Lawsonia intracellularis (L. intracellularis) is a gram-negative, fastidious obligate intracellular bacterium. The pathogen mostly occupies the small and large intestine in pigs and causes porcine proliferative enteropathy (PPE) post weaning. Infection leads to diarrhea, retarded growth and/or sudden death in fattening pigs. A high prevalence of L. intracellularis has been reported throughout the world, contributing to a substantial level of economic loss in the swine industry [1–3]. In some herds, the disease may manifest itself as severe hemorrhagic diarrhea with relative high mortality [4].

The first case of L. intracellularis infection in pigs was described in 1931 [5] and since that time, L. intracellularis has been reported in swine producing countries all over the world. In Denmark, 94% of tested animals were positive by PCR on feces [6], in Sweden 48% of herds were positive in fecal samples tested using nested PCR [7] and in Korea [8] 47% of herds were positive when fecal samples were tested using multiplex PCR. More recent reports indicated a 100% seropositivity in Korea [9], 91% in the USA [10] and 84% in Australia [11]. In Australia L. intracellularis has been estimated to cost the industry in the order of USD 25 per sow annually [12] and direct losses of USD 3 to 11 per affected animal [13].

Highly intensive management of domestic pigs is widely promoted in mainland China, where there is an estimated pig population of approximately 8.3 billion [14]. Severe diarrhea occurs frequently in fattening pigs and pregnant sows, having negative impacts on herd feed conversion rates and herd profitability. Importantly, L. intracellularis infection receives comparatively little attention from animal health authorities compared to highly pathogenic infections such as porcine reproductive and respiratory syndrome (PRRS), classical swine fever (CSF) and Streptococcus suis infection.

In 2008, the first isolate of L. intracellularis was identified from the intestinal mucosa of infected pigs in Southern China [15]. The apparent prevalence of L. intracellularis infected pigs identified by a PCR method was 14% and 16%, in weaners and finishers, respectively in Guangxi province, Southern China [15]. To the best of the authors’ knowledge the prevalence of L. intracellularis infection in pigs in other areas of China has not been reported. Moreover, the major transmission routes are unclear in different stock classes and production systems. In this study our aim was to determine the seroprevalence of L. intracellularis in pigs raised in the major pig-producing provinces in China. An additional aim of our study was to document the association between L. intracellularis positivity and the presence of diarrhea.

A cross-sectional survey carried out between January and May 2011 to estimate the seroprevalence of L. intracellularis infection in Chinese pigs. Seven provinces (Beijing, Hebei, Tianjin, Henan, Hubei, Guangdong and Guanxi) took part in the study. Sampling was carried out using a two-stage cluster design. Two herds from each of the seven provinces were selected at random from a sampling frame of 55 intensive pig cooperatives listed by the Chinese Ministry of Agriculture. Sample size calculations were carried out to determine the appropriate number of individual pigs to be sampled from each of the selected herds. Previous reports have estimated the within-herd prevalence of L. intracellulari infection to be around 13% in weaning pigs and 16% in finishing pigs [15]. Based on a previous report [16], a credible estimate of the intra-cluster correlation coefficient for L. intracellularis infection was 0.06. We assumed an average cluster (i.e. herd) size of 80 pigs. Sample size calculations were carried out on the basis that we wanted to be 95% certain that our final estimate of L. intracellularis prevalence was within 5% of the true population value.

A total of 1060 serum samples were collected from pigs from the 14 herds that took part in the study. This included 147 sera from pre-weaning piglets (2 to 4 weeks of age), 221 sera from weaned piglets (5 to 7 weeks), 279 sera from fattening pigs (8 to 14 weeks), 255 sera from adult sows, and 158 sera from boars. Landrace and Large White (breeds exotic to China) and domestic breeds (Meishan and Jinhua) were represented in the sample. The pig numbers were 332, 287, 211 and 230 for Landrace, Large White, Meishan and Jinhua pigs, respectively. Serum samples were stored at −80°C until assayed.

To identify the association between L. intracellularis seropositivity and the presence of diarrhea, 659 sera were obtained from herds with animals having a recent history of clinical diarrhea while an additional 401 samples were collected from herds without a recent history of diarrhea.

All sera were tested for antibodies against L. intracellularis using a blocking ELISA purchased from Synbiotic Europe SAS (Lyon, France) according to the manufacturer’s instructions. The sensitivity and specificity of this test was 72% and 93%, respectively [17].

The apparent prevalence of L. intracellularis seropositivity was calculated using the number of pigs identified as seropositive divided by the total number of pigs tested. Apparent prevalence estimates were then converted to true prevalence estimates using the approach proposed by Rogan and Gladen [18]. The Rogan and Gladen approach was used in this situation (in preference to more complex Bayesian methods) because the apparent prevalence estimates were relatively high. Confidence limits for the apparent and true prevalence estimates were determined using the Wilson binomial approximation [19, 20]. Confidence intervals for both the apparent and true prevalence estimates were adjusted using the design effect, a function of the intra-cluster correlation coefficient for L. intracellularis infection and the average cluster size. The significance of differences between the true prevalence L. intracellularis seropositivity for pigs in different groups (i.e. location, stock class and breed) was assessed by considering their confidence limits. If the confidence intervals for two groups did not overlap, the inference was that the two prevalence estimates were significantly different at the alpha level of 0.05 [21]. All the calculations were performed using the AusVet epiTools procedures [22].

This study was approved by the Ethical Committee for Animal Experiments at China Agricultural University, Beijing, China.

L. intracellularis antibodies were found in pigs of all ages. The true prevalence of L. intracellularis seropositivity was 77% (95% CI 70 to 83%). The true prevalence of L. intracellularis seropositivity ranged from 68-87% on stock class.

In this study, using a blocking ELISA, the overall true prevalence of L. intracellularis seropositivity was 77% (95% CI 70 to 83%). A higher rate of seroprevalence was found in fattening pigs, sows and boars compared with pre-weaning piglets and weaners. Breeds exotic to China had a higher seroprevalence compared with domestic pig breeds. This is the first report on the infection of L. intracellularis across China. The above evidence indicates that L. intracellularis infection is widespread among intensive pig farms in China.

In this study, none of the pigs had been immunized with L. intracellularis attenuated vaccine, but a high seroprevalence was found in the collected serum samples indicating exposure to the bacterium. The apparent seroprevalence estimates in this study are comparable with those reported in other studies. Apparent seroprevalence was 90% for sows and 56% for fattening pigs in a study conducted in the USA [23], while growing pigs and fattening pigs were found to be 45% and 59% seropositive, respectively, in a Korean study using an immunofluorescent assay as the detection method [9]. In Australia all herds tested positive for L.intracellularis-specific antibodies and the mean within-herd prevalence of positive samples was 84% [11]. Variation in the sensitivity and specificity of the diagnostic methods used to detect the presence of L.intracellularis contributes to the different serological prevalence estimates reported in Europe and Asia [17, 21].

Pigs reared in herds in Central China had true prevalence estimates that were higher compared with pigs reared in herds in the North and South of the country. We speculate that the reason for this finding is that higher seroprevalence rates are associated with intensive pig production units which are common in Central China. The apparent prevalence of L. intracellularis seropositivity was similar for sows and boars and markedly higher than that reported for similar stock classes in other countries. Adult sows with active infection are likely to transmit infection to suckling piglets and infected boars transmit pathogens to sows by artificial insemination. Based on a single assessment of a herd’s diarrhea status we note that the prevalence of seropositivity in herds with a recent history of clinical diarrhea was similar to that of herds without a recent history of diarrhea which would imply some latency of L. intracellularis in sows and boars. The high true prevalence of seropositivity in pre-weaning piglets and boars free of diarrhea does not exclude the possibility that diarrhea was caused by other pathogens. Pig diarrhea is caused by a number of pathogens and mixed infections, including E. coli, Brachyspira hyodysenteriae, Lawsonia intracellularis, and Salmonella spp. [24].

The high true prevalence of L. intracellularis seropositivity identified in this study can be explained by a number of factors. First, inappropriate use of in-feed antibiotics contributes to L. intracellularis infection. In some situations there is a risk of animals developing a resistance to antibiotics such as tetracyclines, sulfonamides, fluroquinalones and zinc bacitracin that are routinely used in feed formulations [25]. A recent study based on data collected from pig herds in Southern China reported that antibiotic excretion rates for sows was 48 mg/day, weaning piglets 19 mg/day, growing pigs 7 mg/day and finishing pigs 1.5 mg/day [26] indicating that the usage of antibiotics and therefore excretion masses are correlated with stock class. High L. intracellularis seropositivity is known to be associated with withdrawal of in-feed medication in fattening pig diets. In this study, the true prevalence of L. intracellularis seropositivity for fattening pigs (81%, 95% CI 69 to 92%) was significantly greater compared with pre-weaning piglets (38%, 95% CI 19 to 57%) and weaners (29%, 95% CI 15 to 44%). To some extent, a low true prevalence of L. intracellularis seropositivity reflects the presence of long-term treatment with antibiotics [9].

A second explanation for the high true prevalence of L. intracellularis seropositivity is that the presence of infection in breeding herds may arise from the introduction of breeding pigs into a herd without application of appropriate quarantine measures. Every year, thousands of European and American breeding pigs are purchased and imported into China by commercial pig farmers. In most situations the risk awareness of the introducing L. intracellularis infection is underestimated and clinical cases of L. intracellularis infection are often misdiagnosed by local veterinary authorities as well as the pig purchaser. The apparent prevalence of L. intracellularis infection in exotic breeding pigs ranged from 26% to 74% in previous reports [6, 27]. One survey indicated that 56% of seropositive pigs in Korea originated from pigs born outside of Korea [9]. Recent reports note that L. intracellularis is spread between herds mainly through the purchase of infected pigs and replacement stock [27, 28]. The importance of transmission by non-porcine vectors is unknown. Following experimental inoculation, histological lesions develop in laboratory mice, rats and hamsters, but not in sparrows or chickens [29].

In addition to the above, transmission from boars is likely to play a role in the spread of L. intracellularis both within and between herds. In this study, the true prevalence in boars could have contributed to infection rates in sows bred by artificial insemination. In the herds that took part in this study, sows were mated using boar semen without antibiotic treatment. The impact of this practice on between-herd spread of infection needs to be clarified.

We conclude that the prevalence of L. intracellularis seropositivity is relatively high in Chinese pigs. The highest true prevalence was observed in sows and boars, suggesting that within a herd these stock classes are a reservoir for infection. The prevalence of L. intracellularis seropositivity in domestic pigs was significantly less than that in imported breeds. A higher seroprevalence was found in pigs in herds in Central and Northern China, which may correspond to the greater use of the intensive production systems in these areas.