Seroprevalence of and Risk Factors for Toxoplasma gondii Infection in Cats in Estonia

Abstract

In Estonia, northeastern Europe, Toxoplasma gondii seroprevalence in humans has not declined, in contrast to many other countries. The reasons for this are unknown. Domestic cats are important hosts in the epidemiology of the parasite, but information on local feline T. gondii infections has been lacking. An epidemiological cross-sectional study was conducted to estimate the seroprevalence of T. gondii and the risk factors associated with seropositivity in cats in Estonia. Surplus from blood samples that had been collected for unrelated diagnostic purposes from 306 pet cats and 184 shelter cats were analyzed for anti–T. gondii immunoglobulin G antibodies by using a direct agglutination test. Two questionnaires were designed to reveal relevant risk factors for seropositivity. The overall seroprevalence of T. gondii in cats in Estonia was 60.8%. Older age, outdoor access, hunting, living outside the city in the countryside, and not being a purebred cat were among the risk factors associated with seropositivity. T. gondii is highly prevalent in domestic cats in Estonia. This suggests that the environment has been contaminated with T. gondii. Seropositivity indicates previous oocyst shedding, and most of the cats had outdoor access. The increase in T. gondii seroprevalence with age indicates acquired infections, and most of the risk factors were lifestyle-related. Cat owners could diminish the risk of T. gondii infection and also limit the spread of the parasite by not allowing their cats to roam free.

Introduction

T oxoplasma gondii is a ubiquitous protozoan parasite that can infect pets, livestock, and humans (Dubey 2010). Domestic cats (Felis catus) and other felids are highly amplifying hosts for the parasite and able to contribute to the environmental reservoir of T. gondii, and thus are of epidemiological importance (Dabritz and Conrad 2010, Gerhold and Jessup 2013). Following the infection and the sexual reproduction of the parasite, cats can shed high numbers of infectious forms of T. gondii in their feces (Dubey 2009). Where the cat is allowed to defecate and how the cat owner disposes of the cat feces determine whether the environment is contaminated (Tenter et al. 2000).

T. gondii is widespread throughout the world, but the seroprevalence of the parasite in domestic cats varies in different regions and countries. Estimates range from 2.1% in China to 89.3% in Colombia (Dubey 2010). In the countries neighboring Estonia, about half of the cats are seropositive—48.4% in Finland (Jokelainen et al. 2012) and 51.6% in Latvia (Deksne et al. 2013). In many hosts, including domestic cats, T. gondii seropositivity increases with age, indicating that most infections are acquired postnatally (Dubey 2010). Risk factors associated with feline T. gondii infections include hunting, outdoor access, and eating raw meat (Jokelainen et al. 2012, Opsteegh et al. 2012, Deksne et al. 2013).

In contrast to the decrease observed in other countries (Jones et al. 2007, Pappas et al. 2009, Villena et al. 2010), there has been no decline in T. gondii seroprevalence in humans in Estonia (Pehk 1994, Birgisdottir et al. 2006, Janson et al. 2013). The reasons for this have not been studied. In wild boar (Sus scrofa), an indicator host species for the spread and prevalence of the parasite, the infection is common and evenly spread in Estonia (Jokelainen et al. 2013). Data on the local prevalence of feline T. gondii infections, as well as on relevant aspects of feline husbandry, have been lacking. The aims of this study were to estimate the prevalence of and evaluate relevant risk factors for T. gondii infections in cats in Estonia.

Materials and Methods

Ethics statement

The samples investigated were surplus from sera and plasmas that had been collected from pet cats and shelter cats for unrelated diagnostic purposes. No blood was drawn solely for this study. The sampling was a convenience sampling, based on the presence of surplus serum or plasma after diagnostic tests, voluntary participation and formal written informed consent of the owners of the pet cats, voluntary participation and oral informed consent of the veterinarian of the animal shelter, and voluntary contribution (storing the surplus from samples) of the veterinarians and laboratory personnel of the clinics. A general consent form, alone or in combination with the written informed consent form specific to this study, was used at the Small Animal Clinic of the Estonian University of Life Sciences, and the written informed consent form specific to this study was used at other clinics. All information collected was treated confidentially. The samples were stored coded.

Study design

A cross-sectional seroepidemiological study was conducted to determine the proportion of cats in Estonia that had been naturally infected with T. gondii. The study population was domestic cats in Estonia, and both pet cats and shelter cats were included in the study.

Sample size calculation

On the basis of the data available from the neighboring countries Finland and Latvia (Jokelainen et al. 2012, Deksne et al. 2013), the seroprevalence of T. gondii in cats in Estonia was expected to be 50%. The number of cats in Estonia is estimated to be 240,000. On the basis of these assumptions, the minimal sample size required to estimate seroprevalence with the desired precision and level of confidence was calculated to be 384 cats (Dean et al. 2015).

Collection of samples

The staffs at small animal clinics and shelters in Estonia were contacted, provided with information about the study, and asked to participate. Information about the study was also provided in a blog (Must and Lassen 2013). The sample collection period was from January 1 to December 31, 2013. The samples were surplus from routinely separated sera and plasmas that had been taken for unrelated diagnostic purposes and stored for this study at −20°C until analysis.

A total of 490 samples were obtained for this study (Table 1). The 306 pet cat samples came from four separate small animal clinics, all of which were located in one town, Tartu. The 184 shelter cat samples came from a shelter located in Tartu.

Table 1.

Toxoplasma gondii Seroprevalence in Cats in Estonia, by Selected Categories

Variable	n	n seropositive	Seroprevalence %	95% Confidence interval (mid-p exact)
Pet cat	306	193	63.1	57.5–68.3
Shelter cat	184	105	57.1	49.8–64.0
Young (<1 year)^*	30	5	16.7	6.9–34.0
Adult (≥1 year)^*	450	286	63.6	59.1–67.9
Male	251	155	61.8	55.6–67.6
Female	224	132	58.9	52.4–65.2
Purebred cat^*	67	25	37.3	26.7–49.3
Persian^§	18	6	33.3	16.1–56.4
British shorthair^§	11	3	27.3	9.2–57.1
Siamese^†	11	8	72.7	42.9–90.8
Scottish fold^§	7	0	0.0	0.0–40.4
Sphynx^#	5	4	80.0	36.0–98.0
Not a purebred cat^*	390	247	63.3	58.4–68.0
From Tartu^*	299	163	54.5	48.9–60.1
Not from Tartu^*	134	89	66.4	58.1–73.9
From Tartumaa	365	210	57.5	52.4–62.5
From another county	66	41	62.1	50.0–72.9
From a town^*	348	192	55.2	49.9–60.3
From countryside^*	84	60	71.4	61.0–80.0
Indoors only^*	128	64	50.0	41.5–58.5
Outdoor access^*	146	106	72.6	64.8–79.2
Receives raw meat	110	75	68.2	59.0–76.2
Receives no raw meat	147	85	57.8	49.7–65.5
Has been hunting^*	134	99	73.9	65.8–80.6
Has never been hunting^*	131	66	50.4	41.9–58.8
Total	490	298	60.8	56.4–65.0

Statistically significant difference in seroprevalence between the two groups.

Statistically significant difference between the seroprevalence in this breed and the overall seroprevalence.

†

Statistically significant difference between the seroprevalence in the Siamese and the British shorthair, the Siamese and the Scottish fold, and the Siamese and other purebred cats.

Statistically significant difference between the seroprevalence in the sphynx and the Scottish fold breeds.

Questionnaires

Two questionnaires were designed for the collection of data for risk factor analyses. The pet cat questionnaires were filled in by the cat owners and included questions on the cat's signalment (age, place of residence, sex, and breed) and lifestyle (diet, outdoor access, and hunting). The shelter cat questionnaires were filled in by shelter veterinarians and covered the cat's sex, breed, and age group, as well as the place where the cat was found or had lived, if known. All cats that had been allowed outdoors at some time in their life, whether supervised or not, were classified as having outdoor access. Cats that had ever hunted were classified as hunters. A few samples arrived without a questionnaire. Moreover, the signalment or lifestyle of some cats was not specified (Table 1).

Serology

The samples were screened for the presence of anti–T. gondii immunoglobulin G (IgG) antibodies using a commercial direct agglutination test (Toxo-Screen DA, bioMérieux, Marcy-l'Étoile, France) according to the manufacturer's instructions. In this method, possible immunoglobulin M (IgM) antibodies are denatured by 2-mercaptoethanol. The samples were diluted to 1:40; this dilution was selected as the cutoff for seropositivity. A four-point scale was used initially to record the results, which were further interpreted as a dichotomous outcome—seropositive or seronegative. All plates included the negative and positive control provided in the kits, at two dilutions 1:40 and 1:4000, and the antigen control. The results were read after 18 h of incubation.

Statistical analyses

Initially, two-by-two tables were used to evaluate simple associations (Dean et al. 2015). Two-tailed p values < 0.05 were considered statistically significant. Stata software 11.0 (StataCorp. 2009) was used for logistic regression analyses. Multivariable models were built stepwise using both forward selection and backward elimination.

Results

The overall seroprevalence estimate of T. gondii in cats in Estonia was 60.8% (Table 1). Specific anti–T. gondii antibodies were present in the serum or plasma of 57.1% of the shelter cats and 63.1% of the pet cats. Adult cats (≥1 year) had a higher seroprevalence (p < 0.001) and 8.7 (95% confidence interval [CI] 3.3–23.2) times higher odds of testing seropositive than younger cats. When the cats' age in years was used as a continuous variable, the odds of testing seropositive increased by 24.1% (95% CI 17.4–31.3%) for every year of age.

In univariable analyses and simple logistic regression analyses, other significant risk factors for seropositivity included not being a purebred cat (odds ratio [OR] = 2.9, 95% CI 1.7–5.0), hunting (OR = 2.8, 95% CI 1.7–4.7), going outdoors (OR = 2.7, 95% CI 1.6–4.4), living in the countryside (OR = 2.0, 95% CI 1.2–3.4), and living outside of Tartu town (OR = 1.7, 95% CI 1.1–2.5). In addition, differences in seroprevalence were noted among cat breeds (Table 1). Two breeds, the Persian and the British shorthair, appeared as significant protective factors (OR = 0.29, 95% CI 0.11–0.79, and OR = 0.22, 95% CI 0.57–0.83, respectively).

The receiver operating characteristic curve of the final logistic regression model, which included age in years and three other variables (Table 2), was evaluated to have a moderate predictive value. The curve extended reasonably well into the upper left-hand corner of the graph, with an area under the curve of 0.83.

Table 2.

The Final Multivariable Logistic Regression Model for Toxoplasma gondii Seropositivity in Domestic Cats in Estonia

	Odds ratio	95% Confidence interval	p value
Age in years	1.28	1.19–1.38	0.000
Not purebred	3.85	1.77–8.40	0.001
Not from Tartu	2.18	1.13–4.20	0.021
Outdoor access	2.31	1.24–4.29	0.008

Probability > χ² = 0.0000.

Area under the receiver operating characteristic curve = 0.8282.

Discussion

The majority of cats in Estonia have encountered T. gondii at least once. Seropositive cats are considered to have already shed oocysts (Dubey et al. 1995, Dubey 2010); thus, our results indicate that the environment is contaminated with T. gondii. Environmental contamination with infectious forms of T. gondii creates a long-lasting risk to other hosts, including humans. This could partly explain the high seroprevalences detected in humans and wild boar (Janson et al. 2013, Jokelainen et al. 2015).

T. gondii seroprevalence in cats in Estonia is significantly higher than recent estimates from Finland and Latvia (Jokelainen et al. 2012, Deksne et al. 2013). The Finnish study used the same method and had the same sample size, whereas a modified indirect enzyme-linked immunosorbent assay was used in the Latvian study, and the sample size was smaller. In both the Finnish and Latvian studies, more purebred cats were included than in our study. Most cats in the Finnish study were healthy, whereas the pet cats included in this study had been sampled because of various health problems. Cats with health issues that required laboratory analysis of their blood sample are therefore overrepresented, and including clinically ill individuals might have had an effect on our estimate of seroprevalence.

T. gondii infection is associated with various health issues, but none of the cats included in this study had been diagnosed with clinical toxoplasmosis. In fact, no clinical toxoplasmosis cases were diagnosed at the Small Animal Clinic of Estonian University of Life Sciences during the study, suggesting that clinical feline toxoplasmosis is underdiagnosed in Estonia.

A commercial test was chosen to detect IgG antibodies, which are long-lasting and relevant for epidemiological studies. The same test has been used in other studies, such as the Finnish study (Jokelainen et al. 2012), yielding comparable data. The test may give false-negative results, for example, during early stages of infection, because it detects only one antibody class, IgG. We decided not to test for anti–T. gondii IgM antibodies due to the limited effect expected on the result (Svobodová et al. 1998). In addition, the cutoff we chose for seropositivity is relatively high (Dubey 2010); we probably underestimated the prevalence of the infection, because viable T. gondii has been isolated from cats with titers lower than 40 (Dubey et al. 2013).

Taking into consideration the limited amount of sample available from some of the cats and the limited benefit expected from using several dilutions (Jokelainen 2013), only one dilution (the cutoff dilution 1:40) was used. Some false-negative results are expected due to the prozone phenomenon that goes undetected when using a single dilution (Jokelainen 2013). Thus, the estimate of seroprevalence we obtained is conservative.

Our convenience sample represents mainly two subgroups of the cat population—pet cats that had been taken to a veterinarian and shelter cats. The difference in seroprevalence between shelter cats and pet cats was not significant. Unfortunately, the exact age of most of the shelter cats was unknown, and thus it remains unclear how different the age distribution was in the two groups. The youngest pet cat included in the study was a 3-month-old kitten; the oldest was 18 years. The mean age of the pet cats was 7.1 years, and the mode was 1 year. Among shelter cats, 12 were regarded as young (less than 12 months old). It is unknown how well the age distribution and the two subgroups sampled represent the local cat population.

Due to the location of the contributing clinics and the shelter, 299 cats were from Tartu and 365 from Tartumaa County. The seroprevalence in cats living in Tartumaa County did not differ significantly from the prevalence in cats from other counties (Table 1). Given that Estonia is a small country and Tartu is not located in its periphery, the results can be considered representative of the whole country. Moreover, there were no geographic differences in the seroprevalence of T. gondii in wild boar, which are good indicators of the spread of the parasite (Jokelainen et al. 2015).

Living in the countryside and living outside Tartu town were among the significant risk factors for seropositivity. The countryside cats and city cats did not appear to have different lifestyles that could easily explain this; for example, the percentage of cats allowed outdoors in these two groups did not differ significantly. The difference in seroprevalence in urban and rural areas might be explained by differences in local reservoirs of the parasite, both in prey animals and in the environment, which serve as local infection sources for the cats. A partial explanation could be that the only university veterinary clinic in Estonia is located in Tartu, so cat owners in Tartu might be relatively well informed about the prevention of feline infectious diseases.

The results of this study confirm that T. gondii infection in cats is usually an acquired infection. Older cats had significantly higher seroprevalence than younger cats, and the seroprevalence and the odds of testing seropositive increased with age. Most of the risk factors associated with seropositivity were lifestyle-related, not characteristics that the cat was born with. The exception, not being a purebred cat, could at least partly be explained by differences in the lifestyles of purebred and non-purebred cats. For example, 28.1% of the purebred cats were allowed outdoors, whereas 59.9% of the non-purebred cats went outdoors (p < 0.001).

In general, cats in Estonia appear to be living a risky lifestyle, because approximately half of them had outdoor access and the possibility to catch prey. Moreover, many of the pet cats received raw meat as part of their diet. Such cats have many opportunities to be infected with T. gondii. Receiving raw meat in the diet was not a significant risk factor in this study, in contrast to in the Finnish study (Jokelainen et al. 2012). Both hunting and access outdoors were relevant risk factors in this study, and outdoor access was included in the final model (Table 2). Outdoor access gives the cats the chance to hunt and eat sources of infection, but also to acquire the infection from the environmental oocyst reservoir. However, 8.6% of the cats with outdoor access reportedly did not hunt, whereas 4.8% of the indoor cats were hunters. Rodents entering residential buildings are also a potential source of feline T. gondii infections.

Domestic cats are important hosts for the success of T. gondii. Luckily, cat owners have several possibilities to prevent their cats from contributing to it. It is not known how well informed Estonian cat owners are about T. gondii, but allowing cats to go outside appears to be common practice in Estonia. If a cat owner knows about T. gondii but allows a cat to go outside, become infected, and shed oocysts, this action can be considered as intentional contamination of the environment with a potentially lethal zoonotic pathogen. If feline infections were better prevented, fewer oocysts would be added to the environmental reservoir, especially near human settlements. Our results support the recommendations (Elmore et al. 2010, Opsteegh et al. 2012) of not allowing cats to roam free and hunt prey. Preventing feline T. gondii infections is important to protect cats from clinical toxoplasmosis (Jokelainen et al. 2012) and to protect public health.

Conclusions

T. gondii antibodies were common in domestic cats in Estonia. This suggests that cats have shed oocysts, which form an environmental reservoir constituting a long-lasting infection source for other hosts. The seroprevalence increased with age, indicating acquired infections. Most of the risk factors were lifestyle-related, not characteristics the cat was born with. Hunting and outdoor access are risk factors that could be avoided. Cat owners could protect their cats from T. gondii infections better and reduce the further spread of the parasite.

Footnotes

Acknowledgments

The cat owners and veterinarians are warmly thanked for their contributions to and interest in this study. The work was partially supported by the Society for the Advancement of Estonian Studies in Canada (Ellen and Eduard Kurvits Fund), the health promotion research program TerVe 3.2.1002.11-0002 EKZE_SS from the Estonian Research Council, the Estonian Science Foundation grant ETF9433, and by project funding M14143VLVP from Strategic Development Fund of the Estonian University of Life Sciences.

Author Disclosure Statement

No competing financial interests exist.

References

Birgisdottir

, Asbjörnsdottir

, Cook

, Gislason

, et al. Seroprevalence of Toxoplasma gondii in Sweden, Estonia and Iceland. Scand J Infect Dis, 2006; 38:625–631.

Dabritz

, Conrad

. Cats and toxoplasma: Implications for public health. Zoonoses Public Health, 2010; 57:34–52.

Dean

, Sullivan

, Soe

. OpenEpi: Open source epidemiologic statistics for public health. www.OpenEpi.com Accessed 2012, 2013, 2014, and 2015.

Deksne

, Petrusēviča

, Kirjušina

. Seroprevalence and factors associated with Toxoplasma gondii infection in domestic cats from urban areas in Latvia. J Parasitol, 2013; 99:48–50.

Dubey

. History of the discovery of the life cycle of Toxoplasma gondii . Int J Parasitol, 2009; 39:877–882.

Dubey

. Toxoplasmosis of Animals and Humans, 2 ^nd ed. Boca Raton: CRC Press, 2010:313 pp.

Dubey

, Lappin

, Thulliez

. Long-term antibody responses of cats fed Toxoplasma gondii tissue cysts. J Parasitol, 1995; 81:887–893.

Dubey

, Darrington

, Tiao

, Ferreira

, et al. Isolation of viable Toxoplasma gondii from tissues and feces of cats from Addis Ababa, Ethiopia. J Parasitol, 2013; 99:56–58.

Elmore

, Jones

, Conrad

, Patton

, et al. Toxoplasma gondii: Epidemiology, feline clinical aspects, and prevention. Trends Parasitol, 2010; 26:190–196.

10.

Gerhold

, Jessup

. Zoonotic diseases associated with free-roaming cats. Zoonoses Public Health, 2013; 60:189–195.

11.

Janson

, Neare

, Hütt

, Orro

, et al. 2013. Zoonotic parasite infections among Estonian veterinarians compared with general population. Oral presentations. Trop Med Int Health, 2013; 18:100.

12.

Jokelainen

. Wild and domestic animals as hosts of Toxoplasma gondii in Finland. Doctoral Dissertation. Helsinki: University of Helsinki, Finland; 2013:112 pp.

13.

Jokelainen

, Simola

, Rantanen

, Näreaho

, et al. Feline Toxoplasmosis in Finland: Cross-sectional epidemiological study and case series study. J Vet Diagn Invest, 2012; 24:1115–1124.

14.

Jones

, Kruszon-Moran

, Sanders-Lewis

, Wilson

. Toxoplasma gondii Infection in the United States, 1999–2004, Decline from the Prior Decade. Am J Trop Med Hyg, 2007; 77:405–10.

15.

Must

, Lassen

. Study: Toxoplasma gondii in Estonian cats. Blog entry. Parasitology in Estonia, http://parasitologyestonia.blogspot.com/2013/01/study-toxoplasma-gondii-in-estonian-cats.html ; Uuring: Toxoplasma gondii Eesti kassidel. Blog entry. Parasitoloogia Eestis, http://parasitoloogiaestis.blogspot.com/2013/01/oppima-toxoplasma-gondii-eesti-kassidel.html 2013.

16.

Opsteegh

, Haveman

, Swart

, Mensink-Beerepoot

, et al. Seroprevalence and risk factors for Toxoplasma gondii infection in domestic cats in The Netherlands. Prev Vet Med, 2012; 104:317–326.

17.

Pappas

, Roussos

, Falagas

. Toxoplasmosis snapshots: Global status of Toxoplasma gondii seroprevalence and implications for pregnancy and congenital toxoplasmosis. Int J Parasitol, 2009; 39:1385–1394.

18.

Pehk

. Toksoplasmoosiuuringud Tartus. Eesti Arst, 1994; 1:20–21.

19.

StataCorp. Stata Statistical Software: Release 11. College Station, TX: StataCorp LP, 2009.

20.

Svobodová

, Knotek

, Svoboda

. Prevalence of IgG and IgM antibodies specific to Toxoplasma gondii in cats. Vet Parasitol, 1998; 80:173–176.

21.

Tenter

, Heckeroth

, Weiss

. 2000. Toxoplasma gondii: From animals to humans. Int J Parasitol, 2000; 30:1217–1258.

22.

Jokelainen

, Velström

, Lassen

. Seroprevalence of Toxoplasma gondii in free ranging wild boars hunted for human consumption in Estonia. Acta Vet Scand, 2015; 57:42.

23.

Villena

, Ancelle

, Delmas

, Garcia

, et al. Congenital Toxoplasmosis in France in 2007: First results from a national surveillance system. EuroSurveillance, 2010; 15:pii=19600.