INTRODUCTION
Toxoplasmosis is caused by Toxoplama gondii and is widespread in humans and animals throughout the world. Domestic cats and other feline species play a key role in the life cycle of the parasite by excreting oocysts into the environment. Many animals serve as an intermediate host and maintain the infection by containing the viable tissue cyst. The most common sources for acquisition of T. gondii infection by humans are ingestion of contaminated raw or undercooked meat, or water containing infective oocysts, or congenitally from mother to foetus during pregnancy [Reference Dubey1–Reference Corea4]. Consumption of non-pasteurized goat's milk is also associated with human toxoplasmosis [Reference Sacks, Roberts and Brooks5, Reference Skinner6] and exposure of women to primary T. gondii infection during pregnancy can result in abortion, and severe disease of the foetus [Reference Tenter, Heckeroth and Weiss7, Reference Dubey8].
Estimates of seroprevalence rates for T. gondii infection vary with geographical area and range from 16% to 40% in England and USA, 50–80% in Europe, and parts of Africa [Reference Dubey1], and 30–65% in the general population elsewhere in the world [Reference Flegr9, Reference Kijlstra and Jongert10]. Likewise, reported seroprevalence rates in women of child-bearing age range from 4% to 77% [Reference Tenter, Heckeroth and Weiss7] and when pregnant, this group as well as immunocompromised individuals are at higher risk of developing serious disease [Reference Tenter, Heckeroth and Weiss7, Reference Negash, Tilahun and Medhin11]. In Ethiopia, seroprevalence rates of 60–96·7% have been documented in different groups of people [Reference Negash, Tilahun and Medhin11, Reference Gebre-Xabier12–Reference Gebremedhin16], but the epidemiology of toxoplasmosis in the East Hararghe Zone of Ethiopia remains unknown. In this paper we present the results of a seroprevalence survey of T. gondii infection in people in three districts of this study area and report statistical analyses of associated risk factors.
MATERIALS AND METHODS
Study area
The study was carried out in Gursum, Babile and Haramaya districts of East Hararghe Zone, Ethiopia. Gursum district (9° 7′–32° N latitude, 42° 17′–42° 38′ E longitude) has an altitude ranging from 1200 to 2950 metres above sea level (masl); Babile district (8° 9′–9° 23′ N latitude, 41° 16′–41° 46′ E longitude) has an altitude ranging from 950 to 2000 masl; and Haramaya district (9° 9′–9° 32′ N latitude, 41° 50′–42° 05′ E longitude) has an altitude ranging from 1600 to 2140 masl. Three health centres were selected based on their accessibility to urban and rural people and provision of various health services to large numbers of the population.
Study design and population
A cross-sectional study on human toxoplasmosis in people from rural and urban areas, visiting the selected health centres was conducted from April 2013 to September 2013. The sample size was determined by random sampling [Reference Thrusfield17] based on the expected prevalence of 74·4%, 95% confidence interval (CI) with 5% precision [Reference Gebre-Xabier12]. The calculated sample size (n = 296) was increased by 20% (n = 354) for better precision and the study subjects were identified by systematic random sampling with the total number based proportionally on the number of people visiting the centres (39 000 in 2012). The study was approved by the Ethical Review Board of Oromia Regional State Health Bureau (ref. no. BEFO/HBTFH/1-8/2297). A questionnaire survey was used to collect data from study participants during blood sample collection and covered socio-demographic characteristics such as age, sex, residence and possible risk factors known to be associated with T. gondii infection (Table 1).
Table 1. Seroprevalence of T. gondii antibodies in relation to demography and risk factors in study districts, East Hararghe Zone, Ethiopia
† Shoats meat (sheep and goat).
‡ Mixed (shoats meat and beef).
§ Both (boiled and raw).
|| Mixed (stream and well water).
* Significant.
Serological tests
Venous blood samples were collected from consenting study participants and sera were stored at −20 °C. T. gondii specific IgG antibodies were detected using a direct agglutination test (DAT; Toxo-Screen DA, bioMérieux SA, France) according to the manufacturer's instructions at screening dilutions of 1/40 and 1/4000; a positive reaction at either dilution was taken as evidence of infection. Serum Toxoplasma IgM antibodies were determined with the immunosorbent agglutination assay (ISAGA; Toxo-screen ISAGA, bioMérieux) according to the manufacturer's protocol and reactions were scored by the ISAGA index of 0–5 (negative); 6–8 (borderline) and 9–12 (positive).
Statistical analysis
Data were recorded in a Microsoft Excel spreadsheet and analysed using Stata v. 11·0 for Windows (Stata Corp., USA). Seroprevalence was expressed as the number of seropositive samples in the total number of samples tested. A χ 2 test was used to determine associations between risk factors and seropositivity; and the strength of these associations were measured by logistic regression analysis; the first level was taken as a reference to determine levels of significance. Non-collinear variables of P ⩽ 0·20 in univariable analysis were entered into a multivariable regression model and goodness-of-fit was assessed using the Hosmer–Lemeshow test [Reference Dohoo, Martin and Stryhn18]. The level of statistical significance was set at P ⩽ 0·05.
RESULTS
A total of 354 people participated in the study (age range 5–70 years, median 30 years) with a higher proportion of females (53·7%) to males (46·3%) (Table 1). Approximately one-quarter (27·1%) of subjects kept cats at home, and consumed raw meat (26·8%). IgG T. gondii seropositivity in all participants was 65·82% (233/354; 95% CI 60·62–70·75), with 66·3% females and 65·2% males. Of the 167 females tested, 8·98% (15/167; 95% CI 0·05–0·14) and 65·9% (110/167; 95% CI 0·58–0·73) were seropositive for T. gondii IgM and IgG, respectively, with varied distribution across the districts (Fig. 1). Likewise, of the 43 pregnant women, 33 (76·4%) and four (9·3%) were positive for specific IgG and IgM antibodies, respectively (Table 2), with a higher IgM seropositivity in the second trimester of pregnancy (Fig. 2). By χ 2 analysis, none of the investigated risk factors showed an association with IgM seropositivity (Table 1).
Fig. 1. Distribution of anti-T. gondii IgG and IgM antibodies of the examined women (n = 167) in the study districts.
Fig. 2. Distribution of anti-T. gondii IgG and IgM antibodies of pregnant women (n = 43) in the study districts.
Table 2. Seroprevalence of anti-T. gondii antibodies in pregnant women
DAT, Direct agglutination test; ISAGA, immunosorbent agglutination assay.
Risk factor analysis
Univariable analysis showed that age, residence, district, source of meat, keeping cats at home, presence of feral cats, and water source were significantly associated with T. gondii IgG seropositivity (Table 3). Accordingly, by multivariable analysis, age, sex and cats kept at home were identified as predictors of IgG seropositivity (Table 4) and variables such as residence and pregnancy in the model were dropped relative to water source and sex due to collinearity. Living in Babile district (OR 2·24, 95% CI 1·25–4·01, P = 0·007) increased the chance of T. gondii infection by twofold compared to other districts and individuals aged from 19 to 35 years (OR 4·32, 95% CI 1·91–9·75, P < 0·001) were more likely to be infected with T. gondii compared to other age groups. On the other hand, consumption of beef meat (OR 5·67, 95% CI 1·00–32·21, P = 0·050) resulted in a sixfold increase in the chance of an individual being infected than consumption of any other meat source. Furthermore, keeping cats at home (OR 2·01, 95% CI 1·11–3·65, P = 0·021) increased by twofold the likelihood of acquiring T. gondii infection in the household, whereas drinking pipe water was almost six times more likely to result in infection (OR 6·07, 95% CI 2·70–16·64, P < 0·001) compared to drinking water from other sources (Table 4). Analysis of the model fitness showed a difference between the observed and predictive value. The Hosmer–Lemeshow χ 2 = 2·84, area under curve (AUC) = 0·7272 and P = 0·9439 indicated that the model fitted the data.
Table 3. Results of univariable logistic regression analysis for predictors of T. gondii IgG seropositivity
OR, Odds ratio; CI, confidence interval.
† Shoats meat (sheep and goat).
‡ Mixed (shoats and beef meat).
§ Both (boiled and raw).
|| Mixed (stream and well water).
* Significant.
Table 4. Multivariable logistic regression analysis of predictors of T. gondii IgG seropositivity
aOR, Adjusted odds ratio; CI, confidence interval.
† Shoats meat (sheep and goat).
‡ Mixed (shoats and beef meat).
§ Mixed (stream and well water).
* Significant.
DISCUSSION
The seroprevalence of toxoplasmosis (anti-T. gondii IgG antibodies) in the human population in Eastern Hararghe, Ethiopia was 65·82% and anti-T. gondii IgM antibodies were detected in 8·98% of women. This rate is consistent with the 60% seroprevalence reported from an earlier study in Ethiopia [Reference Negash, Tilahun and Medhin11] but is lower than the rate range (81·4–96·7%) found in the most vulnerable groups from different parts of the country [Reference Gebre-Xabier12–Reference Gebremedhin16, Reference Walle19]. Nevertheless, this and other studies underpin the fact that the seroprevalence of toxoplasmosis is substantially higher in Ethiopia than other parts of the world, where reported rates range from 9·3% to 43·8% [Reference Berger20–Reference Mwambe24]. Several factors apart from geographical area might account for such differences and include the sensitivity and specificity of the test employed, demographic characteristics and socioeconomic status of the studied populations [Reference Gebre-Xabier12, Reference Alvarado-Esquivel25, Reference Abu-Madi26]. The present study showed increased seropositivity with advancing age, which might reflect the longer exposure of older individuals to infection from various sources leading to lifelong elevated IgG antibody levels as similarly recorded by others [Reference Gebre-Xabier12, Reference Gebremedhin16, Reference Alvarado27, Reference Zemene28].
It was observed that people living in Babile district were twice as likely to be infected with the parasite as those in other districts. The rural population in Babile district are semi-pastoralists, moving frequently with animals seeking pasture and water thus possibly increasing exposure to sources of infection. The district has a relatively hot climate and is at a lower altitude compared to the other districts, these conditions better favour the sporulation and long-term survival of oocysts than colder climates at high altitude [Reference Dubey1]. The observed high seroprevalence in people residing in urban (84·3%) compared to rural (61·3%) areas, might be connected with increased consumption of raw meat in the former (52·9%) than the latter (20·4%), which is facilitated by the availability of retail meat. By contrast, in China no significant association was evident between T. gondii infections and urban living; consumption of raw meat and keeping pet animals are not popular practices in China [Reference Xiao21]. Viable T. gondii is rarely found in beef meat and isolation of the parasite from the tissue of cattle has not to our knowledge been documented [Reference Dubey29]. Nevertheless, in this study, the chance of acquiring infection through consuming raw beef meat was 5·67 times (95% CI 1·00–34·21) higher than for other raw meat sources (P = 0·05), and is probably explained by cultural differences in different regions of Ethiopia. Similarly, several authors have noted an association between seropositivity and raw meat consumption [Reference Dubey1, Reference Negash, Tilahun and Medhin11, Reference Yimer14, Reference Berger20, Reference Chiang22].
The observed difference in seroprevalence of T. gondii between people having or not having domestic cats in the household (76·0% vs. 62·0%, respectively) reached statistical significance (P < 0·05) and suggests high environmental contamination with shedding of infective oocysts by cats. This finding is consistent with some previous studies [Reference Gebremedhin16, Reference Zemene28]; however, other studies have reported the absence of such an association [Reference Gebre-Xabier12, Reference Nijem and Al-Amleh30]. However, the contribution of feral cats to environmental contamination and subsequent exposure of people to the infection should not be discounted. We found a slightly higher IgM response in women (8·98%) compared to previous local reports (2·5–4·2%) [Reference Gebremedhin16, Reference Zemene28], which could be due to the demographic variables and degree of exposure to the source of infection. The finding that 9·3% of pregnant women had anti-T. gondii IgM antibodies signifies the increasing risk of congenital transmission during pregnancy, particularly if the woman acquires the infection for the first time. Such a situation needs to be emphasized to antenatal healthcare practitioners against toxoplasmosis throughout the country. The decrease in anti-T. gondii IgG antibody from the first to second trimester and rise of anti-T. gondii IgM in that order may be attributed to recent infection or reactivation of an existing chronic infection and merits future investigation.
In conclusion, the seroprevalence of toxoplasmosis in the human population in three districts of East Hararghe zone in Ethiopia was considerably high. Water source, age, district and presence of cats at home were found to be risk factors for acquiring T. gondii infection. The moderately high level of IgM seropositivity indicates the presence of current infection and possible occurrence of congenital transmission during pregnancy. Epidemiological studies focusing on congenital toxoplasmosis, and increasing awareness of the disease through education of the people in the study area, are worthy of consideration in the future.
ACKNOWLEDGEMENTS
The financial support of Addis Ababa University, the Ethio-Italian project and the Toxoplasma diagnostic kit support of the Italian National Reference Centre for Toxoplasmosis at the Istituto Zooprofilattico Sperimentale della Sicilia A. Mirri, Palermo, Italy are gratefully acknowledged. The authors also extend their gratitude to the study participants and medical professionals for their collaboration in this study.
DECLARATION OF INTEREST
None.
