Estimating Effect of Obesity on Stroke Using G-Estimation: The ARIC study.

Area of Science:

• Epidemiology
• Cardiovascular Health
• Biostatistics

Background:

• Obesity is a significant risk factor for stroke.
• Previous studies may have underestimated the obesity-stroke relationship due to time-varying confounders.

Purpose of the Study:

• To accurately quantify the relationship between general obesity (GOB) and abdominal obesity (AOB) and stroke risk.
• To adjust for time-varying confounders using G-estimation for a more precise risk assessment.

Main Methods:

• Utilized data from 13,975 participants in the Atherosclerosis Risk in Communities (ARIC) study.
• Defined GOB (BMI ≥ 30 kg/m²) and AOB (waist circumference/waist-to-hip ratio).
• Employed G-estimation and compared results with accelerated failure time models.

Main Results:

• G-estimated hazard ratios (HRs) for GOB and AOB were significantly elevated (e.g., HR 1.60 for GOB, 1.99 for AOB by waist-to-hip ratio).
• Standard models, after adjusting for mediators, showed HRs including the null value.
• G-estimation yielded larger risk estimates, suggesting standard models were biased toward the null.

Conclusions:

• Both general and abdominal obesity are independently associated with increased stroke risk.
• G-estimation provides more accurate estimates by properly handling time-varying confounders.
• Findings highlight the importance of considering obesity in stroke prevention strategies.