Image by Author

When a community has limited resources to prevent disease, knowing which risk factors matter most makes the difference. Population Attributable Fraction (PAF) tells you what percentage of disease cases in a population could be prevented if you eliminated a specific risk factor. This helps public health professionals prioritize interventions by quantifying the impact of removing exposures like smoking, poor diet, or lack of exercise.

PAF answers a simple question: If we could eliminate this risk factor entirely, how much disease would we prevent?

Understanding the Components

PAF requires two pieces of information about your population:

The prevalence of exposure represents the proportion of people exposed to the risk factor. For example, if 25% of adults smoke, your prevalence is 0.25.

The relative risk measures how much more likely exposed individuals are to develop the disease compared to unexposed individuals. A relative risk of 2.5 means smokers are 2.5 times more likely to develop the disease than non-smokers.

Note that this basic approach treats exposure as binary. You’re either exposed or not. Risk factors with multiple levels (like light, moderate, and heavy smoking) require modified calculations that account for each exposure category.

The PAF Formula

The basic formula for PAF is:

Where P is the prevalence of exposure and RR is the relative risk. 

This formula assumes the relationship between exposure and disease is causal and that removing the exposure would eliminate its associated risk.

Calculating PAF: A Worked Example

Let’s say 30% of adults in a population have obesity, and research shows that obese individuals have 2.0 times the risk of developing type 2 diabetes compared to non-obese individuals.

First, express your prevalence as a decimal. Here, 30% becomes P = 0.30.

Next, identify your relative risk from epidemiological studies. For this example, RR = 2.0 (obese individuals have twice the risk of non-obese individuals).

Now plug these values into the formula:

• Calculate the numerator: 0.30 × (2.0 − 1) = 0.30 × 1.0 = 0.30
• Calculate the denominator: 1 + 0.30 × (2.0 − 1) = 1 + 0.30 = 1.30
• Divide to get PAF: 0.30 / 1.30 = 0.231
• Convert to a percentage: 0.231 × 100 = 23.1%

This means that if obesity were completely eliminated from this population, about 23% of type 2 diabetes cases would be prevented. In a city of 100,000 adults with 5,000 diabetes cases, eliminating obesity would prevent about 1,150 of those cases.

Interpreting Your Results

A PAF of 23% means that 23% of all diabetes cases in this population are attributable to obesity under the causal assumptions described above. If you eliminated obesity, you’d prevent roughly 1 in 4 diabetes cases. Higher PAF values indicate that the risk factor accounts for a larger burden of disease. A PAF of 60% suggests that eliminating the exposure would prevent more than half of all cases, making it a high-priority intervention target.

PAF values depend on both prevalence and relative risk. A risk factor with moderate relative risk but high prevalence can have a larger PAF than a factor with high relative risk but low prevalence. This is why addressing common exposures often has greater population impact than targeting rare ones.

Important Limitations

PAF is a useful metric, but it comes with important caveats. PAF calculations assume that the association between exposure and disease is causal. If confounding factors explain the relationship, your PAF will overestimate the preventable burden.

The formula assumes that removing the exposure would immediately eliminate its associated risk. In reality, some diseases have latency periods, and past exposures may continue to affect risk.

PAF doesn’t account for competing risks. If one disease is prevented, individuals may develop different conditions instead.

Multiple risk factors for the same disease can have overlapping effects. Simply adding individual PAFs often exceeds 100%, which is impossible. When risk factors interact or share common pathways, you’ll need more complex calculations to estimate combined effects.

Conclusion

PAF gives you a practical way to estimate the population-level impact of eliminating risk factors. It works best when you’re comparing different interventions or deciding where to allocate prevention resources. By combining prevalence and relative risk data, you can identify which exposures contribute most to disease burden. Understanding PAF helps translate epidemiological findings into public health strategies, though you should consider its assumptions and limitations when making decisions.