Prevalence of cryptosporidian infection in cats in Turin and analysis of risk factors

Abstract

An epidemiological study was carried out to identify factors associated with the risk of cryptosporidian infection in cats. Faecal samples from 200 domestic cats were collected in the small animal clinic at the Faculty of Veterinary Medicine in Turin (north-west Italy). The faecal samples were analysed for the presence of Cryptosporidium species oocysts using a centrifugation concentration floatation method. For each cat, age, sex, breed, indoor/outdoor status, diet, diarrhoea and presence of other enteric parasites were analysed for association with Cryptosporidium species infection. Cryptosporidia oocysts were identified in 49 cats (24.5%) aged from 2 months to 18 years. Statistical analysis revealed that four variables are significantly associated with an increased risk of infection: less than 1 year of age (χ ²=6.5, P=0.01), feeding home-cooked diet (χ ²=6.92, P<0.01), presence of diarrhoea (χ ²=4.34, P<0.037), and presence of other enteric parasites (χ ²=10.31, P<0.01). No statistical differences were found for sex (χ ²=1.56, P=0.21), breed (χ ²=0.78, P=0.38) and outdoor/indoor status (χ ²=1.49, P=0.22). Cryptosporidium species was the parasite most frequently detected in the cats surveyed.

Cryptosporidium species are coccidian parasites that may infect a number of species, including humans, companion and farm animals, wild animals, birds and reptiles (Fayer et al 2000). The first cat with Cryptosporidium species infection was reported in 1979 (Iseki 1979) and since then there have been many case reports of cryptosporidiosis in domestic cats (Poonacha and Pippin 1982, Monticello et al 1987, Lappin et al 1997a, 1997b). In a recent study of potentially zoonotic enteric organisms in cats, Cryptosporidium species was identified as the most prevalent organism (Hill et al 2000). Clinical signs are usually self-limiting in immune-competent individuals (Keusch et al 1995), however, Cryptosporidium species is well recognised as a significant pathogen in immunosuppressed individuals causing severe, watery diarrhoea, malabsorption and weight loss. The objective of this study was to investigate the prevalence of Cryptosporidium species infection in cats in the Turin area and to identify potential risk factors associated with disease.

Materials and methods

Sampling

Two hundred faecal samples were collected from domestic cats that were presented to the small animal clinic at the Faculty of Veterinary Medicine in Turin (north-west Italy). Samples were collected from clinical cases (all animals that passed a faecal sample whilst hospitalised) and cases presenting for routine examination, such as vaccination. Samples were submitted to the laboratory immediately after defecation and were assessed on the same day of collection.

Collection of data

Owners of cats from which samples were collected were asked to fill in a questionnaire in order to collect demographic data and management data that may be associated with increased susceptibility to Cryptosporidium species infection. The following factors were analysed: breed, age, sex, any outdoor access, feeding patterns and diet (home-cooked and commercial diet). Faeces were classified as diarrhoeic or non-diarrhoeic.

Detection of oocysts

The faecal samples were analysed for the presence of Cryptosporidium species oocysts using a centrifugation concentration floatation method (Dubey et al 1990). Five grams of each faecal sample were mixed with 50 ml of Sheather's sugar solution (sugar 500 g; water 320 ml; phenol 6.5 g) and filtered through gauze; the filtrate was centrifuged at 400 g for 10 min. A fixed volume of liquid (500 μl) was pipetted from the top and mounted with a cover glass; 250 fields were examined at ×400 magnification.

Oocysts appeared as spherical shapes with a mean diameter of approximately 4.5 μm. Wet mounts of the concentrate were also examined by light microscopy for the presence of other gastrointestinal parasites. To minimise the risk of false positives or false negatives, each slide was examined by two operators. No discordant samples were observed.

Statistical analysis

Prevalence of infection was calculated by exposure levels to the following risk factors: age (<1 year and >1 year), breed (European vs other breed), sex, diet, free-roaming or kept indoors, presence of diarrhoea, presence of other enteric parasites. Bivariate analysis of the effect of each risk factor was analysed by two-by-two table using the FREQ procedure in the SAS system (SAS 1999). Risk factors whose χ ² test statistics were P<0.25 were included in a multiple logistic regression (unconditional maximum likelihood method, SAS 1999) to test their effects on infection with Cryptosporidium species. The overall significance of the model was evaluated by likelihood-ratio test, whereas significance of each term in the model was evaluated by Wald χ ² (two-tailed significance level α=0.05). Estimated odds ratios (OR) and 95% Wald's confidence intervals were obtained as measures of predictor effects. Only predictors and interaction terms whose effect was significant at the 15% level were retained in the final model. Goodness-of-fit was evaluated by the Hosmer and Lemeshow test.

Results

Cryptosporidium species oocysts were found in 49 of the 200 cats tested (24.5%).

Out of the 200 faecal samples examined, 95 were from female cats (47.5%) and 105 from male cats (52.5%). The median age of the cats was 3.8 years (range: 2 months–18 years); 87 (43.5%) of the cats tested were younger than 1 year. One hundred and seventy-two (86%) cats were mixed-breed and 28 (14%) were purebred (11 Siamese (5.5%) and 17 Persian cats (8.5%)).

Eighty-seven cats (43.5%) lived exclusively indoors; the remainder 113 (56.5%) had outdoor access. Eighty-five cats (42.5%) were fed with a mixed diet; 63 cats (31.5%) were fed exclusively with a commercial diet, and the remainder 52 (26%) with a home-cooked diet. At the time of sampling, 24 cats (12%) were reported to have diarrhoea. Other gastrointestinal parasites were also detected in faeces from 32 cats (16%), including protozoa (Toxascaris leonina (10), Toxocara cati (14), Cystoisospora species (15)) and helminths (Aelurostrongylus abstrusus (3), Dipylidium caninum (4)). Cryptosporidium species oocysts were detected in 15 out of these 32 cats (46.9%) (Table 1).

Table 1.

Prevalence of enteric parasites in 200 cats tested in Turin, Italy

Parasite	Number of positive (%)	Coinfected by Cryptosporidium species
Toxascaris leonina	10 (5)	4
Toxocara cati	14 (7)	5
Cystoisospora species	15 (7.5)	5
Aelurostrongylus abstrusus	3 (1.5)	0
Dipylidium caninum	4 (2)	1

Prevalence of Cryptosporidium species infection in association with the potential risk factors is shown in Table 2. Of the seven risk factors analysed, four were significant at the 25% level and were subsequently included in the logistic regression. The final logistic regression analysis, including age and type of diet as predictors, yielded significant likelihood ratio (χ (χ ²=19.1, df=2, P<0.001), and Hosmer and Lemeshow test showed no evidence of poor fit (χ ²=0.99, df=2, P=0.61).

Table 2.

Logistic regression analysis to determine factors associated with cryptosporidian infection in 200 cats tested in Turin, Italy

Variable	Category	Tested (n=200)	Positive (n=49)		χ 2	Odds ratio	95% CI	P *
No.	%
Sex	Male	105	30	28.6	1.56	1.60	0.79–3.25	0.21
Female	95	19	20	–	–	–	–
Age	<1 year	87	29	33.3	6.5	2.33	1.15–4.73	0.01
>1 year	113	20	17.7	–	–	–	–
Breed	Pure	28	5	17.8	0.78	0.63	0.18–1.85	0.38
Mixed	172	44	90	–	–	–	–
Outdoor	Yes	113	24	49	1.49	0.67	0.33–1.34	0.22
No	87	25	51	–	–	–	–
Feed	Commercial diet	63	8	16	6.92	0.34	0.14–0.82	<0.01
Home/mixed diet	137	41	29.9	–		–	–
Diarrhoea	Yes	24	10	20	4.34	2.51	0.95–6.61	0.037
No	176	39	80
Enteric parasites	Yes	32	15	31	10.31	3.48	1.47–8.23	<0.01
No	168	34	69	–	–	–	–

^* P value for variable; level of significance set at P≤0.05.

Age (<1 year and >1 year), presence of diarrhoea, feed (commercial or home-cooked) and presence of other parasites had a χ ² test result with P<0.25 in two-by-two table analyses and were, therefore, included as ‘predictor’ in the logistic regression model. Of the seven risk factors analysed, four were significantly associated with the excretion of oocysts; significant factors identified were feeding a home-cooked diet (χ ²=6.92, OR 0.34, 95% CI 0.14–0.82, P<0.01), the presence of diarrhoea (χ ²=4.34, OR 2.51, 95% CI 0.95–6.61, P<0.037), the presence of other enteric parasites (χ ²=10.31, OR 3.48, 95% CI 1.47–8.23, P<0.01) and the age category (χ ²=6.5, OR 2.33, 95% CI 1.15–4.73, P=0.01). There was no association of oocysts excretion with sex (χ ²=1.56, OR 1.6, 95% CI 0.79–3.25, P=0.21), breed (χ ²=0.78, OR 0.63, 95% CI 0.18–1.85, P=0.38) and outdoor/indoor status (χ ²=1.49, OR 0.67, 95% CI 0.33–1.34. P=0.22). Results of the analysis of factors potentially associated with the risk of Cryptosporidium species infection in cats are shown in Table 2.

Discussion

The results of the present survey revealed a prevalence of cryptosporidian infection of 24.5% in the cats of the Turin area. A comparison of the prevalence of Cryptosporidium species with the prevalence of other parasites showed that cryptosporidian infections are the most frequently detected.

The prevalence reported in this study is higher than that in other epidemiological studies on faecal specimens. Previous surveys in cats showed that infection rates varied between geographical areas with a range from 3.8% in Japan (Arai et al 1990) to 8.1% (Mtambo et al 1991) and 12.3% (Nash et al 1993) in Scotland.

These differences may, however, also be related to different diagnostic methods used and the number of microscopic fields observed. Abe et al (2002), in a study of Cryptosporidium species infection in dogs, detected oocysts in 3 out of 140 dogs (2.1%) when 20 fields were observed at 400×, and in 9 out of 140 dogs when 200 fields were observed at the same magnification. The duration of oocysts shedding in cats infected with Cryptosporidium species is between 7 and 10 days, but oocysts can be detected in faecal specimens at very low levels and intermittently for up to 5 months after infection (Iseki 1979). In these cases, the observation of many fields at high magnification is required for accurate diagnosis.

It is even possible that the prevalence may be underestimated because the value was the result of examination of a single sample from each cat; the excretion of Cryptosporidium oocysts is intermittent in symptomatic and asymptomatic animals (Current and Garcia 1991, Koulda and Nohynkova 1995), so one negative faecal examination does not mean necessarily that the cat is Cryptosporidium negative.

In the current study, there was no sex or breed association with Cryptosporidium species infection, which is consistent with the findings of McReynolds et al (1999). These authors, furthermore, reported that outdoor cats were approximately five times more likely than indoor cats to be infected with Cryptosporidium species, which is consistent with greater opportunity for cats to prey on infected hosts or become infected through contaminated soil or water.

Our results show a trend for increased prevalence in cats with outdoor access, but the difference was not statistically significant.

Infection was detected in a wide age range, but the risk for Cryptosporidium species infection was found to be increased for cats ≤1 year old, as previously reported (Arai et al 1990). An age-related decrease in susceptibility has also been reported for chickens and in rodent and ruminant models (Lindsay et al 1988, Ortega-Mora and Wright 1994, Rhee et al 1999).

A relatively immature immune system and a lack of immunity from previous exposure are possible explanations for the higher prevalence in kittens.

Our results show that cats infected with other enteric parasites were approximately three times more likely to be Cryptosporidium species positive. No information was obtained regarding the use of de-worming agents within 1 month of sample submission, and therefore the prevalence rate of coinfections may have been underestimated. In some cases the association of Cryptosporidium species with other parasites may be related to the possibility of coinfection at the same time, ie, ingesting a prey infected by Cryptosporidium species and Toxocara cati, Toxascaris leonina or Cystoisospora species.

Another relevant risk factor was represented by the diarrhoea: the probability of presenting with diarrhoea was significantly higher for cats shedding cryptosporidial oocysts than for those that did not excrete the parasite. Diarrhoea resolves in days to weeks in most immunocompetent cats with clinical cryptosporidiosis, and the disease severity is often related to the immunological status of the host (Arai et al 1990, Mtambo et al 1991).

It is necessary for the clinician to establish, if possible, whether the Cryptosporidium species infection may be causing the diarrhoea in the cat that has been presented for treatment. Spain et al (2001), in a prevalence study of enteric zoonotic organisms in cats, reported that diarrhoea is not a reliable predictor of whether a cat is actively shedding enteric organisms. Asahi et al (1991) reported that experimentally infected cats did not develop significant diarrhoea or weight loss during the infections, even though shedding large number of oocysts. Nevertheless, in this study prevalence was significantly higher in diarrhoeic than in non-diarrhoeic cats, suggesting that Cryptosporidium species may have a pathogenic role.

An important aspect of the cryptosporidian infection is the potential zoonotic role of cats; Mtambo et al (1996) reported two distinct-sized Cryptosporidium oocysts in the faeces of both a cat and a lamb infected with the oocysts from the cat; these results have important implications, suggesting that cats may be carriers of other strains of Cryptosporidium.

Although the data available in Italy are not extensive, there are indicators of widespread faecal contamination (Giangaspero 2006); Cryptosporidium has been detected, as well as in humans and animals, in wastewater and surface water (lake, river, and swimming pools) (Brandonisio et al 2000, Di Benedetto et al 2005, Molini et al 2007). Of the 77 Cryptosporidium isolates genotyped, 74 were Cryptosporidium parvum and were mainly detected in livestock but also in humans, dogs and shellfish (Cacciò et al 2001, Giangaspero et al 2005, 2006, Molini et al 2007). Molecular epidemiological studies show that the cat genotype is found also in humans or heterogeneous animal hosts (Bornay-Llinares et al 1999, Cacciò et al 2002). It is actually difficult to speculate about the potential role of cats as reservoirs for human infection; dog ownership (not cat ownership) has been recognised as a risk factor for Cryptosporidium infection in HIV-positive people but is still sensible to consider cats as a possible reservoir for human Cryptosporidium infection (Glaser et al 1998).

The results of this study demonstrate that cryptosporidiosis is a common infection in cats living in Turin city and must be considered an endemic problem in this geographic area.