Farid Boumédiene1, Benoît Marin1, Jaime Luna1,2, Vincent Bonneterre3, William Camu4, Emmeline Lagrange5, Gérard Besson5, Florence Esselin4, Elisa De La Cruz4, Géraldine Lautrette2, Pierre Marie Preux1,6, Philippe Couratier7,8. 1. Inserm U1094, IRD U270, USC1501 INRAE, Univ. Limoges, CHU Limoges, EpiMaCT - Epidemiology of chronic diseases in tropical zone, Institute of Epidemiology and Tropical Neurology, OmegaHealth, Limoges, France. 2. Department of Neurology, Centre de Reference SLA et Autres Maladies du Neurone Moteur, CHU Limoges, Limoges, France. 3. University Grenoble Alpes, CNRS, Grenoble INP, TIMC, 38000, Grenoble, France. 4. Explorations Neurologiques et Centre SLA, CHU et Université de Montpellier, INSERM, Montpellier, France. 5. Department of Neurology, CHU Grenoble-Alpes (Grenoble Teaching Hospital), Grenoble, France. 6. CEBIMER, Centre d'Epidémiologie, de Biostatistique et de Méthodologie de la Recherche, CHU Limoges, Limoges, France. 7. Inserm U1094, IRD U270, USC1501 INRAE, Univ. Limoges, CHU Limoges, EpiMaCT - Epidemiology of chronic diseases in tropical zone, Institute of Epidemiology and Tropical Neurology, OmegaHealth, Limoges, France. philippe.couratier@unilim.fr. 8. Department of Neurology, Centre de Reference SLA et Autres Maladies du Neurone Moteur, CHU Limoges, Limoges, France. philippe.couratier@unilim.fr.
Abstract
OBJECTIVE: To assess spatial aggregates of amyotrophic lateral sclerosis (ALS) incident cases, using a solid geo-epidemiological statistical method, in France. METHODS: This population-based study (2003-2011) investigated 47.1 million person-years of follow-up (PYFU). Case ascertainment of incident ALS cases was based on multiple sources (ALS referral centers, hospital centres and health insurance data). Neurologists confirmed all ALS diagnoses. Exhaustiveness was estimated through capture-recapture. Aggregates were investigated in four steps: (a) geographical modelling (standardized incidence ratio (SIR) calculation), (b) analysis of the spatial distribution of incidence (Phothoff-Winttinghill's test, Global Moran's Index, Kulldorf's spatial scan statistic, Local Moran's Index), (c) classification of the level of certainty of spatial aggregates (i.e. definite cluster; probable over-incidence area; possible over-incidence area) and (d) evaluation of the robustness of the results. RESULTS: The standardized incidence of ALS was 2.46/100,000 PYFU (95% CI 2.31-2.63, European population as reference) based on 1199 incident cases. We identified 13 areas of spatial aggregates: one cluster (stable in robustness analysis), five probable over-incidence areas (2 stable in robustness analysis) and seven possible over-incidence areas (including 4 stable areas in robustness analysis). A cluster was identified in the Rhône-Alpes region: 100 observed vs 54.07 expected cases for 2,411,514 PYFU, SIR: 1.85 (95% CI 1.50-2.25). CONCLUSION: We report here one of the largest investigations of incidence and spatial aggregation of ALS ever performed in a western country. Using a solid methodology framework for case ascertainment and cluster analysis, we identified 13 areas that warrant further investigation.
OBJECTIVE: To assess spatial aggregates of amyotrophic lateral sclerosis (ALS) incident cases, using a solid geo-epidemiological statistical method, in France. METHODS: This population-based study (2003-2011) investigated 47.1 million person-years of follow-up (PYFU). Case ascertainment of incident ALS cases was based on multiple sources (ALS referral centers, hospital centres and health insurance data). Neurologists confirmed all ALS diagnoses. Exhaustiveness was estimated through capture-recapture. Aggregates were investigated in four steps: (a) geographical modelling (standardized incidence ratio (SIR) calculation), (b) analysis of the spatial distribution of incidence (Phothoff-Winttinghill's test, Global Moran's Index, Kulldorf's spatial scan statistic, Local Moran's Index), (c) classification of the level of certainty of spatial aggregates (i.e. definite cluster; probable over-incidence area; possible over-incidence area) and (d) evaluation of the robustness of the results. RESULTS: The standardized incidence of ALS was 2.46/100,000 PYFU (95% CI 2.31-2.63, European population as reference) based on 1199 incident cases. We identified 13 areas of spatial aggregates: one cluster (stable in robustness analysis), five probable over-incidence areas (2 stable in robustness analysis) and seven possible over-incidence areas (including 4 stable areas in robustness analysis). A cluster was identified in the Rhône-Alpes region: 100 observed vs 54.07 expected cases for 2,411,514 PYFU, SIR: 1.85 (95% CI 1.50-2.25). CONCLUSION: We report here one of the largest investigations of incidence and spatial aggregation of ALS ever performed in a western country. Using a solid methodology framework for case ascertainment and cluster analysis, we identified 13 areas that warrant further investigation.
Authors: Chris C Plato; Ralph M Garruto; Douglas Galasko; Ulla-Katrina Craig; Meropi Plato; Anthony Gamst; Jose M Torres; Wigbert Wiederholt Journal: Am J Epidemiol Date: 2003-01-15 Impact factor: 4.897