In a recent article published in the British Journal of Sports Medicine, researchers explored the relationship between non-occupational physical activity (PA) levels and poor cardiovascular diseases (CVDs) and cancer outcomes, including mortality.
Review: Non-occupational physical activity and risk of cardiovascular disease, cancer and mortality outcomes: a dose–response meta-analysis of large prospective studies. Image Credit: Africa Studio / Shutterstock
Cardiovascular diseases (CVDs) and cancers are both leading causes of death globally. Since studies have associated low PA levels with an increased risk of death due to CVDs and site-specific cancers, an accurate estimation of their cause-and-effect relationship is crucial.
Several previous meta-analyses have pursued quality evidence for these associations; however, using varying methodologies that limited the comparability of relative risks across different outcomes. The Global Burden of Disease (GBD) study estimated total PA even from one's occupation to show its impact at the population level for five diseases. Even when occupational PA is measured properly, it remains unexplained whether it has comparable health benefits as non-occupational PA.
This highlights the significance of prevalence estimates for PA levels and the approximation of the dose-response relationship between PA and CVD and cancer outcomes. Modeling studies have confirmed how crucial it is to determine the shape of the dose–response association, which, in turn, is needed to inform public health policies directed toward increasing PA at the population level.
About the study
The present study used a novel harmonization framework to compute the dose-response relationship between non-occupational PA levels and CVD and cancer and all-cause mortality outcomes.
It helped them overcome challenges posed by conventional PA measurement methods. More importantly, it facilitated comparing studies quantifying and reporting PA using varying methods on the activity exposure scale of the marginal metabolic equivalent of task (MET)-hours per week (mMET-hours/week).
The researchers screened Scopus, PubMed, and Web of Science databases to find prospective cohort studies that reported PA levels of adults (≥18 years) with no pre-existing conditions in a minimum of three ordinal exposure levels to report risk estimates for all examined outcomes.
Broadly, there were three outcomes of interest, as follows:
i) cumulative mortality separately for all causes, CVDs and cancer;
ii) Cumulative incidence of all fatal and non-fatal CVDs
iii) Cumulative incidence of all fatal and non-fatal site-specific cancers, e.g., rectal cancer, selected based on prespecified criteria.
Like in all meta-analyses, one reviewer screened each included study, followed by a double check by a second reviewer who worked independently. The team extracted all relevant data from included studies, most importantly, PA exposure and outcome assessments. They extracted information for each PA exposure category to quantify PA volume, the number of cases and participants, plus person-years of follow-up, along with risk estimates.
The researchers used the most adjusted model for risk estimates. They extracted PA exposure data as precisely as documented in the published studies. Finally, they harmonized data on non-occupational PA volume to a common metric, mMET-hours/week, which correctly equated PA volume to time spent at varying PA intensity levels.
All the studies that met the inclusion criteria examined leisure-time PA, alone or in combination with specific types of activity. According to the authors, they so extensively harmonized PA exposure data that it fetched a more extensive evidence base for 17 of 22 health outcomes. Moreover, it is the first meta-analysis to estimate dose–response associations for nine site-specific cancers.
The study results suggested that small increases of up to 17.5 mMET-hours/week in non-occupational PA could be very beneficial. At the population level, 150 min/week of moderate-to-vigorous aerobic PA could help reduce 16% of all premature deaths. It implied that the current PA recommendations of 150 to 300 min/week of moderate aerobic PA, equating to 75 to 150 min/week of vigorous aerobic PA, could fetch nearly maximal benefits. However, the dose–response associations demonstrated that half the recommended PA levels could also help gain appreciable health benefits. Thus, public health messaging should emphasize less on threshold-based recommendations for PA less.
The researchers also observed that all-cause and cancer mortality were similarly associated with recommended levels of moderate-to-vigorous aerobic PA. On the contrary, CVD mortality had stronger associations with narrower CIs for the same PA levels. Likewise, associations were stronger for some cancers, e.g., head & neck cancers and myeloid leukemia, whereas the same for kidney cancer was insignificant. Intriguingly, previous studies focused solely on leisure-time PA to show its harmful effects on prostate cancer, while the current study results showed no such association.
Compared to the GBD study, which assessed approximately 130 MET-hours/week for total PA, this study assessing 17.5 mMET-hours/week for non-occupational PA showed its association with heart disease and stroke was stronger. Overall, the current study results appeared more relevant from a public health perspective.
Overall, the current study showed the dose-dependent associations between increasing non-occupational PA and all-cause, CVD, and cancer mortalities. It appeared to be the strongest for all-cause and CVD mortality and weaker with cancer, depending on the cancer site.
Nonetheless, these inverse non-linear dose–response associations highlighted the significance of non-occupational PA in inactive adults. Even in small bouts, some PA rather than no PA could provide substantial protection against many chronic diseases, including some site-specific cancers and CVDs.
- Garcia L, Pearce M, Abbas A, et al., Non-occupational physical activity and risk of cardiovascular disease, cancer and mortality outcomes: a dose–response meta-analysis of large prospective studies, British Journal of Sports Medicine Published Online First: 28 February 2023. DOI: 10.1136/bjsports-2022-105669, https://bjsm.bmj.com/content/early/2023/01/23/bjsports-2022-105669