Abstract
Background
To examine associations between accelerometer‐derived “weekend warrior” (WW) physical activity pattern (most moderate‐to‐vigorous‐intensity physical activity in 1–2 days) versus moderate‐to‐vigorous‐intensity physical activity spread more evenly with risks of incident first cardiometabolic disease (FCMD) and cardiometabolic multimorbidity (CMM) among hypertension participants.
Methods
UK Biobank participants with hypertension (n=26891) provided a full‐week of accelerometer‐derived physical activity data between 2013 and 2015, processed using GGIR and a machine‐learning classifier. CMM was defined as the occurrence of at least 2 of type 2 diabetes, ischemic heart disease, and stroke. Three activity patterns (WW, regular, and inactive) were compared across multiple thresholds and definitions of active WW, with significance corrected for multiple comparisons at a false discovery rate of 0.05.
Results
Over 7.5 years median follow‐up, 2718 individuals developed FCMD and 219 developed CMM. Using the definition of WW as guideline‐based and ≥50% of total moderate‐to‐vigorous‐intensity physical activity over 1 to 2 days via GGIR, we observed similarly lower risks of incident FCMD (WW versus inactive: hazard ratio [HR], 0.75 [95% CI, 0.69–0.82], P=1.4 × 10−3; regular versus inactive: HR, 0.85 [95% CI, 0.74–0.96], P=2.6 × 10−2) and CMM (WW versus inactive: HR, 0.59 [95% CI, 0.41–0.83], P=4.0 × 10−3). Both active WW and regular physical activity patterns were associated with lower risk of FCMD compared with inactivity, with no statistically significant differences between the 2 active patterns. However, the result of WW versus inactive for CMM showed no statistical significance.
Conclusions
Increased activity, even when concentrated within 1 to 2 days/week may reduce cardiometabolic disease risk in individuals with hypertension.
Keywords: cardiometabolic multimorbidity, physical activity, weekend warrior
Subject Categories: Cardiovascular Disease, Epidemiology, Risk Factors, Hypertension
Nonstandard Abbreviations and Acronyms
- CMM
cardiometabolic multimorbidity
- FCMD
first cardiometabolic disease
- MVPA
moderate‐to‐vigorous‐intensity physical activity
- WW
weekend warrior
Clinical Perspective.
What Is New?
Moderate‐to‐vigorous physical activity, classified by GGIR and a machine‐learning method with consistent results, showed that in individuals with hypertension, the weekend warrior pattern (defined as guideline‐based and ≥50% of total moderate‐to‐vigorous physical activity over 1 to 2 days and median thresholds) was associated with lower risks of first cardiometabolic disease and cardiometabolic multimorbidity.
A more evenly distributed pattern of physical activity was associated with a similarly lower risk of first cardiometabolic disease (eg, type 2 diabetes), whereas the result for cardiometabolic multimorbidity did not reach statistical significance.
What Are the Clinical Implications?
Our findings highlight the weekend warrior pattern as a potential alternative in preventive intervention strategies of cardiometabolic multimorbidity in individuals with hypertension.
Cardiometabolic multimorbidity (CMM), defined as coexistence of at least 2 cardiometabolic diseases among type 2 diabetes, ischemic heart disease, and stroke, has become a major public health challenge. Compared with having a single cardiometabolic disease, CMM is associated with poorer quality of life and an ~2‐ to 4‐fold higher risk of premature death 1 , 2 Among individuals with hypertension, the urgent need for effective lifestyle interventions to stall disease progression in these high‐risk patients. 3
Physical activity is widely recognized as highly beneficial for overall health and is related to a lower risk of CMM. 3 , 4 The World Health Organization and American Heart Association guidelines recommend 150 minutes or more of moderate‐to‐vigorous‐intensity PA (MVPA) per week, yet these guidelines do not offer specific guidance on how this MVPA should be accumulated. 5 , 6 Notably, recent evidence has indicated that “weekend warrior (WW)” physical activity, as measured by wrist‐worn accelerometers, was associated with a lower risk of 264 future diseases, particularly in cardiometabolic conditions, compared with inactivity. 7 , 8 Despite the known benefits of physical activity using self‐reported data on CMM, it remains unclear whether accelerometer‐derived MVPA concentrated within 1 to 2 days per week, often referred to as the active WW pattern, confers similar benefits in preventing first cardiometabolic diseases (FCMD) and CMM compared with more evenly distributed activity in individuals with hypertension. 9
To address this knowledge gap, we used accelerometer data from the UK Biobank to evaluate whether the weekly distribution of MVPA is associated with cardiometabolic outcomes in participants with hypertension. Specifically, we compared activity concentrated within 1 to 2 days per week with a more evenly distributed pattern of MVPA in relation to incident FCMD (type 2 diabetes, ischemic heart disease, or stroke) and subsequent CMM.
METHODS
Data Availability
The data supporting the findings of the current study's findings are available from the corresponding author upon reasonable request. Primary data from the UK Biobank can be accessed via the official website (https://www.ukbiobbank.ac.uk/) by researchers with approved applications.
Ethics Approval and Consent to Participation
Each UK Biobank participant provided written informed consent. The study protocols received ethical clearance from the Northwest Multi‐Centre Research Ethics Committee, the National Information Governance Board for Health and Social Care in England and Wales, and the Community Health Index Advisory Group in Scotland.
Study Population and Data Source
This research used data from the UK Biobank (application number 211772), a large‐scale prospective cohort that recruited >500 000 individuals aged 40 to 69 years across 22 assessment centers in England, Wales, and Scotland between 2006 and 2010. Baseline assessments involved touchscreen questionnaires, underwent physical and functional measurements, and provided biological samples. 10 In a subsequent substudy, 103 712 participants consented to wear a a wrist‐worn triaxial accelerometer (AX3; Axivity) for 7 consecutive days. 11 These devices captured raw acceleration at 100 Hz within a ±8 g dynamic range.
From this substudy, we initially identified 52228 participants with hypertension, defined by self‐report history, use of antihypertensive medication, hospital records, or baseline blood pressure ≥140/90 mm Hg (details shown in Table S1). 3 After excluding individuals with preexisting cardiometabolic conditions, insufficient data quality, or missing covariates, the final analytical cohort comprised 26 891 participants in Figure S1.
Physical Activity Assessment
Raw accelerometer data were processed using the GGIR R package (https://CRAN.R‐project.org/package=GGIR). MVPA was defined as time accumulated in 5‐second epochs with mean acceleration ≥100 mg, a wrist‐accelerometer threshold corresponding to at least moderate intensity. 12 , 13 To reduce misclassification from spurious signals, we additionally identified MVPA in 5‐minute bouts in which at least 80% of epochs met the MVPA threshold. 14 , 15 A valid day required at least 16 hours of wear time from midnight to midnight. As optimal cut‐points for wrist‐worn accelerometers are not definitively established, 16 we assessed multiple thresholds. In the primary analysis, participants were considered active if they achieved either the guideline‐based threshold of ≥150 min/week of MVPA or the sample median of ≥216.0 min/week. 5 , 6 , 17 , 18 We classified participants as active WW (meeting the MVPA threshold and accumulating ≥50% of total weekly MVPA on 1–2 days), active regular (meeting the MVPA threshold but not WW), or inactive (below the MVPA threshold), consistent with prior studies. 17 , 19 We further tested alternative definitions and thresholds in primary analyses. 7 In secondary analyses, MVPA was additionally estimated using a published machine‐learning approach that identifies multiple activity types, including walking, jogging, stationary cycling, and elliptical training (Supplementary methods for details). 20 Total weekly MVPA derived from the machine‐learning approach and from GGIR were moderately correlated. (Pearson r=0.746; Figure S2).
Ascertainment of Cardiometabolic Diseases
Incident cardiometabolic diseases were identified by linking data from primary care, hospital inpatient, and death registry records, as well as self‐reports using International Classification of Diseases, Tenth Revision (ICD‐10) codes. The specific outcomes were type 2 diabetes: E11, ischemic heart disease: I20–I25, and stroke: I60–I64. Incident FCMD was defined as the first diagnosis of any of 3 specific cardiometabolic disease, and incident CMM signified the coexistence of at least 2 of type 2 diabetes, ischemic heart disease, and stroke. 21 , 22 Detailed information is available in the Table S2. Baseline type 2 diabetes was further confirmed based on self‐reported information of medical history, medication use, and glycated hemoglobin concentration ≥48 mmol/mol. 23 Follow‐up period was defined as the end of accelerometer wear until the first relevant event, death, loss to follow‐up, or censoring (October 31, 2022 for England, August 31, 2022 for Scotland, and May 31, 2022 for Wales), whichever occurred first.
Covariates
Detailed information on covariates is shown in Table S3. The analytical model incorporated the following covariates: age, sex, race, assessment center, smoking status, alcohol intake, body mass index, Townsend deprivation index, shift work, employment status, sleep duration, healthy diet score (Table S4), lipid‐lowering medication use, antihypertensive medication use, and family history.
Statistical Analysis
We employed multivariable‐adjusted Cox proportional hazard regression models, with age as the timescale, to estimate the associations between physical activity patterns and risks of FCMD (type 2 diabetes, ischemic heart disease, and stroke) and CMM. Results are presented as hazard ratios (HRs) with 95% CIs. The proportional hazards assumption was verified using Schoenfeld residuals, with no major violations observed (Table S5). Potential multicollinearity was assessed using variance inflation factors (Table S6).
The “Inactive” group served as the reference for primary comparisons. We also directly compared “WW” versus “Regular” patterns. Multiple comparison bias was addressed by controlling the false discovery rate at 0.05. Cumulative incidence was visualized using Kaplan–Meier survival curves, and we calculated absolute risk reduction and the number of person‐years needed to treat at 5 and 10 years. Restricted cubic splines with 3 knots (ie, 10th, 50th, and 90th percentiles) were applied to explore the nonlinear relationship between total MVPA volume and incident cardiometabolic diseases. The reference values were set to the minimum values of physical activity distribution.
Subgroup analyses were conducted to investigate potential moderation by sex (male versus female), age (<65 versus ≥65 years), and body mass index (<30 versus ≥30 kg/m2). Robustness was further tested through several sensitivity analyses: (1) excluding participants with poor health, underweight (body mass index<18.5 kg/m2), or prevalent cancer at baseline; (2) applying a 2‐year lag period to mitigate potential reverse causality; (3) using Fine–Gray models to account for the competing risk of death; (4) calculating E‐values to assess unmeasured confounding; and (5) adjusting for season of accelerometry wear (Figure S3), sedentary time, and blood pressure. All analyses were performed using R version 4.4.0, with 2‐sided P value <0.05 was considered statistically significant.
RESULTS
Population Characteristics
The present study included 26 891 participants with hypertension, with an average age of 64.5 years (SD 7.1) at the time of accelerometer measurement and 48.2% of them were male. The participants were categorized into 3 patterns: active regular pattern, which accounted for 12.7% (3401 individuals); active WW pattern, which accounted for 34.1% (9182 individuals); and inactive pattern, which constituted the largest proportion at 53.2% (14 308 individuals; Table 1). The distribution of daily MVPA for WW activity and regular activity individuals is graphically depicted in Figure 1. In general, the active WW group exhibits notable higher levels of MVPA on their most active 1 to 2 days compared with the remaining 5 days, whereas the regular activity group shows a more uniform distribution of MVPA.
Table 1.
Baseline Characteristics of Participants
| Characteristics | Overall | Active WW* | Active regular* | Inactive* |
|---|---|---|---|---|
| Participants, n | 26891 | 9182 | 3401 | 14308 |
| Follow‐up, y, mean±SD | 7.5 (1.6) | 7.6 (1.5) | 7.6 (1.5) | 7.4 (1.8) |
| Age at recruitment, y, mean±SD | 58.3 (7.1) | 57.2 (7.1) | 56.8 (7.3) | 59.3 (6.9) |
| Age at accelerometer measurement, y, mean±SD | 64.5 (7.1) | 63.4 (7.2) | 63.0 (7.2) | 65.5 (6.9) |
| Assessment center, n (%) | ||||
| England | 24510 (91.1) | 8393 (91.4) | 3096 (91.0) | 13021 (91.0) |
| Scotland | 1319 (4.9) | 475 (5.2) | 177 (5.2) | 667 (4.7) |
| Wales | 1062 (3.9) | 314 (3.4) | 128 (3.8) | 620 (4.3) |
| Sex, male, n (%) | 12973 (48.2) | 5089 (55.4) | 1755 (51.6) | 6129 (42.8) |
| Race, White, n (%) | 26296 (97.8) | 9010 (98.1) | 3299 (97.0) | 13987 (97.8) |
| Body mass index, kg/m2, mean±SD | 27.4 (4.5) | 26.5 (3.8) | 26.3 (4.0) | 28.2 (4.8) |
| Underweight (<18.5) | 77 (0.3) | 31 (0.3) | 14 (0.4) | 32 (0.2) |
| Normal (18.5–23.9) | 5824 (21.7) | 2370 (25.8) | 996 (29.3) | 2458 (17.2) |
| Overweight (24.0–27.9) | 10769 (40.0) | 4049 (44.1) | 1439 (42.3) | 5281 (36.9) |
| Obese (≥28.0) | 10221 (38.0) | 2732 (29.8) | 952 (28.0) | 6537 (45.7) |
| Townsend deprivation index, mean±SD | −1.9 (2.7) | −2.0 (2.7) | −1.6 (2.8) | −2.0 (2.7) |
| Alcohol consumption, units/wk, mean±SD | 14.4 (17.3) | 15.3 (16.5) | 15.5 (17.8) | 13.6 (17.7) |
| Smoking status, n (%) | ||||
| Never | 15254 (56.7) | 5289 (57.6) | 1945 (57.2) | 8020 (56.1) |
| Previous | 10123 (37.6) | 3463 (37.7) | 1285 (37.8) | 5375 (37.6) |
| Current | 1514 (5.6) | 430 (4.7) | 171 (5.0) | 913 (6.4) |
| Healthy diet score, n (%) | ||||
| Poor (0–1) | 5553 (20.7) | 1834 (20.0) | 670 (19.7) | 3049 (21.3) |
| Reasonable (2, 3) | 17871 (66.5) | 6161 (67.1) | 2228 (65.5) | 9482 (66.3) |
| Good (4) | 3467 (12.9) | 1187 (12.9) | 503 (14.8) | 1777 (12.4) |
| Employed, yes, n (%) | 15317 (57.0) | 5667 (61.7) | 2232 (65.6) | 7418 (51.8) |
| Shift worker, yes, n (%) | 1720 (6.4) | 630 (6.9) | 231 (6.8) | 859 (6.0) |
| Family history†, yes, n (%) | 17697 (65.8) | 5924 (64.5) | 2182 (64.2) | 9591 (67.0) |
| Lipid‐lowering medication use, yes, n (%) | 4018 (14.9) | 1177 (12.8) | 386 (11.3) | 2455 (17.2) |
| Antihypertensive medication use, yes, n (%) | 4824 (17.9) | 1413 (15.4) | 527 (15.5) | 2884 (20.2) |
| Blood pressure | ||||
| Systolic, mm Hg, mean±SD | 149.4 (15.0) | 149.5 (14.4) | 149.0 (14.8) | 149.4 (15.4) |
| ≥140 mm Hg, yes, n (%) | 20412 (79.2) | 7152 (80.7) | 2592 (79.6) | 10668 (78.2) |
| Diastolic, mm Hg, mean±SD | 87.5 (8.9) | 87.8 (8.8) | 87.7 (8.7) | 87.4 (9.0) |
| ≥90 mm Hg, yes, n (%) | 10540 (40.9) | 3719 (42.0) | 1363 (41.9) | 5458 (40.0) |
| Accelerometer‐related variables‡ | ||||
| Total weekly MVPA, min, mean±SD | 182.9 (171.5) | 287.4 (136.7) | 389.4 (212.7) | 66.8 (43.9) |
| Sleep duration, hr/d, mean±SD | 6.4 (1.0) | 6.4 (1.0) | 6.4 (0.9) | 6.3 (1.0) |
Data are presented as mean±SD for continuous variables and as frequency and % for categorical variables. Townsend deprivation index is a composite area‐level measure of deprivation based on unemployment, non‐car ownership, non‐home ownership, and household overcrowding; a higher score indicates higher deprivation.
MVPA indicates moderate‐to‐vigorous‐intensity physical activity; and WW, weekend warrior.
Inactive was defined as MVPA below the guideline‐based threshold of 150 min of MVPA/wk. Active WW was defined as at or above the MVPA threshold and ≥50% of total MVPA over 1–2 d. Active regular was defined as at or above MVPA threshold but not active WW.
Family history of diabetes, heart disease, or stroke.
Data analysis was performed using R‐package GGIR, derived from field 90001.
Figure 1. Distribution of MVPA time on top 2 days vs remaining 5 days among active hypertension individuals using guideline‐based activity threshold of 150 minutes or more of MVPA per week.
Depicted is the distribution of daily MVPA on the 2 most active d of the week (blue), vs the remaining 5 d (red), among hypertension individuals with activity above the guideline‐based threshold (ie, ≥150 min MVPA over the wk). A, Active individuals not meeting criteria for weekend warrior activity (regular) are shown. B, Individuals meeting criteria for weekend warrior activity (ie, ≥50% of total MVPA over 1–2 d) are shown. MVPA indicates moderate‐to‐vigorous‐intensity physical activity.
Association of Physical Activity Patterns With the Risks of Cardiometabolic Diseases
During a median follow‐up period of 7.5 years, 2718 individuals were diagnosed with FCMD (1058 incident type 2 diabetes, 2317 incident ischemic heart disease, 688 incident stroke), and 219 individuals were diagnosed with CMM (Tables S7 and S8). The 5‐year cumulative risks of FCMD (except for stroke) and CMM were substantially lower than those observed among inactive individuals for both active WW and active regular pattern (Figure 2 and Figure S4).
Figure 2. The cumulative risks of incident first cardiometabolic disease and cardiometabolic multimorbidity stratified by activity pattern.
Plots depicting the crude cumulative risks of (A) FCMD, (B) CMM, (C) type 2 diabetes, (D) ischemic heart disease, and (E) stroke events, stratified by accelerometer‐derived activity pattern (inactive, red; active regular, blue; active WW, yellow). The number remaining at risk at each y is depicted below each plot. Data analysis was performed using R‐package GGIR, derived from field 90001. Active WW was defined as ≥150 min of MVPA per wk and ≥50% of total MVPA over 1 to 2 d. CMM was defined as the coexistence of 2 or 3 cardiometabolic diseases, including type 2 diabetes, ischemic heart disease, and stroke. CMM indicates cardiometabolic multimorbidity; FCMD, first cardiometabolic disease; MVPA, moderate‐to‐vigorous‐intensity physical activity; and WW, weekend warrior.
Significant associations were observed between the WW pattern and the risks of FCMD (except for stroke) and CMM. These inverse associations remained consistent across multiple thresholds and definitions (eg, the ≥25th and ≥50th percentile thresholds of MVPA with ≥50% of total MVPA over 1 to 2 days, as well as the guideline‐based threshold with ≥50% of total MVPA over 1 to 2 consecutive days; Table 2 and Table S9). We observed similarly lower risks of incident FCMD (WW versus inactive: HR, 0.75 [95% CI, 0.69–0.82], P=1.4 × 10−3; regular versus inactive: HR, 0.85 [95% CI, 0.74–0.96], P=2.6 × 10−2), type 2 diabetes (WW versus inactive: HR, 0.55 [95% CI, 0.46–0.66], P=1.4 × 10−3; regular versus inactive: HR, 0.74 [95% CI, 0.58–0.95], P=4.5 × 10−2), ischemic heart disease (WW versus inactive: HR, 0.80 [95% CI, 0.72–0.90], P=1.4 × 10−3), and CMM (WW versus inactive: HR, 0.59 [95% CI, 0.41–0.83], P=4.0 × 10−3) based on the definition that WW was as guideline based and ≥50% of total MVPA over 1 to 2 days via GGIR (Table 2). The result of WW versus inactive for CMM did not reach statistical significance. In addition, no P values comparing WW versus regular activity were significant at a false discovery rate of 0.05 (all false discovery rate‐corrected P>0.05; Table 2).
Table 2.
Associations of Accelerometer‐Derived Physical Activity Pattern With Incident Cardiometabolic Diseases in Individuals With Hypertension
| Cardiometabolic diseases | Active regular vs inactive | Active WW vs inactive | Active WW vs active regular | |||
|---|---|---|---|---|---|---|
| Hazard ratio (95% CI)* | P value | Hazard ratio (95% CI)* | P value | Hazard ratio (95% CI)* | P value† | |
| Using GGIR algorithm‡ | ||||||
| FCMD | 0.85 (0.74–0.96) | 2.6 × 10−2 | 0.75 (0.69–0.82) | 1.4 × 10−3 | 0.89 (0.78–1.02) | 1.7 × 10−1 |
| Type 2 diabetes | 0.74 (0.58–0.95) | 4.5 × 10−2 | 0.55 (0.46–0.66) | 1.4 × 10−3 | 0.74 (0.56–0.97) | 1.5 × 10−1 |
| Ischemic heart disease | 0.84 (0.72–0.99) | 7.5 × 10−2 | 0.80 (0.72–0.90) | 1.4 × 10−3 | 0.95 (0.80–1.13) | 6.0 × 10−1 |
| Stroke | 1.14 (0.86–1.50) | 5.3 × 10−1 | 0.92 (0.75–1.13) | 4.6 × 10−1 | 0.81 (0.61–1.09) | 2.0 × 10−1 |
| CMM | 0.97 (0.62–1.50) | 8.8 × 10−1 | 0.59 (0.41–0.83) | 4.0 × 10−3 | 0.61 (0.37–1.00) | 1.5 × 10−1 |
| Using machine‐learning§ | ||||||
| FCMD | 0.81 (0.74–0.89) | 1.8 × 10−3 | 0.78 (0.72–0.84) | 1.3 × 10−3 | 0.96 (0.87–1.05) | 5.8 × 10−1 |
| Type 2 diabetes | 0.56 (0.46–0.67) | 1.8 × 10−3 | 0.64 (0.56–0.74) | 1.3 × 10−3 | 1.16 (0.96–1.41) | 3.9 × 10−1 |
| Ischemic heart disease | 0.91 (0.81–1.03) | 1.8 × 10−1 | 0.81 (0.73–0.90) | 1.3 × 10−3 | 0.89 (0.80–1.00) | 2.2 × 10−1 |
| Stroke | 0.94 (0.75–1.18) | 6.0 × 10−1 | 0.87 (0.72–1.04) | 1.4 × 10−1 | 0.92 (0.74–1.14) | 6.1 × 10−1 |
| CMM | 0.70 (0.61–0.81) | 1.8 × 10−3 | 0.73 (0.65–0.82) | 1.3 × 10−3 | 1.04 (0.90–1.21) | 6.2 × 10−1 |
CMM was defined as the coexistence of 2 or 3 cardiometabolic diseases, including type 2 diabetes, ischemic heart disease, and stroke. Inactive was defined as MVPA below the activity threshold of MVPA; active regular was defined as at or above MVPA threshold but not active WW; active WW defined as guideline‐based (≥150 min MVPA/wk) and ≥50% of total MVPA over 1–2 d.
CMM indicates cardiometabolic multimorbidity; FCMD, first cardiometabolic disease; MVPA, moderate‐to‐vigorous‐intensity physical activity; and WW, weekend warrior.
Hazard ratios (95% CI) were calculated in Cox proportional hazards model. Models were adjusted for age, sex, race, assessment, smoking status, alcohol consumption, body mass index, Townsend deprivation index, shift worker, employed, sleep duration, healthy diet score, lipid‐lowering medication use, antihypertensive medication use, and family history.
No P values comparing WW vs regular activity are significant at the prespecified false discovery rate of 0.05.
Data derived from field 90001. Number of events as follows: FCMD (n=2718), CMM (n=219), type 2 diabetes (n=796), ischemic heart disease (n=1654), and stroke (n=498).
Data derived from field 40045. Number of events as follows: FCMD (n=3326), CMM (n=1548), type 2 diabetes (n=979), ischemic heart disease (n=2016), and stroke (n=609).
Regardless of the thresholds defining the active WW group, active regular and active WW were associated with a lower risk of FCMD, such as type 2 diabetes, compared with the inactive group (Table S9). Specifically, consistent results were observed when MVPA was further classified using machine‐learning method (Table 2 and Tables S9 and S10). However, no significant association was found between physical activity patterns and the risk of stroke regardless of the thresholds (Table 2 and Table S9). We further estimated the shape of the dose–response associations between total weekly of MVPA and the outcomes. Similarly, no significant dose–response associations was observed between total MVPA and the risk of stroke (P for overall=0.23; Figure S5), potentially attributed to the limited number of cases of stroke. Table 3 and Tables S11 and S12 shows estimated 5‐ and 10‐year absolute risk of FCMD and CMM by pattern across multiple thresholds and definitions of active WW.
Table 3.
Five‐ and 10‐Year Risk Estimates by Physical Activity Pattern for FCMD and CMM
| 5‐y risk estimates | 10‐y risk estimates | |||
|---|---|---|---|---|
| Absolute risk reduction (95% CI) | NNT | Absolute risk reduction (95% CI) | NNT | |
| Using GGIR algorithm* | ||||
| FCMD | ||||
| Active regular vs inactive | 2.19 (1.32–3.05) | 46 | 3.19 (2.11–4.27) | 31 |
| Active WW vs inactive | 2.37 (1.75–2.99) | 42 | 3.71 (2.94–4.47) | 27 |
| CMM | ||||
| Active regular vs inactive | 0.32 (0.04–0.60) | 315 | 0.36 (0.04–0.69) | 275 |
| Active WW vs inactive | 0.46 (0.27–0.65) | 218 | 0.61 (0.39–0.83) | 163 |
| Using machine learning† | ||||
| FCMD | ||||
| Active regular vs inactive | 1.67 (0.94–2.39) | 60 | 2.53 (1.64–3.42) | 40 |
| Active WW vs inactive | 1.82 (1.21–2.44) | 55 | 2.64 (1.88–3.40) | 38 |
| CMM | ||||
| Active regular vs inactive | 1.43 (0.93–1.92) | 70 | 2.37 (1.75–2.99) | 42 |
| Active WW vs inactive | 1.33 (0.89–1.76) | 75 | 2.12 (1.58–2.67) | 47 |
CMM was defined as the coexistence of 2 or 3 cardiometabolic diseases, including type 2 diabetes, ischemic heart disease, and stroke. Active WW was defined as guideline‐based (≥150 min MVPA/wk) and ≥50% of total MVPA over 1 to 2 d; inactive was defined as MVPA below the activity threshold of MVPA; active regular was defined as at or above MVPA threshold but not active WW.
CMM indicates cardiometabolic multimorbidity; FCMD, first cardiometabolic disease; MVPA, moderate‐to‐vigorous‐intensity physical activity; NNT, number of person‐years needed to treat; and WW, weekend warrior.
Data derived from field 90001.
Data derived from field 40045.
Subgroup Analysis and Sensitivity Analysis
Subgroup analyses stratified by sex (male versus female), age (<65 versus ≥65 years), and body mass index (<30 versus ≥30 kg/m2) were shown in Tables S13–S18. Two definitions of active WW were used: ≥150 min/week MVPA with ≥50% over 1 to 2 days and ≥50th percentile threshold of MVPA with ≥50% over 1–2 days. Notably, significant sex‐specific interactions with physical activity patterns were observed for the risk of FCMD (P for interaction <0.05; Table S13).
Sensitivity analyses generally concurred with the primary findings (Table S19–S26). The results of FCMD and CMM remained generally consistent with the main analyses when we excluded individuals who had prevalent cancer and when we included the season of accelerometry wear and diastolic blood pressure and systolic blood pressure as additional covariates in the models (Tables S20, S23, and S25). The results of the analysis additionally adjusting for sedentary time did not materially change, although the association between the active regular pattern and the risk of FCMD was attenuated and lost statistical significance in active WW group using alternative definition of the WW pattern (≥50th percentile threshold of MVPA with ≥50% over 1–2 days; Table S24). Similar results were observed when we excluded individuals who had the outcomes within the first 2 years of follow‐up, excluded individuals who were in poor health or underweight (body mass index <18.5 kg/m2), used Fine and Gray models, and restricting hypertension to blood pressure/medication criteria (Tables S19, S22, and S26). In addition, we provided E‐values for all significant associations in Table S27. More than half of all E‐values for significant associations in the main analysis had an HR of >1.50.
DISCUSSION
In this prospective cohort of adults with hypertension, we examined whether the weekly pattern of accelerometer‐measured activity was associated with incident cardiometabolic outcomes. Overall, accumulating sufficient MVPA was associated with lower risks of FCMD and CMM, regardless of whether that activity was concentrated within 1 to 2 days or distributed more evenly across the week. Results were broadly consistent across accelerometer‐processing approaches, alternative WW definitions, and multiple analytic thresholds. We did not observe evidence that stroke risk differed meaningfully between WW and regular activity patterns.
Our findings indicated that engagement in physical activity, both WW and regular patterns may similarly optimize cardiometabolic risk across a broad spectrum of outcomes, with the exception of stroke. Consistent with our results, Dos Santos et al. reported similarly lower rates of all‐cause, cardiovascular disease, and cancer mortality with self‐reported WW versus regular pattern. 24 Importantly, although our primary analysis applied the most common definition of active WW proposed using wrist‐based accelerometers (ie defined as guideline based and ≥50% of total MVPA over 1 or 2 days), which clearly separated individuals with a tendency to be far more active on 1 or 2 days of the week compared with the remaining 5, our observations regarding the risk of FCMD remained consistent when applying up to 6 alternative definitions and thresholds among participants with hypertension. In addition, these findings were consistent when MVPA was further classified using a validated machine‐learning method. These data supported efforts to increase physical activity for cardiometabolic diseases such as type 2 diabetes and ischemic heart disease may be effective in participants with hypertension, even when such efforts are concentrated into 1 to 2 days per week.
The definition of hypertension in our main study was based on a combination of blood pressure measurements, medication use, self‐reported hypertension, and ICD‐10 codes. Although clinical guidelines and research studies may adopt varying hypertension definitions, which could limit the external validity given the cohort composition, we observed similar results to the main analyses when we restricted hypertension to blood pressure/medication criteria only.
WW pattern was associated with lower risks of incident FCMD (except for stroke) and CMM among participants with hypertension during a period of 7.5 years. These observations support the conclusions of previous investigations focused on physical activity concentrated in 1 to 2 days 7 , 17 and further extend prior work reporting improved FCMD and CMM with increasing moderate and vigorous activity derived from questionnaires. 9
Growing studies have reported that the WW pattern accounted for the highest proportion among physical activity patterns. 7 , 17 , 25 , 26 Inconsistent with previous studies, our research found that the active WW pattern appeared uncommon among participants with hypertension. Given that engaging in physical activity concentrated within 1 to 2 days per week may be more feasible for certain contemporary lifestyles, and this pattern demonstrates apparent benefit in reducing the risks of FCMD and CMM, our results advocate for effective management of WW physical activity pattern as a potential strategy in mitigating the risks of FCMD and CMM among participants with hypertension. In addition, the findings of our study were inconsistent with those of a previous study that showed an inverse association between activity pattern and the incidence of stroke in the general population, although associations with stroke were no longer significant when the active WW was defined as the median threshold of MVPA per week. 17 Neither engagement in most MVPA concentrated within 1 to 2 days of the week nor distributing MVPA more evenly throughout the week had a beneficial impact on stroke in individuals with hypertension.
Physiologically, although individuals with hypertension often exhibit impaired vascular function and heightened cardiovascular responses to exercise, evidence suggests that sufficient physical activity volume, regardless of temporal distribution, can induce favorable cardiovascular adaptations. 27 Engagement in MVPA enhances endothelial health by reducing total peripheral resistance and improving nitric oxide bioavailability, processes critical to blood pressure regulation and cardiometabolic health. 28 These mechanisms provide biological plausibility for our observation that adults with hypertension experience reduced cardiometabolic risk by adhering to either a WW or a regular physical activity pattern.
Limitations
Several limitations warrant consideration. First, UK Biobank participants are predominantly of White European ancestry and are generally healthier than the broader population, which may limit generalizability even if exposure–outcome associations are less susceptible to this selection pattern. Second, physical activity assessed over a single 7‐day period and therefore may not fully capture long‐term behavioral patterns; however, prior work suggests that 1‐week accelerometer measures show good reproducibility at the population level 29 , 30 Third, confounders were assessed at the initial baseline visit rather than at the accelerometry assessment, although the analyses supported the stability (Figure S6). Fourth, despite the use of 2 validated approaches to classify MVPA, measurement error may still differ across activity types. Fifth, the limited number of stroke cases might have hindered the primary analyses. Finally, as with all observational studies, causality cannot be established, although the E‐value analyses suggest that substantial unmeasured confounding would be required to fully explain the observed associations.
CONCLUSIONS
Physical activity played comparable roles in developing FCMD and CMM. These results advocate for effective management of WW pattern and regular physical activity pattern as a potential strategy in mitigating the risks of FCMD and CMM in individuals with hypertension.
Sources of Funding
This study is funded by the Scientific Research Project of the Liaoning Province Education Department (grant number JYTMS20230092) and Joint Fund Project General Support Program Project of Liaoning Provincial Department of Science and Technology (grant number 2024‐MSLH‐591).
Disclosures
None.
Supporting information
Tables S1–S27
Figures S1–S6
Supplemental Methods
STROBE Checklist
Acknowledgments
This research has been conducted using the UK Biobank Resource under Application Number 211772. The authors would like to thank all the participants and professionals contributing to the UK Biobank.
This article was sent to Charalambous C. Charalambous, PhD, Assistant Editor, for review by expert referees, editorial decision, and final disposition.
Supplemental Material is available at https://www.ahajournals.org/doi/suppl/10.1161/JAHA.125.045185
For Sources of Funding and Disclosures, see page 9.
References
- 1. Zhang H, Zhou XD, Shapiro MD, Lip GYH, Tilg H, Valenti L, Somers VK, Byrne CD, Targher G, Yang W, et al. Global burden of metabolic diseases, 1990‐2021. Metabolism. 2024;160:155999. doi: 10.1016/j.metabol.2024.155999 [DOI] [PubMed] [Google Scholar]
- 2. Joseph JJ, Rajwani A, Roper D, Zhao S, Kline D, Odei J, Brock G, Echouffo‐Tcheugui JB, Kalyani RR, Bertoni AG, et al. Associations of cardiometabolic multimorbidity with all‐cause and coronary heart disease mortality among black adults in the Jackson Heart Study. JAMA Netw Open. 2022;5:e2238361. doi: 10.1001/jamanetworkopen.2022.38361 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Xie H, Li J, Zhu X, Li J, Yin J, Ma T, Luo Y, He L, Bai Y, Zhang G, et al. Association between healthy lifestyle and the occurrence of cardiometabolic multimorbidity in hypertensive patients: a prospective cohort study of UK biobank. Cardiovasc Diabetol. 2022;21:199. doi: 10.1186/s12933-022-01632-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Khurshid S, Weng LC, Nauffal V, Pirruccello JP, Venn RA, Al‐Alusi MA, Benjamin EJ, Ellinor PT, Lubitz SA. Wearable accelerometer‐derived physical activity and incident disease. NPJ Digit Med. 2022;5:131. doi: 10.1038/s41746-022-00676-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. World Health Organization . Global Recommendations on Physical Activity for Health. World Health Organization; 2010. [PubMed] [Google Scholar]
- 6. Arnett DK, Blumenthal RS, Albert MA, Buroker AB, Goldberger ZD, Hahn EJ, Himmelfarb CD, Khera A, Lloyd‐Jones D, McEvoy JW, et al. 2019 ACC/AHA guideline on the primary prevention of cardiovascular disease: a report of the American College of Cardiology/American Heart Association task force on clinical practice guidelines. Circulation. 2019;140:e596–e646. doi: 10.1161/cir.0000000000000678 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Kany S, Al‐Alusi MA, Rämö JT, Pirruccello JP, Churchill TW, Lubitz SA, Maddah M, Guseh JS, Ellinor PT, Khurshid S. Associations of “weekend warrior” physical activity with incident disease and cardiometabolic health. Circulation. 2024;150:1236–1247. doi: 10.1161/circulationaha.124.068669 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Cao Z, Min J, Xu C. Accelerometer‐derived “weekend warrior” physical activity pattern and incident type 2 diabetes. Cardiovasc Diabetol. 2025;24:130. doi: 10.1186/s12933-025-02676-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Liu BP, Zhu JH, Wan LP, Zhao ZY, Wang X, Jia CX. The impact of physical activity intensity on the dynamic progression of cardiometabolic multimorbidity: prospective cohort study using UK biobank data. JMIR Public Health Surveill. 2023;9:e46991. doi: 10.2196/46991 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Sudlow C, Gallacher J, Allen N, Beral V, Burton P, Danesh J, Downey P, Elliott P, Green J, Landray M, et al. UK Biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age. PLoS Med. 2015;12:e1001779. doi: 10.1371/journal.pmed.1001779 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Doherty A, Jackson D, Hammerla N, Plötz T, Olivier P, Granat MH, White T, van Hees VT, Trenell MI, Owen CG, 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 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Hildebrand M, Van Hees VT, Hansen BH, Ekelund ULF. Age group comparability of raw accelerometer output from wrist‐ and hip‐worn monitors. Med Sci Sports Exerc. 2014;46:1816–1824. doi: 10.1249/mss.0000000000000289 [DOI] [PubMed] [Google Scholar]
- 13. Menai M, van Hees VT, Elbaz A, Kivimaki M, Singh‐Manoux A, Sabia S. Accelerometer assessed moderate‐to‐vigorous physical activity and successful ageing: results from the Whitehall II study. Sci Rep. 2017;8:45772. doi: 10.1038/srep45772 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Khurshid S, Weng LC, Al‐Alusi MA, Halford JL, Haimovich JS, Benjamin EJ, Trinquart L, Ellinor PT, McManus DD, Lubitz SA. Accelerometer‐derived physical activity and risk of atrial fibrillation. Eur Heart J. 2021;42:2472–2483. doi: 10.1093/eurheartj/ehab250 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Rowlands AV, Yates T, Davies M, Khunti K, Edwardson CL. Raw accelerometer data analysis with GGIR R‐package: does accelerometer brand matter? Med Sci Sports Exerc. 2016;48:1935–1941. doi: 10.1249/mss.0000000000000978 [DOI] [PubMed] [Google Scholar]
- 16. Thompson D, Batterham AM, Peacock OJ, Western MJ, Booso R. Feedback from physical activity monitors is not compatible with current recommendations: a recalibration study. Prev Med. 2016;91:389–394. doi: 10.1016/j.ypmed.2016.06.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Khurshid S, Al‐Alusi MA, Churchill TW, Guseh JS, Ellinor PT. Accelerometer‐derived “weekend warrior” physical activity and incident cardiovascular disease. JAMA. 2023;330:247–252. doi: 10.1001/jama.2023.10875 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. National Health Service . Health Survey for England, 2022 Part 1. Accessed February 24, 2025. https://digital.nhs.uk/data‐and‐information/publications/statistical/health‐survey‐for‐england/2022‐part‐1.
- 19. Liu HY, Liu YY, Li QM, Liu L, Wen TJ, Gong TT, Wu QJ, Gao SY. Associations of accelerometer‐derived ‘weekend warrior’ physical activity pattern with incident sarcopenia and falls. Age Ageing. 2026;55:afaf370. doi: 10.1093/ageing/afaf370 [DOI] [PubMed] [Google Scholar]
- 20. Walmsley R, Chan S, Smith‐Byrne K, Ramakrishnan R, Woodward M, Rahimi K, Dwyer T, Bennett D, Doherty A. Reallocation of time between device‐measured movement behaviours and risk of incident cardiovascular disease. Br J Sports Med. 2021;56:1008–1017. doi: 10.1136/bjsports-2021-104050 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Cai Z, Chen G, Zhao W, Wei Z, Wang X, Huang Z, Zheng H, Wu K, Liu Y, Lan Y, et al. Increased risk of cardiometabolic disease in ideal weight adults with history of overweight/obesity in China: a prospective cohort study. J Am Heart Assoc. 2024;13:e033610. doi: 10.1161/jaha.123.033610 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Lu Y, Li G, Ferrari P, Freisling H, Qiao Y, Wu L, Shao L, Ke C. Associations of handgrip strength with morbidity and all‐cause mortality of cardiometabolic multimorbidity. BMC Med. 2022;20:191. doi: 10.1186/s12916-022-02389-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Wang M, Collings PJ, Jang H, Chen Z, Shi Q, Ho HS, Luo S, Au Yeung SL, Kim Y. Prospective associations between muscle strength and genetic susceptibility to type 2 diabetes with incident type 2 diabetes: a UK biobank study. BMC Med. 2025;23:93. doi: 10.1186/s12916-024-03819-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Dos Santos M, Ferrari G, Lee DH, Rey‐López JP, Aune D, Liao B, Huang W, Nie J, Wang Y, Giovannucci E, et al. Association of the “weekend warrior” and other leisure‐time physical activity patterns with all‐cause and cause‐specific mortality: a nationwide cohort study. JAMA Intern Med. 2022;182:840–848. doi: 10.1001/jamainternmed.2022.2488 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Lin F, Lin Y, Chen L, Huang T, Lin T, He J, Lu X, Chen X, Wang Y, Ye Q, et al. Association of physical activity pattern and risk of Parkinson's disease. NPJ Digit Med. 2024;7:137. doi: 10.1038/s41746-024-01135-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Min J, Cao Z, Duan T, Wang Y, Xu C. Accelerometer‐derived ‘weekend warrior’ physical activity pattern and brain health. Nat Aging. 2024;4:1394–1402. doi: 10.1038/s43587-024-00688-y [DOI] [PubMed] [Google Scholar]
- 27. Chang KY, Pham J, Liu Z, Kao Y, Brothers RM, Hwang CL. Blood pressure and vascular function response to maximal graded exercise tests in young men with masked hypertension. Am J Physiol Regul Integr Comp Physiol. 2025;329:R459–R467. doi: 10.1152/ajpregu.00007.2025 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Banks NF, Rogers EM, Stanhewicz AE, Whitaker KM, Jenkins NDM. Resistance exercise lowers blood pressure and improves vascular endothelial function in individuals with elevated blood pressure or stage‐1 hypertension. Am J Physiol Heart Circ Physiol. 2024;326:H256–H269. doi: 10.1152/ajpheart.00386.2023 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Keadle SK, Shiroma EJ, Kamada M, Matthews CE, Harris TB, Lee IM. Reproducibility of accelerometer‐assessed physical activity and sedentary time. Am J Prev Med. 2017;52:541–548. doi: 10.1016/j.amepre.2016.11.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Ahmadi MN, Gill JMR, Stamatakis E. Association of Changes in physical activity and adiposity with mortality and incidence of cardiovascular disease: longitudinal findings from the UK Biobank. Mayo Clin Proc. 2022;97:84. doi: 10.1016/j.mayocp.2021.11.026 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Tables S1–S27
Figures S1–S6
Supplemental Methods
STROBE Checklist
Data Availability Statement
The data supporting the findings of the current study's findings are available from the corresponding author upon reasonable request. Primary data from the UK Biobank can be accessed via the official website (https://www.ukbiobbank.ac.uk/) by researchers with approved applications.