The use of standardized incidence and prevalence rates in epidemiological studies on multiple sclerosis. A meta-analysis study

Objective: To demonstrate whether or not the age and sex adjustment of incidence and prevalence rates in multiple sclerosis (MS) could allow more reliable comparison between epidemiological studies performed in different areas of the world and to establish if the latitude gradient theory could be confirmed after the standardization for age and sex distribution.

Methods: A meta-analysis of population-based incidence and prevalence studies on MS from 1980 through 1998 using the terms 'multiple sclerosis', 'prevalence' and 'incidence' in the bibliographic databases MEDLINE and EMBASE was performed. We included studies that reported the diagnostic criteria, number of cases and the population studied, the date of the study, the latitude, and the age- and sex-specific crude incidence and prevalence rates. According to the inclusion criteria, 69 of 127 papers on prevalence and 22 of 70 papers on incidence were considered for age adjustment and 27 prevalence and 8 incidence studies for sex adjustment. The mean incidence and prevalencerates and the 95% confidence intervals age- and sex-adjusted to the World and the European standard populations were calculated.

Results: The Spearman rank correlation and the multiple regression analyses indicated that age adjustment to standard populations could overcome the limitations in comparing the crude prevalence and incidence rates of different epidemiological studies on MS. When the mean crude and age- and sex-adjusted prevalence and age-adjusted incidence rates were stratified by latitude (from south to north), the latitudinal gradient, which was highly significant for the crude rates, became less remarkable for the age- and sex-adjusted prevalence rates and not significant for the age-adjusted incidence rates.

Conclusions: The crude incidence and prevalence rates in epidemiological studies on MS should be age- and sex-adjusted to a common standard population to permit a more reliable comparison among studies performed in different countries. Our findings support the opinion that the latitude does not play a key role in determining the onset of MS. Whenever possible, the crude incidence and prevalence rates should be adjusted to the ethnic origin and migration characteristics.