To evaluate the association between worldwide ALS incidence rates and age, using a dose-response meta-analysis. We reviewed Medline and Embase up to July 2016 and included all population-based studies of newly-diagnosed cases, using multiple sources for case ascertainment. A dose-response meta-analysis was performed. A meta-regression investigated potential sources of heterogeneity. Of 3254 articles identified in the literature, we included 41 incidence studies covering 42 geographical areas. Overall, the fit between observed and predicted age-specific rates was very good. The expected variation of ALS incidence with age was characterized, in each study, by a progressive increase in the incidence from the 40s leading to a peak in the 60s or 70s, followed by a sharp decrease. Cochran's Q test suggested a significant heterogeneity between studies. Overall, estimated patterns of ALS age-specific incidence (at which the peak was reached) were similar among subcontinents of Europe and North America: peak of ALS incidence ranged in these areas between 6.98 and 8.17/100,000 PYFU, which referred to age in the range 71.6-77.4 years. The relationship between age and ALS incidence appeared different for Eastern Asia which was characterized by a peak of ALS incidence at 2.20/100,000 PYFU around 75 years of age. This study confirms the consistency of the age-specific ALS incidence pattern within different subcontinents. Age-specific incidence appears lower in Eastern Asia as compared to Europe and North America.
Keywords: Age-specific; Amyotrophic lateral sclerosis; Epidemiology; Ethnic groups; Incidence.