Article Text


Erysipelas in laying hens is associated with housing system
  1. H. Eriksson, DVM1,2,
  2. A.-K. Nyman, MSc, PhD1,
  3. C. Fellström, DVM, PhD, Professor2 and
  4. P. Wallgren, DVM, PhD, Professor1,2
  1. 1Department of Animal Health and Antimicrobial Strategies, National Veterinary Institute, SE-751 89 Uppsala, Sweden
  2. 2Department of Clinical Sciences, Swedish University of Agricultural Sciences, Box 7054, SE-750 07 Uppsala, Sweden
  1. E-mail for correspondence: helena.eriksson{at}


Following the change from conventional cages to non-cage housing systems and furnished cages, which in Sweden was finalised by 2005, problems caused by Erysipelothrix rhusiopathiae increased in laying hen flocks. This study aimed to investigate possible associations between housing systems for laying hens and outbreaks of erysipelas. Also, sera from 129 flocks in different housing systems, collected during 2005–2007, were analysed for the presence of antibodies to E rhusiopathiae using an indirect ELISA test. Antibodies were detected in all housing systems. The mean flock absorbance values from free-range flocks were significantly higher than corresponding values from other housing systems. Data on the Swedish laying hen population were compared with the recorded number of erysipelas outbreaks during 1998–2011. Outbreaks occurred on 15 farms with indoor litter-based systems (n=87 farms in 2011). No outbreak was diagnosed on farms with flocks in conventional or furnished cages. The results indicate that the risk for an outbreak was higher in free-range systems than in indoor litter-based systems, and lowest for flocks housed in cages. Absence of erysipelas in the majority of subsequent flocks on the affected farms suggested that proper measures, including vaccination, were undertaken.

Statistics from


The bacterium Erysipelothrix rhusiopathiae has recently been reported to cause disease outbreaks (erysipelas) in laying hen flocks in non-cage housing systems (Mazaheri and others 2005, Kaufmann-Bart and Hoop 2009, Eriksson and others 2010, Stokholm and others 2010). Such systems include indoor litter-based systems as well as free-range systems with access to outside pens (Anonymous 2006).

The Swedish Animal Welfare Ordinance of 1988 states that hens kept for the production of eggs must be housed in systems that fulfil the birds' needs of nests, perches and dust-baths (Anonymous 2009). Therefore, housing systems for Swedish laying hens have been changed from conventional cages to furnished cages (ie, cages with nests, perches and dust-baths) and non-cage systems. The process was most intense during the years 2001–2004, and by 2005 almost all flocks were housed in the new systems (Brasch and Nilsson 2008). Since 1998, erysipelas has been identified as a disease problem associated with high mortality and decreased egg production in flocks housed in indoor litter-based and free-range systems in Sweden, and a study on the causes of mortality of Swedish laying hens during 2001–2004 indicated a higher risk of erysipelas in laying hens in litter-based housing systems than in conventional and furnished cages (Fossum and others 2009). This observation is of general interest, since the use of conventional cages has been banned in the EU since 2012 (CEC 1999).

Reports on serological responses to natural exposure of laying hens to E rhusiopathiae are scarce, but experimental infection of chickens have shown that the presence of antibodies detected by a growth agglutination (GA) test correlates with exposure to E rhusiopathiae (Takahashi and others 1994). When employing the GA test on serum samples from laying hens housed in cages, it was suggested that exposure to E rhusiopathiae was common (Takahashi and others 2000). High prevalence of antibodies to E rhusiopathiae has also been demonstrated in serum samples from laying hens in New Zealand with an ELISA test (Kurian and others 2012).

The overall objective of this study was to examine associations between housing systems and the occurrence of E rhusiopathiae infections in laying hen flocks. This was done by comparing the occurrence of erysipelas outbreaks diagnosed in commercial Swedish laying hen flocks in different housing systems from 1998 to 2011. In addition, the presence of antibodies to E rhusiopathiae in serum collected from laying hens from different housing systems at slaughter during the years 2005–2007 was investigated.

Material and methods


Data collection

Information on the number of flocks diagnosed with erysipelas per housing system per year was compiled from data available at the National Veterinary Institute (SVA). Housing systems were divided into three categories: cage systems (conventional and furnished cages), indoor litter-based systems and free-range systems. Data included all flocks diagnosed with erysipelas at SVA during the years 1998–2011. Diagnosis was based on postmortem examination with pathological findings consistent with E rhusiopathiae infection; hepatomegaly, splenomegaly and occasionally valvular endocarditis, necrotic hepatitis and necrotic splenitis (Fossum and others 2009) and subsequent isolation of E rhusiopathiae from spleen or liver (Eriksson and others 2009). In addition, flocks where E rhusiopathiae infection had been found during necropsy at regional laboratories in Sweden were included, provided that the results were confirmed at SVA by demonstration of E rhusiopathiae.

Data on the laying hen population in Sweden (percentage of hens per housing system) were collected for the years 1999, 2001, 2003, 2005 and 2007 from the Swedish Board of Agriculture (Brasch and Nilsson 2008), and for the year 2011 from the Swedish Association of Egg Producers (Anonymous 2012).

Serological investigation


Serum samples from 129 Swedish laying hen flocks were included in the study. The flocks were sampled at slaughter during the years 2005–2007 within the Swedish surveillance programme for avian influenza based on guidelines from the EC (CEC 2005). Four to 10 samples per flock were available for serological analysis. Information on housing systems for the sampled flocks was collected, and the flocks were assorted as housed in furnished cages, indoor litter-based or free-range systems. The distribution of flocks between years and housing system is shown in Table 1.


Distribution per year and housing system of Swedish laying hen flocks investigated for antibodies to Erysipelothrix rhusiopathiae using an indirect ELISA method

Detection of serum antibodies to E rhusiopathiae

Antibodies to E rhusiopathiae were analysed by adapting an indirect ELISA system developed for detection of antibodies to E rhusiopathiae in pigs (Wallgren and others 2000). Briefly, the following procedures were performed. To obtain antigen E rhusiopathiae, serotype 1 was grown on horse blood agar (SVA) in 37°C for 18—hours. The growth was harvested in 2—ml PBS without Ca and Mg, pH 7.4 (SVA). The suspension was then treated in an ultrasonic disintegrator (Measuring Scientific Equipment, London, England) for five minutes per 8—ml cell suspension. The solution was centrifuged at 12,000—g for 20—minutes at 4°C (RC2B, Sorvall, Newtown, USA) and the liquid phase was used as antigen.

Before analysis, the antigen was diluted 1:20,000 in coating carbonate buffer (disodium carbonate 1.59—g/l, sodium hydrogen carbonate 2.93—g/l, sodium azide 0.2—g/l, SVA) and 100—μl of the solution was transferred to each well in a microtiter plate (Greiner M129A, Greiner GmBH, Frickenhausen, Germany) which was incubated at 20°C overnight. After three subsequent washings with PBS supplemented with 0.1 per cent Tween 20 (PBS-T; SVA), 100—μl of serum diluted 1:100 in PBS-T was added to double wells, and the plates were incubated at 20°C for two hours. The plates were then washed three times with PBS-T before 100—μl of conjugate (horseradish peroxidase conjugated rabbit antibody to chicken/turkey IgG (Zymed Laboratories, San Francisco, California, USA), diluted 1:5 000 in PBS-T) was added to each well, and the plates incubated for one hour at 37°C. After three washes with PBS-T, 100—μl of SVANOVIR Substrate Solution (Svanova Biotech AB, Uppsala, Sweden) was added to each well. The reaction was stopped after 10—minutes in room temperature by adding 100—μl 2M H2SO4 to each well. The absorbance was read at 450—nm in an ELISA reader (Tecan Sunrise, Tecan Nordic AB, Mölndal, Sweden).

When adapting the ELISA, sera from 56 grandparent chickens aged 7–9—weeks, from three different flocks kept under stringent biosecurity measures were analysed in order to get an indication on absorbance values in a naive population. The mean absorbance value for these sera was 0.06±0.01 (range 0.043–0.095) which corresponds to a cut-off value A450=0.2 (the mean absorbance value+4 SDs+0.1).

Statistical analysis

Descriptive statistics was used to present the results of the outbreaks of erysipelas and the distribution of the flock absorbance values. A mean absorbance value was calculated for each sampled flock based on the individual absorbance values. The associations between flock absorbance values and housing system (furnished cages, indoor litter-based or free-range) were investigated using univariable linear regression analysis. The statistical analysis was performed using Stata Software (Release V.11.2; College Station, Texas, USA: StataCorp LP).



The laying hen population per year and housing system is presented in Fig 1. During this period, the dominant system of conventional cages was changed to other production systems. In 2011, there were 194 farms with hens in indoor litter-based systems, 87 farms with furnished cages and 84 free-range farms. Between 1998 and 2011, erysipelas was diagnosed in 51 flocks in total (Fig 2). Of these, 22 flocks were from 15 farms with indoor litter-based systems, and 29 flocks were from 21 free-range farms. Erysipelas was not diagnosed in any flock in cages during the period studied.


Percentage of the laying hen population (number of hens) in different housing systems in Sweden during the years 1999–2011


Number of erysipelas outbreaks diagnosed in Swedish laying hen flocks in different housing systems during the years 1998–2011. No outbreak was recorded in a laying hen flock in conventional or furnished cages

Erysipelas was diagnosed in more than one flock on three of the 15 farms with indoor litter-based systems. Two of the farms that had several houses on the production sites had outbreaks in four flocks each. On the third farm, the disease reoccurred in a flock in the same house two years after the first disease outbreak.

On three of the 21 free-range farms, the disease was diagnosed more than one time. On one farm, the disease reoccurred several times during the period studied (1998, 1999, 2001, 2004, 2004 and 2009), both in subsequent flocks in the same house and in flocks housed in another house on the farm. On the second free-range farm, outbreaks reoccurred in a subsequent flock in the same house, erysipelas was at that time also diagnosed in another house. On the third farm, the disease was diagnosed in two flocks simultaneously housed in adjacent houses.

Levels of serum antibodies to E rhusiopathiae

The flock absorbance values for all sampled flocks are presented in Fig 3, and show an approximate normal distribution. The overall mean flock absorbance value was 0.79±0.20 (n=129), which was significantly higher (P<0.001) than the absorbance value obtained in the seronegative grandparent chickens (0.06±0.01). For flocks housed in furnished cages the mean absorbance value was 0.76±0.13 (n=40), for flocks in indoor litter-based systems 0.75±0.21 (n=39) and for free-range flocks 0.88±0.22 (n=50) (Fig 4). Free-range flocks had significantly higher (P<0.005) absorbance values compared with the other housing systems. No statistical difference could be observed when comparing the absorbance values of samples from flocks in furnished cages and flocks in indoor litter-based systems.


Histogram of the flock absorbance values obtained using an indirect ELISA method for detecting antibodies to Erysipelothrix rhusiopathiae for 129 Swedish laying hen flocks


Box-and-whisker plot of flock absorbance values obtained using an indirect ELISA method for detecting antibodies to Erysipelothrix rhusiopathiae in laying hen flocks in different housing systems. The box shows the lower quartile, median and the upper quartile, and the whiskers shows the upper and lower adjacent values (1.5 interquartile range from the lower and upper quartile, respectively). The two dots for furnished cages are outside values


The European egg production has recently undergone large changes of housing systems due to welfare demands for laying hens and the ban of conventional cages within EU from 2012 (CEC 1999). This process was finished in Sweden already in 2005. During and after the switch of housing systems in Sweden, infections with E rhusiopathiae has been diagnosed in laying hens in indoor litter-based and free-range systems (Fossum and others 2009, Eriksson and others 2010), an observation, that due to the new legislation within EU, ought to be of general interest.

In the present study, a clear association between housing system and outbreaks of erysipelas was found, as no outbreak was recorded in a laying hen flock in conventional or furnished cages during the period studied, an observation that is in line with previous results (Fossum and others 2009). As only one report of an erysipelas outbreak in a flock housed in conventional cages (Mutalib and others 1993) and no report of any outbreak in furnished cages was found in the literature, it appears that housing of hens in cages constitute a decreased risk for outbreaks of erysipelas. Despite this, the flock absorbance values from these flocks indicated an exposure to E rhusiopathiae, and antibodies to E rhusiopathiae have also been demonstrated in hens in cages by others (Takahashi and others 2000, Kurian and others 2012).

The transmission and pathogenesis of E rhusiopathiae in laying hen flocks has not yet been clearly defined (Bricker and Saif 2008). Thus, the epidemiology of erysipelas in laying hens needs to be scrutinised to explain the differences in outbreak incidence between housing systems. However, once introduced to a flock, E rhusiopathiae is thought to spread mainly through the faecal-oral route. Differences in occurrence of erysipelas outbreaks between different housing systems may be attributed to the fact that laying hens housed in litter-based systems are continuously exposed to faecally contaminated litter and equipment in contrast with hens in cages.

The majority of the farms affected by erysipelas only experienced the disease once, which indicates that appropriate disease preventing measures were undertaken following diagnosis, as well as onwards. In order to minimise the risk for transmission of E rhusiopathiae between flocks, farmers were recommended to revise and sharpen their biosecurity routines. In addition, vaccination of subsequent flocks was strongly recommended on all affected farms, as not even thorough cleaning and disinfection of the house was expected to eliminate the bacteria to a low enough level. For farms with more than one flock, vaccination could also be a measure to prevent outbreaks in previously unaffected flocks on the farm during an outbreak of erysipelas.

To our knowledge, there is no previously published study where the risk for an outbreak of erysipelas in flocks housed in indoor litter-based systems has been compared with the risk in free-range flocks. Unfortunately, the switch from production in conventional cages is not documented in detail in Sweden. Hence, the populations at risk in the different housing systems were not known, and no statistics could be calculated. However, by scrutinising the recorded number of outbreaks and the available population data in our study, free-range farms, with 21 affected farms during the years 1998 to 2011 and a total number of 84 farms in 2011, seem to be at a higher risk for an outbreak of erysipelas than flocks in the indoor litter-based systems with 15 affected farms during the same period, and a total number of 194 farms in 2011.

Free-range flocks are kept under conditions with an increased risk for exposure to E rhusiopathiae present in the wild fauna, and there is also a risk for a buildup of contamination by E rhusiopathiae in the outside pen. It has been believed that the bacterium could survive indefinitely in soil, although experimental studies have indicated a maximum survival time of around three months depending on environmental conditions (Szynkiewicz 1964, Wood 1973). In the future, a general vaccination of free-range flocks may be considered, but further documentation of the risk for outbreaks in free-range flocks are desirable. Such studies ought to include investigations on the presence of E rhusiopathiae in the wild fauna as well as studies on the survival of the bacterium in the environment of laying hens.

When employing the cut-off value defined from young chickens it was evident that the laying hens had been exposed to E rhusiopathiae. However, since E rhusiopathiae is a ubiquitous microbe (Wang and others 2010), adult birds are likely to have been exposed to the bacterium and absorbance levels above the cut-off value are not necessarily indicating the development of disease. Increasing levels of antibodies to E rhusiopathiae with age should, therefore, be regarded as normal, and has also been documented in other species, such as pigs and human beings (Wallgren and others 2000), and also recently described in laying hens (Kurian and others 2012).

Thus, adult birds are assumed to have significant levels of antibodies to E rhusiopathiae, and the results obtained indicated true serological responses to the bacterium in laying hens in all housing systems, despite a relatively limited number of outbreaks of erysipelas diagnosed per year. Subclinical infections and the fact that some strains of E rhusiopathiae are apathogenic to chickens (Takahashi and others 1994) could be explanations for this. Yet, comparing mean absorbance values may be interesting since the pathogen load may differ between housing systems.

By contrast with a study on antibodies to E rhusiopathiae in poultry in New Zealand, which did not detect any association between housing system and presence of antibodies (Kurian and others 2012), we observed significant differences in flock absorbance values between housing systems. In agreement with the distribution of outbreaks between the different housing systems, the highest flock absorbance values were obtained in free-range flocks which also supported the theory that free-range flocks are more at risk for exposure to E rhusiopathiae.

Interestingly, the flock absorbance values for flocks in furnished cages did not differ significantly from that of flocks housed in indoor litter-based housing systems. As no outbreak of erysipelas was reported in any flock in cages during the period studied, a lower level of antibodies to E rhusiopathiae than in the other production systems could have been expected. However, previous serological studies have shown that flocks housed in cages also are exposed to E rhusiopathiae (Takahashi and others 2000, Kurian and others 2012).

In conclusion, our study shows that laying hens, commonly, are exposed to E rhusiopathiae. The probability of an outbreak of erysipelas is dependent on the housing system. Flocks in free-range systems are at a higher risk than flocks in indoor litter-based systems, and flocks in cages appear to be at the lowest risk. Considering these results, the prophylactic measures may have to diverge between housing systems and vaccination is one way to prevent erysipelas in flocks at risk.


The authors wish to thank Sigbrit Mattsson for excellent technical work in the laboratory. The Albert Hjarre foundation is acknowledged for funding of the serological analysis.


View Abstract


  • Provenance: not commissioned; externally peer reviewed

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.