Industrial Dairy Cattle Farms in Hungary Source of Coxiella burnetii Infection in Humans

Introduction

Q fever is a zoonotic disease distributed worldwide, with the exception of New Zealand. The causative agent of the disease is Coxiella burnetii, which is an obligate intracellular bacterium. The main reservoirs of the bacterium are sheep, goats, and cattle (Eldin et al. 2017), although many authors have reported that C. burnetii can also be found in wild and marine mammals, reptiles, birds, and ticks (Gyuranecz et al. 2012, Eldin et al. 2017). C. burnetii infection causes Q fever in humans. The primary animal reservoirs, which can shed C. burnetii in their urine, feces, milk, and birth products, are mainly responsible for human infections (Angelakis and Raoult 2010). Many human Q fever outbreaks have been associated with small ruminants (Ghaoui et al. 2019) and, therefore, more research studies can be found about C. burnetii infections in sheep and goats than in dairy cattle.

The main clinical sign in Coxiella-infected small ruminants is abortion during late pregnancy. However, infected animals can also shed the bacteria mostly with the placenta and birth fluids after a normal pregnancy (Ghaoui et al. 2019). Placental membranes can contain huge number of bacteria that are excreted during parturition. After drying, C. burnetii can be aerosolized and infect humans by inhalation (Schimmer et al. 2012). We found a well-designed study about bovine coxiellosis (Guatteo et al. 2006), but the knowledge about spread among dairy cattle and transmission to humans in large intensive dairy units is limited. Information about Q fever in industrial dairy cattle farms, especially in the central European region, is not so detailed as in other regions of Europe. The aims of this study were to determine the importance of Q fever in dairy cattle farms as a factor posing a high zoonotic risk to humans, and to estimate the seroprevalence of C. burnetii in different occupational groups of farm workers.

Materials and Methods

Human blood samples collected from 70 dairy farm workers between February and July 2020 were tested for the presence of antibodies to C. burnetii. The study was approved by the Hungarian Scientific Ethics Committee and all subjects provided their informed consent. The eight dairy cattle farms included in the survey were located in different parts of Hungary, equally distributed between two large statistical geographic regions of the country (Transdanubia, Great Plain, and North). All dairy units had between 600 and 1,000 milking cows, which had previously been found to be enzyme-linked immunosorbent assay (ELISA) and PCR positive for C. burnetii by bulk tank milk testing. Industrial dairy farms were included of this study based on willingness to provide a human blood sample and data from all workers to the authors. Participation in the study was voluntary and we encouraged farms to participate in this research but many of them refused this due to the zoonotic potential of the disease and for fear of being subjected to potential official restrictions. The five occupational groups were categorized into three risk groups based on their possible close contact with C. burnetii infected animals, mostly placenta and other birth products.

The examined group consisted of veterinarians (n = 8) and inseminators (n = 12), as a high-risk occupational group, animal caretakers (n = 26), as a medium-risk group, and herd managers (n = 7) and milking parlor workers (n = 17), as a low-risk group. The population under study consisted of 13 women and 57 men, aged between 19 and 64 years. Questionnaires were used to record the participants' demographic data, occupation, length of employment at the farm, and any symptoms of a potential previous Q fever (any fever with headache, pneumonia, myalgia, hepatitis, and swollen lymphnodes) (Supplementary Table S1). Serum samples were tested for the presence of IgG reacting with phase I and phase II antigens of C. burnetii strain Nine Mile using a commercially available immunofluorescence assay (Focus Diagnostics, Cypress, CA). The test was performed according to the manufacturer's instructions in a reference laboratory at the National Public Health Centre (Budapest, Hungary). Cutoff values of 1:16 serum dilutions were applied as screening procedure, followed by defining the end titers of seropositive samples.

Risk groups were compared by pairwise Fisher exact tests. p Values were adjusted for multiple comparisons using false discovery rate (FDR) correction. The level of significance was set to 0.05. Statistical analysis was performed in R 4.0.2 (R Core Team 2020). The study was performed in accordance with the principles of the Declaration of Helsinki. No other approval from institutional ethics committee was required for this study.

Results

IgG phase I antibodies were detected in 53 out of the 70 (75.7%) serum samples tested. In 59 out of the 70 individuals tested (84.3%), anti-C. burnetii IgG phase II was detected. Among the IgG I seropositive individuals, there were 8 out of 8 veterinarians (100%), in 12 out of 12 inseminators (100%), in 22 out of 26 animal caretakers (84.6%), in 8 out of 17 parlor workers (47%), and in 3 out of 7 herd managers (42.8%). IgG phase II antibodies to C. burnetii were found in 8 out of 8 veterinarians (100%), in 12 out of 12 inseminators (100%), in 26 out of 26 animal caretakers (100%), in 8 out of 17 milking parlor workers (47%), and in 5 out of 7 herd managers (71.4%) (Table 1). The titers of IgG antibodies showed a wide variation, whereas only a few very high values >1:256 were observed among them (Table 2).

There was a correlation between the occupation and the seropositivity rate. Applying Spearman's rank correlation, we found a statistically significant correlation between the length of employment and the percentage of positivity (r = 0; p < 0.001) among the pooled groups. There was no correlation between age and the seropositivity rate. The seropositivity rate was 37.5% in farm staff employed at dairy farms for <1 year, 83.3% in farm workers with a length of employment between 2 and 5 years, and 94.7% in staff employed at dairy farms for >5 years (Table 3). We found a confirmed case of acute Q fever in one veterinarian during the past few years. The percentage of IgG positivity was found to be higher in men (89.4%) than in women (61.5%). Considering IgG phase I, high-risk and medium-risk groups had significantly higher chance of seropositivity than the low-risk group (p = 0.0001 and p = 0.0099, respectively). No statistically significant difference was found between the high-risk and medium-risk groups in the occurrence of IgG phase I seropositivity (p = 0.1213). Regarding IgG phase II, the occurrence of seropositivity was significantly higher in the high-risk and medium-risk groups than in the low-risk group (p = 0.0005 and p = 0.0002, respectively). No significant difference was found between high-risk and medium-risk groups in IgG phase II seropositivity (p = 1.0000).

Discussion

The number of annually reported acute human Q fever infections in Hungary ranged between 28 and 48 from 2015 to 2019 (ECDC report; European Centre for Disease Prevention and Control 2020). Although some of these cases were associated with farm workers (Balla, unpublished data), unfortunately there are no official reports about them. A recent study has found that C. burnetii infection rapidly increased in Hungarian dairy farms due to the growing number of animals in dairy units and farm structures moving toward concentration. The prevalence of C. burnetii was found to be 97.6% based on ELISA and PCR test findings of the bulk tank milk (Dobos et al. 2020a). Samples of retained placenta from Hungarian dairy herds showed 65.2% C. burnetii positivity by PCR and 57.1% positivity by immunohistochemistry (Dobos et al. 2020b). A larger herd size could pose a risk because of the increased chance of C. burnetii introduction or the presence of a larger susceptible population of cows. Cattle shed C. burnetii mainly with birth products such as the placenta, amniotic fluids, and vaginal mucus (Guatteo et al. 2006b). These bacteria are also shed in the milk, feces, and urine, but the most important sources of Coxiella transmission from animals to humans are birth products (Guatteo et al. 2006). As the surveyed large dairy farms have ∼600–1,000 calvings per year, the chance of farm workers to acquire Coxiella infection is much higher than in smaller family farms.

Many studies have reported that C. burnetii was found in parturient bovine placentae (Hansen et al. 2011, Botta et al. 2019). However, C. burnetii is generally inactive in the fetal membranes, and human Q fever outbreaks are rarely associated with cattle, but infected placental membranes provide a huge opportunity for the dissemination of viable organisms to the environment to infect both humans and cattle (Luoto and Huebner 1950, Hansen et al. 2011). This study has demonstrated a high prevalence of C. burnetii antibodies in dairy farm workers. All workers had daily contacts with dairy cattle, and although we found different seropositivity rates among different occupational groups, it is difficult to compare our serological results with other findings from different countries, because the pooled groups were not entirely the same and they have used different screening tests and cutoff values. However, the 84.2% prevalence of IgG phase II and the 75.7% prevalence of IgG phase I antibodies to C. burnetii found in Hungarian farm workers were much higher than the seropositivity rates demonstrated in English farmers (27%) (Thomas et al. 1995) or in Polish farmers (17.8%) (Cisak et al. 2003). A recent study has found 24.1% seropositivity among Ecuadorian farm workers (Echeverría et al. 2019). A southern Italian survey conducted in Sicily found 21.4% and 25% prevalence of anti-Coxiella antibodies in female and male farm workers, respectively (Fenga et al. 2015). Very low prevalence of C. burnetii IgM immunoglobulin (4.6%) was detected in livestock workers in Trinidad (Adesiyun et al. 2011).

A large-scale study detected an 11% seroprevalence rate of anti-C. burnetii IgG among humans coming into contact with dairy cattle in Denmark, but only 3% of the surveyed Danish dairy farmers were seropositive (Bosnjak et al. 2010). The seroprevalence rate of dairy cattle farm residents was the highest (72.1%) in the Netherlands of all values reported all over the world (Schimmer et al. 2014). In this study, we found 100% seropositivity among veterinarians and inseminators. We detected the highest prevalence in veterinarians, which is consistent with research findings from many other countries. In Denmark, 36% of the veterinarians tested had antibodies while only 2% of inseminators were seropositive (Bosnjak et al. 2010). Among veterinarians, the seroprevalence of antibodies to C. burnetii was 13.5% in Japan (Abe et al. 2001), 9.5% in Austria (Nowotny et al. 1997), and 22.2% in the United States (Whitney et al. 2009). According to an Estonian study, the prevalence of C. burnetii antibodies was 9.62% among veterinary professionals, but among farm animal veterinarians, the seroprevalence was significantly higher, 17.39% (Neare et al. 2019). The prevalence of C. burnetii seropositivity found in Dutch livestock veterinarians was 69.2% (Wielders et al. 2015). Many international studies clearly indicate that farm veterinarians are the most important occupational risk group (Wielders et al. 2015, Brom et al. 2013). The reason why Hungarian farm veterinarians had such a high C. burnetii infection rate was probably their intensive contact with highly infected dairy farms. Veterinarians work with sick animals during parturition, remove retained fetal membranes, treat metritis, flush out the uterus, and they are exposed to contact with infected placenta and birth products on a daily basis. Inseminators of large industrial dairy herds often have the same duties as veterinarians and, therefore, they constitute the second most important occupational risk group.

We found 100% seropositivity among animal caretakers, who are also at a high risk of becoming infected with C. burnetii because of their close contact with infected cattle. They also have a close contact with bedding materials, which are also a source of Coxiella transmission from animals to humans (Guatteo et al. 2006). We found 47% seropositivity among milking parlor workers and 71.4% among herd managers. These two occupational groups had less contact with animals and mostly with birth products. In milk, sporadic shedding of Coxiella is the most common kinetic pattern (Guatteo et al. 2007). As the milking machine is a relatively closed system, the milk is probably not a common source of Coxiella transmission from cattle to milking parlor workers. Herd managers are regularly present on dairy farms to manage the milk production, but sometimes they assist with calvings or act as substitutes for inseminators; in such cases they have a higher chance of becoming infected with C. burnetii. In 2 out of the 70 subjects (2.8%) examined in this study, the significantly elevated titers of IgG phase I antibodies to C. burnetii ≥1:1024 titers, indicating the likelihood of chronic Q fever and further clinical examinations (Dupont et al. 1994).

Conclusion

In conclusion, this study has demonstrated that IgG phase I and phase II antibodies to C. burnetii are higher in Hungarian dairy farm workers than those described in several international seroepidemiological studies among different occupational groups in other European countries. Veterinarians are the occupational group most exposed to infection, but inseminators and animal caretakers are at a similarly high risk of infection in industrial dairy farms.

The high Coxiella burden in dairy farms underlines the importance of controlling the development of chronic Q fever among occupationally exposed people, such as dairy farm workers as well as the need for implementation some preventive measurements. The prevalence of C. burnetii was found very high among dairy cattle farms according to our recent study, and we found similar high human seroprevalence of C. burnetii among dairy farm workers as we have previously assumed. Our study also has demonstrated that high C. burnetii seroprevalence among dairy farm workers correlated with the high prevalence of C. burnetii in Hungarian dairy herds. It would be highly advisable to set up an obligatory national Q fever monitoring system in Hungarian dairy farms in both human workers and the animals in the future.