Time spent exercising and obesity: An application of Lewbel's instrumental variables method

Area of Science:

• Public Health
• Epidemiology
• Exercise Science

Background:

• Understanding the relationship between physical activity and body mass is crucial for public health initiatives.
• Previous studies may have underestimated the impact of physical activity due to measurement errors in time-use data.
• The potential for joint determination between physical activity and body mass requires advanced statistical methods.

Purpose of the Study:

• To investigate the precise role of physical activity in determining body mass using the American Time Use Survey.
• To address and correct for measurement error in single-day time-use data versus averaged daily time use.
• To explore the bidirectional relationship between physical activity and body mass.

Main Methods:

• Utilized data from the American Time Use Survey.
• Employed Lewbel's instrumental variables estimator to account for endogeneity between physical activity and body mass.
• Exploited first-stage heteroskedasticity for identification, avoiding traditional exclusion restrictions.

Main Results:

• Averaging 30 minutes of transportation-related physical activity (biking or walking) per day reduced men's Body Mass Index (BMI) by 1.5.
• No significant effect of physically active leisure on men's BMI was observed in the sample.
• Thirty minutes of daily physical activity, regardless of type (transportation or leisure), reduced women's BMI by 1.

Conclusions:

• Transportation-related physical activity is an effective strategy for weight management in men.
• Physical activity demonstrates a significant impact on reducing BMI for women, irrespective of activity type.
• Accurate measurement of physical activity, accounting for daily variations, is essential for understanding its true effect on body mass.