Introduction

The atypical pathogens Mycoplasma pneumoniae (syn. Schizoplasma pneumoniae), Chlamydophila pneumoniae (Chlamydia pneumoniae), and Legionella pneumophila cause mild, moderate, or severe acute respiratory tract infections (ARTIs) in children, although Legionella pneumophila infection is more common in adults. These infections occur worldwide [1–3]. Cyclical outbreaks of M. pneumoniae infections can be expected on average every 3–7 years, but at any given time may account for as many as 40% of community-acquired pneumonia cases [3]. The prevalence of C. pneumoniae in children with ARTIs varies from 0 to 44% [4].

Studies of possible associations between the epidemiology of atypical pathogens and meteorological conditions (e.g., temperature, humidity, rainfall, amount of solar radiation and wind velocity) are few. However, it was recently reported that the community incidence of pneumonia due to M. pneumoniae increased weekly by 16.9% for every 1°C increase in the average temperature, and by 4.1% for every 1% increase in relative humidity [5]. Community-acquired M. pneumoniae or C. pneumoniae infections affect mainly preschool- and school-aged children and young adults. Few studies have reported the frequency of M. pneumoniae and C. pneumoniae infections in infants [6].

Clinically it is difficult to distinguish M. pneumoniae from C. pneumoniae infections and hence laboratory tests are essential in identifying these pathogens. Serological detections, although commonly used, are complicated by false negative results in the early acute phase of infection, and the difficulty in obtaining convalescent serum during hospital stays of one week or less. Polymerase chain reaction (PCR) is suitable for rapid diagnosis of these infections, even when the colonization rate is only 1-2% of the population [7–9]. Combining both PCR and serology seems to be a more reliable diagnostic approach.

The incidence rate of childhood ARTIs due to these pathogens is very different from one country to another [2–6]. Our purpose was to use PCR to identify and determine the percentage of ARTIs that were due to M. pneumoniae and C. pneumoniae, especially in infants, and investigate the relatedness between their epidemiology and meteorological conditions in Suzhou, Jiangsu Province, China. In addition we compared the clinical characteristics of M. pneumoniae and C. pneumoniae infections, host immune state, radiographic appearance, and tidal breathing measurements.

Materials and methods

Study population

From 1 January 2006 to 31 December 2006, 1598 consecutive children with ARTIs admitted to Children’s Hospital affiliated with Soochow University were enrolled and evaluated prospectively. These children were hospitalized because of prolonged fever (>3 d), severe symptoms of cough, wheeze, tachypnea, and chest retractions. The clinical outcome and diagnosis for all children were obtained after discharge from the hospital. The discharge diagnosis was based on standard clinical criteria made by attending physicians. Upper respiratory tract infection was diagnosed if a child had nasal obstruction, nasal discharge, fever, or sore throat. Lower respiratory tract infection was diagnosed when wheeze, tachypnea, chest retractions, abnormal auscultatory findings, and radiologic evidence of a lower respiratory tract infection were present. Children were excluded from the study if they had proven chronic lung disease, immunodeficiency, congenital heart disease, or bronchopulmonary dysplasia. The Institutional Review Boards of Suzhou University approved the study protocol, and the parents or legal guardians of each child gave informed written consent.

Methods

Sample collection

Nasopharyngeal aspirate (NPA) samples were obtained from each patient within 24 hours of admission using a sterile plastic catheter introduced into the lower part of the pharynx via the nasal cavity. The samples were immediately transported to the Laboratory of Molecular Biology of our hospital for detection of M. pneumoniae and C. pneumoniae using PCR. Seven common respiratory virus (respiratory syncytial virus, influenza A and B, parainfluenza 1, 2, 3 and adenovirus ) using direct immunofluorescence and human metapneumovirus using RT-PCR described previously [10]. Blood samples were also obtained at admission and immediately sent to the Department of Biochemical Laboratory for routine blood, C-reactive protein, humoral and cell immunity, and alanine transaminase tests.

DNA extraction

Each NPA sample was diluted in 2 mL of normal saline before centrifugation at 500 × g for 10 minutes. The resultant cell pellet was resuspended and then centrifuged at 12 000 × g for 5 minutes, followed by extraction of DNA from a 400-μL sample using DNA-EZ Reagents (Sangon Biotech, USA) in accordance with the manufacturer’s instructions. A final 200 μL of DNA was eluted and divided into 2 aliquots for PCR and stored at −20°C.

Detection of M. pneumoniae gene by real-time PCR

Real-time PCR was performed to identify the P1 adhesion protein gene of M. pneumoniae, as described previously [11]. The forward and reverse primers were 5’-CCA ACC AAA CAA CAA CGT TCA-3’ and 5’-ACC TTG ACT GGA GGC CGT TA-3’, respectively, and the probe sequence was 5’-TCA ACT CGA ATA ACG GTG ACT TCT TAC CAC TG-3’. The fluorescent reporter dye at the 5’ end was 6-carboxyfluorescein (FAM) and the quencher at the 3’ end was 6-carboxytetramethylrhodamine (TAMRA). The PCR reactions consisted of a 21-μL PCR mast mixture (Shenyou Biotechnology, Shanghai, China) including primers and probes combined with 3 μL of the sample DNA and 1 U GoTaq DNA Polymerase (Promgea, Wisconsin, USA). Real-time PCR was performed using an iCycler iQ5 real-time PCR detection system (Bio-Rad, Hercules, CA, USA) and cycling conditions were: 2 min at 37°C; 10 min at 94°C, and 40 cycles of 10 s at 94°C, 30 s at 55°C, and 40 s at 72°C.

Quantification curves were plotted using several concentrations of control plasmids containing the target gene.

Detection of C. pneumoniae gene by nested-PCR

A different aliquot of extracted DNA from the NPA sample was used for detecting the gene of the major outer membrane protein of C. pneumoniae by nested touchdown PCR, as described previously [12]. The external and internal primers were: Cpex-F 5’-TTA CAA GCC TTG CCT GTA GG-3’; Cpex-R 5’-GCG ATC CCA AAT GTT TAA GGC-3’; Cpin-F 5’-TTA TTA ATT GAT GGT ACA AT A-3’; and Cpin-R 5’ATC TAC GGC AGT AGT ATA GTT-3’. Amplifications were performed in a thermal cycler (GeneAmp PCR System 9600, Applied Biosystems). For the first amplification, 0.4 μM of each CPex-F and Cpex-R primer and 2 U GoTaq DNA polymerase (Promgea, Madison, WI, USA) were included in these touchdown PCR reactions. The annealing temperature was lowered 1°C every 2 cycles, from 65°C until touching down at 55°C, at which temperature 20 more cycles were performed. The denaturation and extension temperature were constant at 94°C and 72°C, respectively. One microliter of the products of the first round of amplification by external primers were added into the PCR reactions of the second round amplification using internal primers. The second PCR cycling conditions were: 2 min at 95°C; and 30 cycles of 1 min at 94°C, 1 min at 50°C, and 1 min at 72°C. The nested PCR products were separated via 1.5% agarose gel electrophoresis and visualized using ethidium bromide staining.

Serology testing for M. pneumoniae and C. pneumoniae

The presence of specific IgM and IgG antibodies against M. pneumoniae were investigated in serum samples of patients using a commercial ELISA kit (Serion ELISA classic M. pneumoniae IgG/IgM, Institute Virion\Serion, Germany). IgA and IgG antibodies against C. pneumoniae were detected with Serion ELISA classic C. pneumoniae IgA/IgG Kits. Evidence of acute M. pneumoniae infection was defined as either a single positive serum IgM (cutoff 13 U/mL) or a 4-fold increase in IgG in convalescent serum obtained a week after admission. Acute C. pneumoniae infection was defined as either a single positive serum IgA (cutoff 3 U/mL) or a 4-fold increase in IgG in convalescent serum.

Diagnostic criteria for M. pneumoniae and C. pneumoniae acute infection

Acute infection due to M. pneumoniae or C. pneumoniae was confirmed when any of the following was determined: NPA samples proved positive via PCR, serum samples positive for IgM or IgA, respectively, or IgG increasing 4-fold from the acute to the convalescent phases.

Tidal breathing measurements

Measurements of tidal breathing in most infants were performed during natural and quiet sleep as assessed by behavioral criteria [13]. Tidal breathing flow-volume loops were obtained and analyzed with a commercially available pediatric pulmonary function device (ECO Medics, V’max 26, Switzerland). Tidal breathing measurements included flow and volume signals, and were assessed by professional staff. Lung function was graded as normal, mild, moderate, severe, or extremely severe according to three-component score system which could reflect underlying obstructive airways disease [14, 15] as shown in Table 1.

Table 1 Three-component score system for assessing tidal lung function
Full size table

Meteorological data

The primary investigator obtained data regarding monthly mean temperatures, relative humidity, rainfall, solar radiation, and mean wind velocity from the local weather bureau for lat 31.19° N, long 120.37° E. Suzhou has a subtropical climate and the monthly data for meteorological variables were shown in Figure 1.

Figure 1
figure 1

The monthly data for meteorological variables in Suzhou, China.

Full size image

Statistical analyses

Categorical data were analyzed using the Cochran-Mantel-Haenszel statistic or the chi-squared (χ2) or Fisher’s exact tests. The continuous variables were compared using analysis of variance (ANOVA). The Kruskal-Wallace test was used if the data were abnormal in distribution or nonparametric. The associations between meteorological conditions and the prevalence of the 2 pathogens were evaluated using Spearman’s rank correlations. A P-value < 0.05 was considered statistically significant. All analyses were performed using the Statistical Package for SAS for windows, version 8.2 (SAS, USA).

Results

Epidemiology

Of the 1598 children admitted to our hospital for ARTIs, 957 (59.9%) were boys, and the ratio of boys to girls was 1.5:1. The mean age was 26.4 ± 28.3 months (range: 1 month to 13 years). Of all the children, 817 (51.1%) were aged < one year, 616 (38.5%) were 1–5 years, and 165 (10.3%) were >5 years old.

Twenty-three cases of M. pneumoniae infection and 6 cases of C. pneumoniae infection had co-infections with other viruses (data not shown), and cases of M. pneumoniae or C. pneumoniae coinfection with viruses were excluded in this study. ARTIs were caused by M. pneumoniae or C. pneumoniae in 295 (18.5%) of the patients. The mean age of these children was 33.1 ± 36.3 months. There were 171 boys, and the ratio of boys to girls was 1.4:1. Of the 295 infected with either or both of these pathogens, 199 (12.5%) were due to M. pneumoniae alone (mean age 35.8 ± 37.3 months), 81 (5.1%) were due to C. pneumoniae alone (mean age 27.3 ± 34.5 months), and 15 (5.1%) were co-infected (mean age 28.0 ± 29.7 months). The infection rate due to M. pneumoniae only was significantly and positively associated with increasing age (χ2 = 34.76, P < 0.0001). With regard to demographic data, no significant difference in gender or prematurity was found among the 3 etiological groups (Table 2).

Table 2 The prevalence and demographic data of children hospitalized with M. pneumoniae and C. pneumoniae infection
Full size table

Seasonality and correlations with meteorological conditions

Both M. pneumoniae and C. pneumoniae infections occurred throughout the year. The M. pneumoniae infection rate reached a maximum in summer (P < 0.0001) and a minimum in winter (P = 0.0001), whereas the C. pneumoniae infection rate was lowest in autumn (P = 0.02; Table 3). The peak number of M. pneumoniae infections occurred in July (23.2% or 32/138) and were the lowest (4.3% or 5/117) in December. C. pneumoniae infections were also high in July (12.3% or 17/138; Figure 2).

Table 3 The seasonality and correlations with climatic conditions of M. pneumoniae and C. pneumoniae infection in hospitalized children
Full size table
Figure 2
figure 2

The seasonality of M. pneumoniae and C. pneumoniae infections in hospitalized children. MP, M. pneumoniae; CP, C. pneumoniae.

Full size image

Rates of M. pneumoniae infection in hospitalized children in Suzhou correlated strongly with monthly mean temperature, and weakly with monthly mean wind velocity (Table 3). There was no significant correlation between meteorological conditions and C. pneumoniae infections.

Clinical manifestations

There was no significant difference in terms of clinical manifestations among the groups except for fever (χ2 = 7.824, P = 0.02; Table 4). Children co-infected with M. pneumoniae and C. pneumoniae were found to have fever more frequently. There were no significant differences in length of hospital stay among the different groups. All patients were cured or improved.

Table 4 Clinical manifestations of children hospitalized with M. pneumoniae and C. pneumoniae infection
Full size table

Laboratory findings

There were no significant differences in laboratory findings (routine blood, C-reactive protein level, humoral and cell immunity, alanine transaminase, radiographic data, and tidal lung function test) among the 3 groups (Table 5).

Table 5 No significant differences in laboratory findings of children hospitalized with M. pneumonia and C. pneumoniae infection
Full size table

Discussion

M. pneumoniae and C. pneumoniae are two common atypical pathogens that cause ARTIs in children worldwide. Infections may be endemic or epidemic. In the present study, the total rate of M. pneumoniae and C. pneumoniae infections in hospitalized children with ARTIs in Suzhou, China for the year 2006 was 18.4% (295/1598), in which 12.5% and 5.1% were caused only by M. pneumoniae or C. pneumoniae, respectively. In other recent studies in which real-time PCR was used for identification, Defilippi et al. [16] in Genoa, Italy reported an infection rate for M. pneumoniae in children of 11.8% and in Tokyo, Japan Hamano-Hasegawa et al. [17] found a rate of 14.8%. The rates of M. pneumoniae infection found via real-time PCR tests in Suzhou, China in the years 2007 and 2008 were 5.39% and 6.36%, respectively (unpublished results). Hosker et al. [18] reported an epidemic of M. pneumoniae infection in Hong Kong with a rate of 15% to 20% that occurred from 1986 to 1988, which was similar to our study.

Apparent seasonality of M. pneumoniae infection was observed in this study, with a peak level in summer and the highest rate in July. When the disease is endemic, seasonality may not be a factor, but when it is an epidemic more cases occur in the summer or early autumn [3, 16]. We observed a similar trend in this study. Onozuka et al. [5] found a strong correlation between temperature and relative humidity and the number of M. pneumoniae pneumonia cases in hospitalized children younger than 15 years in Japan, while another study in Germany did not show any correlation [19]. However, we presume that climatic conditions in our area are similar to Japan, and different from the northwestern region of Germany where the range of temperatures across the seasons is not as great.

Regarding C. pneumoniae infections, the 6.0% rate in children with ARTIs found in the present study is similar to the 6.7% (4/112) found in a study performed by Kurz et al. [20] in Vienna, Austria, and lower than the 9.3% reported by Schmidt et al. [21] for Greifswald, Germany. No seasonality or correlation with climatic conditions was found in our study. Studies conducted in various countries have shown that M. pneumoniae and C. pneumoniae infections are most common in school-aged children, followed by children from 1 to 5 years old, but rare in infants [16, 21–24]. In contrast, we found that M. pneumoniae (10.8%) and C. pneumoniae (5.9%) infections are also common in infants. The high population density in China, which increases exposure and the chance of infection, may be an important influence in acquiring these infections.

The use of PCR in the present study provided rapid and specific diagnosis of M. pneumoniae and C. pneumoniae infections. A recent retrospective study comparing the diagnostic value of PCR to the indirect particle agglutination antibody test for M. pneumoniae infection in children from NPA samples found that PCR provided a more rapid diagnosis, particularly in young children (P = 0.003), immunocompromised patients (P = 0.019), and during the early stage of the disease [25]. When PCR is combined with serological tests the yield increases further [26].

In the present study, the clinical manifestations and laboratory findings did not differ between M. pneumoniae and C. pneumoniae. Cough and fever were the most common symptoms. Fever was noticed more frequently in children with co-infection, although co-infection did not increase the severity of clinical manifestations. The rate of wheezing in M. pneumoniae or C. pneumoniae infections was higher than that reported by Esposito et al. [27], and the rates of dyspnea and tachypnea were lower. The reason for this may be that the present study included upper respiratory infection, bronchitis, and asthma exacerbation within the definition of ARTIs.

The 8.8% (26/295) incidence of M. pneumoniae or C. pneumoniae infections with asthma exacerbation in this study indicates a possible role of infection by these atypical pathogens in asthma exacerbation. Recently, it was reported that in asthma M. pneumoniae and C. pneumoniae appeared to be involved more with asthma persistence than exacerbation [28].

In previous studies, no significant difference was found between M. pneumoniae and C. pneumoniae infections with regard to white blood cell count, C-reactive protein, percentage of neutrophils, platelets, or radiographic data [16, 27, 29]. To the best of our knowledge, this study is the first to analyze the humoral and cell immunity status and tidal lung function of children with M. pneumoniae and C. pneumoniae infections. Although no significant difference in immunity status existed between the patients infected with the two pathogens, some studies showed a depressed humoral and cellular immunity in children with M. pneumoniae infection compared to healthy children [30, 31]. Because of the difficulty in distinguishing M. pneumoniae infection from C. pneumoniae on the basis of clinical manifestations or laboratory findings, PCR tests are very useful [32].

The limitations of our study include a probable bias, as the study was conducted only in hospitalized children and not outpatients. In addition, a study period of only one year is relatively short to assess the epidemiology of M. pneumoniae and C. pneumoniae infection.

Conclusion

M. pneumoniae infections in children demonstrated a definite seasonality, a strong correlation with the temperature, and a weak correlation with wind velocity. Consistent with former studies, ours indicates that M. pneumoniae and C. pneumoniae are important in children with ARTIs who are younger than 5 years old, and especially in infants. Clinical manifestations and laboratory findings were informative, but real-time PCR and nested-PCR provide adequate rapid and specific diagnosis of M. pneumoniae and C. pneumoniae infections.