Associations of "Weekend Warrior" Physical Activity With Incident Disease and Cardiometabolic Health

Abstract

Background: Achievement of guideline-recommended levels of physical activity (≥150 minutes of moderate-to-vigorous physical activity per week) is associated with lower risk of adverse cardiovascular events and represents an important public health priority. Although physical activity commonly follows a "weekend warrior" pattern, in which most moderate-to-vigorous physical activity is concentrated in 1 or 2 days rather than spread more evenly across the week (regular), the effects of physical activity pattern across a range of incident diseases, including cardiometabolic conditions, are unknown.

Methods: We tested associations between physical activity pattern and incidence of 678 conditions in 89 573 participants (62±8 years of age; 56% women) of the UK Biobank prospective cohort study who wore an accelerometer for 1 week between June 2013 and December 2015. Models were adjusted for multiple baseline clinical factors, and P value thresholds were corrected for multiplicity.

Results: When compared to inactive (<150 minutes moderate-to-vigorous physical activity/week), both weekend warrior (267 total associations; 264 [99%] with lower disease risk; hazard ratio [HR] range, 0.35-0.89) and regular activity (209 associations; 205 [98%] with lower disease risk; HR range, 0.41-0.88) were broadly associated with lower risk of incident disease. The strongest associations were observed for cardiometabolic conditions such as incident hypertension (weekend warrior: HR, 0.77 [95% CI, 0.73-0.80]; P=1.2×10^-27; regular: HR, 0.72 [95% CI, 0.68-0.77]; P=4.5×10^-28), diabetes (weekend warrior: HR, 0.57 [95% CI, 0.51-0.62]; P=3.9×10^-32; regular: HR, 0.54 [95% CI, 0.48-0.60]; P=8.7×10^-26), obesity (weekend warrior: HR, 0.55 [95% CI, 0.50-0.60]; P=2.4×10^-43, regular: HR, 0.44 [95% CI, 0.40-0.50]; P=9.6×10^-47), and sleep apnea (weekend warrior: HR, 0.57 [95% CI, 0.48-0.69]; P=1.6×10^-9; regular: HR, 0.49 [95% CI, 0.39-0.62]; P=7.4×10^-10). When weekend warrior and regular activity were compared directly, there were no conditions for which effects differed significantly. Observations were similar when activity was thresholded at the sample median (≥230.4 minutes of moderate-to-vigorous physical activity/week).

Conclusions: Achievement of measured physical activity volumes consistent with guideline recommendations is associated with lower risk for >200 diseases, with prominent effects on cardiometabolic conditions. Associations appear similar whether physical activity follows a weekend warrior pattern or is spread more evenly throughout the week.

Keywords: accelerometry; cardiovascular disease; physical activity; prevention; weekend warrior.

Figure 1.

Study design Depicted is an overview of study design. In 89,573 UK Biobank participants providing one week of accelerometer data, we classified physical activity into three patterns: weekend warrior, regular activity, and inactive, using the guideline-based threshold of ≥150 minutes of moderate-to-vigorous physical activity (MVPA)/week,^, and performed association testing with 678 incident diseases. TDI: Townsend Deprivation Index.

Figure 2.

Distribution of MVPA time according to activity pattern Panel A: Distribution of total moderate-to-vigorous physical activity (MVPA) minutes by activity pattern group measured with accelerometer (gray = inactive, blue = weekend warrior, red = regular activity). Panel B: Distribution of MVPA minutes in top 2 days (green) versus remaining 5 days (brown). Upper panel depicts the weekend warrior group and the bottom panel depicts the regular activity pattern group.

Figure 3.

Associations between physical activity pattern and incident disease Incident disease phenome-wide association study comparing people with regular activity versus inactive pattern in the top panel and weekend warrior versus inactive pattern in the bottom panel. In each plot, every point depicts the negative log10 p-value for the association between accelerometer-measured activity group and one incident disease (grouped by category on the x-axis) in Cox proportional hazards models adjusted for adjusted for age at physical activity measurement, sex, ethnic background, tobacco use, Townsend Deprivation Index, alcohol intake, educational attainment, employment status, self-reported health, and diet quality, with darker shaded points meeting a significance level of p<0.01 (corresponding approximately to FDR 0.01, blue dotted line). Upward-facing triangles represent higher risk (hazard ratios >1), while downward-facing triangles represent lower risk (hazard ratio <1). In each plot, the top association (i.e., smallest p-value) for each disease category is annotated. Additional conditions were also annotated if -log10(p) was higher than 15. Some annotation names are shortened for readability. For graphical display purposes, the “other” category is not displayed. The plot is truncated at -log(p) of 20 (with diseases having larger values listed at 20). Full results are available in Supplementary Table B.

Figure 4.

Cumulative risk of cardiometabolic conditions stratified by activity pattern Depicted is the 5-year cumulative incidence of four exemplar cardiometabolic conditions (see label above each plot), stratified by activity pattern (gray = inactive, orange = weekend warrior, green = regular). The number remaining at-risk at each year is depicted below each plot.