Introduction

Streptococcus pneumoniae (pneumococcus) is a common cause of invasive diseases, such as pneumonia, meningitis, and bacteremia. Since the global introduction of pneumococcal conjugate vaccines (PCV), the incidence of invasive pneumococcal disease (IPD) has declined [1–4]. However, this disease remains among the leading causes of morbidity and mortality in children under five years old and the elderly, particularly in low- and middle-income countries [5]. Global estimates indicate that pneumococcus caused 8.9 million cases of pneumonia in children aged 1–59 months and 294,000 deaths among children aged 1–59 months worldwide in 2015 [6]. Although conjugate vaccines are available, Brazil continues to experience a high annual pneumonia rate, primarily based on the number of hospitalizations for pneumonia in children during their first year of life, reaching 3,433 per 100,000 inhabitants [7]. Pneumococcal vaccine coverage in Brazil in 2017 was approximately 92%, with rates reaching 93.7% in southern Brazil [8].

Nasopharyngeal pneumococcal colonization (PNC) is a prerequisite for pneumococcal disease and facilitates transmission within the community [9, 10]. Various factors influence pneumococcal carriage, including age, nursery attendance, number of siblings, and vaccination [11, 12]. Viruses may also affect the prevalence of pneumococcal colonization. Recent studies have reported an increase in pneumococcal density when respiratory viruses were simultaneously present in the nasopharynx [13–16]. Increased density promotes pneumococcal invasion and the severity of IPD [12]. However, no specific value was established to reliably predict the possibility of pneumococcal pneumonia [17–20].

These viruses primarily cause acute respiratory infections (ARI) during childhood. During viral infection, PNC is associated with increased viral activity and progression to ARI [20–23]. Interactions between respiratory viruses and pneumococci increase the severity of respiratory illnesses [13, 14, 21, 24]. Additionally, the role of various viruses in creating conditions for bacterial superinfections, such as pneumococcal pneumonia, was previously documented [25, 26]. Epidemiological studies have shown that an increase in certain viruses among children is associated with higher rates of pediatric admission owing to IPD [24, 26, 27]. Furthermore, mechanistic studies have revealed that viruses increase bacterial adhesion, translocation, and persistence [28–31].

The relationship between viral infections and pneumococcal diseases has been thoroughly examined. However, there information regarding the interaction between respiratory viruses and pneumococcal carriages in healthy children is lacking [14, 15]. Several studies have explored the heightened severity of infectious diseases when these pathogens interact [13–16, 21, 31]. However, there is a notable gap in data concerning children who are carriers of both pneumococcal and viral pathogens. This study aimed to examine the relationship between respiratory viruses, pneumococcal carriage, and the development of respiratory diseases in Brazilian children aged 18–59 months.

Materials and methods

Results

In 2017, Veranópolis City had an estimated 781 children aged 18–59 months. A total of 229 children were included in this cross-sectional study (Fig 1). Five children were excluded from the medical record evaluation, resulting in 224 children included in the cohort study. The study population comprised equal numbers of females and males. Siblings were not clustered in the analyses because of the small number of participants under this condition (only 10 participants).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Flow diagram of children aged 18–59 months, Veranópolis/RS, Brazil, between 2018 and 2019.

ICF, informed consent form.

https://doi.org/10.1371/journal.pone.0316444.g001

The prevalence of pneumococcal carriage was 63.7% (146/229, CI 95%: 57.4–69.8%) and 49.3% (113/229, CI 95%: 42.9–55.8%) for respiratory viruses. Respiratory viruses were more frequently found among pneumococcal carriers (54.4% (80/146) vs. non-carriers [39.7% (33/83); p = 0.033]). CI95%: 46.7–62.7%). Co-occurrence rate of 34.9% (80/229, CI 95%: 29.0–41.3%). As shown in Table 1, children who tested positive for viral infection, with or without the co-occurrence of pneumococcus, were younger than pneumococcal carriers (31 and 33 months vs. 42 months; p < 0.001). There was a significant relationship between viral and bacterial co-occurrence and male sex (63.8%; p = 0.004), whereas the absence of the investigated organisms was associated with females (Table 1).

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Demographic characteristics of the population (n = 229) according to identified microorganisms, Veranópolis/RS, Brazil, 2018–2019.

https://doi.org/10.1371/journal.pone.0316444.t001

Additionally, multinomial logistic regression analysis of certain demographic data and the presence of pneumococcal carriage, respiratory virus, and co-occurrence was conducted (S2 Table). With each passing month of age, the likelihood of acquiring a respiratory virus decreased by 4% (p = 0.020; OR 0.96, 95% CI 0.92; 0.99), and 5% for co-occurrence (p = 0.003; OR 0.95, 95% CI 0.92; 0.98). The male sex was associated with a higher co-occurrence rate (p = 0.001; OR 3.62, 95% CI 1.64; 7.98). The likelihood of co-occurrence was higher in fall (p = 0.018; OR 5.3, 95% CI 1.33; 21.09), compared with that in spring, which had the highest number of children without microorganisms detected. The symptoms associated with each virus at the time of collection were investigated using two cutoffs: Ct <30 (only high levels) and <35 (any level), as detailed in Table 2. The only symptom associated with the occurrence of any virus was sore throat with any hRV Ct value (p < 0.05).

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Frequency of clinical symptoms in children with detection of respiratory viruses by real-time PCR using the cut-off levels cycle threshold (CT), Ct- value of < 30 (only high levels) and < 35 (any level), Veranópolis/RS, Brazil, between 2018 and 2019.

https://doi.org/10.1371/journal.pone.0316444.t002

Various viruses were detected in children with pneumococcal infections. A relationship was observed between total hRV (alone + co-detection) 39% (57/146, p = 0.011) and the detection of any respiratory virus 54.8% (80/146, p = 0.040) with pneumococcal carriage (Table 3). Additionally, the presence of hBOV alone was associated with the absence of pneumococcal carriage (10.8%; p = 0.016). All samples tested negative for INF A and B, as well as for hPIV 1, 2, and 3.

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Frequency of respiratory viruses in children with and without pneumococcal carriage, Veranópolis/RS, Brazil, between 2018 and 2019.

https://doi.org/10.1371/journal.pone.0316444.t003

Among the eligible participants for the cohort study, 19.2% (43/224) required medical care within 14 days. Co-occurrence was present in 32.5% (14/43) of children who needed medical care. However, no statistically significant correlation was found between pneumococcal carriage, respiratory virus detection, or co-occurrence in the medical appointments of the participants 14 days after specimen collection (Table 4).

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Incidence of medical appointments, clinical symptoms and diagnosis in 14 days of medical record analysis according to identified microorganism, Veranópolis/RS, Brazil, between 2018 and 2019.

https://doi.org/10.1371/journal.pone.0316444.t004

No relationships were found between the frequency of pneumococcal carriage, respiratory virus detection, or co-occurrence and different clinical symptoms and diagnoses of the participants 14 days after the collection of specimens (Table 4). The frequency of each respiratory virus and the symptoms of the children after 14 days of medical care are detailed in S4 Table.

Seasonality could not be reliably determined because of the lack of homogeneity in the sample collection across seasons, as detailed in S5 Table. Samples were collected based on the availability of volunteers, with fewer collections occurring during summer, likely because of school vacations.

Discussion

This study represents an initial investigation in Latin America to explore the relationship between respiratory viruses and pneumococcal infections in healthy children. We found a significant relationship between respiratory viruses, particularly rhinoviruses and pneumococci in children aged 18–59 months in Southern Brazil. However, no association was found with disease development 14 days after exposure to these agents. The prevalence of respiratory viruses was higher in younger children, a finding consistent with previous literature reporting a high viral prevalence in children younger than our study population (< 1 year) [38, 39]. Although our study did not include children under one year of age, we observed a similar relationship between younger age and respiratory viruses in children aged 18–59 months.

Children who were "too sick" were excluded from this study. The recruitment process allowed for the inclusion of children with mild respiratory symptoms as reported by their guardians. Consequently, 108 (47.2%) and 79 (34.5%) children exhibited coryza/congestion and coughing, respectively. Despite the frequency of these symptoms in our study, they were not associated with the detection of the viruses studied. According to the literature, these symptoms are commonly associated with respiratory viruses. Notably, lower Ct levels are often associated with more severe respiratory symptoms and mortality [40]. However, in our study, high Ct levels were observed, which is consistent with the clinical condition of the study group (considered healthy) and the lack of association with symptoms.

An association between hRV and sore throat was identified independent of Ct values, and some studies have demonstrated an association between hRV and sore throat [41, 42]. For instance, in a study conducted on individuals experiencing sore throats, hRV was the most commonly detected pathogen [43]. Moreover, an experimental study involving intranasal inoculation of hRV observed scratchy/sore throat symptoms appearing 10–12 h after the induced inoculation [44]. Notably, these symptom observations were not derived from a medical examination; however, some subjectivity was involved. In the present study, INF and hPIV were not detected. The prevalence of these viruses is typically low (<3%) among asymptomatic children [15, 16] because they interact with their hosts and produce symptoms [45, 46].

No relationship was observed between a specific virus and pneumococcal colonization. However, the detection of more than one virus, especially hRV, is associated with pneumococcal carriage. These findings are consistent with those of Demuri (2018) [15] and Howard (2019) [16], who also reported a relationship between the presence of respiratory viruses and pneumococcal colonization in samples with more than one respiratory virus. In both cited studies, hRV was the most frequent virus involved in the co-occurrence of pneumococcal carriage and significantly influenced the increase in pneumococcal density [15, 16]. There is considerable diversity among studies regarding the age ranges of healthy children. For instance, one study focused on children under three years old [15], while another included children aged four to seven years [16]. Additionally, a study conducted in the Democratic Republic of Congo in children aged 2–60 months reported pneumococcus and viral co-occurrences in 30% of the study population. However, this study did not analyze the co-occurrence of pneumococci and each virus individually [47]. Furthermore, some studies have reported an association between pneumococcus and hRV infection in healthy children. A study conducted in western Australia demonstrated a positive correlation between hRV and pneumococcal carriage in children older than one year [48].

Currently, there is no evidence suggesting an antagonistic relationship between hBOV and pneumococcal carriage. Co-infection with hBOV and pneumococcus has been reported [49–51]. A Brazilian study investigating hBOV-related community-acquired pneumonia in children caused found that pneumococcus was the main bacteria in co-infection cases, with pneumococcus detected in 33% (7/21) of co-infections [52]. Similarly, a Japanese study observed a positive correlation between the co-detection of pneumococcus and hBOV, where 85.7% (12/14) of children tested positive exclusively for both pathogens in otitis media fluid. However, this correlation was not observed in the nasopharynx, as the proportion of pneumococcus was similar between groups of individuals positive and negative for hBOV and other pathogenic bacteria. Notably, pneumococci were the only pathogens present in all but one of the co-detections in hBOV-positive cases, whereas in the nasopharynx, it was mixed with other pathogenic bacteria [53]. This suggests that these pathogens migrated to their target site (otitis media fluid), with the nasopharynx not being the site where the etiological agent was identified. It is essential to highlight that in children without serious diseases, these microorganisms may play an antagonistic role in the colonization of the nasopharynx, potentially suggesting the protection of pneumococcus in the presence of hBOV alone. However, interpretation of the results of both studies was limited by the small number of patients included. Nevertheless, these studies collectively underscore the relevance of pneumococcus and respiratory virus coinfection in disease pathogenesis. Our findings, such as the absence of INF and hPIV in the sampled population, likely reflect the predominance of these viruses in children with more severe symptoms. INF, in particular, is well known for its association with co-infection and increased risk of subsequent severe pneumococcal disease [54, 55]. Further studies are needed to elucidate the mechanisms underlying the interactions between hBOV and pneumococci.

We examined the clinical outcomes of the population 14 days after specimen collection to gain a deeper insight into the relationships between these microorganisms. We focused on medical consultations, but exclusively recorded the symptoms and diagnoses of interest. The importance of viral infections in predisposing individuals to subsequent bacterial infections is well-established. Among healthy children, studies have demonstrated that the presence of respiratory viruses and pneumococci at high densities is a risk factor for subsequent infections [15, 16, 21]. However, our results suggest harmonious coexistence between viruses and pneumococci. Notably, this population had high vaccination coverage for the prevention of pneumococcal diseases. Therefore, the protection conferred by the vaccine may have prevented the development of disease, regardless of the presence of viral agents. In populations with insufficient vaccination coverage, the relationship with the presence of viruses may differ and warrants further investigation.

These observations suggest that pneumococcal carriage is affected by various factors that are poorly understood, with the participation of viruses likely to play a prominent role. Studies investigating the relationship between respiratory viruses and pneumococcal colonization should be approached with caution. There is considerable diversity across studies, including variations in the age of the children studied, specimen collection methods, microbiological and molecular techniques used for pneumococcal detection, viruses studied, and the definition of outcomes (presence or density of pneumococci). Although this diversity may hinder direct comparisons among study results, it underscores the need for additional research that can offer new perspectives on this issue.

This study had unique characteristics and limitations that merit consideration. This is one of the pioneering investigations analyzing the co-occurrence of pneumococcus and various respiratory viruses among children aged 18–59 months, a population that is predominantly vaccinated against pneumococcal diseases. A limitation of our study, similar to other studies, is the use of PCR to detect respiratory viruses in healthy children. Interpretation of the results can be challenging, as detection may indicate prolonged shedding of the virus resulting from a previous episode [16, 56]. Furthermore, although comprehensive, our viral panel did not include other agents that could potentially contribute to pneumococcal colonization, such as seasonal coronaviruses.

In summary, our findings revealed a positive relationship between pneumococcal carriage and multiple respiratory viruses, particularly rhinoviruses, in a population with high vaccination coverage. However, despite this association, there was no increased risk of subsequent development of respiratory disease during the 14-day follow-up period. This field of investigation remains open to further studies that may offer insights into this complex issue.

Supporting information

Acknowledgments

We extend our sincere appreciation to all the staff at the Instituto Moriguchi for their cooperation and collaboration. We also thank the Veranópolis Health Department and Hospital São Peregrino Lazziozi for their assistance in collecting information from medical records. Special thanks go to the staff at the Federal University of Health Science of Porto Alegre, particularly Cristiane Bündchen, for providing statistical assistance. We would like to thank Editage (www.editage.com) for English language editing. We are grateful to all the children who participated in the study as well as to the nurses, pediatricians, and community health workers at the Public and Private Health Systems of Veranópolis.