You are here

  • Home
  • Archive
  • Volume 58, Issue 4
  • Joint associations of device-measured physical activity and abdominal obesity with incident cardiovascular disease: a prospective cohort study

Article Text

Article menu
Original research
Joint associations of device-measured physical activity and abdominal obesity with incident cardiovascular disease: a prospective cohort study
  1. Miguel Adriano Sanchez-Lastra1,2,3,
  2. Ding Ding4,5,
  3. Borja Del Pozo Cruz6,7,8,
  4. Knut Eirik Dalene9,
  5. Carlos Ayán2,3,
  6. Ulf Ekelund1,9,
  7. Jakob Tarp10
  1. 1 Department of Sports Medicine, Norwegian School of Sports Sciences, Oslo, Norway
  2. 2 Department of Special Didactics, University of Vigo Faculty of Education and Sports Sciences, Pontevedra, Spain
  3. 3 Wellness and Movement Research Group (WellMove), Galicia Sur Health Research Institute (IIS Galicia Sur), SERGAS-UVIGO, Vigo, Spain
  4. 4 Prevention Research Collaboration, The University of Sydney School of Public Health, Sydney, New South Wales, Australia
  5. 5 The University of Sydney Charles Perkins Centre, Camperdown, New South Wales, Australia
  6. 6 Department of Sports Science and Clinical Biomechanics, University of Southern Denmark Centre for Active and Healthy Ageing, Odense, Denmark
  7. 7 University of Cadiz Faculty of Education Sciences, Puerto Real, Spain
  8. 8 Biomedical Research and Innovation Institute of Cádiz (INiBICA) Research Unit, University of Cádiz Puerta del Mar University Hospital, Cádiz, Spain
  9. 9 Department of Chronic Diseases, Norwegian Institute of Public Health, Oslo, Norway
  10. 10 Department of Clinical Epidemiology, Aarhus University & University Hospital, Aarhus, Denmark
  1. Correspondence to Dr Miguel Adriano Sanchez-Lastra, Department of Special Didactics, University of Vigo Education Sciences and Sports Faculty, Pontevedra 36005, Spain; misanchez{at}uvigo.gal

Abstract

Objective To examine the joint associations between physical activity and abdominal obesity with the risk of cardiovascular disease (CVD) events.

Methods We included 70 830 UK Biobank participants (mean age±SD=61.6 ± 7.9 years; 56.4% women) with physical activity measured by wrist-worn accelerometers and without major chronic diseases. Participants were jointly categorised into six groups based on their physical activity level (tertiles of total volume and specific intensity levels) and presence or absence of abdominal obesity based on measured waist circumference. Associations with incident CVD (fatal and non-fatal events) were determined using proportional subdistribution hazard models with multivariable adjustment.

Results After excluding events during the first 2 years of follow-up, participants were followed for a median of 6.8 years, during which 2795 CVD events were recorded. Compared with the low abdominal adiposity and highest tertile of physical activity, abdominal obesity was associated with higher risk of incident CVD, especially in those with low levels of vigorous-intensity physical activity (HR 1.42, 95% CI 1.22 to 1.64). Approximately 500 min per week of moderate-to-vigorous intensity and approximately 30–35 min of vigorous-intensity physical activity offset the association of abdominal obesity and the risk of having a CVD event.

Conclusion Physical activity equivalent to approximately 30–35 min of vigorous intensity per week appears to offset the association between abdominal obesity and incident CVD. About 15 times more physical activity of at least moderate intensity is needed to achieve similar results.

  • Public health
  • Cardiovascular Diseases
  • Risk factor
  • Physical activity
  • Body Mass Index

Data availability statement

The UK Biobank database used in this study can be accessed by researchers on application (https://www.ukbiobank.ac.uk/register-apply/).

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

https://doi.org/10.1136/bjsports-2023-107252

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    WHAT IS ALREADY KNOWN ON THIS TOPIC

    • Abdominal obesity is associated with a higher risk of developing and dying from cardiovascular disease (CVD). Whether physical activity can reduce or fully offset this association is unclear.

    WHAT THIS STUDY ADDS

    • We found that approximately 500 min of moderate-to-vigorous intensity physical activity or 30–35 min of vigorous-intensity physical activity per week, measured by wrist-worn accelerometers, fully offset the association of abdominal obesity with incident CVD.

    HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

    • Encouraging people to participate in vigorous-intensity physical activity (eg, climbing stairs at a fast pace) for short periods on most days of the week appears to be a promising strategy to counteract the effects of abdominal obesity on CVD outcomes in middle-aged adults.

    Introduction

    Cardiovascular diseases (CVDs) are the leading causes of deaths globally but are to a large degree preventable through modifiable risk factors such as smoking, alcohol consumption, poor diets, obesity and physical inactivity.1

    Obesity is associated with higher risk of CVD, however, susceptibility to obesity-related CVD depends largely on body fat distribution.2 Further, individuals with higher abdominal adiposity (ie, abdominal obesity) are at higher risk of CVD morbidity and mortality independently of total adiposity.2–4 This association is largely mediated by impaired glucose and lipid metabolisms, which are often seen with increasing ectopic fat deposition.5 6 Lifestyle behaviours, such as physical activity, reduce the risk of CVD via the same metabolic pathways.7 It is, therefore, plausible that higher levels of physical activity may counteract the deleterious association between abdominal adiposity and CVD.

    Accelerometer-measured physical activity captures sporadic and incidental periods of physical activity, which cannot be measured by self-report instruments.8 It provides valid information on light-intensity physical activity (LPA) which may be beneficial for cardiometabolic health and longevity.9 At the other end of the intensity spectrum, recent research suggests vigorous-intensity physical activity (VPA) is particularly relevant for reducing the risk for incident CVD.10 11 However, whether different intensities of physical activity modify the association between abdominal and total adiposity and the risk for incident CVD remains unclear.

    Using accelerometer-measured physical activity in a large prospective cohort of adults from the UK, we aimed to examine the joint and stratified association between abdominal adiposity and different intensities of physical activity with incident CVD. As a secondary aim, we repeated the analysis replacing abdominal adiposity with total adiposity.

    Methods

    This study followed the Strengthening the Reporting of Observational Studies in Epidemiology reporting guidelines (online supplemental table 1). The statistical analysis and presentation is consistent with the CHecklist for statistical Assessment of Medical Papers statement.12

    Supplemental material

    Data source and study population

    We used data from the UK Biobank (application number 29717), a cohort of more than 500 000 participants (5.5% response rate), aged 37–73 years, recruited across the UK, who provided written informed consent. Baseline data collection, conducted between 2006 and 2010 acros 22 assessment centres, included questionnaires, physical measurements and biological sampling, and were linked to electronic registries. Between 2013 and 2015, physical activity was measured by accelerometers in a subsample of participants (n=103 613).13

    Exposures

    Adiposity

    Trained clinical staff performed all anthropometric measurements at baseline following standardised protocols.13 We determined abdominal adiposity using two proxies. First, we classified participants as having abdominal obesity based on the WHO’s waist circumference (WC) cut-offs (≥88 cm for women and ≥102 cm for men).14 Second, to account for the high correlation between WC and the body mass index (BMI, weight(kg)/height (m2)),6 we derived a measure of abdominal adiposity that is statistically independent of total adiposity (BMI) by regressing WC on BMI. Using the residuals from this regression, we categorised participants into low, medium and high ‘WC-for-BMI’ groups, mirroring the sex-specific distribution of normal weight, overweight and obesity based on BMI.15 Finally, we used the BMI as a proxy of total adiposity. Additional descriptive information on the creation of these measures of adiposity is included in online supplemental table 2.

    Physical activity

    Participants wore an Axivity AX3 accelerometer (Axivity, York, UK) on their dominant wrist for 24 hours during seven consecutive days. Data were collected at a sampling rate of 100 Hz and a dynamic range of ±8 g. Individual participant data were extracted from 5 s epoch time series. Exclusions criteria included <72 hours of wear time, missing data in the included 1-hour periods, implausible acceleration values (average vector magnitude acceleration >100 mg) or uncalibrated data.16 Total physical activity was extracted from measured average vector magnitude in mg. Time spent in LPA and moderate-to-vigorous physical activity (MVPA) was determined through a machine-learning approach.16 17 The validation sample for this approach had insufficient VPA for the algorithm to identify a useful characterisation pattern. Therefore, time spent in VPA was estimated from the average proportion of time spent in accelerations above 400 mg.18 Participants were categorised into tertiles (low, medium, high) for each activity measure (further details in online supplemental table 2).

    Incident CVD

    We defined incident CVD events as the first non-fatal cardiovascular event (International Statistical Classification of Diseases and Related health Problems 10th edition (ICD-10) codes I20–I25, I50 and I60–I64) from hospital inpatient records, primary or secondary diagnoses, in addition to death by cardiovascular event (primary) (ICD-10 codes I00–I99),19 obtained from national registries. Events occurring after finishing the accelerometer measurement were considered incident. Participants were followed to incident CVD, death, withdrawal from the study, or 30 September 2021 for England, 31 July 2021 for Scotland or 31 March 2016 for Wales, whichever occurred first.

    Covariates

    We identified confounders a priori based on a directed acyclic graph.20 Detailed information on covariates is shown in online supplemental table 2. Self-reported sociodemographic covariates included ethnicity, education, living with partner and employment status. The Townsend Index was used as a marker of area-level socioeconomic status.21 Self-reported lifestyle covariates included a dietary quality index, alcohol intake and smoking history. History of depression, hypertension and statins medication were derived from a combination of questionnaire data and verbal interview. To identify undiagnosed hypertensive patients, clinical blood pressure measurements were considered alongside self-reported data. Type 2 diabetes cases were identified using a previously developed algorithm.22

    Analytical sample

    To reduce the risk of reverse causality, participants with major chronic diseases at baseline (eg, neurodegenerative, immunological or systemic diseases), underweight (BMI <18.5 kg/m2), unable to walk, with mobility limitations or with a history of cancer or CVD before the accelerometer assessment were excluded (online supplemental table 2). Following these criteria, a total of 22 451 participants were excluded (online supplemental figure 2). Of the remaining participants, 68 090 had complete data on exposures, outcomes and covariates. We retrieved missing data on covariates for an additional 2740 participants by performing multiple imputation using chained equations with 20 datasets generated.

    Statistical analyses

    Descriptive statistics included mean and SD for normally distributed continuous variables, median and range for skewed continuous variables, and percentages for categorical data.

    First, we examined the independent associations of abdominal obesity and physical activity with incident CVD using the lowest WC and highest physical activity groups as the reference, respectively. Then we examined the joint associations of abdominal obesity (two categories based on WHO cut-off points and three based on ‘WC-for-BMI’ residuals) and physical activity tertiles (resulting in six and nine distinct categories, respectively), with incident CVD. The group with the lowest level of adiposity and highest level of physical activity was used as the reference. In addition, we examined the joint associations between total adiposity (three categories of BMI) and physical activity (resulting in nine BMI-physical activity categories) with incident CVD.

    Subdistribution HRs with 95% CIs for incident CVD were calculated using Fine-Gray models to account for non-CVD deaths as a competing event. Models were left-truncated to start the follow-up time 2 years after the accelerometer assessments (2013–2015). Age was used as the underlying timescale, and models were adjusted for covariates at baseline (2006–2010), sequentially. We employed models 1, 1b and 1c for independent associations analyses, while models 1–3 were used for joint and stratified associations analyses. Model 1 was adjusted for age at accelerometer assessment (as timescale) and sex. Models 1b and 2 were adjusted as model 1 plus ethnicity, education, living with partner, employment, Townsend Index, diet quality, alcohol intake, smoking and depression. Models 1b for WC and BMI exposures were also adjusted for MVPA. Models 1b and 2 were adjusted as model 1 plus ethnicity, education, living with partner, employment, Townsend Index, diet quality, alcohol intake, smoking and depression. Model 1b for WC and BMI exposures was also adjusted for MVPA. Models 1b and 2 for WC based on WHO cut-offs were additionally adjusted for BMI in continuous form.6 We considered model 2 as the main model. Prevalent type 2 diabetes, hypertension and use of statins medication were considered mediators in the causal pathway from abdominal obesity and/or physical activity to CVD and we did not adjust for them in the main model. However, these could be conceptually considered as confounders or mediators. To address this, we created models 1c and 3, which were adjusted as models 1b and 2, respectively, but included type 2 diabetes, hypertension and statins medication as confounders in an alternative scenario (online supplemental figure 1). Proportional hazards assumption was verified using log-log plots, and interactions between adiposity and physical activity variables were assessed through likelihood-ratio tests.

    Several supporting and sensitivity analyses were performed. First, we examined the associations between abdominal obesity and incident CVD stratified by tertiles of physical activity. Second, we repeated the joint associations analysis between abdominal obesity, physical activity and CVD stratified by sex. Third, we repeated the analysis based on WHO cut-offs without adjustment for the BMI. Finally, we also repeated the primary analysis left-truncating the models to exclude the first 5 years of follow-up. The analyses were performed using Stata V.17 statistical software (StataCorp).

    Patient and public involvement

    There was no patient or public involvement in the planning, conceptualisation, research design, analysis, interpretation or composition of the findings. Results will be disseminated via institutional websites, press releases and the UK Biobank participant resource.

    Equity, diversity and inclusion statement

    Our research team comprises six men and one woman from various countries in Europe and Australia. The study population is diverse in terms of age, gender, demographics and comorbidities. However, the predominantly white composition of the cohort resulted in underrepresentation of other ethnic groups. Individuals from lower socioeconomic backgrounds and marginalised communities may also have lower representation. Methodological constraints required excluding individuals with specific comorbidities and mobility limitations from analyses.

    Results

    We included 70 830 participants (56% women) with a mean (SD) age at the time of accelerometer assessment of 61.6 (7.9) years. Median follow-up time was 6.8 years during which 2795 CVD events occurred (2558 non-fatal and 237 fatal). Descriptive characteristics of the participants across WC categories are provided in table 1. Online supplemental table 3 presents a summary of the weekly physical activity levels (median and range) per activity tertiles and WC categories.

    View this table:
    Table 1

    Descriptive characteristics of the included participants by sex and group of waist circumference

    Abdominal obesity and low physical activity were independently associated with higher risk of incident CVD. Adjusting for BMI, abdominal obesity was associated with a 13% higher risk of CVD (HR 1.13, 95% CI 1.01 to 1.26). The HRs (95% CI) for high versus low physical activity ranged from 1.04 (0.95 to 1.15) for LPA to 1.25 (1.13 to 1.39) for VPA (table 2).

    View this table:
    Table 2

    Independent associations between physical activity, adiposity categories and risk of incident cardiovascular disease

    Joint associations between abdominal adiposity and physical activity with CVD

    The highest risk of incident CVD was consistently observed in those with abdominal obesity and low physical activity, with HRs (95% CI) of 1.33 (1.14 to 1.54) for abdominal obesity and low total activity and 1.42 (1.22 to 1.64) for abdominal obesity and low VPA (figure 1). In contrast, high LPA (median=2090 min/week), high MVPA (median=512 min/week) and high VPA (median=32 min/week) mitigated the association between abdominal obesity and CVD (HRs (95% CI): 1.01 (0.85 to 1.20), 1.02 (0.85 to 1.22) and 1.04 (0.85 to 1.28), respectively). A high level of total physical activity did not completely offset the association with abdominal obesity.

    Figure 1
    Figure 1

    Joint associations between waist circumference (WC) categories from clinical cut-offs, physical activity tertiles and incident cardiovascular disease. Ref: reference. Figure shows weekly volumes of (A) total physical activity, (B) light-intensity, (C) moderate-to-vigorous intensity and (D) vigorous-intensity physical activity. Median (range) of physical activity levels for high, medium and low tertiles were, for total physical activity, 35.9 (31.0–97.2), 27.7 (24.7–31.0) and 21.4 (5.9–24.7) mg, respectively. For light-intensity min/week, 2794.3 (2390.4–5799.5), 2089.1 (1808.1–2390.3) and 1489.0 (28.0–1808.0). For moderate-to-vigorous intensity min/week, 511.9 (349.2–2999.7), 247.1 (164.1–349.1) and 87.5 (0.0–164.0). For vigorous-intensity min/week, 31.7 (13.1–756.7), 6.1 (2.8–13.0) and 1.0 (0.0–2.7). Results are from main model, adjusted for age (as timescale), sex, body mass index, ethnicity, education, living with partner, employment, Townsend Index, diet quality, alcohol intake, smoking and depression.

    The magnitude of the associations diminished after adjustment for type 2 diabetes, hypertension and statins medication as potential mediators, but the association pattern was unchanged (online supplemental figure 3).

    Analysis of joint physical activity and WC-for-BMI categories supported the findings based on WHO cut-off points (figure 2). The highest risk was observed in the group with high WC and low MVPA (HR 1.64; 95% CI 1.39 to 1.94). Compared with the high physical activity and low WC-for-BMI, having medium or low levels of total physical activity, MVPA and VPA was associated with higher risk for incident CVD irrespective of abdominal adiposity. Having a higher WC-for-BMI was not associated with increased risk of CVD in those with high levels of MVPA (median=512 min/week) and VPA (median=32 min/week). Model 3 exhibited similar patterns, although less pronounced associations.

    Figure 2
    Figure 2

    Joint associations between categories of waist circumference (WC)-for-body mass index residuals, physical activity tertiles and incident cardiovascular disease. Ref: reference. Figure shows tertiles of weekly volumes of (A) total physical activity, (B) light-intensity, (C) moderate-to-vigorous intensity and (D) vigorous-intensity physical activity. Median (range) of physical activity levels for high, medium and low tertiles were, for total physical activity, 35.9 (31.0–97.2), 27.7 (24.7–31.0) and 21.4 (5.9–24.7) mg, respectively. For light-intensity min/week, 2794.3 (2390.4–5799.5), 2089.1 (1808.1–2390.3) and 1489.0 (28.0–1808.0). For moderate-to-vigorous intensity min/week, 511.9 (349.2–2999.7), 247.1 (164.1–349.1) and 87.5 (0.0–164.0). For vigorous-intensity min/week, 31.7 (13.1–756.7), 6.1 (2.8–13.0) and 1.0 (0.0–2.7). Results are from main model, adjusted for age (as timescale), sex, body mass index, ethnicity, education, living with partner, employment, Townsend Index, diet quality, alcohol intake, smoking and depression.

    Joint associations between general adiposity and physical activity with CVD

    A clear dose–response association between BMI and risk of CVD was found in all strata of physical activity in joint associations between physical activity and general adiposity as measured by BMI (online supplemental figure 4). High levels of physical activity attenuated but did not offset the association from having overweight or obesity. Being normal weight was associated with higher risk in those with low levels of total physical activity (HR 1.21; 95% CI 1.02 to 1.43) and low VPA (HR 1.24; 95% CI 1.06 to 1.47). Obesity combined with low MVPA was associated with the highest risk (HR 2.03; 95% CI 1.75 to 2.36).

    Supporting and sensitivity analyses

    Results from the analyses stratified by physical activity levels aligned with joint associations analyses, indicating that high levels of total physical activity, LPA, MVPA and VPA mitigated the risk associated with abdominal obesity and high WC-for-BMI (online supplemental tables 4 and 5). Whereas medium levels of VPA also offset the association in people with abdominal obesity and high WC-for-BMI, individuals with medium MVPA and LPA remained at 17%–27% higher risk. The patterns for the joint associations between WC and physical activity with incident CVD were similar in men and women. However, the effect sizes appeared greater in magnitude in women compared with men (online supplemental tables 6 and 7). Repeating the analysis of WHO cut-offs without adjustment for BMI accentuated associations and high levels of physical activity no longer offset the association between abdominal obesity and CVD (online supplemental figure 5). The sensitivity analysis excluding the first 5 years of follow-up showed similar patterns but generally attenuated associations between higher WC and risk of CVD (online supplemental figure 6).

    Discussion

    Our results suggest that physical activity of any intensity can mitigate or even offset the association between abdominal obesity and incident CVD. Approximately 30–35 min per week of VPA appear to effectively offset the association of abdominal obesity with incident CVD.

    Although high levels of LPA appear to attenuate the risk, the magnitude of associations were markedly stronger for MVPA and, particularly, for VPA. Indeed, as much as approximately 500 min per week of MVPA appear needed to fully offset the association with incident CVD.

    Our results extend previous observations examining associations between VPA and risk for incident CVD10 11 by modelling these associations in combination with proxies of abdominal adiposity. The observation that relatively small amounts of VPA offset the association was consistent both when analysing abdominal adiposity using WHO’s WC cut-offs and when using WC-for-BMI residuals. We also found that being categorised in the lowest tertile of any physical activity intensity was associated with an increased risk of incident CVD, even when abdominal adiposity was low. Encouraging people with low levels of physical activity to engage in small amounts of VPA may be more appealing compared with large amounts of MVPA to reduce the risk of incident CVD.

    We observed that high levels of physical activity were beneficial regardless of BMI strata. However, higher general adiposity, particularly obesity (BMI>30 kg/m2), was associated with higher risk of incident CVD irrespective of physical activity. This finding corroborates previous observations from self-reported physical activity data and CVD mortality in women.23 However, recent evidence suggests that high levels of self-reported physical activity may eliminate the risk of CVD mortality in people with grade I obesity (BMI 30–35 kg/m2), but not in those with grade II (≥35 kg/m2).15 Insufficient statistical power prevented us from analysing associations by obesity grades. Further investigation is needed to determine the extent to which physical activity can mitigate CVD risk across different levels of obesity.

    Measurements of WC do not distinguish between subcutaneous and visceral adipose tissue, both metabolically active.24 During physical activity, subcutaneous adipose tissue in the upper body exhibits high lipolytic rates,25 while the lower body contributes minimally,26 and fatty acids used by active muscles mainly originate from subcutaneous abdominal fat.27 The BMI includes lower-body subcutaneous fat tissue that may act as a metabolic buffer and protect other tissues from lipotoxicity caused by excess lipids and ectopic fat.28 Moreover, individual variation in visceral adipose tissue exists for individuals with similar BMI, potentially explaining the variability in CVD risk profiles.2 Thus, it is plausible that physical activity may more effectively mitigate the association between abdominal adiposity and incident CVD compared with total adiposity, particularly when the latter is assessed by BMI. This may also clarify the stronger associations observed between joint WC-physical activity categories and CVD without BMI adjustment (online supplemental figure 5) compared with the analysis with BMI adjustment (figure 1).

    Our results may inform future physical activity recommendations. Current recommendations, primarily based on self-reported physical activity, suggest that 150–300 min of MVPA or 75–150 min of VPA or a combination of both is associated with substantial health benefits. This recommendation considers the same health benefits for half the amount of VPA compared with MVPA. Our results, as well as those from others,11 29 challenge this concept. We found that approximately 15 times more MVPA was needed to mitigate the association between abdominal adiposity with incident CVD compared with VPA. This difference may be explained by the higher intensity physical activity leading to improvements in cardiorespiratory fitness, a strong marker of cardiovascular health that has been shown to modify the association between obesity and mortality, also known as the fat-but-fit paradox.30–32 Previous studies have demonstrated that small bouts of VPA can improve cardiorespiratory fitness and have positive cardiovascular effects that may be greater than those from moderate-intensity activity.33–36 Altogether, our results challenge the notion that substituting half the amount of MVPA with VPA maintains equal health benefits. Future guidelines may prioritise promoting the achievement of small amounts of VPA as a more feasible and time-effective strategy to counteract the CVD risk from abdominal obesity.

    Our study’s main strength lies in the inclusion of a large sample with WC and BMI measurements performed by clinical staff, reducing self-report bias.We categorised both central (WC) and total adiposity (BMI) based on WHO cut-offs, easily and routinely assessed by clinicians, enabling direct implementation of our findings in clinical practice. To investigate the independent associations of WC with CVD, we adjusted for BMI to isolate the specific contribution of abdominal adiposity to CVD risk.6 Furthermore, we performed additional analyses using a variable based on WC-for-BMI residuals that confirmed the robustness of our findings. We measured physical activity by wrist-worn accelerometers, which is more accurate than self-report. However, physical activity levels derived from wrist-worn accelerometers are substantially higher than that from waist worn accelerometers.37 Thus, absolute levels of time spent in moderate and vigorous-intensity and their associations with clinical outcomes are not directly comparable between studies employing different monitor placements.

    Limitations

    Limitations of our study include the use of WC and the BMI as proxy measures of abdominal and total adiposity, respectively. Also, we used a single baseline assessment of adiposity and covariates, as well as a single accelerometer measurement that was performed about 5 years later. In subsamples of 16 230 and 15 854 participants with repeated WC and BMI assessments in 2014, we observed relatively stable adiposity measures, with 84% remaining in the same WC category and 78% in the same BMI category compared with their baseline measurements. However, we were unable to adjust for potential changes in exposures and covariates. The single measurement of physical activity also limits inference on individual variability of physical activity over time. To mitigate the risk of reverse causation, we excluded participants with prevalent diseases, mobility limitations and events occurring within the initial 2 years of follow-up (5 years in the sensitivity analysis). However, this may not completely eliminate this bias.38 We also performed multivariate adjustment for numerous potential confounders, but residual bias from unmeasured or insufficiently measured confounders, such as smoking volume, intensity or duration, may still be present. Additionally, UK Biobank is not representative of its source population and there is evidence of healthy volunteer selection bias,39 which may be accentuated in the subgroup with accelerometer measurements.10

    Implications for clinical practice and health policy

    People living with obesity face difficulties in attaining and maintaining weight loss and in physical activity participation.40 Our results imply that public health messaging should focus on achieving small amounts of VPA for reducing the risk of incident CVD, instead of only focusing on ‘a number on a scale’. This may be especially attractive for those with limited time or motivation for more structured exercise.36

    Future directions

    Future studies should use more precise measures of adiposity, such as DXA scans or MRI, to further explore how physical activity, preferably long-term patterns, can reduce the detrimental effect of excess adipose tissue on CVD risk. Studying the role of intramuscular fat and lean mass tissue’s quality and volume in these associations could also provide valuable insights.41 Additional research may want to include accelerometers attached to other body parts (eg, waist, thigh) to determine potential differences in the amount of physical activity needed to mitigate the risk for incident CVD associated with adiposity compared with the results from wrist-worn devices.

    Conclusion

    Physical activity equivalent to approximately 30–35 min of vigorous intensity per week, measured by wrist-worn accelerometers, appears to offset the association between abdominal obesity and incident CVD. A considerably greater amount of physical activity of at least moderate intensity, approximately 500 min per week, is needed to achieve similar results.

    Data availability statement

    The UK Biobank database used in this study can be accessed by researchers on application (https://www.ukbiobank.ac.uk/register-apply/).

    Ethics statements

    Patient consent for publication

    Ethics approval

    This study involves human participants and was approved by North West - Haydock Research Ethics Committee (ID 21/NW/0157). Participants gave informed consent to participate in the study before taking part.

    Acknowledgments

    This research received no financial assistance. MAS-L 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 (Application ID 2009254). BDPC was supported by the Government of Andalusia, Research Talent Recruitment Programme (EMERGIA 2020/00158).

    References

    1. ↵
      1. World Health Organization
      . Cardiovascular diseases (CVDs). 2021. Available: https://www.who.int/news-room/fact-sheets/detail/cardiovascular-diseases-(cvds) [Accessed 27 Sep 2022].
    2. ↵
      2. Powell-Wiley TM ,
      3. Poirier P ,
      4. Burke LE , et al
      . Obesity and cardiovascular disease: a scientific statement from the American heart association. Circulation 2021;143:E984–1010. doi:10.1161/CIR.0000000000000973
      OpenUrlCrossRef
    3. ↵
      2. Bray GA ,
      3. Heisel WE ,
      4. Afshin A , et al
      . The science of obesity management: an endocrine society scientific statement. Endocr Rev 2018;39:79–132. doi:10.1210/er.2017-00253
      OpenUrlCrossRefPubMed
    4. ↵
      2. Rao G ,
      3. Powell-Wiley TM ,
      4. Ancheta I , et al
      . Identification of obesity and cardiovascular risk in ethnically and racially diverse populations: a scientific statement from the American heart association. Circulation 2015;132:457–72. doi:10.1161/CIR.0000000000000223
      OpenUrlFREE Full Text
    5. ↵
      2. Longo DL ,
      3. Heymsfield SB ,
      4. Wadden TA
      . Mechanisms, pathophysiology, and management of obesity. N Engl J Med 2017;376:254–66. doi:10.1056/NEJMra1514009
      OpenUrlCrossRefPubMed
    6. ↵
      2. Ross R ,
      3. Neeland IJ ,
      4. Yamashita S , et al
      . Waist circumference as a vital sign in clinical practice: a consensus statement from the IAS and ICCR working group on visceral obesity. Nat Rev Endocrinol 2020;16:177–89. doi:10.1038/s41574-019-0310-7
      OpenUrlPubMed
    7. ↵
      2. Thyfault JP ,
      3. Bergouignan A
      . Exercise and metabolic health: beyond skeletal muscle. Diabetologia 2020;63:1464–74. doi:10.1007/s00125-020-05177-6
      OpenUrlCrossRefPubMed
    8. ↵
      2. Prince SA ,
      3. Adamo KB ,
      4. Hamel ME , et al
      . A comparison of direct versus self-report measures for assessing physical activity in adults: a systematic review. Int J Behav Nutr Phys Act 2008;5:56. doi:10.1186/1479-5868-5-56
      OpenUrlCrossRefPubMed
    9. ↵
      2. Chastin SFM ,
      3. De Craemer M ,
      4. De Cocker K , et al
      . How does light-intensity physical activity associate with adult cardiometabolic health and mortality? Systematic review with meta-analysis of experimental and observational studies. Br J Sports Med 2019;53:370–6. doi:10.1136/bjsports-2017-097563
      OpenUrlAbstract/FREE Full Text
    10. ↵
      2. Dempsey PC ,
      3. Rowlands AV ,
      4. Strain T , et al
      . Physical activity volume, intensity, and incident cardiovascular disease. Eur Heart J 2022;43:4789–800. doi:10.1093/eurheartj/ehac613
      OpenUrl
    11. ↵
      2. Ahmadi MN ,
      3. Clare PJ ,
      4. Katzmarzyk PT , et al
      . Vigorous physical activity, incident heart disease, and cancer: how little is enough Eur Heart J 2022;43:4801–14. doi:10.1093/eurheartj/ehac572
      OpenUrlCrossRef
    12. ↵
      2. Mansournia MA ,
      3. Collins GS ,
      4. Nielsen RO , et al
      . A checklist for statistical assessment of medical papers (the CHAMP statement): explanation and elaboration. Br J Sports Med 2021;55:1009–17. doi:10.1136/bjsports-2020-103652
      OpenUrlFREE Full Text
    13. ↵
      1. The UK Biobank
      . UK Biobank: protocol for a large-scale prospective epidemiological resource; 2007. Available: https://www.ukbiobank.ac.uk/wp-content/uploads/2011/11/UK-Biobank-Protocol.pdf [Accessed 23 Sep 2019].
    14. ↵
      1. World Health Organization
      . Waist circumference and waist-hip ratio: report of a WHO expert consultation, Geneva, 8-11 December 2008; 2011.
    15. ↵
      2. Sanchez-Lastra MA ,
      3. Ding D ,
      4. Dalene KEE , et al
      . Physical activity and mortality across levels of adiposity: a prospective cohort study from the UK Biobank. Mayo Clin Proc 2021;96:105–19. doi:10.1016/j.mayocp.2020.06.049
      OpenUrlCrossRefPubMed
    16. ↵
      2. Doherty A ,
      3. Jackson D ,
      4. Hammerla N , et al
      . Large scale population assessment of physical activity using wrist worn accelerometers: the UK biobank study. PLoS ONE 2017;12:e0169649. doi:10.1371/journal.pone.0169649
    17. ↵
      2. Walmsley R ,
      3. Chan S ,
      4. Smith-Byrne K , et al
      . Reallocation of time between device-measured movement Behaviours and risk of incident cardiovascular disease. Br J Sports Med 2022;56:1008–17. doi:10.1136/bjsports-2021-104050
      OpenUrlAbstract/FREE Full Text
    18. ↵
      2. Hildebrand M ,
      3. Van Hees VT ,
      4. Hansen BH , et al
      . Age group comparability of raw accelerometer output from wrist- and hip-worn monitors. Med Sci Sports Exerc 2014;46:1816–24. doi:10.1249/MSS.0000000000000289
      OpenUrlCrossRefPubMedWeb of Science
    19. ↵
      2. Bosco E ,
      3. Hsueh L ,
      4. McConeghy KW , et al
      . Major adverse cardiovascular event definitions used in observational analysis of administrative databases: a systematic review. BMC Med Res Methodol 2021;21:1–18. doi:10.1186/s12874-021-01440-5
      OpenUrlCrossRefPubMed
    20. ↵
      2. Textor J ,
      3. van der Zander B ,
      4. Gilthorpe MS , et al
      . Robust causal inference using directed acyclic graphs: the R package “Dagitty" Int J Epidemiol 2016;45:1887–94. doi:10.1093/ije/dyw341
      OpenUrlCrossRefPubMed
    21. ↵
      2. Townsend P ,
      3. Phillimore P ,
      4. Beattie A
      . Health and deprivation: inequality and the North. Croom Helm Ltd, 1988.
    22. ↵
      2. Eastwood SV ,
      3. Mathur R ,
      4. Atkinson M , et al
      . Algorithms for the capture and adjudication of prevalent and incident diabetes in UK biobank. PLoS One 2016;11:e0162388. doi:10.1371/journal.pone.0162388
    23. ↵
      2. Hu FB ,
      3. Willett WC ,
      4. Li T , et al
      . Adiposity as compared with physical activity in predicting mortality among women. N Engl J Med 2004;351:2694–703. doi:10.1056/NEJMoa042135
      OpenUrlCrossRefPubMedWeb of Science
    24. ↵
      2. Bjørndal B ,
      3. Burri L ,
      4. Staalesen V , et al
      . Different adipose depots: their role in the development of metabolic syndrome and mitochondrial response to hypolipidemic agents. J Obes 2011;2011:490650. doi:10.1155/2011/490650
      OpenUrlCrossRefPubMed
    25. ↵
      2. Arner P ,
      3. Kriegholm E ,
      4. Engfeldt P , et al
      . Adrenergic regulation of lipolysis in situ at rest and during exercise. J Clin Invest 1990;85:893–8. doi:10.1172/JCI114516
      OpenUrlCrossRefPubMedWeb of Science
    26. ↵
      2. Horowitz JF ,
      3. Klein S
      . Oxidation of nonplasma fatty acids during exercise is increased in women with abdominal obesity. J Appl Physiol (1985) 2000;89:2276–82. doi:10.1152/jappl.2000.89.6.2276
      OpenUrlCrossRefPubMedWeb of Science
    27. ↵
      2. Horowitz JF
      . Fatty acid mobilization from adipose tissue during exercise. Trends Endocrinol Metab 2003;14:386–92. doi:10.1016/s1043-2760(03)00143-7
      OpenUrlCrossRefPubMedWeb of Science
    28. ↵
      2. Neeland IJ ,
      3. Poirier P ,
      4. Després JP
      . The cardiovascular and metabolic heterogeneity of obesity: clinical challenges and implications for management. Circulation 2018;137:1391–406. doi:10.1161/CIRCULATIONAHA.117.029617
      OpenUrlAbstract/FREE Full Text
    29. ↵
      2. Wang Y ,
      3. Nie J ,
      4. Ferrari G , et al
      . Association of physical activity intensity with mortality: a national cohort study of 403 681 US adults. JAMA Intern Med 2021;181:203–11. doi:10.1001/jamainternmed.2020.6331
      OpenUrl
    30. ↵
      2. Babu AS ,
      3. Arena R ,
      4. Myers J
      . Post-COVID era: time to re-introduce "cardiorespiratory fitness" as a vital sign. EClinicalMedicine 2022;51:101546. doi:10.1016/j.eclinm.2022.101546
      OpenUrl
    31. ↵
      2. Tarp J ,
      3. Grøntved A ,
      4. Sanchez‐Lastra MA , et al
      . Fitness, fatness, and mortality in men and women from the UK biobank: prospective cohort study. JAHA 2021;10. doi:10.1161/JAHA.120.019605
    32. ↵
      2. Ortega FB ,
      3. Ruiz JR ,
      4. Labayen I , et al
      . The fat but fit paradox: what we know and don’t know about it. Br J Sports Med 2018;52:151–3. doi:10.1136/bjsports-2016-097400
      OpenUrlFREE Full Text
    33. ↵
      2. Allison MK ,
      3. Baglole JH ,
      4. Martin BJ , et al
      . Brief intense stair climbing improves cardiorespiratory fitness. Med Sci Sports Exerc 2017;49:298–307. doi:10.1249/MSS.0000000000001188
      OpenUrlCrossRef
    34. ↵
      2. Jenkins EM ,
      3. Nairn LN ,
      4. Skelly LE , et al
      . Do stair climbing exercise “snacks” improve cardiorespiratory fitness Appl Physiol Nutr Metab 2019;44:681–4. doi:10.1139/apnm-2018-0675
      OpenUrl
    35. ↵
      2. Franklin BA ,
      3. Eijsvogels TMH ,
      4. Pandey A , et al
      . Physical activity, cardiorespiratory fitness, and cardiovascular health: a clinical practice statement of the American society for preventive cardiology part II: physical activity, cardiorespiratory fitness, minimum and goal intensities for exercise training, prescriptive methods, and special patient populations. Am J Prev Cardiol 2022;12:100425. doi:10.1016/j.ajpc.2022.100425
      OpenUrl
    36. ↵
      2. Stamatakis E ,
      3. Ahmadi MN ,
      4. Gill JMR , et al
      . Association of Wearable device-measured vigorous intermittent lifestyle physical activity with mortality. Nat Med 2022;28:2521–9. doi:10.1038/s41591-022-02100-x
      OpenUrlCrossRef
    37. ↵
      2. Loprinzi PD ,
      3. Smith B
      . Comparison between wrist-worn and waist-worn accelerometry. J Phys Act Health 2017;14:539–45. doi:10.1123/jpah.2016-0211
      OpenUrl
    38. ↵
      2. Tarp J ,
      3. Hansen BH ,
      4. Fagerland MW , et al
      . Accelerometer-measured physical activity and sedentary time in a cohort of US adults followed for up to 13 years: the influence of removing early follow-up on associations with mortality. Int J Behav Nutr Phys Act 2020;17:39. doi:10.1186/s12966-020-00945-4
      OpenUrlPubMed
    39. ↵
      2. Fry A ,
      3. Littlejohns TJ ,
      4. Sudlow C , et al
      . The representativeness of the UK biobank cohort on a range of sociodemographic, physical, lifestyle and health-related characteristics. J Epidemiol Community Health 2016;70:A26. doi:10.1136/jech-2016-208064.41
    40. ↵
      2. Wharton S ,
      3. Lau DCW ,
      4. Vallis M , et al
      . Obesity in adults: a clinical practice guideline. CMAJ 2020;192:E875–91. doi:10.1503/cmaj.191707
      OpenUrlFREE Full Text
    41. ↵
      2. Zamboni M ,
      3. Gattazzo S ,
      4. Rossi AP
      . Myosteatosis: a relevant, yet poorly explored element of sarcopenia. Eur Geriatr Med 2019;10:5–6. doi:10.1007/s41999-018-0134-3
      OpenUrlCrossRef

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Footnotes

    • UE and JT are joint senior authors.

    • Twitter @adrianosanlas, @DrMelodyDing

    • Contributors Study concept and design: MAS-L, JT and UE. Acquisition of data: MAS-L, UE and JT. Statistical analysis of the data: MAS-L and JT. Interpretation of the data: all authors. Drafting the manuscript: MAS-L, UE and JT. Critical revision of the manuscript for important intellectual contet: all authors. Study supervision: UE, JT. MAS-L is the study guarantor.

    • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

    • Competing interests None declared.

    • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.