frican Horse Sickness (AHS) is a vector-borne viral disease of horses, mules, and donkeys. The clinical signs and lesions occur as a result of increased vascular permeability and are characterized by an impairment of the respiratory and circulatory system (Radostits et al., 2007). AHS was first recognized in southern Africa, with the first outbreak recorded in 1719, when more than 1700 animals died. Although endemic in equines in Sub-Saharan Africa, outbreaks have also been recorded in North Africa, the Middle and Near East and southern Europe (MacLachlan and Guthrie 2010). AHS virusis transmitted by the biting midges Culicoides. Culicoidesimicolais the field vector of AHS virus although Culicoidesbolitinos has been shown to play an important role in the transmission of AHS virus in the cooler upper highlands of South Africa (Meiswinkel and Paweska 2003). The highest incidence of the disease usually occurs in the late summer and early autumn in years when the climatic conditions favor an abundance of Culicoidesmidges (Coetzer and Guthrie 2004).

Clinically, the disease is characterized by anacute pulmonary form, a cardiac form or sub-acute form, mixed form and a mild form known as horse sickness fever (Upadhyaya, 2011). AHS is one of the viral diseases characterized by up to 95, 50and 10% mortality rates in horses, mules, and donkeys, respectively (OIE, 2008).

In a study performed in horses and donkeys from southern Darfur and Khartoum states, the sera were analyzed by passive haemagglutination test and serum neutralization test, the overall seroprevalence was 42.64% and 27.75%, respectively (Ihsan 2004) However, these studies were conducted in few states, and small sample sizes were investigated. Hence, there was a gap of research information and published studies which point out the seroprevalence of AHS and investigate its association with potential risk factors. Therefore, this study was conducted to fill the information gap regarding the seroprevalence and potential risk factors associated with the disease in Sudan.

The study was designed as a cross-sectional observational study with a multistage sampling technique. Four states; Northern, River Nile, Khartoum, and Darfur were randomly selected from the whole country. Then from each state, four localities were selected. Finally, animals were investigated by visiting markets, farms and villages.

The blood was taken aseptically from the jugular vein into sterile vacutainers without anti coagulants, allowed to clot for 1-2 hours at room temperature, stored vertically overnight at 4 o C and then centrifuged at 4000 rpm for 5 minutes. The sera were taken in sterile bijou bottles and inactivated in the water bath at 56 o C for 30 minutes, allowed to cool and then stored at -20 o C until assayed in the laboratory.

A pre-tested structured questionnaire with the primary objective of elucidating the multi-factorial background of AHS was conducted in an interactive manner with every individual owner of horses and donkeys. The format was designed to investigate individual animal characteristics, management, and environmental risk factors.

Blood samples were examined at The Central Laboratory, Ministry of Higher Education and Scientific Research, Khartoum. Competitive Enzyme-Linked Immunosorbent Assay (c-ELISA) was used to detect the presence of specific antibodies against the AHS virus in the collected sera samples following manufacturer's protocol.

i. Preparation of reagents Washing solution: one part of the concentrate washing solution which provided in the kit was diluted in 24 parts of distilled or deionized water (40 ml of concentrate solution and 960 ml of water) and remained stable at +4°C. Controls (+ve) and (-ve): were ready to use and didn't need any preparation.

Conjugate and substrate were ready to use and didn't need any preparation. 1. All reagents were brought to room temperature before use. 2. 100 µl of the positive control was dispensed into two wells and 100 µl of the negative control into two wells. Sera samples were diluted 1/5 in provided diluents. This step was made directly in the wells of the plate by adding 80 µl of the diluents and 20 µl of serum samples and the plate was shaken carefully for homogenization. The plate was covered and incubated for 1 hour at 37°C.

The content of the plate was thrown out by a brusque turnover of the plate to avoid the possible mixture of the content from one well to another.

A volume of 300 µl of washing solution was dispensed on each well.

The plate was shaken delicately, avoiding the contamination between wells. The plate was turned over to empty the wells.

The process was repeated five times as indicated on the instructions of the kit.

Before empty the content of the last washing step, the next reagent was ready to use. The plate was not maintained dry more than strictly needed.

After the last step of washing the plate was shaken and turned over on the absorbent filter paper. 4. The plate was washed five times as described in step 3. 5. 100 µ / well of the conjugate was added. Incubated for 30 minutes at 37°C. 6. The plate was washed five times as described in step 3. 7. 100 µl/well of the substrate solution was dispensed by using a multi-channel pipette. Incubated for 10 minutes at room temperature. 8. 100 µl well of the stop solution was dispensed and dispensing of bubbles was avoided. 9. The plate was read at 405 nm using spectrophotometer. 10. Validation criteria:

The ELISA test validation was checked for each plate based on two criteria set by the manufacturer for the mean optical density (OD) of the positive and negative control. The OD of the positive control was less than 0.2 and the OD of the negative control was higher than 1.0.

Blocking percentage (BP) of each sample was calculated based on OD value applying the following formula: Samples showed BP value lower than 45% were considered to be negative for antibodies of AHS virus. Samples showed BP value higher than 50% were considered as positive for antibodies of AHS virus.

Samples with BP value between 45% and 50% were considered doubtful, and they were retested. If the result is the same, another extraction was made and tested 2 weeks later.

Samples with BP value between 45% and 50% were considered doubtful and they were retested. If the result is the same, another extraction was made and tested 2 weeks later.

All data collected were entered into Microsoft excel spreadsheet. For analysis of the data, SPSS version 21 software was used. Data were analyzed descriptively in the first step, using the frequency table and cross tabulation. Then the association of the potential risk factors with the seroprevalence of AHS at the individual level was analyzed using the Chi-square test. The level of significance was set at P<0.05. For the investigation of the association between the seroprevalence in response to individual animal characteristics, management, and environmental potential risk factors, a multivariate analysis was performed in which logistic regression was used. The strength of association between the risk factors and the sero-prevalence of AHS was quantified using the odds ratio (OR) and the level of significance was set at p ? 0.05.

In a total of 920 equines (590 horses and330 donkeys) sampled and examined, the overall seroprevalence of AHS was72.2 %. Within states, the highest seroprevalence of AHS was reported in Khartoum and Darfur states (81%) and (78.8%), respectively. Table 1shows the univariate analysis of the association of potential risk factors with the seroprevalence of AHS. The table shows the risk factors which have been investigated, number of animals tested, number of animals positive with their percent (%), the value of Chi-square (? 2 ) and the P value.

The seroprevalence of AHS showed statistically significant association with state (? 2 = 47.434, p<0.001), species (? 2 = 50.163, p <0.001), sex (? 2 = 26.206, p<0.001), housing (? 2 = 26.477, p<0.001), vaccination (? 2 = 44.466, p<0.001), breed (? 2 = 57.256, p<0.001), the presence of water bodies (? 2 = 26.271, p<0.001), presence of cullicoides (? 2 = 42.658, p<0.001), presence of ticks (? 2 = 23.608, p<0.001), activity of animals (? 2 = 41.435, p<0.001), awareness of owners (? 2 = 25.639, p<0.001), age (? 2 = 20.186, p<0.001), health score(? 2 = 12.038, p<0.001),pregnancy status(? 2 = 3.249, p = 0.0355), and infection with other diseases(? 2 = 14.637, p<0.001). However, risk factors of presence of other animals and pervious infection with AHS did not show statistically significant associations (p>0.05).

The final multivariate model revealed that equines in Khartoum state were almost five times more likely to be sero-positive compared with equines in Northern state (OR = 4.909, p = 0.017), local equines were two times and half more likely to be sero-positive compared with cross equines(OR = 2.532, p = 0.004), horses were 3.8times more likely to be sero-positive compared with donkeys (OR = 3.776, p = 0.017), equines raised in areas with water bodies were two times more likely to be sero-positive compared with those in dry areas (OR = 2.172, p = 0.033), and nonvaccinated equines were 17 times more likely to be sero-positive compared with vaccinated ones (OR =17.298, p<0.001).

IV.

The current study indicatedan overall seroprevalence of AHS in equines at the study states of 72.2 %. This result was higher than those reported in Khartoum and South Darfur states ( Furthermore, there was a significant variation of seroprevalence between the two species of equines (horses and donkeys).A higher seroprevalence was observed in horses as compared with donkeys, and the difference was statistically significant (p ? 0.05). This finding is in agreement with a couple of previous studies (Alemayehu and Benti, 2009;Yeshitila and Bekele, 2017). Also, by the OIE (2008) reports which stated that among equines horses were the most susceptible to AHS with a mortality rate of 50-95% followed by mules with a mortality rate around 50% and donkeys with a mortality rate of 5-10%. However, our result differed with the findings of other previous studies conducted by Tesfaye et al. (2012) and Ende et al. (2013) in Ethiopia.

The current study further revealed that there isa statistically significant variation (p ? 0.05) of seroprevalence between the local and cross breed. A higher seroprevalence was observed in the local compared with the cross breed. However, this significant difference could be attributed to the fact that the cross horses receive high care by raising them in safe stables that protect them from the infestation of vectors and vaccinated annually against the disease.

Our findings indicated a statistically significant variation between areas with water bodies and dry (p ? 0.05). A higher seroprevalence of AHS was observed in equines raised in areas with water bodies, and this result is in close agreement with the results reported by Coetzer and Guthrie (2004), Demissie (2013) and Radostitis et al. (2007) who explained that AHS endemic areas are more likely to be low lying, flooded by rain water, warm and marsh regions that create favorable environment for multiplication of Culicoides and mechanical vectors.

Furthermore, our final multivariate model revealed a statistically significant (p ? 0.05) variation between non-vaccinated and vaccinated equines. A higher seroprevalence of AHS was observed in nonvaccinated equines. Lack of vaccination is the most strong risk factor (OR =17.298, P <0.001) among the risk factors investigated and found statistically significantly associated with AHS. However, the observed association could be confounded with poor management and lack of knowledge of equines owners about the disease and the importance of vaccination as a protective tool against AHS. Unfortunately, both risk factors were not investigated.

Regarding sex, our findings showed significant variation (p ? 0.05) in the se roprevalence between sexes in the univariate analysis, with a higher seroprevalence in males than in females. This result is in close agreement with the result reported in Ethiopia by Yeshitila and Bekele (2017). However, the risk factor of sex did not remain statistically significant in the final multivariate model.

In the current study, we have used the odds ratio to quantify the strength of association between potential risk factors and the seroprevalence of AHS. To our best knowledge, this is the first study to quantify the strength of association between potential risk factors and AHS in Sudan. However, although the discussed risk factors were statistically significantly associated with AHS in the final multivariate model, it is difficult to consider them as necessarily causally related and should be interpreted in light of the causal criteria that have been proposed by Thrusfield (2005).

V.

The present study indicated that AHS seroprevalence was highly prevalent in equines in the study states of Sudan. Furthermore, the seroprevalence of AHS was statistically significantly associated with the vaccination status of equines, geographic location of state, species, breed, and presence of water bodies. These identified risk factors should be carefully considered when control strategies for AHS are implemented.