Related Experiment Videos
Hypothesis concerning the U-shaped relation between body mass index and mortality
D B Allison1, M S Faith, M Heo
1Obesity Research Center, St. Luke's/Roosevelt Hospital Center, Columbia University College of Physicians and Surgeons, New York, NY, USA.
American Journal of Epidemiology
|August 15, 1997
Summary
The U-shaped link between body mass index (BMI) and mortality may be influenced by how body composition is measured. Using body composition measures instead of BMI may provide clearer insights into adiposity and mortality risks.
Area of Science:
- Epidemiology
- Public Health
- Biostatistics
Background:
- A U- or J-shaped relationship between Body Mass Index (BMI) and mortality is frequently observed.
- Previous research suggests confounding factors like smoking and pre-existing conditions may explain elevated mortality at lower BMIs.
- This study proposes measurement error, specifically the limitations of BMI in differentiating fat mass from fat-free mass, as an additional explanation.
Purpose of the Study:
- To investigate whether nonspecific measurement error, due to BMI's inability to distinguish fat mass from fat-free mass, can explain the observed U-shaped association between BMI and mortality.
- To demonstrate that even with a hypothetical model where mortality is directly linked to body composition (fat mass and fat-free mass), a U-shaped BMI-mortality curve can still emerge.
Main Methods:
- Utilized body composition data from 1,137 healthy adults.
- Developed a hypothetical BMI-mortality model where the logit of death increased linearly with fat mass and decreased linearly with fat-free mass.
- Simulated scenarios where percentage of fat increased monotonically with BMI.
Main Results:
- The simulation demonstrated that a U-shaped association between BMI and mortality could be recovered even under the specified hypothetical model.
- This suggests that the observed U-shaped curve might arise from the limitations of BMI as a measure of adiposity, rather than solely from confounding factors.
- The differential health impacts of fat mass versus fat-free mass may be obscured by using BMI.
Conclusions:
- The study provides evidence that measurement error related to BMI's composition (fat vs. fat-free mass) can contribute to the U-shaped BMI-mortality association.
- Recommends prioritizing body composition measures over BMI in future epidemiologic studies examining adiposity and mortality.
- Highlights the importance of precise measurement tools for understanding complex health-related associations.