Seropositive response to FeLV or FIV was significantly higher in sick compared to healthy cats. Sick cats were 5 times more likely to test positive for FeLV p27 antigen, and 2 times more likely to be positive for FIV antibody. These observations are similar to studies carried out in the United Kingdom and Canada [12, 17]. Since both FeLV and FIV are immunosuppressive in nature, it is likely that FeLV- or FIV-infected cats become predisposed to opportunistic or secondary infections .
The probability of seropositive response for FeLV or FIV was significantly higher among male cats. This is consistent with some epidemiological findings previously reported for retrovirus screening [13, 19]. Danner et al.  reported that the sex of a cat significantly influences risk for FIV but not FeLV. On the other hand, Bandecchi et al. , observed no significant association between sex and seropositivity to FeLV or FIV. These variations might be related to difference in the type of cat populations being studied. For example, Danner et al.  exclusively sampled feral cats, while Bandecchi et al.  sampled owned cats. In contrast, our study included owned and shelter cats that probably had varying medical and behavioural characteristics .
In our study, cats with aggressive behaviour were 2 times more likely to test positive for FeLV or FIV, as compared with non-aggressive cats. A relationship between aggressive behaviours and seropositivity to FIV has been reported by several authors [19, 23, 24]. However, the high prevalence of FeLV observed in aggressive cats in our study contradicts the earlier notion that FeLV is primarily a disease of friendly or socialized cats . Thus, there is the need to re-consider social behaviour in assessing FeLV transmission patterns [19, 22]. It is possible that demographic factors that influence serological status to FeLV or FIV may change over time or with geography, climate and other factors.
FIV seropositive status was more frequent in intact male and female cats, compared to their neutered counterparts. On the other hand, positive response to FeLV testing was higher in intact males and neutered females. Reports of the relationship between sex and risk for feline retroviruses vary considerably. For example, in a study involving all parts of North America, FeLV or FIV seropositive outcomes were more frequent in intact females compared to spayed females, and in castrated males compared to intact males . However, when only cats from the United State were considered, positive response to FeLV or FIV was observed more frequently among neutered male and female cats . Greater seropositive status with respect to retrovirus infections in intact cats might be explained by their frequent involvement in territorial aggression and free-roaming behaviours that could increase risk for contact with infected cats. Although, in our study, seropositive test and odds of FeLV or FIV were greater among cats with outdoor access, this relationship was not significant statistically as reported recently in Canadian and German studies [17, 19]. It is possible that lack of history on the original lifestyle of cats (indoor or outdoor) before their arrival at the shelter may affect our conclusion on the relationship between retrovirus seropositive response and lifestyle of cat population sampled. High prevalence of FeLV or FIV among free-roaming sexually intact cats might explain why the American Association of Feline Practitioners (AAFP) and The European Advisory Board for Cat Diseases (ABCD) recommended neutering as a means of reducing the frequency of feline retrovirus infections [27, 28].
In our study, cats living in multi-cat households were most likely to be seropositive for FeLV, followed by cats living in shelters. FeLV seropositive status was least common among cats from single cat households. Conversely, seropositivity to FIV was not significantly influenced by type of household. This finding agrees with Fromont et al. , who also observed that prevalence of FeLV is more likely to be affected by population density compared to FIV. Overcrowding associated with multi-cat households often results in stress, poor hygiene, and increased direct contact among cats. FeLV transmission may likely be facilitated in these circumstances, since the virus is transmitted predominately by sharing of food and water containers [5, 29].
The age of the tested cat was significantly associated with FeLV and FIV. Seropositive response to FeLV was 2.5 times higher in young cats, whereas seropositive response to FIV was 1.8 times higher in adult cats. This finding is in agreement with Levy et al. , and consisted with previous observations that increased susceptibility to FeLV was higher among young cats while susceptibility to FIV increases as cats grow older .
Although 4.3% of the sampled cat population was positive for both FeLV and FIV, we could not demonstrate any association between the two viruses statistically. There are conflicting opinions about the epidemiological relationship between FeLV and FIV. Some authors argued that FeLV and FIV occur independently [12, 31], and others reported significant associations [19, 32]. The debate notwithstanding, co-infection of cats with FeLV and FIV could lead to more negative health outcomes, compared to single infection with either virus .
It has been recommended that asymptomatic cats tested positive for FeLV p27 antigen should be re-tested within 6-12 weeks. This is because of low positive predictive value of single ELISA-based assays particularly in population with low seropositive rate [28, 34]. No follow-up test was performed in the population we sampled, due to logistical problems and lack of willingness on the part of cat owners. In addition, cats in animal shelters usually remain for brief time periods as a result of adoptions and limited shelter space. Caution is therefore needed to avoid overgeneralizing the present findings.