Dose‒response associations of self-reported and device-measured physical activity with major adverse cardiovascular events in people with prevalent diseases.

Dose‒response associations of self-reported and device-measured physical activity with major adverse cardiovascular events in people with prevalent diseases - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Official websites use .gov A .gov website belongs to an official government organization in the United States. Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites. Search Log in Dashboard Publications Account settings Log out Search… Search NCBI Primary site navigation Search Logged in as: Dashboard Publications Account settings Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide PERMALINK Copy As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice J Sport Health Sci . 2025 Dec 24;15:101116. doi: 10.1016/j.jshs.2025.101116 Search in PMC Search in PubMed View in NLM Catalog Add to search Dose‒response associations of self-reported and device-measured physical activity with major adverse cardiovascular events in people with prevalent diseases Miguel Adriano Sanchez-Lastra Miguel Adriano Sanchez-Lastra a Department of Special Didactics, Faculty of Education and Sports Sciences, University of Vigo, Pontevedra 36005, Spain b Well-Move Research Group, Galicia-Sur Health Research Institute (SERGAS-UVIGO), Vigo 36213, Spain c Oslo Research Centre for Physical Activity and Population Health (ORC-PAPH), Department of Sports Medicine, Norwegian School of Sport Sciences, Oslo 0806, Norway Find articles by Miguel Adriano Sanchez-Lastra a, b, c, ⁎ , Borja del Pozo Cruz Borja del Pozo Cruz d Department of Sport Sciences, Faculty of Medicine, Health, and Sports, Universidad Europea de Madrid, Villaviciosa de Odón, Madrid 28670, Spain Find articles by Borja del Pozo Cruz d , Jakob Tarp Jakob Tarp c Oslo Research Centre for Physical Activity and Population Health (ORC-PAPH), Department of Sports Medicine, Norwegian School of Sport Sciences, Oslo 0806, Norway Find articles by Jakob Tarp c , Knut Eirik Dalene Knut Eirik Dalene e Department of Chronic Diseases, Norwegian Institute of Public Health, Oslo 0473, Norway Find articles by Knut Eirik Dalene e , Tormod S Nilsen Tormod S Nilsen f Department of Physical Performance, Norwegian School of Sport Sciences, Oslo 0806, Norway Find articles by Tormod S Nilsen f , Ding Ding Ding Ding g Prevention Research Collaboration, Sydney School of Public Health, The University of Sydney, Camperdown, NSW 2006, Australia h Charles Perkins Centre, The University of Sydney, Camperdown, NSW 2050, Australia Find articles by Ding Ding g, h , Ulf Ekelund Ulf Ekelund c Oslo Research Centre for Physical Activity and Population Health (ORC-PAPH), Department of Sports Medicine, Norwegian School of Sport Sciences, Oslo 0806, Norway e Department of Chronic Diseases, Norwegian Institute of Public Health, Oslo 0473, Norway Find articles by Ulf Ekelund c, e Author information Article notes Copyright and License information a Department of Special Didactics, Faculty of Education and Sports Sciences, University of Vigo, Pontevedra 36005, Spain b Well-Move Research Group, Galicia-Sur Health Research Institute (SERGAS-UVIGO), Vigo 36213, Spain c Oslo Research Centre for Physical Activity and Population Health (ORC-PAPH), Department of Sports Medicine, Norwegian School of Sport Sciences, Oslo 0806, Norway d Department of Sport Sciences, Faculty of Medicine, Health, and Sports, Universidad Europea de Madrid, Villaviciosa de Odón, Madrid 28670, Spain e Department of Chronic Diseases, Norwegian Institute of Public Health, Oslo 0473, Norway f Department of Physical Performance, Norwegian School of Sport Sciences, Oslo 0806, Norway g Prevention Research Collaboration, Sydney School of Public Health, The University of Sydney, Camperdown, NSW 2006, Australia h Charles Perkins Centre, The University of Sydney, Camperdown, NSW 2050, Australia ⁎ Corresponding author. [email protected] Received 2025 May 3; Revised 2025 Oct 5; Accepted 2025 Nov 6; Collection date 2026 Dec. © 2026 Published by Elsevier B.V. on behalf of Shanghai University of Sport. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). PMC Copyright notice PMCID: PMC13091526  PMID: 41453572 Highlights • In individuals with a history of cardiovascular disease (CVD), type 2 diabetes (T2DM), or cancer survivors, higher levels of physical activity were associated with lower incidence of major adverse cardiovascular events and CVD mortality, with similar dose–response patterns across conditions. • Self-reported data suggested that current physical activity guidelines may be insufficient for risk reduction, and no association with CVD mortality was observed in those with T2DM. • Device-measured physical activity was more strongly associated with lower cardiovascular risk at lower activity levels than self-reported activity, particularly among people with T2DM and cancer survivors. Keywords: Exercise, Cardiovascular disease, Cancer, Type 2 diabetes mellitus, Mortality Abstract Purpose To investigate the dose–response associations of self-reported and device-measured physical activity with major adverse cardiovascular events (MACE) and cardiovascular disease (CVD) mortality in individuals with a history of CVD, type 2 diabetes mellitus (T2DM), or cancer. Methods We analyzed data from 34,183 UK Biobank participants (CVD: n = 8703; T2DM: n = 11,623; cancer: n = 13,857; age = 60.0 ± 7.1 years, mean ± SD; 45.3% women). Physical activity was assessed using self-reports and wrist-worn accelerometers in a subsample ( n = 6077). Dose‒response associations with MACE and CVD mortality were evaluated using Fine-Gray competing-risk models with penalized splines, adjusting for confounders. Results During a median follow-up of 13.1 years, 8729 MACE and 982 CVD deaths were recorded. Self-reported physical activity showed non-linear inverse associations with MACE, with the optimal dose (nadir) observed at ≈4000 MET-min/week in CVD (hazard ratio (HR) = 0.95) and cancer (HR = 0.92) and ≈6000 MET-min/week in T2DM (HR = 0.91), compared to participants in the 5th centiles of activity. For CVD mortality, the optimal dose was observed at ≈5000 MET-min/week in CVD (HR = 0.86) and cancer (HR = 0.62), respectively. Device-measured moderate-to-vigorous intensity physical activity showed stronger dose‒response associations with MACE for T2DM (optimal dose observed at ≈200 min/week, HR = 0.88) and cancer (optimal dose at ≈700 min/week, HR = 0.27, wide confidence intervals). Conclusion Physical activity could reduce MACE risk in people with a history of CVD, T2DM, or cancer. While a similar shape of dose-response patterns was observed across these groups, self-reported data suggested the minimal effective dose might exceed current World Health Organization recommendations. However, device-measured data supported existing guidelines for people with T2DM and cancer survivors. Graphical abstract Open in a new tab 1. Introduction Cardiovascular diseases (CVDs) remain the leading cause of mortality globally. However, CVD risk can be reduced by mitigating modifiable behavioral and metabolic risk factors, including tobacco and alcohol consumption, physical inactivity, unhealthy diets, obesity, hypertension, hypercholesterolemia, and hyperglycemia. 1 While the protective effects of physical activity against major adverse cardiovascular events (MACE) are well-established in the general population, 2 evidence on the dose‒response relationship in those living with pre-existing chronic conditions is limited. Due to this lack of evidence, current physical activity guidelines, including those from the World Health Organization (WHO), recommend similar activity levels for individuals with a history of CVD, type 2 diabetes mellitus (T2DM), or cancer, despite disease-specific pathophysiological factors (e.g., chronic inflammation, metabolic dysregulation, and treatment effects) 3 , 4 , 5 that may affect the dose‒response association between physical activity and MACE. 6 , 7 As a result, individuals with cardiometabolic diseases or cancer survivors could require a higher volume of physical activity to achieve comparable cardiovascular benefits to those without such health conditions. With over 1 billion individuals globally affected by these conditions, 8 , 9 , 10 specifying optimal physical activity thresholds for disease management is critical to refine public health recommendations and reduce disease burden. Moderate-to-vigorous physical activity (MVPA) is associated with lower risk of MACE in the general population and people with prevalent CVD, though dose‒response patterns seem to differ by health status. While non-linear associations have been shown in those without prevalent CVD, 11 studies in people with prevalent CVD have reported linear, 11 L-shaped, 12 or J-/U-shaped relationships. 12 , 13 Studies in people with diabetes suggest physical activity can attenuate but not completely offset the diabetes-associated excess risk of coronary heart disease, 14 potentially requiring physical activity above current WHO-recommended levels to achieve substantial risk reductions. 15 While physical activity is associated with lower all-cause and CVD mortality in cancer survivors, 16 , 17 evidence on the dose‒response associations between post-diagnosis physical activity and MACE is still inconclusive. 18 Clarifying these dose‒response relationships is crucial for defining minimal and optimal physical activity thresholds to reduce MACE risk in populations living with chronic diseases. Robust evidence on these associations will inform tailored guidelines, potentially enhancing public health strategies, reducing CVD burden, and improving outcomes for individuals with chronic diseases. This study therefore aimed to clarify inconsistencies in previous findings by examining dose–response associations with incidence of MACE and CVD mortality using both self-reported and device-measured physical activity in people with prevalent CVD, T2DM, and cancer survivors. 2. Methods This prospective cohort study is reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines ( Supplementary Table 1 ). 2.1. Data source and population Data were from the UK Biobank (Application Number 29717)—a cohort of more than 500,000 individuals (5.5% response rate), aged 37–73 years, recruited from various regions across the UK, all of whom provided written informed consent. Baseline data collection (between 2006 and 2010 across 22 assessment centers) included questionnaires, physical assessments, and biological sampling, which were subsequently linked to electronic registries. Between 2013 and 2015, a subset of participants ( n = 103,660, when participants were aged 43–79) underwent accelerometer assessments of physical activity. Ethical approval was granted by the North-West Research Ethics Committee. Further cohort details are published elsewehere. 19 2.2. Study groups and analytical sample We defined 3 different groups based on their history of chronic diseases at baseline: CVD, T2DM, and cancer survivors (see Supplementary Table 2 for additional details on definitions, data sources, and international disease codes). These groups were pre-specified based on a pragmatic balance of public-health relevance, data quality, and statistical power because these conditions are highly prevalent 8 , 9 , 10 and strongly linked to higher risk of MACE, 3 , 4 , 5 , 20 are robustly ascertained in the UK Biobank (detailed self-reported information on medical conditions, hospital inpatient, and cancer registry data), and sufficient sample size and event counts are available in the dataset to support precise, disease-specific dose–response estimation. Prevalent CVD was defined as a history of myocardial infarction, stroke, angina, or heart failure, algorithmically derived from baseline self-reported and clinical data combined with hospital admissions records. Prevalent T2DM was identified using the Eastwood algorithm (based on diagnosis history, demographics, and medication data from questionnaire/interview) 21 or through a measured hemoglobin A1c level ≥ 48 mmol/mol. Cancer survivors had ≥5 years of survival post-malignant neoplasm diagnosis (excluding non-melanoma skin cancer). 22 Participants were excluded if they were classified into more than one of these 3 groups to avoid conflating heterogeneous pathophysiological and therapeutic contexts which would complicate the interpretation of dose–response associations. Those with history of in-situ neoplasms or neoplasms of uncertain or unknown behavior were also excluded. Participants newly diagnosed with cancer, CVD, or T2DM between baseline and the accelerometer measurements were included in this analysis if they had been diagnosed with only one of these conditions. To mitigate potential reverse causality, in addition to left-censoring the first 2 years of follow-up, we excluded participants with less than 2 years of follow-up ( n = 1962) and those with other major chronic diseases at baseline (neurological, immunological, or systemic diseases), those who were underweight (body mass index (BMI) < 18.5 kg/m 2 , as underweight status may reflect underlying illness), those unable to walk, and those with mobility limitations (details in Supplementary Table 3 ). Applying these exclusion criteria and excluding those with more than one of the conditions under study at baseline ( n = 3863) resulted in the removal of 55,349 participants from the study. Participants classified in the disease groups without complete information on exposures, outcomes, or covariates were also excluded ( n = 1839, Supplementary Fig. 1 ). 2.3. Exposures Self-reported physical activity was assessed using an adapted version of the International Physical Activity Short Form Questionnaire (IPAQ-SF), a widely used questionnaire with acceptable reliability and validity across diverse populations. 23 Participants reported the frequency of engaging in walking, moderate-, and vigorous-intensity physical activities lasting ≥10 min per session in a typical week. Weekly physical activity was expressed as metabolic equivalents (METs) in minutes per week (MET-min/week), comprising walking (3.3 METs), moderate activities (4.0 METs), and vigorous activities (8.0 METs). Following IPAQ-SF protocol, 24 participants with incomplete or missing responses regarding activity frequency or duration were excluded ( n = 117,183), as well as those reporting over 960 min of daily physical activity (combined walking and MVPA). Time spent in each category was truncated at 180 min per day. Participants were additionally excluded if their reported combined time for walking, MVPA, sleep (missing values replaced with 8 h per day), and total screen time (sum of leisure time TV and computer use) exceeded 24 h per day. As a result, a total of 34,183 participants with self-reported physical activity data were finally included. The IPAQ-SF is known to overestimate total physical activity compared with accelerometer-based measures. 25 To address this, we included device-measured data available in a subsample of UK Biobank participants ( n = 103,660) using Axivity AX3 accelerometers (Axivity, York, UK) worn on the dominant wrist continuously for 7 days (100 Hz sampling rate; ±8 g range). Individual data were extracted from 5-s epoch time series. Exclusion criteria included implausible average vector magnitude acceleration (>100 m g ), uncalibrated data, <72 h wear time, or missing data in each 1-h period of the 24-h cycle ( n = 7008). 26 As a result, a total of 6077 participants classified in the different disease groups had valid data. Time spent in MVPA and light physical activity (LPA) was derived from a machine learning tool, 27 and total PA was the measured average vector magnitude in mg ( Supplementary Table 3 ). 2.4. Outcomes We defined incident MACE as the first occurrence—after baseline measurements for self-report and accelerometer wear, respectively—of a non-fatal major cardiovascular disease diagnosis (International Classification of Diseases-10th Edition (ICD-10) Codes I20‒I25, I50, and I60‒I64) recorded in hospital inpatient data, either as a primary or secondary diagnosis. Mortality from CVD was analyzed only using self-reported physical activity due to the low number of events in the subsample with accelerometer measurements ( n = 27 CVD deaths in the CVD group, n = 25 in the T2DM group, and n = 20 among cancer survivors). It was defined as deaths attributed to CVD as primary cause (ICD-10 Codes I00‒I99) occurring after baseline measurements of self-reported physical activity. Participants were followed until they experienced an incident CVD event, died, exited the study, or until the recommended censoring dates from the UK Biobank at the time of analysis (10/31/2022 for England, 08/31/2022 for Scotland, and 03/31/2022 for Wales), whichever came first. 2.5. Covariates Detailed information on covariates is shown in Supplementary Table 3 . We used directed acyclic graphs to identify confounders before conducting any analyses ( Supplementary Figs. 2–4 ). Self-reported sociodemographic covariates included ethnicity, living with partner, employment status, and education, and the Townsend index was used as an additional marker of socioeconomic status. 28 Self-reported lifestyle covariates included a diet quality index based on meeting healthy eating targets related to food types, 29 alcohol intake, and smoking history. History of depression and hormonal treatment medication for prostate and breast cancers were derived from a combination of questionnaire data and a verbal interview. BMI was calculated from measured weight and height (kg/m 2 ). Duration of disease was defined as years since first diagnosis of CVD, T2DM, or cancer, and ascertained from self-reports, hospital inpatient data, and cancer registry records. For self-reported physical activity analyses, duration was calculated up to the baseline assessment date, and for device-measured analyses, up to the end date of accelerometer wear. The number of pre-existing cancer diagnoses was extracted from cancer registries. Mastectomy and lumpectomy operations were extracted from verbal interview data. Chemotherapy was extracted from hospital inpatient records. 2.6. Statistical analysis For each of the study groups, we described the sample by tertiles of self-reported physical activity levels (i.e., MET-min/week). For each study group, we tested the dose‒response associations between physical activity (self-reported and device-based) and MACE incidence or CVD mortality using Cox penalized splines regression models to identify the minimal effective dose (lowest dose showing a statistically significant effect) and optimal dose (lowest statistically significant risk). Data were trimmed at the 95th percentile of the exposure distribution to avoid the overinfluence of extreme values, and we used the 5th percentile of the exposure distribution (after trimming) for each condition separately as the reference. We tested and compared models with 2–5 degrees of freedom, and the model with the lowest Akaike information criterion was used in each case. The Fine-Gray subdistribution method 30 was employed in all models to account for competing risks (i.e., deaths due to causes other than the event of interest). Results are provided as subdistribution hazard ratios (HRs) with 95% confidence intervals (95%CIs). We adjusted all our models for age, sex, ethnicity, education, Townsend deprivation index, smoking, alcohol use, dietary pattern, BMI, employment status, and depression. For participants with CVD and T2DM, models were additionally adjusted for years with the disease. For those with cancer, additional adjustments included chemotherapy, hormonal treatment (prostate and breast cancers), whether a mastectomy/lumpectomy was performed (women), years since the diagnosis of the first cancer, and the number of prevalent cancers. Three sensitivity analyses were performed: (a) excluding BMI adjustment due to its potential mediating role in the associations under study; (b) censoring participants with events observed within the first 5 years of follow-up to minimize potential residual bias from reverse causation; and (c) repeating the analysis in cancer survivors restricting the group to those who survived >5 years from their last (instead of first) cancer diagnosis. Analyses were performed using R statistical software (Version 4.3.1; R Foundation for Statistical Computing, Vienna, Austria). 3. Results 3.1. Self-reported physical activity, MACE, and CVD mortality The study included 34,183 participants (8703 with CVD; 11,623 with T2DM; and 13,857 cancer survivors), with 8729 MACE events and 982 CVD deaths recorded over a median follow-up ranging between 10.0 and 13.4 years across the study groups. Table 1 shows the main characteristics of participants in each of the study groups. Supplementary Tables 4–6 show descriptive data on disease status stratified by self-reported physical activity tertiles for each study group, suggesting that baseline characteristics varied considerably across physical activity tertiles. Supplementary Tables 7 and 8 show descriptive statistics of the types of prevalent diseases present within participants with prevalent CVD and cancer survivors. Table 1. Baseline characteristics and follow-up data of the participants in each study group. Prevalent CVD Prevalent T2DM Cancer survivors Sample size ( n ) 8703 11,623 13,857 Self-reported PA (MET-min/week) 1902 (902–3810) 1508 (652–3159) 1804 (866–3564) Self-reported physical activity analysis Follow-up years to MACE (median) 10.0 13.2 13.4 MACE ( n ) 4532 2526 1671 Follow-up years to CVD mortality (median) 13.2 13.5 13.6 CVD mortality events ( n ) 452 354 176 Device-based MVPA (min/week) 214.7 (104.5–378.1) 172.5 (76.1–327.0) 221.1 (104.5–374.4) Device-based PA analysis ( n ) 1353 1818 2906 Follow-up years (median) 7.5 7.7 7.8 MACE ( n ) 403 237 206 Age (year) 61.5 ± 6.4 59.2 ± 7.27 59.7 ± 7.2 Females 2457 (28.2) 3999 (34.4) 9045 (65.3) Ethnicity White 8417 (96.7) 10,333 (88.9) 13,554 (97.8) Asian 152 (1.8) 662 (5.7) 91 (0.7) Black 68 (0.7) 388 (3.3) 96 (0.7) Others/mixed 66 (0.8) 240 (2.1) 117 (0.8) Education No qualifications 1893 (21.8) 2072 (17.8) 1977 (14.3) Below college/university degree 4427 (50.9) 5954 (51.2) 6839 (49.3) College/university degree 2383 (27.4) 3597 (30.9) 5041 (36.4) Townsend deprivation index a 2.0 ± 1.4 1.9 ± 1.4 2.2 ± 1.4 Smoking Never 3775 (43.4) 5617 (48.3) 7623 (55.0) Previous 4067 (46.7) 4864 (41.8) 5197 (37.5) Current 861 (9.9) 1142 (9.8) 1037 (7.5) Alcohol use Never 339 (3.9) 736 (6.3) 462 (3.3) Former 331 (3.8) 552 (4.7) 395 (2.9) Current, <3 times/week 3833 (44.0) 6149 (52.9) 6603 (47.6) Current, ≥3 times/week 4200 (48.3) 4186 (36.0) 6397 (46.2) Dietary pattern b Not meeting any targets 1497 (17.2) 2397 (20.6) 2135 (15.4) Meeting 1 target 3306 (38.0) 4560 (39.2) 5296 (38.2) Meeting 2 or 3 targets 3900 (44.8) 4666 (40.1) 6426 (46.4) Body mass index (kg/m 2 ) 28.2 ± 4.3 30.9 ± 5.6 26.8 ± 4.3 Employment status Unemployed 569 (6.5) 817 (7.0) 782 (5.6) Employed 3607 (41.4) 6164 (53.0) 6588 (47.5) Retired 4527 (52.0) 4642 (39.9) 6487 (46.8) Depression (%) 408 (4.7) 590 (5.1) 666 (4.8) Years since first diagnosis c 7.3 ± 6.9 5.0 ± 6.5 8.6 ± 7.5 Hormonal treatment (prostate/breast cancer) NA NA 2103 (15.2) Mastectomy and/or lumpectomy (%) NA NA 5183 (37.4) Chemotherapy (%) NA NA 532 (3.8) Number of prevalent cancer diagnoses (median) NA NA 1.0 Open in a new tab Notes: Data are shown as mean ± SD, median (IQR), or n (%); percentages might not add up to 100% due to rounding. a For Townsend deprivation index, lower scores indicate more affluence. b Dietary pattern is based on meeting or not three healthy eating targets related to food types: (a) ≤3 weekly servings of red meat and ≤1 servings/week of processed meat; (b) ≥2 servings per week of fish including at least 1 with oily fish; and (c) ≥5 servings per day of fruits and vegetables. c Years since first diagnosis of CVD, T2DM, or cancer for the each of the corresponding study groups. Abbreviations: CVD = cardiovascular disease; IQR = interquartile range; MACE = major adverse cardiovascular events; MET = metabolic equivalent of task; MVPA = moderate-to-vigorous physical activity; NA = not available; T2DM = type 2 diabetes mellitus. Self-reported physical activity showed non-linear inverse associations with MACE incidence among participants with CVD up to an optimal dose of ≈4000 MET-min/week (lowest HR = 0.95; Fig. 1 ). Similarly, we found significant non-linear dose‒response associations between self-reported physical activity and a lower incidence of MACE for participants with T2DM (optimal dose at ≈6000 METs-min/week, lowest HR = 0.91; Fig. 1 ), which showed a slight upward trend at the highest activity levels (J-shape). For cancer survivors, the association was L-shaped with an optimal dose at ≈4000 MET-min/week (lowest HR = 0.92; Fig. 1 ). Confidence intervals generally widened at the upper end of the physical activity distribution, particularly beyond ≈6000 MET-min/week, indicating less precise estimates. Fig. 1. Open in a new tab Dose–response associations of self-reported physical activity with incidence of MACE and CVD mortality in participants with history of CVD ( n = 8703; 4532 MACE; 452 CVD deaths), T2DM ( n = 11,623; 2526 MACE; 354 CVD deaths), and cancer survivors ( n = 13,857; 1671 MACE; 176 CVD deaths). Results show subdistribution hazard ratios (solid line and associated 95% confidence interval band, dotted lines). Reference values were set at the 5th percentile of exposure distribution. For MACE risk, these were 447.8 MET-min/week (CVD), 405.6 MET-min/week (T2DM), and 429.5 MET-min/week (cancer survivors). For CVD mortality, the corresponding values were 444.2 MET-min/week (CVD), 405.6 MET-min/week (T2DM), and 426.2 MET-min/week (cancer survivors). Models were adjusted for age, sex, ethnicity, education, Townsend deprivation index, smoking, alcohol use, dietary pattern, body mass index, employment status, and depression. For participants with CVD and T2DM, models were additionally adjusted for years since diagnosis. For those with cancer, additional adjustment included chemotherapy, hormonal treatment (prostate and breast cancers), whether a mastectomy or lumpectomy was performed, years since diagnosis of first cancer, and number of prevalent cancers. The vertical dotted lines represent the level of weekly physical activity recommended by the World Health Organization. Dose‒response associations were assessed with penalized splines with 2–5 degrees of freedom and data trimmed at the 95th centile of the exposure distribution of interest. Participants with less than 2 years of follow-up were left-censored. For participants with prevalent cancer, we also excluded those who survived less than 5 years after the first cancer diagnosis. Fine and Gray models were conducted in all cases to account for competing events. CVD = cardiovascular disease; HR = subdistribution hazard ratio; IPAQ = International Physical Activity Questionnaire; MACE = major adverse cardiovascular events; MET = metabolic equivalent of task; T2DM = type 2 diabetes mellitus. We found associations of similar shape but of a larger magnitude for the association between self-reported physical activity and the risk of CVD mortality ( Fig. 1 ) among those with CVD (optimal dose at ≈5000 METs-min/week, HR = 0.86) and cancer (optimal dose at ≈5000 METs-min/week, HR = 0.62). Among those with T2DM, we observed an imprecise null effect, with very wide confidence intervals at higher activity levels. 3.2. Device-measured physical activity and MACE A total of 6077 participants with accelerometer data were included (1353 with CVD, 1818 with T2DM, and 2906 cancer survivors), with 846 MACE events recorded over a median follow-up of 7.5–7.8 years ( Table 1 ). For participants with CVD, the MVPA dose–response curve was essentially flat around unity, indicating no associations between device-measured MVPA and the incidence of MACE ( Fig. 2 ). Similar results were found for LPA and total physical activity ( Supplementary Fig. 5 ). For participants with T2DM, higher MVPA was associated with lower risks of MACE (optimal dose observed at ≈200 min/week, HR = 0.88; non-linear association) ( Fig. 2 ). Higher total physical activity (optimal dose at ≈40 m g , HR = 0.55; linear association) and LPA (optimal dose at ≈2000 min/week, HR = 0.68; non-linear association) were also associated with lower risks of MACE ( Supplementary Fig. 6 ). In cancer survivors, MVPA ( Fig. 2 ) showed a monotonic inverse association with MACE risk (HR = 0.27 at the upper tail of the distribution, ≈700 min/week). Similar results were found for total physical activity (optimal dose at ≈40 m g , HR = 0.47), but no association was detected for LPA ( Supplementary Fig. 7 ). Across these analyses, confidence intervals were wide at the highest physical activity levels, reducing precision and increasing uncertainty. Fig. 2. Open in a new tab Dose–response associations of accelerometer-derived weekly moderate-to-vigorous intensity physical activity with incidence of MACE in participants with history of CVD ( n = 1353; 403 events), T2DM ( n = 1818; 237 events), and cancer survivors ( n = 2906; 206 events). Results show subdistribution hazard ratios (solid line and associated 95% confidence interval band, dotted lines). Reference values were set at the 5th percentile of exposure, corresponding to 33.61 min/week (CVD), 25.35 min/week (T2DM), and 33.11 min/week (cancer survivors) min/week. Models were adjusted for age, sex, ethnicity, education, Townsend deprivation index, smoking, alcohol use, dietary pattern, body mass index, employment status, and depression. For participants with CVD and T2DM, models were additionally adjusted for years since diagnosis. For those with cancer, additional adjustment included chemotherapy, hormonal treatment (prostate and breast cancers), whether a mastectomy or lumpectomy was performed, years since diagnosis of first cancer, and number of prevalent cancers. The vertical dotted lines represent the level of weekly physical activity recommended by the World Health Organization. Dose‒response associations were assessed with penalized splines with 2–5 degrees of freedom and data trimmed at the 95th centile of the exposure of interest distribution. Participants with less than 2 years of follow-up were left-censored. Fine and Gray models were conducted in all cases to account for competing events. CVD = cardiovascular disease; HR = subdistribution hazard ratio; MVPA = moderate to vigorous physical activity; T2DM = type 2 diabetes mellitus. 3.3. Sensitivity analyses Removing BMI from the covariate structure did not alter the shape or magnitude of the associations of interest ( Supplementary Figs. 8 and 9 ). However, excluding the first 5 years of follow-up and removing participants who experienced events during this period weakened the associations between self-reported physical activity and incident MACE ( Supplementary Fig. 10 ). Similarly, the association between self-reported physical activity and CVD mortality was attenuated, remaining statistically significant only in participants with cancer ( Supplementary Fig. 11 ). Restricting the group of cancer survivors to those who survived more than 5 years from their last cancer diagnosis yielded similar results to the analysis in those surviving more than 5 years since their initial cancer diagnosis ( Supplementary Fig. 12 ). 4. Discussion Self-reported physical activity data indicated non-linear dose‒response associations between physical activity and MACE across CVD, T2DM, and cancer survivors, with minimal effective doses around ≈1500 MET-min/week, exceeding current international guidelines (≈600 MET-min/week). Device-based analyses showed stronger associations in T2DM and cancer survivors at 100–150 min/week of MVPA but no association in participants with CVD. Taken together, both methods suggested protective effects of physical activity in T2DM and cancer survivors, albeit at different dose ranges, while associations in people with a history of CVD were weaker or absent. 4.1. Self-reported physical activity and MACE risk Self-reported physical activity showed non-linear dose‒response associations with MACE, yielding modest risk reductions (5%–10%) at ∼4000 METs-min/week for prevalent CVD and T2DM and at ∼6000 METs-min/week for cancer survivors, plateauing at higher activity levels. This modest effect size may reflect disease-specific pathophysiology (endothelial dysfunction, chronic inflammation, metabolic disturbances) or treatment effects, which may attenuate the cardiovascular benefits of physical activity. 6 , 7 These findings highlight the importance of addressing these underlying conditions through multi-faceted prevention strategies. They also underscore that physical activity remains a valuable intervention with potential for improving health outcomes in individuals living with non-communicable chronic diseases. Individuals with previous CVD face a 20% higher 5-year recurrent event risk than those without CVD, 3 , 31 and people with diabetes are at double the risk of developing CVD regardless of other risk factors. 4 , 32 Cancer survivors experience a 35% higher risk of CVD compared to those without cancer. 5 Therefore, modest risk reductions could translate into meaningful gains in long-term health at the population level for these vulnerable groups. Our findings using self-reported data suggest that individuals with a history of CVD, T2DM, and cancer may require physical activity levels exceeding current international recommendations to achieve significant reductions in MACE and CVD mortality risk. 2 This aligns with previous research in people with T2DM 15 and CVD. 11 , 33 Specifically, self-reported data showed that the minimal effective dose required to lower MACE risk in these populations was ≈1500 MET-min/week (≈430 min/week of brisk walking). In the T2DM group, the curve suggested a J-shaped pattern, with increased risk at very high activity levels. This should be interpreted with caution, as very few participants reported such high activity, resulting in limited power and wide confidence intervals. Whether this could also reflect a true detrimental effect of overexertion in some individuals with T2DM cannot be ruled out and warrants further study. In the CVD group, the observed effect was modest (lowest HR = 0.95). The weaker associations in this group may be explained by differences in activity patterns, disease status, functional limitations, or medication effects that may attenuate the effects of physical activity. 34 It has also been hypothesized that individuals with a history of CVD may need more physical activity to elicit a beneficial hemodynamic response. 35 Together, these findings suggest that associations between physical activity and cardiovascular outcomes may be weaker or more complex in individuals with prevalent CVD. From a clinical perspective, given this small relative risk reduction, given the high baseline MACE risk in people with CVD, any absolute benefit would likely be modest. 4.2. Device-measured physical activity and MACE risk Our study also advances current knowledge by using device-measured physical activity to address the limitations of self-reported measures, which are prone to recall and social desirability biases. Previous research on physical activity and CVD risk in individuals with chronic disease history has relied primarily on either self-reported data 11 , 12 , 15 , 33 , 36 , 37 or device-measured activity, 38 , 39 , 40 , 41 with most studies focusing on mortality. In contrast, we included both self-reported and device-measured physical activity for a nuanced comparison with MACE risk and extended the analysis to also include cancer survivors. For the CVD group, our analysis of device-measured MVPA revealed no associations with MACE, consistent with the attenuated results from self-report; this may reflect the more complex health status of CVD patients (e.g., comorbidity and medication burden), specific activity patterns typical of this group (e.g., shorter/unstructured bouts, or rehabilitation-oriented vs. non-ambulatory activities), or limited statistical power. By contrast, stronger associations were found with MACE in people with T2DM and cancer survivors, with a minimal effective dose between 100 and 150 min/week of MVPA (e.g., brisk walking). These device-measured findings suggest that physical activity may be more beneficial at lower levels of activity than indicated by self-reported data in these populations, as previously suggested in apparently healthy participants for all-cause and CVD mortality. 42 In cancer survivors, the results suggested large reductions at very high MVPA volumes (≈73% lower risk at ∼700 min/week); however, these estimates should be interpreted cautiously. Values at this extreme were based on a few participants and events and had wide confidence intervals. Furthermore, those able to sustain such volumes may differ systematically from those unable (e.g., fewer treatment sequelae, greater baseline fitness, healthier co-behaviors). From a practical perspective, such MVPA volumes may not be feasible for most survivors given common post-treatment effects (e.g., fatigue, deconditioning, fluctuating functional capacity). 43 Therefore, emphasis in this population should be placed on achieving at least ∼100–150 min/week and, as tolerated, progressing toward and beyond current recommendations (∼150–300 min/week) rather than aiming for extremely high MVPA levels. It is noteworthy that, in supplementary analyses, LPA was inversely associated with MACE in people with T2DM but not in cancer survivors. This contrast may reflect factors unique to cancer survivorship, such as persistent fatigue, treatment-related sequelae (including treatment-induced cardio toxicity), or reduced physiological reserve. 16 Whether such factors blunt the cardiovascular benefits of lower-intensity activity across clinical populations warrants further investigation. 4.3. Integrating self-report and device-measured findings Altogether, our findings from both self-reported and device-measured data indicate certain similarities in the shape of the dose‒response patterns across individuals with CVD, T2DM, and cancer. Consistent physical activity recommendations across these groups could simplify public health messaging and promote broader dissemination. However, disease-specific differences—particularly the weaker or absent associations in people with CVD—suggest that further evidence is needed before firm conclusions can be made about whether guidelines should differ between chronic disease populations. While results from self-report suggest that exceeding current guidelines might be necessary for achieving significant risk reductions in these populations, the results from device-measured data support current recommendations for individuals with T2DM and cancer survivors. The observed differences between self-reported and device-measured results should be interpreted cautiously, as these methods are not directly comparable. Self-reports like the IPAQ-SF typically capture leisure-time MVPA accumulated over weeks, whereas accelerometers provide high-resolution acceleration signals over a short period. In addition, the IPAQ-SF primarily reflects leisure-time activity, while accelerometers capture total daily activity across domains (occupational, transport, domestic, and leisure-time physical activity). Consequently, accelerometer-derived MVPA cannot be directly benchmarked against WHO guideline’s thresholds, which were developed largely from self-reported data. This discrepancy highlights why accelerometer data will play an increasingly important role in informing future guidelines. 42 Additionally, differences in study entry and follow-up times between the analyses using self-reported and device-measured physical activity further complicate direct comparisons, as MACE risk increases over time, and new cases may develop while others die during follow-up. In particular, the longer follow-up in the self-reported cohort may have allowed more time for disease progression, treatment changes, or lifestyle modifications, which could attenuate observed associations compared to the shorter follow-up in the accelerometer subset. Conversely, accelerometers may better capture habitual activity during a narrower window, yielding stronger associations but over a shorter time frame. Future studies using device-measured physical activity in more heterogeneous samples are needed to confirm whether different and specific recommendations are needed for those living with chronic diseases. 4.4. Strengths and limitations This study has several strengths, including its large sample size, prospective design, focus on three major chronic disease populations, the use of both self-reported and device-based physical activity data, and the application of advanced statistical methods (e.g., directed acyclic graphs to guide confounder adjustment, and the use of competing-risk models). However, several limitations should be considered. First, reverse causation is a particular concern in groups with prevalent disease. Although we mitigated this by excluding participants with mobility limitations, other major prevalent diseases, and events within the initial 2 years of follow-up, and by adjusting for confounders, we cannot fully rule out reverse causation and residual confounding related to health status. This interpretation is further supported by the attenuation of associations observed when excluding events in the first 5 years of follow-up in sensitivity analyses, which suggests potential residual reverse causality. In addition, residual confounding from the “healthy adherer” effect, whereby more active individuals may also engage in other favorable health behaviors not fully accounted for in our models, cannot be ruled out. Second, exposure misclassification is possible. Single physical activity assessments could not capture variability over time, which may be accentuated in people with prevalent diseases with evolving health status and treatments. Also, while self-reported questionnaires are prone to recall and social desirability bias, device-based measures are also subject to measurement error. Wrist-worn accelerometers may under detect or misclassify activities with limited wrist motion (e.g., cycling, resistance training) and signal-processing choices (e.g., gravity removal, non-wear detection, epoch setting), which can materially affect activity metrics. 26 Although our derivation of MVPA from accelerometer data used a machine learning tool that reduces reliance on traditional cut-points, it still inherits conceptual limitations. Because thresholds are often derived from convenience samples with relatively homogeneous fitness levels, the same acceleration may represent physiologically light activity in fitter individuals yet demanding activity in less fit individuals. Future work should consider individualized or multi-sensor metrics (e.g., heart rate–acceleration fusion) to better capture physiological load. The proliferation of MVPA thresholds, alongside newer machine learning approaches, further complicates interpretability and comparability across studies. Third, at high physical activity volumes, data were sparse and confidence intervals were wide; therefore, estimates in this range were imprecise and carried greater uncertainty, and should be interpreted cautiously in both self-reported and device-measured analyses. In addition, the small sample size and limited number of events in our device-measured subsample remain a constraint. Fourth, generizability is limited. The UK Biobank cohort is subject to healthy volunteer selection bias, 44 potentially more pronounced in the subgroup with accelerometer data. Participation in the accelerometer study required an additional visit and a week-long commitment to device wear, which likely selected for individuals who were more mobile, motivated, and health-conscious than the total UK Biobank sample, further limiting generalizability. We excluded people with CVD, T2DM, and cancer multimorbidity to enhance disease-specific interpretability, but this reduces generalizability, as multimorbidity is common in real-world populations. Future studies should assess whether the associations observed here extend to patients with overlapping chronic conditions. Similarly, as cancer encompasses diseases with diverse etiology, prognosis, and treatment sequelae, collapsing survivors into a single group may mask potential differences in associations across cancer types. However, the limited number of events precluded cancer type-specific analyses, so future studies should investigate whether associations vary by cancer type. Additionally, the definition of cancer survivorship as ≥5 years post-diagnosis introduces heterogeneity, as health status and physical capacity could differ markedly between those closer to diagnosis and those many years beyond. This variation may also have affected the observed dose–response associations. 5. Conclusion People with history of CVD, T2DM, or cancer could reduce their risk of MACE through physical activity. While similarities were observed in the shape of the dose‒response patterns of physical activity and MACE risk across these groups, self-reported data indicated that the minimal effective dose might exceed current WHO recommendations. However, device-measured data supported the existing guidelines for individuals with T2DM and cancer survivors. Authors’ contributions MASL, BdPC, and UE participated in the design of the study, data collection, data reduction/analysis, and interpretation of results; JT, DD, TSN, and KED participated in the design of the study and contributed to the interpretation of results. All authors contributed to the manuscript writing. All authors have read and approved the final version of the manuscript, and agree with the order of presentation of the authors. Declaration of competing interest The authors declare that they have no competing interests. Acknowledgments Data availability The UK Biobank database can be accessed by researchers on application ( https://www.ukbiobank.ac.uk/register-apply/ ). Acknowledgments This research received no specific funding. MASL was funded by the Spanish Ministry of Universities under application 33.50.460A.752 and by the European Union NextGenerationEU/PRTR through a Margarita Salas contract of the University of Vigo. DD was funded by a National Health and Medical Research Council Emerging Leader Grant. This research has been conducted using the UK Biobank Resource under Application Number 29717. This work uses data provided by patients and collected by the NHS as part of their care and support. Footnotes Peer review under responsibility of Shanghai University of Sport. Supplementary materials associated with this article can be found in the online version at doi:10.1016/j.jshs.2025.101116 . Supplementary materials mmc1.pdf (3.1MB, pdf) Download video file (4.9MB, mp4) References 1. World Health Organization. Cardiovascular diseases (CVDs). 2021. Available at: https://www.who.int/news-room/fact-sheets/detail/cardiovascular-diseases-(cvds) . [accessed 27.08.2022] 2. World Health Organization . World Health Organization; Geneva: 2020. Guidelines on Physical Activity and Sedentary Behaviour. [ Google Scholar ] 3. Wood D.A. WHF roadmap on secondary prevention of CVD. Glob Heart. 2024;19:9. doi: 10.5334/gh.1294. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 4. Dal Canto E., Ceriello A., Rydén L., et al. Diabetes as a cardiovascular risk factor: An overview of global trends of macro and micro vascular complications. Eur J Prev Cardiol. 2019;26:25–32. doi: 10.1177/2047487319878371. [ DOI ] [ PubMed ] [ Google Scholar ] 5. Florido R., Daya N.R., Ndumele C.E., et al. Cardiovascular disease risk among cancer survivors: The atherosclerosis risk in communities (ARIC) study. J Am Coll Cardiol. 2022;80:22–32. doi: 10.1016/j.jacc.2022.04.042. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 6. Aleman B.M., Moser E.C., Nuver J., et al. Cardiovascular disease after cancer therapy. EJC Suppl. 2014;12:18–28. doi: 10.1016/j.ejcsup.2014.03.002. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 7. Gao J., Pan X., Li G., Chatterjee E., Xiao J. Physical exercise protects against endothelial dysfunction in cardiovascular and metabolic diseases. J Cardiovasc Transl Res. 2022;15:604–620. doi: 10.1007/s12265-021-10171-3. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 8. Lindstrom M., DeCleene N., Dorsey H., et al. Global burden of cardiovascular diseases and risks collaboration, 1990–2021. J Am Coll Cardiol. 2022;80:2372–2425. doi: 10.1016/j.jacc.2022.11.001. [ DOI ] [ PubMed ] [ Google Scholar ] 9. GBD 2021 Diabetes Collaborators Global, regional, and national burden of diabetes from 1990 to 2021, with projections of prevalence to 2050: A systematic analysis for the global burden of disease study 2021. The Lancet. 2023;402:203–234. doi: 10.1016/S0140-6736(23)01301-6. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 10. GBD 2019 Cancer Risk Factors Collaborators The global burden of cancer attributable to risk factors, 2010–19: A systematic analysis for the global burden of disease study 2019. The Lancet. 2022;400:563–591. doi: 10.1016/S0140-6736(22)01438-6. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 11. Bakker E.A., Lee D.C., Hopman M.T.E., et al. Dose–response association between moderate to vigorous physical activity and incident morbidity and mortality for individuals with a different cardiovascular health status: A cohort study among 142,493 adults from the Netherlands. PLoS Med. 2021;18 doi: 10.1371/journal.pmed.1003845. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 12. Cabanas-Sánchez V., Duarte Junior M.A., Lavie C.J., Celis-Morales C., Rodríguez-Artalejo F., Martínez-Gómez D. Physical activity and cause-specific cardiovascular mortality among people with and without cardiovascular disease: A cohort study of 0.6 million US adults. Mayo Clin Proc. 2024;99:564–577. doi: 10.1016/j.mayocp.2023.05.028. [ DOI ] [ PubMed ] [ Google Scholar ] 13. Keteyian S.J., Leifer E.S., Houston-Miller N., et al. Relation between volume of exercise and clinical outcomes in patients with heart failure. J Am Coll Cardiol. 2012;60:1899–1905. doi: 10.1016/j.jacc.2012.08.958. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 14. Fortuin-De Smidt M.C., Sewe M.O., et al. Physical activity attenuates but does not eliminate coronary heart disease risk amongst adults with risk factors: EPIC-CVD case-cohort study. Eur J Prev Cardiol. 2022;29:1618–1629. doi: 10.1093/eurjpc/zwac055. [ DOI ] [ PubMed ] [ Google Scholar ] 15. Tarp J., Luo M., Sanchez-Lastra M.A., et al. Leisure-time physical activity and all-cause mortality and cardiovascular disease in adults with type 2 diabetes: Cross-country comparison of cohort studies. J Sport Health Sci. 2024;13:212–221. doi: 10.1016/j.jshs.2023.10.004. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 16. Friedenreich C.M., Stone C.R., Cheung W.Y., Hayes S.C. Physical activity and mortality in cancer survivors: A systematic review and meta-analysis. JNCI Cancer Spectr. 2019;4:pkz080. doi: 10.1093/jncics/pkz080. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 17. Lee H.H., Lee H., Bhatt D.L., et al. Changes in physical activity and incident cardiovascular events in cancer survivors. Eur Heart J. 2023;44:4997–5000. doi: 10.1093/eurheartj/ehad677. [ DOI ] [ PubMed ] [ Google Scholar ] 18. Li Q., Pan X., Li X., Huang W. Association of physical activity intensity with all-cause mortality in cancer survivors: A national prospective cohort study. Cancers (Basel) 2022;14:5760. doi: 10.3390/cancers14235760. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 19. The UK Biobank . 2007. UK Biobank: Protocol for a Large-Scale Prospective Epidemiological Resource. https://www.ukbiobank.ac.uk/wp-content/uploads/2025/01/Main-study-protocol.pdf Available at: [accessed 09.08.2019] [ Google Scholar ] 20. Emery J., Butow P., Lai-Kwon J., Nekhlyudov L., Rynderman M., Jefford M. Management of common clinical problems experienced by survivors of cancer. The Lancet. 2022;399:1537–1550. doi: 10.1016/S0140-6736(22)00242-2. [ DOI ] [ PubMed ] [ Google Scholar ] 21. Eastwood S.V., Mathur R., Atkinson M., et al. Algorithms for the capture and adjudication of prevalent and incident diabetes in UK Biobank. PLoS One. 2016;11 doi: 10.1371/journal.pone.0162388. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 22. Marzorati C., Riva S., Pravettoni G. Who is a cancer survivor? A systematic review of published definitions. J Cancer Educ. 2017;32:228–237. doi: 10.1007/s13187-016-0997-2. [ DOI ] [ PubMed ] [ Google Scholar ] 23. Craig C.L., Marshall A.L., Sjöström M., et al. International physical activity questionnaire: 12-country reliability and validity. Med Sci Sports Exerc. 2003;35:1381–1395. doi: 10.1249/01.MSS.0000078924.61453.FB. [ DOI ] [ PubMed ] [ Google Scholar ] 24. IPAQ Research Committee. Guidelines for Data Processing and Analysis of the International Physical Activity Questionnaire (IPAQ)-short and Long Forms . Available at: https://sites.google.com/view/ipaq/score . [accessed 23.08.2025] 25. Lee P.H., Macfarlane D.J., Lam T.H., Stewart S.M. Validity of the international physical activity questionnaire short form (IPAQ-SF): A systematic review. Int J Behav Nutr Phys Act. 2011;8:115. doi: 10.1186/1479-5868-8-115. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 26. Doherty A., Jackson D., Hammerla N., et al. Large scale population assessment of physical activity using wrist worn accelerometers: The UK Biobank study. PLoS One. 2017;12 doi: 10.1371/journal.pone.0169649. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 27. Willetts M., Hollowell S., Aslett L., Holmes C., Doherty A. Statistical machine learning of sleep and physical activity phenotypes from sensor data in 96,220 UK Biobank participants. Sci Rep. 2018;8:7961. doi: 10.1038/s41598-018-26174-1. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 28. Townsend P., Phillimore P., Beattie A. Croom Helm Ltd.; London: 1988. Health and Deprivation: Inequality and the North. [ Google Scholar ] 29. Sanchez-Lastra M.A., Ding D., Del Pozo Cruz B., et al. Joint associations of device-measured physical activity and abdominal obesity with incident cardiovascular disease: A prospective cohort study. Br J Sports Med. 2024;58:196–203. doi: 10.1136/bjsports-2023-107252. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 30. Fine J.P., Gray R.J. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc. 1999;94:496–509. [ Google Scholar ] 31. Kerr A.J., Broad J., Wells S., Riddell T., Jackson R. Should the first priority in cardiovascular risk management be those with prior cardiovascular disease? Heart. 2009;95:125–129. doi: 10.1136/hrt.2007.140905. [ DOI ] [ PubMed ] [ Google Scholar ] 32. Emerging Risk Factors Collaboration. N Sarwar, Gao P., Seshasai S.R., et al. Diabetes mellitus, fasting blood glucose concentration, and risk of vascular disease: A collaborative meta-analysis of 102 prospective studies. The Lancet. 2010;375:2215–2222. doi: 10.1016/S0140-6736(10)60484-9. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 33. Jeong S.W., Kim S.H., Kang S.H., et al. Mortality reduction with physical activity in patients with and without cardiovascular disease. Eur Heart J. 2019;40:3547–3555. doi: 10.1093/eurheartj/ehz564. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 34. Mikus C.R., Boyle L.J., Borengasser S.J., et al. Simvastatin impairs exercise training adaptations. J Am Coll Cardiol. 2013;62:709–714. doi: 10.1016/j.jacc.2013.02.074. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 35. Green D.J., Hopman M.T.E., Padilla J., Laughlin M.H., Thijssen D.H. Vascular adaptation to exercise in humans: Role of hemodynamic stimuli. Physiol Rev. 2017;97:495–528. doi: 10.1152/physrev.00014.2016. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 36. Stewart R.A.H., Held C., Hadziosmanovic N., et al. Physical activity and mortality in patients with stable coronary heart disease. J Am Coll Cardiol. 2017;70:1689–1700. doi: 10.1016/j.jacc.2017.08.017. [ DOI ] [ PubMed ] [ Google Scholar ] 37. Moholdt T., Wisløff U., Nilsen T.I., Slørdahl S.A. Physical activity and mortality in men and women with coronary heart disease: A prospective population-based cohort study in Norway (the HUNT study) Eur J Cardiovasc Prev Rehabil. 2008;15:639–645. doi: 10.1097/HJR.0b013e3283101671. [ DOI ] [ PubMed ] [ Google Scholar ] 38. Ahn H.J., Choi E.K., Rhee T.M., et al. Accelerometer-derived physical activity and the risk of death, heart failure, and stroke in patients with atrial fibrillation: A prospective study from UK Biobank. Br J Sports Med. 2024;58:427–434. doi: 10.1136/bjsports-2023-106862. [ DOI ] [ PubMed ] [ Google Scholar ] 39. Cao Z., Min J., Chen H., et al. Accelerometer-derived physical activity and mortality in individuals with type 2 diabetes. Nat Commun. 2024;15:5164. doi: 10.1038/s41467-024-49542-0. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 40. Cao Z., Min J., Hou Y., Si K., Wang M., Xu C. Association of accelerometer-derived physical activity with all-cause and cause-specific mortality among individuals with cardiovascular diseases: A prospective cohort study. Eur J Prev Cardiol. 2025;32:20–29. doi: 10.1093/eurjpc/zwae248. [ DOI ] [ PubMed ] [ Google Scholar ] 41. Liang Y.Y., He Y., Huang P., et al. Accelerometer-measured physical activity, sedentary behavior, and incidence of macrovascular and microvascular events in individuals with type 2 diabetes mellitus and prediabetes. J Sport Health Sci. 2025;14 doi: 10.1016/j.jshs.2024.100973. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 42. Ekelund U., Sanchez-Lastra M.A., Dalene K.E., Tarp J. Dose‒response associations, physical activity intensity and mortality risk: A narrative review. J Sport Health Sci. 2024;13:24–29. doi: 10.1016/j.jshs.2023.09.006. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 43. Turner R.R., Steed L., Quirk H., et al. Interventions for promoting habitual exercise in people living with and beyond cancer. Cochrane Database Syst Rev. 2018;9 doi: 10.1002/14651858.CD010192.pub3. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 44. Fry A., Littlejohns T., Sudlow C., Doherty N., Allen N. The representativeness of the UK Biobank cohort on a range of sociodemographic, physical, lifestyle and health-related characteristics. J Epidemiol Community Health (1978) 2016;70:A26.1–A26.6. doi: 10.1136/jech-2016-208064.41. [ DOI ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials mmc1.pdf (3.1MB, pdf) Download video file (4.9MB, mp4) Data Availability Statement The UK Biobank database can be accessed by researchers on application ( https://www.ukbiobank.ac.uk/register-apply/ ). Articles from Journal of Sport and Health Science are provided here courtesy of Shanghai University of Sport ACTIONS View on publisher site PDF (1.1 MB) Cite Collections Permalink PERMALINK Copy RESOURCES Similar articles Cited by other articles Links to NCBI Databases Cite Copy Download .nbib .nbib Format: AMA APA MLA NLM Add to Collections Create a new collection Add to an existing collection Name your collection * Choose a collection Unable to load your collection due to an error Please try again Add Cancel Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top