Paper
Paper Coxiella burnetii seroprevalence and risk factors on commercial sheep farms in The Netherlands B. Schimmer, M. M. A. de Lange, J. L. A. Hautvast, P. Vellema, Y. T. H. P. van Duynhoven Coxiella burnetii seroprevalence was assessed on Dutch dairy and non-dairy sheep farms using ELISA. Risk factors for seropositivity on non-dairy sheep farms were identified at farm and sheep level by univariate and multivariate multilevel analyses. Based on 953 dairy and 5671 non-dairy serum samples, sheep seroprevalences were 18.7 per cent and 2.0 per cent, respectively, and 78.6 per cent and 30.5 per cent at farm level. Significant risk factors for nondairy sheep farms were farm location in the south of the country, sheep kept on marginal grounds, one or several supply addresses for ewes during 2007–2009 and wearing farm boots and/or outfit by professional visitors. On sheep level, risk factors included among others farm location in the south of the country, lamb breeding as main farm purpose, goat density within 10 km farm radius, use of windbreak curtain or windshields, and presence of ≥6 stillborn lambs in 2009. Farm location in the south of the country and goat density suggests that infected goats have played a role in the transmission to non-dairy sheep. Other risk factors suggest introduction of the bacterium through sheep supply and professional visitors. Biosecurity measures should be strengthened, including avoiding infection during handling of stillborn lambs and birth products in the lambing period.
Introduction
Coxiella burnetii infections in sheep are generally not considered a veterinary concern due to lack of significant impact on animal health. In The Netherlands, approximately 1,1 million sheep were kept in 2010. Most sheep are kept for meat production next to a relatively small dairy sheep sector for the production of dairy products. Besides, sheep can also be kept for breeding purposes or nature management. The highest sheep density is found in the coastal provinces. In The Netherlands, dairy sheep are usually milked twice a day during several months a year, and most of them are kept outdoors part of the year. Non-dairy sheep are generally kept outside, except for a few weeks around lambing that usually occurs in a stable. During the protracted Dutch Q fever epidemic in 2007–2009, abortion waves at dairy goat and dairy sheep farms were considered the primary source of human infections (Roest and others 2011). The transmission of C. burnetii from non-dairy sheep farms to human beings seemed not as important as for dairy small ruminant farms (goat and sheep), although direct contact with non-dairy sheep was identified as the most probable source for several human Q fever clusters in The Netherlands during this period (Koene and others 2011, Whelan and others 2012) and internationally (Lyytikainen and others Veterinary Record (2014) B. Schimmer, MD, M. M. A. de Lange, Msc, Y. T. H. P. van Duynhoven, PhD, National Institute for Public Health and the Environment, Centre for Infectious Disease Control Netherlands, P.O. Box 1, Bilthoven 3720 BA, The Netherlands J. L. A. Hautvast, MD, PhD, Department of Primary and Community Care, Academic Collaborative Centre for Public Health AMPHI, Radboud University Nijmegen Medical Centre,
doi: 10.1136/vr.102155 P.O. Box 9101, Nijmegen 6500 HB, The Netherlands P. Vellema, DVM, PhD, Dip.ECSRHM, Department of Small Ruminant Health, Animal Health Service (GD), P.O. Box 9, Deventer 7400 AA, The Netherlands E-mail for correspondence: [email protected] Provenance: not commissioned; externally peer reviewed Accepted March 2, 2014
1998, Grilc and others 2007, Porten and others 2006). While sourcefinding studies during the Q fever epidemic in The Netherlands identified positive non-dairy sheep farms, no elevated risk for human Q fever was observed in the vicinity of such farms (van der Hoek and others 2012). The low sheep seroprevalence of 2.4 per cent in 2008 found in a convenience sample indicated that non-dairy sheep might play a minor role in the Dutch Q fever epidemic (Van den Brom and others 2013). In other European countries, recent seroprevalences on sheep level ranged between 3.1 per cent and 18.9 per cent and ranged between 3.0 per cent and 74.0 per cent on herd level. In 2008, the C. burnetii seroprevalence was also studied on dairy goat farms with 7.8 per cent of individual goats testing positive (ELISA) with the highest seroprevalence in the southeastern part of the country (Van den Brom and others 2013). In the same year, the herd seroprevalence on Dutch dairy cattle farms was 78.6 per cent (ELISA) and 56.6 per cent (PCR) based on a submitted bulk tank milk (BTM) sample by each farm (Muskens and others 2011). Representative data on seroprevalence of C. burnetii infections in dairy and non-dairy sheep are scarce. Moreover, risk factors for C. burnetii seropositivity in sheep have been rarely studied (Thomas and others 1995, Ruiz-Fons and others 2010, Van den Brom and others 2013). The aim of the current study was to estimate the C. burnetii seroprevalence in a representative sample of non-vaccinated dairy and non-dairy sheep at farms with over 100 sheep in The Netherlands. Furthermore, we identified risk factors for seropositivity at farm and sheep level at non-dairy sheep farms in order to update control measures and to provide targeted advice for the Dutch sheep industry.
Materials and methods
Study design and sampling strategy
In this cross-sectional study in 2009–2011, we approached 33 dairy sheep farms and 1344 non-dairy sheep farms with at least 100 breeding ewes in November 2008, according to the national identification and registration database, for serum sampling and a farm questionnaire (Figs 1 and 2). A farm with a minimum of 100 ewes was chosen as it was considered to be a commercial sheep farm. In the early stage July 5, 2014 | Veterinary Record
Paper
33 dairy sheep farms with >100 ewes approached
19 non-participating farms (56.2%)
14 participating farms (43.8%) (953 samples)
No farm questionnaire 2 farms (120 samples)
Farm questionnaire 12 farms (833 samples)
FIG 1: Study participation of invited commercial dairy sheep farms with ≥100 ewes, The Netherlands, 2009–2010
1344 non-dairy sheep farms > 100 ewes approached, 32 non-eligible after initial contact
Paper Coxiella burnetii seroprevalence and risk factors on commercial sheep farms in The Netherlands B. Schimmer, M. M. A. de Lange, J. L. A. Hautvast, P. Vellema, Y. T. H. P. van Duynhoven Coxiella burnetii seroprevalence was assessed on Dutch dairy and non-dairy sheep farms using ELISA. Risk factors for seropositivity on non-dairy sheep farms were identified at farm and sheep level by univariate and multivariate multilevel analyses. Based on 953 dairy and 5671 non-dairy serum samples, sheep seroprevalences were 18.7 per cent and 2.0 per cent, respectively, and 78.6 per cent and 30.5 per cent at farm level. Significant risk factors for nondairy sheep farms were farm location in the south of the country, sheep kept on marginal grounds, one or several supply addresses for ewes during 2007–2009 and wearing farm boots and/or outfit by professional visitors. On sheep level, risk factors included among others farm location in the south of the country, lamb breeding as main farm purpose, goat density within 10 km farm radius, use of windbreak curtain or windshields, and presence of ≥6 stillborn lambs in 2009. Farm location in the south of the country and goat density suggests that infected goats have played a role in the transmission to non-dairy sheep. Other risk factors suggest introduction of the bacterium through sheep supply and professional visitors. Biosecurity measures should be strengthened, including avoiding infection during handling of stillborn lambs and birth products in the lambing period.
Introduction
Coxiella burnetii infections in sheep are generally not considered a veterinary concern due to lack of significant impact on animal health. In The Netherlands, approximately 1,1 million sheep were kept in 2010. Most sheep are kept for meat production next to a relatively small dairy sheep sector for the production of dairy products. Besides, sheep can also be kept for breeding purposes or nature management. The highest sheep density is found in the coastal provinces. In The Netherlands, dairy sheep are usually milked twice a day during several months a year, and most of them are kept outdoors part of the year. Non-dairy sheep are generally kept outside, except for a few weeks around lambing that usually occurs in a stable. During the protracted Dutch Q fever epidemic in 2007–2009, abortion waves at dairy goat and dairy sheep farms were considered the primary source of human infections (Roest and others 2011). The transmission of C. burnetii from non-dairy sheep farms to human beings seemed not as important as for dairy small ruminant farms (goat and sheep), although direct contact with non-dairy sheep was identified as the most probable source for several human Q fever clusters in The Netherlands during this period (Koene and others 2011, Whelan and others 2012) and internationally (Lyytikainen and others Veterinary Record (2014) B. Schimmer, MD, M. M. A. de Lange, Msc, Y. T. H. P. van Duynhoven, PhD, National Institute for Public Health and the Environment, Centre for Infectious Disease Control Netherlands, P.O. Box 1, Bilthoven 3720 BA, The Netherlands J. L. A. Hautvast, MD, PhD, Department of Primary and Community Care, Academic Collaborative Centre for Public Health AMPHI, Radboud University Nijmegen Medical Centre,
doi: 10.1136/vr.102155 P.O. Box 9101, Nijmegen 6500 HB, The Netherlands P. Vellema, DVM, PhD, Dip.ECSRHM, Department of Small Ruminant Health, Animal Health Service (GD), P.O. Box 9, Deventer 7400 AA, The Netherlands E-mail for correspondence: [email protected] Provenance: not commissioned; externally peer reviewed Accepted March 2, 2014
1998, Grilc and others 2007, Porten and others 2006). While sourcefinding studies during the Q fever epidemic in The Netherlands identified positive non-dairy sheep farms, no elevated risk for human Q fever was observed in the vicinity of such farms (van der Hoek and others 2012). The low sheep seroprevalence of 2.4 per cent in 2008 found in a convenience sample indicated that non-dairy sheep might play a minor role in the Dutch Q fever epidemic (Van den Brom and others 2013). In other European countries, recent seroprevalences on sheep level ranged between 3.1 per cent and 18.9 per cent and ranged between 3.0 per cent and 74.0 per cent on herd level. In 2008, the C. burnetii seroprevalence was also studied on dairy goat farms with 7.8 per cent of individual goats testing positive (ELISA) with the highest seroprevalence in the southeastern part of the country (Van den Brom and others 2013). In the same year, the herd seroprevalence on Dutch dairy cattle farms was 78.6 per cent (ELISA) and 56.6 per cent (PCR) based on a submitted bulk tank milk (BTM) sample by each farm (Muskens and others 2011). Representative data on seroprevalence of C. burnetii infections in dairy and non-dairy sheep are scarce. Moreover, risk factors for C. burnetii seropositivity in sheep have been rarely studied (Thomas and others 1995, Ruiz-Fons and others 2010, Van den Brom and others 2013). The aim of the current study was to estimate the C. burnetii seroprevalence in a representative sample of non-vaccinated dairy and non-dairy sheep at farms with over 100 sheep in The Netherlands. Furthermore, we identified risk factors for seropositivity at farm and sheep level at non-dairy sheep farms in order to update control measures and to provide targeted advice for the Dutch sheep industry.
Materials and methods
Study design and sampling strategy
In this cross-sectional study in 2009–2011, we approached 33 dairy sheep farms and 1344 non-dairy sheep farms with at least 100 breeding ewes in November 2008, according to the national identification and registration database, for serum sampling and a farm questionnaire (Figs 1 and 2). A farm with a minimum of 100 ewes was chosen as it was considered to be a commercial sheep farm. In the early stage July 5, 2014 | Veterinary Record
Paper
33 dairy sheep farms with >100 ewes approached
19 non-participating farms (56.2%)
14 participating farms (43.8%) (953 samples)
No farm questionnaire 2 farms (120 samples)
Farm questionnaire 12 farms (833 samples)
FIG 1: Study participation of invited commercial dairy sheep farms with ≥100 ewes, The Netherlands, 2009–2010
1344 non-dairy sheep farms > 100 ewes approached, 32 non-eligible after initial contact
