Clustering of ALS in France: results of the BMAALS study.

International audienceBackground:The BMAALS study aims to improve scientific knowledge about the possible link between ALS and cyanotoxin L-BMAA. We report here the results of the geoepidemiological step of this survey in identifying all incident cases on the area under consideration and searching for clusters.Objective:To identify the presence of clusters of ALS among 10 French departments of France (58 254 km 2) covering 5.2 million people followed for 9 years.Methods:Multiple sources of information were used to ensure complete case ascertainment of individuals diagnosed with ALS between 2003 and 2011 according to Airlie House criteria in the following areas of France (i) Limousin region (Corrèze, Creuse, and Haute-Vienne departments); (ii) Rhone –Alpes region (Isere, Savoie, Haute Savoie, Drôme, and Ardèche departments); (iii) Languedoc Roussillon region (Pyrenees Orientales and Herault departments). Multiple sources of case were (i) French national ALS centers; (ii) hospitals and clinics; (iii) health insurance structures; and (iv)private neurologists. Crude incidence per 100,000 inhabitants was assessed using estimates of population from the Insee. Previously applied methods were used for cluster detection(1): Standardized Incidence Ratios (SIR) calculation used the overall incidence in the area as the reference. I Moran and Kulldorf statistics were also performed.Results:Between 1 January 2003 and 31 December 2011, 1211 patients were diagnosed as having probable, probable laboratory supported or definite ALS in the area under study.The average annual crude incidence rate for those cases was 2.57 (95% CI: 2.43 – 2.71) per 100,000 person year of follow-up (PYFU). Standardized on 2010 European population the incidence was 2.50 (95% CI: 2.37 – 2.65) per 100,000 PYFU. Thirteen clusters of ALS were identified within thearea. The median SIR value was 3.3 (95% CI: 1.2 – 7.1).Discussion and conclusion:We report here the widest study of ALS incidence and clustering ever conducted in France using a thorough methodology. We based our calculation onthe incidence assessed in a wide area of France covering 5.2 million inhabitants. The clusters that were identified will be explored for their relation with L-BMAA exposure during the next steps of the BMAALS study