INTRODUCTION
Syphilis, a sexually transmitted disease (STD) caused by Treponema pallidum, currently remains a worldwide public health problem, which reflects the failure of syphilis control programmes and the lack of political advocacy [Reference Lin1–Reference Tong3]. Demographic, behavioural and health-related characteristics that increase the risk of syphilis infection have been studied extensively, and co-infection with other STDs is one of the most reliable indicators of syphilis risk. Blocker et al. reviewed 30 studies and concluded that infection with syphilis increased the risk of human immunodeficiency virus (HIV) infection with an odds ratio of 8·5 for men and 3·3 for women [Reference Blocker, Levine and St Louis4]. Some studies have suggested that co-infection with syphilis increases the risk of infection with hepatitis C virus (HCV) [Reference Wandeler5, Reference Witt6]. Association between infection with syphilis and hepatitis B virus (HBV) has also been observed [Reference McQuillan7], with some variation in different population subgroups [Reference Tien8, Reference Tsakris9].
At present, studies about other factors associated with syphilis infection, such as risk-taking behaviours and medical conditions, which potentially affect the immune system and susceptibility, are limited. Syphilis and HBV infection are both endemic in China [Reference Lin1, Reference Liang10], thus further research on the association between syphilis and HBV exhibits great value both academically and practically. Here, we conduct a retrospective case-control study to identify behavioural factors and medical conditions associated with syphilis infection.
METHODS
Study population and data collection
A retrospective case-control study was performed in Zhongshan Hospital, Medical College of Xiamen University from January 2013 to June 2014. A total of 17 585 inpatients were screened for syphilis infection by serological tests. Patients excluded from the study were 251 returning patients, 28 patients with undefined test results and two patients with unqualified blood samples. Of the remaining 17 304 patients, 675 had positive serological tests for syphilis and were ultimately diagnosed with syphilis based on clinical presentation, clinical data and medical history. Eleven patients with congenital syphilis were excluded leaving 664 patients to be included in subsequent analysis. Eight hundred age- and sex-matched controls were selected at random from the 16 629 non-syphilis inpatients at the hospital (Fig. 1). The study design was approved by the Institutional Ethics Committee of Zhongshan Hospital, Medical College of Xiamen University, and was found to be in compliance with national legislation and the Declaration of Helsinki guidelines.
Selection criteria of the study subjects.
All study participants were interviewed by physicians and data were collected on demographic and geographical variables (e.g. ethnicity, city of origin), health history and behavioural variables including tobacco use and alcohol consumption. Non-local residents were regarded as migrants. Current co-infection with other STDs (e.g. HIV, HBV infection) and liver and renal functions were assessed clinically.
Clinical diagnosis
According to the European guideline on the management of syphilis [Reference Janier11], syphilis was clinically diagnosed in this study by combining the serodiagnosis and disease history (including clinical characteristics and/or the patient's sexual history), in line with our previous studies [Reference Tong2, Reference Tong3]. Laboratory tests for syphilis were performed using the Treponema pallidum particle agglutination assay (Fujirebio, Japan), and an automated chemiluminescence immunoassay (Boson Biotechnology, China) according to the manufacturers’ instructions and as previously reported [Reference Qiu12]. The fluorescent treponemal antibody absorption (Euroimmun Medizinische Labordiagnostika, Germany) test was used to discriminate the discordant results. Enzyme-linked immunosorbent assay (ELISA) for HBV was performed for determination of hepatitis B surface antigen (HBsAg), antibody to HBsAg (anti-HBs), hepatitis B e antigen (HBeAg), antibody to HBeAg (anti-HBe) and antibody to hepatitis B core antigen (anti-HBc) (In Tec, China). Subjects were considered HBV-positive if they had a positive test for HBsAg and/or anti-HBc [Reference Weinbaum13]. ELISA methods were also performed for HCV antigens (In Tec, China) and for exposure to HIV (HIV1 + 2 antigens/antibodies; Beijing Wantai Biological Pharmacy Enterprise Co. Ltd, China). Serum alanine transaminase (ALT), aspartate transaminase (AST), total bilirubin, blood urea nitrogen (BUN) and creatinine were quantitated using a Roche Cobas 8000 automatic biochemistry analyser (Roche Diagnostics, F. Hoffmann-La Roche Ltd, Switzerland).
Statistical analysis
All statistical analyses were performed using SPSS v. 19·0 for Windows (SPSS Inc., USA). Univariate and multivariate logistic regression analyses were performed and crude and adjusted odds ratios (ORs) and 95% confidence intervals (CIs) were calculated. A two-sided P value < 0·05 was considered to be statistically significant.
RESULTS
Description of syphilis patients
From January 2013 to June 2014, 675 inpatients at Zhongshan Hospital were diagnosed with syphilis, showing a prevalence of 3·9% (675/17 304). After excluding 11 congenital syphilis cases, 664 syphilis patients were enrolled for further analysis. The median age of syphilis patients in the study was 58 years (range, 20–93 years), including 417 males (62·8%) and 247 females (ratio 1·69:1). However, the male:female ratio varied by 10-year age group as follows: 0·44:1 (age 20–29), 0·56:1 (age 30–39), 1·14:1 (age 40–49), 2·18:1 (age 50–59), 2·18:1 (age 60–69), 2·53:1 (age 70–79), 2·38:1 (age 80–89), 0·5:1 (age 90–93) (Fig. 2).
Age and gender of 664 syphilis patients.
Univariate analysis
The characteristics of the 664 syphilis patients were compared to the characteristics of non-syphilis patients. Migrants between cities in Fujian Province had increased odds of syphilis infection, compared to local residents. An increased odds of syphilis infection was also evident for married participants and divorced/widowed participants, compared to those unmarried. Apparently, tobacco use was significantly associated with syphilis infection, while alcohol consumption was not. No significant association was observed between syphilis infection and ethnicity, employment status or drug use (Table 1). Interestingly, an increased odds of syphilis infection was observed for participants with elevated BUN levels (>8·2 mmol/l). Participants with hypertension or female participants with a reproductive history also had increased probability of syphilis infection. However, other renal function parameters, liver function parameters, history of surgery, transfusion or diabetes mellitus were not significantly associated with syphilis infection (Table 1).
Univariate analysis for association with syphilis infection
OR, Odds ratio; n.a., not applicable; M, male; F, female; ALT, alanine transaminase; AST, aspartate transaminase; BUN, blood urea nitrogen; sCr, serum creatinine; HIV, human immunodeficiency virus; HBV, hepatitis B virus; HCV, hepatitis C virus.
a The sum of numbers in some subgroups may not add up to the total number of subjects because of missing values.
b Fisher's exact test P value was 0·454.
c Data for the history of vaginal delivery or abortion are for women only.
d Data for hepatitis B virus infection are for HIV-negative patients only.
e Data for hepatitis C virus infection are for HIV-negative patients only.
Of 1464 subjects, HIV status was determined to be negative for 1213 study subjects but 251 patients were lacking HIV test results. There were 1165 HIV-negative study subjects tested for HBV and 1114 HIV-negative study subjects tested for HCV. HBV infection was associated with increased risk of syphilis infection (OR 2·073, P < 0·001), while HCV infection was not (Table 1).
Further multiple-level investigation was performed on the associations between syphilis infection and tobacco use, as well as alcohol consumption by stratifying the participants into different groups (Table 2). Results indicated that current smokers had increased odds of syphilis infection, while no significant difference was observed between former smokers and non-smokers. Furthermore, whether current or former smokers who consumed tobacco for >20 pack-years or for >15 years were associated with increased probability of syphilis infection. Compared to no alcohol consumption, the duration or frequency of alcohol use was not associated with syphilis infection. It should be noted that current or former drinkers who consumed alcohol for >15 years had more than threefold increased odds of syphilis infection compared to those with 1–15 years of alcohol consumption (data not shown). Similarly, participants who consumed alcohol (>28 drinks/week) had a twofold increased odds of syphilis infection compared to those consuming 1–28 drinks/week (data not shown). An intensive study of the association between syphilis and serological reactivity to HBV is also shown in Table 2. Compared to the pattern negative for all HBV serological markers, an isolated anti-HBc + pattern, an anti-HBs + /anti-HBc + pattern, an HBsAg + /anti-HBe + /anti-HBc + pattern, and an anti-HBs + /anti-HBe + /anti-HBc + pattern were all associated with the increased probability of syphilis infection, with P values 0·004, 0·002, 0·015 and 0·039, respectively.
Multiple-level analysis about the association between syphilis and tobacco use, alcohol use and HBV serological data
HBV, hepatitis B virus; OR, odds ratio; HBsAg, hepatitis B surface antigen; anti-HBs, antibody to HBsAg; HBeAg, hepatitis B e antigen; anti-HBe, antibody to HBeAg; anti-HBc, antibody to hepatitis B core antigen.
a The sum of numbers in some subgroups may not add up to the total number of subjects because of missing values.
b No tobacco exposure indicates never having smoked cigarettes. Former tobacco exposure indicates having smoked cigarettes daily or irregularly in the past and having quit smoking for ⩾3 months. Current tobacco exposure indicates having smoked cigarettes daily or irregularly in the past 3 months.
c Non-smokers acknowledged smoking for 0 years. Current or former smokers who acknowledged smoking <1 year were recorded as smoking for 1 year.
d Non-smokers acknowledged no tobacco use. Pack-years are calculated by multiplying the average number of packs (1 pack = 20 cigarettes) smoked per day by the number of years a person smoked. Current or former smokers who consumed tobacco at <1 pack-year were recorded as having consumed 1 pack-year.
e No alcohol exposure was defined as never having consumed ⩾1 drink per month. Former alcohol exposure was defined as having consumed ⩾1 drink or more daily or irregularly per month in the past and having stopped drinking for ⩾3 months. Current alcohol exposure was defined as having consumed ⩾1 drink daily or irregularly per month in the past 3 months.
f Non-drinkers were regarded as drinking for 0 year. Current or former drinkers who having been drinking for <1 year were regarded as drinking for 1 year.
g Non-drinkers were regarded as having consuming no alcohol. A drink-equivalent (hereafter called ‘drink’) was defined as one 12-oz beer, one 6-oz glass of wine, one 3-oz mixed drink, or one 1·5-oz shot of liquor. Current or former drinkers who consumed <1 drink on average per week were regarded as having consumed 1 drink per week.
h Others including 1 (HBsAg-, anti-HBs-, HBeAg-, anti-HBe+, anti-HBc-), 5 (HBsAg+, anti-HBs-, HBeAg-, anti-HBe-, anti-HBc+), 1 (HBsAg-, anti-HBs+, HBeAg-, anti-HBe+, anti-HBc-), 2 (HBsAg+, anti-HBs+, HBeAg-, anti-HBe-, anti-HBc+), 2 (HBsAg+, anti-HBs+, HBeAg+, anti-HBe-, anti-HBc+), 11 (HBsAg+, anti-HBs+, HBeAg-, anti-HBe+, anti-HBc+), 5 (HBsAg+, anti-HBs-, HBeAg+, anti-HBe+, anti-HBc+).
Multivariate analysis
Multivariate analysis was performed on 490 syphilis patients and 652 controls, 322 participants were eliminated because of missing values (e.g. lack of HIV results) (Table 3). Migrants between cities in Fujian Province, relative to local residents, remained positively associated with syphilis infection [adjusted odds ratio (aOR) 1·368, 95% CI 1·048–1·785]. An increased odds of syphilis infection was observed in current smokers (aOR 1·607, 95% CI 1·177–2·195). Elevated BUN levels (>8·2 mmol/l) resulted in increased odds in favour of syphilis infection (aOR 1·798, 95% CI 1·188–2·673). Associations between syphilis and different serological patterns of HBV infection were noted. Compared to the HBV-negative pattern, an isolated anti-HBc + pattern, an anti-HBs + /anti-HBc + pattern, an HBsAg + /anti-HBe + /anti-HBc + pattern and an anti-HBs + /anti-HBe + /anti-HBc + pattern were all positively associated with the probability of syphilis infection.
Multivariable analysis for association with syphilis infection
aOR, Adjusted odds ratio; CI, confidence interval; HBV, hepatitis B virus; HBsAg, hepatitis B surface antigen; anti-HBs, antibody to HBsAg; HBeAg, hepatitis B e antigen; anti-HBe, antibody to HBeAg; anti-HBc, antibody to hepatitis B core antigen; BUN, blood urea nitrogen.
* Others including 1 (HBsAg-, anti-HBs-, HBeAg-, anti-HBe+, anti-HBc-), 5 (HBsAg+, anti-HBs-, HBeAg-, anti-HBe-, anti-HBc+), 1 (HBsAg-, anti-HBs+, HBeAg-, anti-HBe+, anti-HBc-), 2 (HBsAg+, anti-HBs+, HBeAg-, anti-HBe-, anti-HBc+), 2 (HBsAg+, anti-HBs+, HBeAg+, anti-HBe-, anti-HBc+), 11 (HBsAg+, anti-HBs+, HBeAg-, anti-HBe+, anti-HBc+), 5 (HBsAg+, anti-HBs-, HBeAg+, anti-HBe+, anti-HBc+).
DISCUSSION
To date, syphilis infection is increasingly prevalent in China. Although individual risk factors for syphilis infection have been well described (unprotected sex, etc.), some potential determinants are less well studied. In this study, we comprehensively evaluated the factors that may influence the likelihood of syphilis infection regarding risk-taking behaviours and medical conditions. Through rigorous analyses, our research indicated that migration, tobacco use, elevated BUN levels and HBV infection were associated with syphilis infection.
Xiamen city is one of the coastal urban regions in China where the syphilis burden is greatest [Reference Tucker and Cohen14]. Our study showed a prevalence of syphilis of 3·9%, nearly as high as the prevalence in other high-risk groups, such as female sex workers (2·4–3·2%) [Reference Wang15], male students who have sex with men (5·0%) [Reference Xu16], and males with commercial sex partners (5·3%) [Reference Yang17]. The male:female ratio of syphilis patients was 1·69:1, higher than reported in the national survey [Reference Chen18], and disproportionate in different age groups. Migration between cities in the same province increased the risk of syphilis infection. Migrants reported more sexual activity, and appeared to be less knowledgeable about STDs and less aware of voluntary counselling and testing services reported in previous studies [Reference Pan19].
HBV infection was prevalent in all subjects, but significantly more common in syphilis than in non-syphilis study subjects. Syphilis and HBV are primarily transmitted by sexual contact, vertical transmission, or unsafe injections (i.e. blood transfusions or dialysis) [Reference Workowski20]. Syphilis and HBV are also more prevalent in the same high-risk subpopulations, including men who have sex with men, drug users, and persons with multiple sexual partners [Reference Workowski20]. Because it is an ulcerative infection, syphilis may also directly facilitate acquisition of HBV [Reference Tien8].
China is one of the countries where HBV is most highly endemic globally, with an estimated 120 million people carrying HBsAg [Reference Liang10]. Meanwhile, syphilis has been ranked as the third most prevalent notifiable infectious disease in China, with >0·4 million new infections each year (www.stats.gov.cn). HBV serological tests of 1165 HIV-negative subjects in this study demonstrated an association between syphilis and HBV infection. Furthermore, there are several types of HBV infection status in clinical practice. Compared to an HBV-negative status, syphilis infection was positively associated with specific HBV serological reactivity. Therefore, this study suggests that HBV-syphilis co-infection may become or already be a serious public health problem.
Syphilis patients are also often co-infected with HIV [Reference Blocker, Levine and St Louis4]. The latter can lead to immunosuppression, which could potentially increase the risk of developing syphilis [Reference Nnoruka and Ezeoke21]. However, the HIV infection rate in this study was fairly low, and the association between syphilis and HIV infection was not analysed. Although HCV-syphilis co-infection was common in previous studies [Reference Wandeler5, Reference Witt6], no association between syphilis and HCV infection was observed in the present study, again, possibly because of a low prevalence of HCV infection. HIV/syphilis/HBV/HCV tests are compulsory in China for inpatients and blood donors to prevent the spread of infectious diseases [Reference Lin22]. This practice also identifies individuals infected with HIV/syphilis/HBV/HCV.
Smoking has been considered a marker of high-risk for syphilis infection, especially when associated with increased sexual activity, more lifetime sexual partners, less condom use, and earlier age at first intercourse and pregnancy [Reference Hansen23, Reference Wolf and Freedman24]. Consistent with this, current smoking and duration and frequency of smoking were associated with higher risk of syphilis infection in the present study. In addition, higher risk of syphilis infection was associated with duration and frequency alcohol consumption. As early as 1901, a study concluded that alcohol use was associated with newly diagnosed syphilis infection [25]. It may also be associated with risky sexual behaviour or casual selection of sex partners [Reference Zetola26], as well as suppression of the immune system or promotion of other biological changes that increase susceptibility to STDs [Reference Cook and Clark27].
In the current study, elevated BUN level was positively associated with syphilis infection, suggesting that reduced renal function may contribute to syphilis infection. Kidney disease may affect general immunity, resulting in persistent systemic inflammation and acquired immunosuppression, which facilitate the infection [Reference Kurts28]. Conversely, syphilis may negatively impact renal function. Renal manifestations of syphilis have been recognized since the 18th century [Reference Thompson29]. A variety of renal conditions have been reported in secondary syphilis, ranging from nephrotic syndrome to acute nephritic syndrome, rapidly progressive glomerulonephritis and renal failure [Reference Singh and Romanowski30]. Nevertheless, BUN can be elevated by many extra-renal factors, such as high protein diet, hypovolaemia, congestive heart failure and gastrointestinal haemorrhage [Reference Beier31].
The retrospective case-control study allowed a better identification of associated factors than a cross-sectional study. However, for associations reported here, it is not known whether factors associated with syphilis infection were present before subjects were exposed to and became infected with T. pallidum. Thus, no causal relationships can be inferred. Furthermore, the current study is hospital-based and nearly 22·0% (322/1464) of the study subjects were excluded from multivariate analysis because of missing data, raising the concern of selection bias. Another limitation is that the sample size may be too small, and statistical power too low, to compare syphilis prevalence in patient subgroups with different patterns of HBV seroreactivity. Additionally, this study was conducted in a highly endemic area for both syphilis and HBV, further research is needed to explore the syphilis-HBV co-infection in populations with moderate or low prevalence.
In conclusion, this study demonstrated a high prevalence of syphilis in the study subjects and a high rate of syphilis-HBV co-infection in HIV-negative patients. Individuals who migrate between cities are at higher risk of infection with syphilis than non-migrants, suggesting that political advocacy, counselling and testing services could be valuable. Abstinence from smoking and limited consumption of alcohol may reduce the risk of infection with syphilis. In addition, syphilis infection may be more prevalent in individuals with elevated BUN levels. To prevent the spread of infectious diseases, inpatients and blood donors should continue to be tested for exposure to and/or infection with HIV, syphilis, HBV and HCV at the same time.
ACKNOWLEDGEMENTS
We acknowledge contributions from colleagues in the clinical laboratory, Zhongshan Hospital, Medical College of Xiamen University. This work was supported by the National Natural Science Foundation (grant nos. 81271335, 81201360, 81301501, 81471967, 81471231, 81401749); the Key Projects in Fujian Province Science and Technology Programme (grant nos. 2013D025, 2014D021); the Projects of Xiamen Science and Technology Programme (grant nos. 3502Z 20 144 022, WSK2010–01, 3502Z20139001, 3502Z 20 134 050); the Key Project of Cultivating Young Talent in Fujian Province's Health System (grant nos. 2013-ZQN-ZD-35, 2014-ZQN-ZD-34); the National Science Foundation for Distinguished Young Scholars of Fujian (grant no. 2014D001); the Youth Foundation Project of Fujian Provincial Health Department (grant nos. 2013-2-90, 2014-2-68); and the 863 Programme (grant no. 2014AA 021 401). The funders played no role in the study design, data collection, or analyses, the decision to publish, or manuscript preparation.
DECLARATION OF INTEREST
None.
