The Limitations of ‘Proximity’ in Epidemiology Studies

Uni Blake
Posted August 11, 2023

With Pennsylvania expected to release a pair of important health studies next week, it’s essential to understand the proper use of “proximity” or nearness as researchers try to identify potential links between various exposures and health outcomes in epidemiology.

One common approach is to gauge exposure based on proximity and density to potential sources.

It sounds simple – the closer you are to a source, the higher your exposure, right? Yet, while using proximity as a surrogate for exposure can be helpful, it has its limitations and there are potential pitfalls in solely relying on it. These include:

Assuming homogeneity – Proximity does not guarantee uniform exposure levels. Just because someone is physically close to a potential source of exposure doesn't mean all individuals are equally exposed. Various factors, such as duration of exposure and a person’s activities, vary significantly.  Factors like occupation and spatial variability (due to wind patterns and topographical features) can all play a significant role in determining exposure levels.

Misclassifying exposure – Relying solely on proximity can lead to exposure misclassification. Being close to a potential source does not automatically translate into being exposed. On the flip side, individuals further away from a source might have had greater exposure due to factors like wind patterns or environmental contamination.

Temporal uncertainty – Timing matters. Proximity may not account for the temporal aspect of exposure accurately. Someone might have been close to a source in the past, but exposure could have ceased long before the study began, leading to inaccurate conclusions about their current health outcomes.

Reverse causality – Correlation is not causation. Proximity does not always mean a causal relationship between the source and the health outcome. For example, individuals with health conditions may be more likely to stay close to healthcare facilities, creating a false association between proximity and the disease.

Confounding factors – Proximity is often linked to other factors influencing the study's outcomes. For example, socioeconomic status and age can be associated with living close to specific sources. Failing to account for these confounding variables, among others, may skew results.

The limitations of proximity studies serve as a reminder of the complex nature of exposure-outcome investigations. These studies, inherently by design, capture only a fraction of the complexity at play. While anchored in the premise that proximity implies exposure, the studies could oversimplify reality. As we anticipate the release of the Pennsylvania research, it’s important to recognize that proximity studies require scrutiny and supplementary research to comprehensively understand the interplay between exposure, health outcomes, and the complex factors involved.

So, what is the takeaway? Neglecting to consider these complexities in interpreting the studies could lead to misinterpretations and incorrect policy decisions.