Why population attributable fractions can sum to more than one

Area of Science:

• Epidemiology
• Public Health

Background:

• Population attributable fractions (PAFs) estimate disease prevention by reducing risk factors.
• PAFs for multiple risk factors can sum to over 100%, an illogical result.

Purpose of the Study:

• To explain why population attributable fractions (PAFs) can sum to more than 1.
• To demonstrate the limitations of PAFs in scenarios with multiple risk factors.

Main Methods:

• Analysis of a hypothetical disease model with sequential risk factor elimination.
• Examination of how overlapping risk factor impacts contribute to PAF summation.

Main Results:

• PAFs can exceed 100% because individuals with multiple risk factors are counted in multiple prevention scenarios.
• PAF estimates assume mutually exclusive risk factor elimination, which is often unrealistic.
• Sequential attributable fractions (SAFs) provide upper and lower bounds for prevented cases.

Conclusions:

• Clearer understanding of PAF assumptions is needed to avoid overestimating disease burden reduction.
• Utilizing SAF limits or multivariable PAFs can yield more realistic estimates.
• Accurate burden estimation is crucial for effective public health resource allocation.