Z, Zhu Z, Liang Y, Huang W, et al. Normal-weight central obesity and risk of cardiovascular and microvascular events in adults with prediabetes or diabetes: Chinese and British cohorts. Diabetes Metab Res Rev. 2023;39(8):e3707. Epub 20230731. pmid:37525502
- 30. Ho FK, Zhou Z, Petermann-Rocha F, Para-Soto S, Boonpor J, Welsh P, et al. Association between device-measured physical activity and incident heart failure: a prospective cohort study of 94 739 UK biobank participants. Circulation. 2022;146(12):883–91. pmid:36036153
- 31. Hanlon P, Nicholl BI, Jani BD, Lee D, McQueenie R, Mair FS. Frailty and pre-frailty in middle-aged and older adults and its association with multimorbidity and mortality: a prospective analysis of 493 737 UK Biobank participants. Lancet Public Health. 2018;3(7):e323–32. pmid:29908859
- 32. Barnett K, Mercer SW, Norbury M, Watt G, Wyke S, Guthrie B. Epidemiology of multimorbidity and implications for health care, research, and medical education: a cross-sectional study. Lancet. 2012;380(9836):37–43. pmid:22579043
- 33. Mekary RA, Willett WC, Hu FB, Ding EL. Isotemporal substitution paradigm for physical activity epidemiology and weight change. Am J Epidemiol. 2009;170(4):519–27. pmid:19584129
- 34. Gilchrist SC, Howard VJ, Akinyemiju T, Judd SE, Cushman M, Hooker SP, et al. Association of sedentary behavior with cancer mortality in middle-aged and older US adults. JAMA Oncol. 2020;6(8):1210–7. pmid:32556069
- 35. Dempsey PC, Strain T, Winkler EAH, Westgate K, Rennie KL, Wareham NJ, et al. Association of accelerometer-measured sedentary accumulation patterns with incident cardiovascular disease, cancer, and all-cause mortality. J Am Heart Assoc. 2022;11(9):e023845. pmid:35470706
- 36. Jochem C, Wallmann-Sperlich B, Leitzmann MF. The influence of sedentary behavior on cancer risk: epidemiologic evidence and potential molecular mechanisms. Curr Nutr Rep. 2019;8(3):167–74. pmid:30887424
- 37. Giurgiu M, Timm I, Ebner-Priemer UW, Schmiedek F, Neubauer AB. Causal effects of sedentary breaks on affective and cognitive parameters in daily life: a within-person encouragement design. Npj Ment Health Res. 2024;3(1):64. pmid:39706901
- 38. Diaz KM, Murdock ME, Clark AW, Kumar S, Jerez V, Serafini MA, et al. Breaking up prolonged sedentary behavior to improve cardiometabolic health (BREAK2): protocol for a dose-finding adaptive randomization trial. BMC Public Health. 2025;25(1):1929. pmid:40414835
- 39. Henson J, Yates T, Edwardson CL, Khunti K, Talbot D, Gray LJ, et al. Sedentary time and markers of chronic low-grade inflammation in a high risk population. PLoS One. 2013;8(10):e78350. pmid:24205208
- 40. Damiot A, Pinto AJ, Turner JE, Gualano B. Immunological implications of physical inactivity among older adults during the COVID-19 pandemic. Gerontology. 2020;66(5):431–8. pmid:32585674
- 41. Vandercappellen EJ, Koster A, Savelberg HHCM, Eussen SJPM, Dagnelie PC, Schaper NC, et al. Sedentary behaviour and physical activity are associated with biomarkers of endothelial dysfunction and low-grade inflammation-relevance for (pre)diabetes: the Maastricht Study. Diabetologia. 2022;65(5):777–89. pmid:35119485
- 42. Malaikah S, Willis SA, Henson J, Sargeant JA, Yates T, Thackray AE, et al. Associations of objectively measured physical activity, sedentary time and cardiorespiratory fitness with adipose tissue insulin resistance and ectopic fat. Int J Obes (Lond). 2023;47(10):1000–7. pmid:37491534
- 43. Rynders CA, Blanc S, DeJong N, Bessesen DH, Bergouignan A. Sedentary behaviour is a key determinant of metabolic inflexibility. J Physiol. 2018;596(8):1319–30. pmid:28543022
- 44. Anderson E, Durstine JL. Physical activity, exercise, and chronic diseases: a brief review. Sports Med Health Sci. 2019;1(1):3–10. pmid:35782456
- 45. Letts ESK, Di Cristofaro N, Tucker P, Cairney J, Kobsar D, Timmons BW, et al. Development and accuracy of a novel machine learning model to detect toddlers’ physical activity and sedentary time using accelerometers: little movers activity analysis. medRxiv. 2025.
- 46. Hunter RF, Murray JM, Coleman HG. The association between recreational screen time and cancer risk: findings from the UK Biobank, a large prospective cohort study. Int J Behav Nutr Phys Act. 2020;17(1):97. pmid:32746843
- 47. Christofoletti M, Sandreschi PF, Quadros EN, d’Orsi E, Rech CR, Manta SW, et al. Physical activity and sedentary behavior as multimorbidity discriminators among elderly Brazilians: a cross-sectional study. Sao Paulo Med J. 2021;139(4):372–9. pmid:34190868