Seroprevalence of Bartonella in Eastern China and analysis of risk factors
© Sun et al; licensee BioMed Central Ltd. 2010
Received: 13 November 2009
Accepted: 20 May 2010
Published: 20 May 2010
Bartonella infections are emerging in the Zhejiang Province of China. However, there has been no effort to date to explore the epidemiology of these infections in this region, nor to identify risk factors associated with exposure to Bartonella. The aim of this study was to investigate the seroprevalence of Bartonella in both patients bitten by dogs and blood donors (for control) in Eastern China, and to identify risk factors associated with exposure to Bartonella. As no previous data for this region have been published, this study will provide baseline data useful for Bartonella infection surveillance, control, and prevention.
Blood samples were collected from industrial rabies clinic attendees and blood donors living in eight areas of the Zhejiang Province of China, between December 2005 and November 2006. An indirect immunofluorescent antibody test was used to determine the presence of Bartonella in these samples. Risk factors associated with Bartonella exposure were explored using Chi-square tests and logistic regression analysis of epidemiological data relating to the study's participants.
Bartonella antibodies were detected in 19.60% (109/556) of blood samples. Seroprevalence varied among the eight areas surveys, ranging from over 32% in Hangzhou to only 2% in Jiangshan (X2 = 28.22, P < 0.001). We detected a significantly higher prevalence of Bartonella antibodies in people who had been bitten by dogs than in blood donors (X2 = 13.86, P < 0.001). Seroprevalence of Bartonella was similar among males (18.61%, n = 317) and females (20.92%, n = 239).
Bartonella antibodies were encountered in people living across Zhejiang Province and the seropositivity rate among those exposed to dog bites was significantly higher than that among blood donors, indicating that dog bites may be a risk factor for Bartonella infection.
The genus Bartonella comprises of fastidious, Gram-negative hemotropic bacteria that infect blood erythrocytes and endothelial cells of their hosts . More than 20 species or subspecies of Bartonella have been described, of which at least 12 are recognized as human pathogens . These species exploit a variety of different mammals as reservoir hosts and arthropods as vectors, and provoke a broad spectrum of manifestations in humans.
According to clinical reports, the number of patients with Bartonella infections in Zhejiang Province is the second largest in China. However, to date, there have been no efforts to explore the epidemiology of these infections in this region, nor to identify risk factors associated with exposure to Bartonella. Directly detecting Bartonella in humans is problematic, but detection of antibodies as an indication of exposure to the agent has been widely used in previous studies. In this study, we adopted the latter approach for the first time to investigate human exposure to Bartonella in the Zhejiang Province of China.
Patient blood samples (n = 205 from rabies clinics; n = 351 from normal blood donors) were collected from eight locations (Hangzhou, Tiantai, Longyou, Huzhou, Jiangshan, Chun'an, Longyou, and Linhai) of the Zhejiang Province of China from December 2005 to November 2006. The aims of our study were explained to all participants upon enrollment, and their consent was obtained prior to inclusion in this study. All enrolled participants provided information upon inclusion in the study, with regard to their age, gender, place of residence, and whether they had been exposed to dog bites.
Experimental research reported in this study has been performed with the approval of the ethics committee of Zhejiang Provincial Center for Disease Control and Prevention. Human research was carried out in compliance with the Helsinki Declaration.
A commercially available Bartonella henselae (B. henselae)-based indirect immunofluorescence antibody (IFA) test kit (Euroimmun, Order No. FI219b-1005G) was used to assess the presence of IgG antibodies against B. henselae in human samples. After the samples were diluted 1:100, 1:320, 1:1000, 1:3200, and 1:10000 in phosphate-buffered saline PBS-Tween buffer (provided in the test kit), the IFA assay was conducted following the manufacturer's protocol. Positive and negative controls (provided in the test kit) were also used. Immunofluorescence was observed using an epifluorescence microscope at magnifications of 40× and 200×. According to the manufacturer's guidelines, a titer of 1:320 was considered indicative of an infection, and IgG antibody titers of 1:1000 or higher were judged as an indicator for acute infection.
Chi-square tests and logistic regression analysis of Bartonella seroprevalence among sites, gender, age groups, and exposure to dog bites were conducted using the SPPS, version 11.0 statistical package (Chicago, IL, USA). The dependent variable in the logistic regression was assigned as the serological status of patients and the independent variables were site, gender, age, exposure, and site × exposure. The method of logistic regression used was forward-conditional. The stepwise probability was set to 0.05 for entry and 0.10 for removal. the classification cut-off was 0.5 and the maximum number of iterations was 20. Omnibus tests of model coefficients were also conducted.
Results of positive reaction at different dilution titers
Bitten by dogs (n)
Control group (n)
Zhejiang Province site locations and seroprevalence incidence.
Bartonella Positive (n)
Positive rate (%)
The Chi-square value in omnibus tests of model coefficients was determined to be 60.326 (P < 0.05). Furthermore, the overall correct percentage was found to be 81.5%. Variables in the equation included site, exposure, and site×exposure. The wald of site×exposure, site, and exposure were determined to be 25.477 (P = 0.001), 20.090 (P = 0.005), and 16.174 (P = 0.024), respectively.
Various Bartonella species, including B. henselae, have previously been detected in domestic cats , small mammals, and Haemaphysalis longicornis and Ixodes sinensis ticks  in the Zhejiang Province of China. Of note, transmisson of Bartonella-related infection from any of these sources to humans is feasible. Therefore, it is necessary to study the prevalence of Bartonella in people throughout the Zhejiang Province and to analyze relevant risk factors of infection. In our study, the seroprevalence of Bartonella was found to be 14.81% among healthy patients, a percentage similar to the recently reported percentage (14.28%) recorded in the Yunnan Province in Southwestern China , but much lower than the percentage reported in Beijing (34.5%) . Furthermore, Kikuchi et al.  previously reported a B. henselae specific IgG seroprevalence of 3.1% among adult patients with cardiovascular disease and 10.9% in a high-risk population of healthy veterinary students in Japan, of whom only 0.8% had positive results of serologic testing for B. henselae specific IgM. The bases for these differences are not clear, but are likely to be multi-factorial.
Due to the inherent serological cross-reactivity among Bartonella species in IFA assays [8, 9], the observed seropositivity to B. henselae antigen may actually represent a previous or current infection with another Bartonella species. According to our analysis (carried out according to the manufacturer's instruction), cross reactivity with other Bartonella species cannot be ruled out; however, this kit has more than 84% specificity and 88% sensitivity for B. henselae IgG. As a result, our data suggest that detected antibodies were against all Bartonella species, while the majority of them were against B. henselae.
The fact that patients with Bartonella antibodies were detected across all eight study sites of the Zhejiang Province indicates that the general population in this region is at risk of exposure to Bartonella. Of note, the prevalence varied between sites within the province, although Hangzhou, Tiantai, and Longyou had the highest prevalence of positive reactions. Furthermore, the exposure rate due to dog bites, which was identified as a significant determinant of seroprevalence, also varied between sites, and was significantly higher in Hangzhou than other sites, and significantly lower in Jiangshan than other sites. Moreover, site×exposure fit the equation in logistic regression analysis, indicating that site and exposure results correlated to each other.
We also found that seroprevalence of Bartonella in individuals bitten by dogs was significantly higher than that among unbitten blood donors. However, the explanation of this finding is not altogether obvious, as domestic cats, rather than dogs, are considered a primary reservoir host for Bartonella. It is established that dogs serve as hosts for Ctenocephalides felis, the suspected vector of Bartonella. Therefore, dogs may acquire and maintain infected and infectious fleas, thereby, indirectly acting as a source of infection. Alternatively, people living in proximity to dogs may also be more likely to live in proximity to cats. However, it remains a possibility that dog bites themselves may transmit Bartonella infection, as Bartonella DNA has been detected in dog saliva .
Previously published studies have indicated that B. henselae infections are more common in children than adults. However, we did not see a significantly higher seropositivity among children than adults in our studies. In contrast, in our study the highest seropositivity was observed in the oldest (>59 years) patient age group. Others have reported similar findings .
Bartonella infections can be difficult to diagnose clinically, particularly as many of the clinical presentations associated with them have nonspecific symptoms, such as fever. Thus, reliance on clinical reports from physicians to determine the public health burden of infection is not feasible. In contrast, serological surveys of individuals offer a means of quantifying the level of exposure to Bartonella among the population. In this study, we used this approach to demonstrate that exposure to this pathogen is both common and widespread across the Zhejiang Province of China, although prevalence was higher in some parts of the province than others. Furthermore, it appeared that people bitten by dogs were significantly more likely to have Bartonella antibodies than blood donors without exposure to bites. This information should serve to help those in relevant settings to consider Bartonella infection in the differential diagnosis of their patients.
We thank Guojing Si, Fengxia Meng, Haixia Wu, Dongmei Li, Dongsheng Ren, Jun Wang, Baoxin Cui, and Xiaoran Yang for their help. We also extend appreciation to the Hangzhou Center for Disease Control and Prevention (CDC), Wenzhou CDC, Tiantai CDC, Jindong CDC, Longquan CDC, Linhai CDC, Chun'an CDC, Shangyu CDC, Longyou CDC, Ouhai CDC, Jiande CDC and Jiangshan CDC for their help.
This study was supported by the National Natural Science Foundation of China (NSFC) grant 30371246 and the Project of the State Scientific & Technological Development of the 10th Five Year Plan grant 2003BA712A09-02.
- Boulouis HJ, Chang CC, Henn JB, Kasten RW, Chomel BB: Factors associated with the rapid emergence of zoonotic Bartonella infections. Vet Res. 2005, 36: 383-410. 10.1051/vetres:2005009.View ArticlePubMedGoogle Scholar
- Henn JB, Gabriel MW, Kasten RW, Brown RN, Theis JH, Foley JE, Chomel BB: Gray foxes (Urocyon cinereoargenteus) as a potential reservoir of a Bartonella clarridgeiae-like bacterium and domestic dogs as part of a sentinel system for surveillance of zoonotic arthropod-borne pathogens in northern California. J Clin Microbiol. 2007, 45: 2411-2418. 10.1128/JCM.02539-06.View ArticlePubMedPubMed CentralGoogle Scholar
- Guiming Fu, Jimin Sun, Qiyong Liu, Tianci Yang, Zhangyao Ren, Gangqiang Ding: Epidemiology investigation of Bartonella henselae in cats from Zhejiang Province. Chinese Journal of Vector Biology and Control. 2008, 19: 138-140.Google Scholar
- Sun J, Liu Q, Lu L, Ding G, Guo J, Fu G: Coinfection with four genera of bacteria (Borrelia, Bartonella, Anaplasma, and Ehrlichia) in Haemaphysalis longicornis and Ixodes sinensis ticks from China. Vector Borne Zoonotic Dis. 2008, 8: 791-796. 10.1089/vbz.2008.0005.View ArticlePubMedGoogle Scholar
- Hui Yang, Heming B, Falian Y, Binbin Y: Serological survey on Bartonella infection in Yunnan. Chinese Journal of Natural Medicine. 2007, 9: 277-280.Google Scholar
- Yang XR, Liu QY, Cui BY, Wang LX, Peng ZH, Ren DS: Using direct enzyme linked immunosorbent assay for the detection of IgG antibody on Bartonella henselae among healthy people in Changping, Beijing. Zhonghua Liu Xing Bing Xue Za Zhi. 2007, 28: 688-691.PubMedGoogle Scholar
- Kikuchi E, Maruyama S, Sakai T, Tanaka S, Yamaguchi F, Hagiwara T: Serological investigation of Bartonella henselae infection in clinically cat scratch disease-suspected patients, patients with cardiovascular diseases and healthy veterinary students in Japan. Microbiol Immunol. 2002, 46: 313-316.View ArticlePubMedGoogle Scholar
- Dalton MJ, Robinson LE, Cooper J, Regnery RL, Olson JG, Childs JE: Use of Bartonella antigens for serologic diagnosis of cat-scratch disease at a national referral center. Arch Intern Med. 1995, 155: 1670-1676. 10.1001/archinte.155.15.1670.View ArticlePubMedGoogle Scholar
- La Scola B, Raoult D: Serological cross-reactions between Bartonella quintana, Bartonella henselae, and Coxiella burnetii. J Clin Microbiol. 1996, 34: 2270-2274.PubMedPubMed CentralGoogle Scholar
- Duncan AW, Maggi RG, Breitschwerdt EB: Bartonella DNA in dog saliva. Emerg Infect Dis. 2007, 13: 1948-1950.View ArticlePubMedPubMed CentralGoogle Scholar
- Pandak N, Dakovic-Rode , Cabraja I, Kristof Z, Kotarac S: Prevalence of Bartonella henselae antibodies in children and blood donors in Croatia. Infection. 2009, 37: 166-167. 10.1007/s15010-008-8113-0.View ArticlePubMedGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2334/10/121/prepub