Examining associations between physical activity and cardiovascular mortality using negative control outcomes

Area of Science:

• Epidemiology
• Cardiovascular Health
• Public Health

Background:

• Observational studies indicate physical inactivity is a significant risk factor for cardiovascular disease (CVD).
• Inferring causality is challenging due to potential biases like reverse causation and residual confounding.
• Negative control outcomes can help isolate the true association between physical activity and CVD mortality.

Purpose of the Study:

• To utilize a negative control outcome to assess confounding in the association between physical activity and CVD mortality.
• To determine the extent to which biases and confounders explain the observed link between physical activity and CVD mortality.

Main Methods:

• A large sample of 104,851 participants from English and Scottish health surveys were followed for approximately 9.4 years.
• Cardiovascular disease (CVD) mortality and accidental deaths (negative control outcome) were recorded.
• Hazard ratios (HR) and 95% confidence intervals (CI) were calculated for the association between physical activity and mortality outcomes.

Main Results:

• Accidental death, used as a negative control, shared similar confounding structures (age, sex, smoking, illness) with CVD mortality.
• Physical activity showed an inverse association with CVD mortality (HR=0.75, 95% CI: 0.70, 0.80).
• The association between physical activity and accidental death was in the same direction but less pronounced (HR=0.86, 95% CI: 0.69, 1.07), with no linear dose-response.

Conclusions:

• The inverse association between physical activity and CVD mortality is likely causal but may be overestimated in previous studies.
• Negative control methodologies offer a valuable approach to enhance causal inference in physical activity research.
• Further research using negative controls can refine our understanding of physical activity's true impact on health outcomes.