Conference Coverage

How Accurate Are International Prevalence Estimates for Neurologic Conditions?


SAN DIEGO—Epidemiologic studies of neurologic conditions have inconsistent methodologies that result in differing estimates of disease prevalence, according to research presented at the 65th Annual Meeting of the American Academy of Neurology. Countries of similar socioeconomic status arrive at differing prevalence estimates for a given condition because of these methodologic variations, which include age categories, ascertainment methods, and diagnostic criteria.

More than 90% of the variation in point estimates of most neurologic conditions resulted from study heterogeneity, said Nathalie Jette, MD, Associate Professor of Neurology at the University of Calgary in Canada. Study heterogeneity varied from a low of 67% for Duchenne muscular dystrophy, 84% for amyotrophic lateral sclerosis (ALS), and 90% or more for all other neurologic conditions studied.

A Systematic Review and Meta-Analysis of Epidemiologic Studies
Dr. Jette and colleagues systematically reviewed population-based studies of the prevalence of 15 neurologic conditions, including Alzheimer’s disease, ALS, epilepsy, multiple sclerosis (MS), Parkinson’s disease, and traumatic brain injury (TBI). Two investigators independently reviewed abstracts, full articles, and abstracted data. Duplicate records were removed. Data abstraction included study location, dates of data collection, sources of case ascertainment, diagnostic criteria, and estimates of incidence and prevalence.

Nathalie Jette, MD

The researchers evaluated study heterogeneity descriptively and used software to calculate the proportion of variation in point estimates that resulted from heterogeneity between studies, rather than from sampling error. Heterogeneity was expressed in percentages.

As part of a meta-regression analysis, the investigators compared prevalence estimates for dementia, epilepsy, and MS based on four data sources (door-to-door visits, hospital or clinic assessments, telephone or mail surveys, and other sources). To allow for comparison between conditions, the team calculated prevalence ratios.

Data Sources Affected Prevalence Estimates
Overall, prevalence studies were performed more often than incidence studies, but more than 90% of the investigations of spinal cord injury and TBI were incidence studies. “Depending on the condition, certain data sources were significantly associated with the prevalence estimate,” said Dr. Jette. None of the data sources appeared to have a consistent association with higher or lower prevalence estimates across the three conditions, however.

“There is a need to establish widely accepted standards for epidemiologic research on neurologic conditions, including recommendations regarding common data elements that should be collected, suggested methods of ascertainment, and reporting of results,” concluded Dr. Jette.

—Erik Greb
Senior Associate Editor

Suggested Reading
Centers for Disease Control and Prevention (CDC). Epilepsy in adults and access to care—United States, 2010. MMWR Morb Mortal Wkly Rep. 2012;61(45):909-913.

Centers for Disease Control and Prevention (CDC). Nonfatal traumatic brain injuries from sports and recreation activities—United States, 2001–2005. MMWR Morb Mortal Wkly Rep. 2007;56(29):733-737.

Next Article: