Image: iStockphoto.com/gorodenkoff
Authors: Walter A Rocca
Editor's Choice
eNeurologicalSci, REVIEW ARTICLE | Volume 38, March 2025, 100539
Open Access
DOI: https://doi.org/10.1016/j.ensci.2024.100544
The practice of modern neurology is based on research evidence. Research evidence is constructed by teams of investigators throughout the world, from Tokyo to Buenos Aires, from Sidney to Vancouver, from New York to Rome. These investigators use research methods and tools to design and conduct their studies. Neuroepidemiology is the science used to conduct clinical studies and population studies and to construct medical research evidence. Thus, neuroepidemiology is the architecture of medical research evidence. In particular, epidemiologic methods are used to describe the frequency and distribution of neurological diseases in human populations, to discover risk and protective factors, to study outcomes of diseases, and to measure the response to treatments.
In the past 40 years, the field of neuroepidemiology has evolved rapidly and is now recognized among practicing neurologists, researchers, public health officials, and policymakers both in higher income countries and in lower income countries. However, for young neurologists and young physicians in many countries outside of North America and Europe, access to clinical research training remains problematic, even in 2024. Young researchers must master research methodology and learn how to apply the methods to collect and analyze original data and to address clinically relevant questions. In a series of three lectures, we provide a brief introduction to neuroepidemiology. The first lecture provides some important definitions and a classification of methods. In addition, it provides a brief introduction to the design of population surveys to measure the burden of neurological disease.
There are three fundamental tools available in neuroepidemiology. In a descriptive study, we are considering the population as an aggregate of individuals and we study the frequency of diseases and the distribution of diseases by time, place, and personal characteristics. The commonly used measures are mortality rates, prevalence percentages, and incidence rates. In analytic or observational studies, we compare two groups of persons to test a causal hypothesis. We are studying the possible association between an independent variable X and a dependent variable Y. In the case-control study, we compare individuals with a disease to individuals free of that disease and look back in their medical history to identify risk or protective factors.
In both types of observational study, we do not have control over the assignment of individuals to the groups. Individuals are exposed or not exposed to a risk factor and develop and do not develop a disease independent of our study. We are only observing but we cannot intervene to change the course of nature. Therefore, case-control studies and cohort studies can be considered experiments of nature (not of investigators).
In an experimental study, we again compare two groups of individuals with or without an independent variable X; however, in this design, we have control over the assignment to the variable (intervention). The classic experimental design is the randomized, double-blinded clinical trial of a new medication.
The key steps involved in designing a population survey, the most common tool in descriptive epidemiology- First need to define the study population. The study population may be simply defined using geographic boundaries (a city, a province, or an entire state) and time boundaries (a single day or a time window). Second, we need to develop diagnostic criteria. The criteria must be clearly specified with inclusion and exclusion rules, so that the study can be replicated in other populations. The third step is to ascertain the cases. This can be done more passively using existing information (e.g., medical records or special registries) or more actively by contacting directly the individuals in the population (direct contact surveys). The ascertainment will be different for a prevalence study in which we count only individuals affected by the disease at one point in time versus an incidence study in which we count the new cases of a disease developing over a time window.
VIEW PDF ⧉