From the Division of Infectious Diseases, Department of Medicine, School of Medicine, and Department of Epidemiology, School of Public Health, University of North Carolina at Chapel Hill, Chapel Hill, NC.
Correspondence: William C. Miller, Division of Infectious Diseases, Department of Medicine, School of Medicine, and Department of Epidemiology, School of Public Health, CB#7030, 2146 Bioinformatics, University of North Carolina at Chapel Hill, Chapel Hill, NC. E-mail: firstname.lastname@example.org.
Infectious diseases present particular challenges to epidemiologists. Unlike the “chronic diseases,” infections usually have a single, identifiable etiologic agent. The challenge for most infectious disease studies is not to find the causal agent, but to address the more proximal (and in some cases, more distal) causes of disease.
Proximal factors include those that influence the probability of exposure to the infectious agent (for example, through close human-to-human contact, a vector, or an environmental source). A “cause” in this context reflects the probability of exposure to the pathogen. Furthermore, the strength of an association for a proximal risk factor may vary with the prevalence of the infection. An important risk factor in a high-prevalence environment may be unimportant in a low-prevalence environment, or vice versa.
These proximal causes highlight the dependent nature of most infectious diseases. Infection in one person may directly affect the probability of infection in another person. Pathogens may be transmitted in high-prevalence settings simply because exposure to the pathogen is increased—a phenomenon easily demonstrated in mathematical modeling. These “dependent happenings” require methods that account for the lack of independence among individuals, adding complexity to the study of infectious diseases.
Distal factors arise when exposure to a pathogen does not necessarily progress to disease. This “distal” causal pathway (progression from pathogen exposure to infection to disease) represents the complex interplay between pathogen and host. The organism may possess virulence that increases the probability of disease given exposure. A host may possess innate or acquired immunity. An agent may resist immunity. For example, Neisseria gonorrhoeae has many variant strains, which limits acquired immunity to gonorrhea even with repeated infections.
The host's environment further modifies the interaction between pathogen and host, sometimes affecting both proximal and distal portions of the causal pathway. For example, social factors (such as socioeconomic status, housing, and population density) may influence both the exposure to the infectious agent and the probability of developing disease if exposed. Limited income can lead to crowded housing and thus increased exposure to Mycobacterium tuberculosis. At the same time, low income can impair nutritional status—thus increasing susceptibility to active TB given infection.
Disentangling these proximal and distal causal pathways is one of the great challenges facing infectious disease epidemiologists. In designing studies, researchers must carefully consider where in the causal pathway a potential risk factor is likely to act, as well as the biologic characteristics of the organism. To address these challenges, epidemiologists have begun to develop new tools, sometimes borrowed from other disciplines. An example is the study of social or sexual networks within populations, which can help predict transmission probability and acquisition risk. Similarly, spatial epidemiology can provide insights at the population level that may be missed when the unit of analysis is the individual.
Infectious disease epidemiology has also been strengthened by mathematical modeling. Careful mathematical models can elucidate the spread of epidemics. Traditional deterministic models are useful, but newer methods (including pairwise models, models of network structure, and spatial models) are widening our understanding of disease transmission through populations. Of course, such mathematical models depend heavily on empirical data, and subsequent empirical studies are needed to confirm (or refute) model predictions.
It is because of these complexities that the interface between general epidemiology and infectious disease epidemiology is so important. Each can learn from the other. As Epidemiology's Special Editor for Infectious Diseases, I intend to provide a venue for that creative interface. Towards that end, I encourage infectious disease authors to consider the hypothesized causal pathway for both the infection and disease under evaluation, and to incorporate that information into their study design and analysis. As with other content areas at the journal, state-of-the-art methods are valued. I also encourage the development or adaptation of methods to the unique problems of infectious diseases. Empirical studies and mathematical modeling studies both have essential roles. Of course, we also welcome papers using traditional methods that make significant contributions. With this new emphasis on infectious diseases, my intention is that the challenges of infectious epidemiology will become more widely appreciated by epidemiologists in general—and that infectious disease epidemiology will benefit in the process. I look forward to working with the international epidemiology community to make this a reality.