Tion has been reported in several eco-climatic areas: arid in Western Africa and Arabic Peninsula [1,6], sub-humid in Eastern Africa [7,8], wet forests in central Africa [5], dam and irrigated agricultural land under hot climatic conditions in Egypt, Mauritania and Sudan [9?1] and recently humid highlands in Madagascar [3,12]. The SCR7 web respective roles of direct and vectorial transmissions remain unevaluated in both human and cattle and probably vary among these eco-climatic areas. Madagascar experienced two major Rift Mikamycin IA chemical information Valley fever (RVF) outbreaks: 1990?1 in the eastern-coast and central highlands and 2008?9 in the south, the north and the highlands [13?15]. The last outbreaks occurred in two epidemic waves during the two successive rainy seasons of 2007?8 and 2008?9. Following the first wave, passive surveillance and emergency response were developed. Sentinel surveillance in herds were set up with field veterinarians [16]. This sentinel surveillance allowed the early detection of the second wave of outbreak inPLOS Neglected Tropical Diseases | DOI:10.1371/journal.pntd.July 14,2 /Rift Valley Fever Risk Factors in Madagascarcattle at the end of 2008 and thus the implementation of local control measures to prevent the spreading of RVF outside the region [15,16]. At the end of the epidemic, about 700 suspected human cases were recorded from which 26 were fatal. About 400 human and cattle samples were received for laboratory analyses and RVF infection was confirmed or considered as probable in 86 human and 46 ruminant samples [15]. Following the 2008?9 epidemics, studies showed a wide and heterogeneous spread of RVFV infection both in human and cattle [15,17] suggesting that some areas were more favorable than others to transmission [17]. Madagascar has a large variety of eco-climatic patterns, including semi-arid in the south, tropical in the west and on the eastern-coast, and temperate in the central highlands [18]. Apart from the highlands [3,12,19], RVF epidemiology is poorly understood in this country [15,17]. Since 2007, a human syndromic-based surveillance system has been developed which has allowed the detection of the first case of RVF in humans in 2008 [15]. Besides, retrospective investigations suggested that RVFV circulated among livestock since December 2007 [15], revealing a dearth in veterinary surveillance. The main difficulty to implement veterinary surveillance in Madagascar is the lack of basic means to collect and communicate veterinary information [20]. Thus, the identification of at-risk environments is essential to optimize the available resources by targeting RVF surveillance. In addition, there is a need to provide insight into the role of the two transmission routes and better adapt available control measures. Herein, the objectives of our study were: (i) to identify the environmental factors and areas favorable to RVFV transmission to both cattle and human and (ii) to identify human behaviors favoring human infections in Malagasy contexts.Materials and MethodsTo achieve these goals, we characterized the environments of Malagasy communes using a Multiple Factor Analysis (MFA). Then we analyzed cattle and human serological data using a Generalized Linear Mixed Models (GLMMs), with the individual serological status (cattle or human) as the response, and MFA factors, as well as potential other risk factors (covariates), as explanatory variables.Cattle and human datasetsThe cattle dataset contained results of a national cr.Tion has been reported in several eco-climatic areas: arid in Western Africa and Arabic Peninsula [1,6], sub-humid in Eastern Africa [7,8], wet forests in central Africa [5], dam and irrigated agricultural land under hot climatic conditions in Egypt, Mauritania and Sudan [9?1] and recently humid highlands in Madagascar [3,12]. The respective roles of direct and vectorial transmissions remain unevaluated in both human and cattle and probably vary among these eco-climatic areas. Madagascar experienced two major Rift Valley fever (RVF) outbreaks: 1990?1 in the eastern-coast and central highlands and 2008?9 in the south, the north and the highlands [13?15]. The last outbreaks occurred in two epidemic waves during the two successive rainy seasons of 2007?8 and 2008?9. Following the first wave, passive surveillance and emergency response were developed. Sentinel surveillance in herds were set up with field veterinarians [16]. This sentinel surveillance allowed the early detection of the second wave of outbreak inPLOS Neglected Tropical Diseases | DOI:10.1371/journal.pntd.July 14,2 /Rift Valley Fever Risk Factors in Madagascarcattle at the end of 2008 and thus the implementation of local control measures to prevent the spreading of RVF outside the region [15,16]. At the end of the epidemic, about 700 suspected human cases were recorded from which 26 were fatal. About 400 human and cattle samples were received for laboratory analyses and RVF infection was confirmed or considered as probable in 86 human and 46 ruminant samples [15]. Following the 2008?9 epidemics, studies showed a wide and heterogeneous spread of RVFV infection both in human and cattle [15,17] suggesting that some areas were more favorable than others to transmission [17]. Madagascar has a large variety of eco-climatic patterns, including semi-arid in the south, tropical in the west and on the eastern-coast, and temperate in the central highlands [18]. Apart from the highlands [3,12,19], RVF epidemiology is poorly understood in this country [15,17]. Since 2007, a human syndromic-based surveillance system has been developed which has allowed the detection of the first case of RVF in humans in 2008 [15]. Besides, retrospective investigations suggested that RVFV circulated among livestock since December 2007 [15], revealing a dearth in veterinary surveillance. The main difficulty to implement veterinary surveillance in Madagascar is the lack of basic means to collect and communicate veterinary information [20]. Thus, the identification of at-risk environments is essential to optimize the available resources by targeting RVF surveillance. In addition, there is a need to provide insight into the role of the two transmission routes and better adapt available control measures. Herein, the objectives of our study were: (i) to identify the environmental factors and areas favorable to RVFV transmission to both cattle and human and (ii) to identify human behaviors favoring human infections in Malagasy contexts.Materials and MethodsTo achieve these goals, we characterized the environments of Malagasy communes using a Multiple Factor Analysis (MFA). Then we analyzed cattle and human serological data using a Generalized Linear Mixed Models (GLMMs), with the individual serological status (cattle or human) as the response, and MFA factors, as well as potential other risk factors (covariates), as explanatory variables.Cattle and human datasetsThe cattle dataset contained results of a national cr.