SEARCH

SEARCH BY CITATION

Keywords:

  • Acquisition;
  • climate;
  • cystic fibrosis;
  • Pseudomonas aeruginosa ;
  • seasonality

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Transparency Declaration
  9. References

Pseudomonas aeruginosa, the principal respiratory pathogen in cystic fibrosis (CF) patients, is ubiquitous in the environment. Initial P. aeruginosa isolates in CF patients are generally environmental in nature. However, little information regarding seasonality of P. aeruginosa acquisition is available. We conducted a retrospective study to evaluate the seasonality of initial P. aeruginosa acquisition in young children with CF in the USA using the Cystic Fibrosis Foundation National Patient Registry from 2003 to 2009. Additionally, we assessed whether seasonal acquisition varied by climate zone. A total of 4123 children met inclusion criteria and 45% (n = 1866) acquired P. aeruginosa during a mean 2.0 years (SD 0.2 years) of follow up. Compared with winter, increased P. aeruginosa acquisition was observed in summer (incidence rate ratio (IRR): 1.22; 95% CI: 1.07–1.40) and autumn (IRR: 1.34; 95% CI: 1.18–1.52), with lower acquisition observed in spring (IRR: 0.81; 95% CI: 0.70–0.94). Seasonal variations in P. aeruginosa acquisition rates in the temperate and continental climate zones were similar to those in the overall cohort. In contrast, no significant seasonal effect was observed in the dry climate zone. In a corresponding analysis, no seasonal difference was observed in the rate of acquisition of Staphylococcus aureus, another common CF respiratory pathogen. These results provide preliminary support that climatic factors may be associated with initial P. aeruginosa acquisition in CF patients. Investigation and identification of specific risk factors, as well as awareness of seasonal variation, could potentially inform clinical recommendations including increased awareness of infection control and prevention strategies.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Transparency Declaration
  9. References

Infectious diseases often exhibit temporal and seasonal variation in incidence. These patterns depend upon the specific pathogen and host, the mode of transmission and the environmental characteristics [1, 2]. Pseudomonas aeruginosa is a ubiquitous environmental organism and the most significant pathogen in cystic fibrosis (CF) lung disease. The prevalence of P. aeruginosa infection among CF patients increases with age, with positive respiratory tract cultures reported for c.20–30% of infants, 30–40% of children aged 2–10 years, 60% of adolescents and 80% of adults in the USA [3]. Earlier initial P. aeruginosa acquisition has been associated with increased morbidity and mortality [4-6]. Patients with CF are initially infected with environmental isolates of P. aeruginosa [7, 8]. Regardless of treatment, patients commonly have a period of intermittent isolation of P. aeruginosa from the respiratory tract before becoming chronically infected [9-11]. Eventually there appears to be clonal selection of a specific P. aeruginosa genotype that then undergoes genetic adaptation within the CF airway [8, 12].

Although the seasonal acquisition of P. aeruginosa in CF patients has been investigated [13, 14], these studies have been limited to small sample sizes in geographically homogeneous areas. Insight into seasonal variations in the rate of P. aeruginosa acquisition throughout the USA could elucidate climatic factors driving host–pathogen interactions and inform recommendations for monitoring respiratory cultures in young children with CF. The objectives of this study were therefore to: (i) evaluate the seasonal acquisition of P. aeruginosa in young children with CF in the USA and (ii) determine whether seasonal acquisition varied by climate zone.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Transparency Declaration
  9. References

We performed a retrospective study using the Cystic Fibrosis Foundation (CFF) National Patient Registry data from 1 January 2003 to 31 December 2009. The Registry contains detailed demographic and clinical information on CF patients treated at all US CFF-accredited centres, who comprise >80% of diagnosed US CF patients. Data are entered at each clinical encounter, including results of respiratory cultures. Since 2003, CFF clinical care guidelines have included obtaining quarterly cultures (i.e. four times per year) [15].

The study population included all patients in the Registry born in 2003 or later who had at least one respiratory culture recorded before 2 years of age and whose first recorded culture was negative for P. aeruginosa. Therefore, only children up to 6 years of age were included in the study. The primary outcome was initial P. aeruginosa acquisition (i.e. based on the first culture from which P. aeruginosa was isolated). The primary exposure was season of P. aeruginosa acquisition, defined as spring (March–May), summer (June–August), autumn (September–November) and winter (December–February).

Seasonality within the US is geographically variable; therefore, we evaluated potential seasonal differences within climate zones using the revised Köppen–Geiger Climate classification [16]. Briefly, in this classification scheme a total of five broad regions (Tropical, Dry, Temperate, Continental and Polar) are defined based on meteorological variables. We linked locations from individual-level zip code data from the CFF Registry, taken as the zip code in the year in which P. aeruginosa was acquired or the year of the last clinical visit recorded in the Registry (for those remaining P. aeruginosa free), to the Köppen–Geiger climate classification using ArcGIS version 10.1 (ESRI, Redlands, CA, USA). In the US CF population, four climate zones are represented (tropical, dry, temperate and continental) (Fig. 1). Due to the minimal number of patients residing in the tropical zone (n = 37), seasonal variability in this zone was not evaluated; these patients were retained in the overall seasonal analysis.

image

Figure 1. Climate zones of the continental USA based on the revised Koppen–Geiger climate classification (adapted from Peel et al. [16]).

Download figure to PowerPoint

Descriptive statistics were produced. Student's t-tests with unequal variances were used to compare continuous variables and chi-square tests compared categorical variables between acquisition status.

Incidence of P. aeruginosa was calculated for each season over the study period. The denominator for these rates was the number of persons under observation and at risk for initial P. aeruginosa acquisition during each season, and the numerator consisted of the cases occurring during that season. Due to the important role that culture frequency may play in observed incidence, the average number of cultures per person for each of the seasons was also evaluated.

To evaluate the association of season with P. aeruginosa incidence, we used Poisson log-linear regression with quasi-likelihood and included the number of individuals at risk as the offset term. Winter season was the baseline for all comparisons, as we hypothesized that P. aeruginosa incidence would be lowest during this season. Results are presented as incidence rate ratios (IRRs) and corresponding 95% CIs. These analyses were then repeated for each climate zone separately.

We performed three sensitivity analyses. We evaluated whether the seasonal associations observed in the entire cohort were seen in individuals with different cystic fibrosis transmembrane conductance regulator mutation classes (ΔF508 homozygous, heterozygous and other), and age at date of culture (<2 or ≥2 years). We also limited the Poisson analyses by defining the persons at risk in each season as only those who had a culture recorded in the Registry during that season. Finally, hypothesizing that Staphylococcus aureus would not exhibit a seasonal pattern of acquisition, we replicated the above analyses with season of initial acquisition of Saureus as the primary outcome of interest, among patients in the Registry born in 2003 or later who had at least one respiratory culture recorded before 2 years of age and whose first recorded culture was negative for S. aureus.

The study was approved by the CFF Registry Committee and the University of Washington Institutional Review Board. All analyses were conducted using STATA 12.0 (StataCorp, College Station, TX, USA).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Transparency Declaration
  9. References

The study cohort consisted of 4123 patients. Of these, 45% (n = 1866) acquired P. aeruginosa and 55% (n = 2257) remained P. aeruginosa-negative during a mean of 2.0 years (SD = 0.2 years) of observation (Fig. 2). A total of 32 698 cultures were performed during follow-up in these individuals. The majority of cultures (90%) were oropharyngeal; however, other culture types included sputum (expectorated or induced) (8%) and bronchoscopic (1%). The demographic and clinical characteristics of the cohort are presented in Table 1. Children acquiring P. aeruginosa tended to be diagnosed later, were less likely to be diagnosed by newborn screening and were more likely to be ΔF508 homozygous compared with those that did not acquire P. aeruginosa.

Table 1. Characteristics of the study cohorts, by Pseudomonas aeruginosa and Staphylococcus aureus acquisition status
 Pseudomonas aeruginosa (n = 4123)Staphylococcus aureus (n = 3196)
P. aeruginosa acquired (n = 1866)P. aeruginosa negative (n = 2257)S. aureus acquired (n = 1987)S. aureus negative (n = 1209)
  1. SD, standard deviation.

  2. a

    p <0.05.

Male (%)52a4752a48
Race/ethnicity (%)
Non-Hispanic White84a8484a82
Black6557
Other101122
Mean age (SD) at diagnosis, months2.2 (4.0)2.3 (4.2)2.5 (4.3)2.3 (4.4)
Identified by newborn screening (%)35a5034a53
Cystic fibrosis transmembrane conductance regulator mutation
ΔF508 homozygous53a4150a46
ΔF508 heterozygous36443742
Other11151312
image

Figure 2. Flow chart of study cohort for evaluation of acquisition of Pseudomonas aeruginosa (Pa). [This figure was corrected on 04/07/2013, after original online publication 25/06/2013. In the bottom right hand box “Acquired Pa during follow-up”, “n = 2265” has been corrected to “n = 1866”.]

Download figure to PowerPoint

The overall P. aeruginosa incidence rate during the study period was 16.5/1000 person-months (95% CI: 15.7–17.2) and the median age of P. aeruginosa acquisition was 19 months. Figure 3 displays P. aeruginosa incidence for each season during the study period. The highest rates of P. aeruginosa acquisition were observed in the summer and autumn for each year, with peak P. aeruginosa incidence occurring in autumn for all study years except 2008. The overall incidence rates in winter, spring, summer and autumn, respectively, were 15 (95% CI: 13–16), 12 (95% CI: 11–13), 18 (95% CI: 17–20) and 20/1000 person-months (95% CI: 18–22). Figure 3 also displays the average number of cultures in each season. It can be seen that the number of cultures per person per season steadily increased over the study period. At the beginning of the study period in 2003, the average number of cultures per patient per season was 0.4, rising to approximately 1 in 2009, probably reflecting improved adherence to CFF guidelines regarding quarterly cultures. The average culture rate from 2003 to 2009 was similar for all seasons (0.9 cultures/person-season).

image

Figure 3. Seasonal incidence rates for Pseudomonas aeruginosa acquisition among young children with cystic fibrosis in the USA from 2003 to 2009. Whiskers represent 95% confidence intervals.

Download figure to PowerPoint

The results of Poisson regression models evaluating the association between P. aeruginosa incidence and season of acquisition are presented in Table 2. Compared with winter, significantly higher P. aeruginosa incidence was observed in summer (IRR: 1.22; 95% CI: 1.08–1.39) and autumn (IRR: 1.34; 95% CI: 1.18–1.52), whereas significantly lower P. aeruginosa incidence was observed in spring (IRR: 0.81; 95% CI: 0.70–0.93).

Table 2. Seasonal incidence of Pseudomonas aeruginosa and Staphylococcus aureus acquisition among children <6 years of age with cystic fibrosis in the USA from 2003 to 2009
 Overall IRR (95% CI)Climate zone
Dry IRR (95% CI)Temperate IRR (95% CI)Continental IRR (95% CI)
  1. IRR, incidence rate ratio.

  2. From Poisson regression models with the number of individuals at risk as the offset term. Winter season is the reference season. Results in bold are statistically significant (p <0.05).

Pseudomonas aeruginosa
WinterRef.Ref.Ref.Ref.
Spring 0.81 (0.70–0.93) 0.69 (0.40–1.19)0.83 (0.67–1.04)0.80 (0.64–1.00)
Summer 1.22 (1.08–1.39) 0.91 (0.55–1.50)1.14 (0.94–1.39) 1.36 (1.12–1.64)
Autumn 1.34 (1.18–1.52) 1.11 (0.69–1.78) 1.21 (1.01–1.47) 1.55 (1.28–1.88)
Staphylococcus aureus
WinterRef.Ref.Ref.Ref.
Spring0.99 (0.87–1.12)0.96 (0.61–1.51)0.99 (0.81–1.20)0.99 (0.82–1.19)
Summer0.98 (0.87–1.11)0.82 (0.51–1.31)0.96 (0.79–1.17)1.06 (0.89–1.27)
Autumn1.11 (0.99–1.25)1.13 (0.74–1.74)1.04 (0.86–1.25)1.15 (0.97–1.38)

The distribution of patients by climate zone was as follows: 289 (7%) in dry, 1679 (43%) in temperate and 1913 (49%) in continental climate zones. The overall P. aeruginosa incidence in these regions was 14 (95% CI: 12–17), 17 (95% CI: 16–18) and 17/1000 (95% CI: 16–18) person-months for dry, temperate and continental climate zones, respectively. A minimal number of patients that changed zip code also changed climate zone (n = 54) while under observation. The seasonal variation in P. aeruginosa acquisition in the temperate and continental climate zones were similar to the overall cohort (Table 2). In contrast, a significant seasonal effect was not observed in the dry climate zone.

In sensitivity analyses, the seasonal pattern of P. aeruginosa acquisition seen in the entire cohort was also seen in each cystic fibrosis transmembrane conductance regulator mutation subgroup (ΔF508 homozygous, heterozygous or other), among those with age at date of culture <2 or ≥2 years, and when the persons at risk in each season were defined as only those who had a culture recorded in the Registry during that season (data not presented).

Finally, we repeated the season of acquisition analyses for S. aureus. This study population differed slightly from that for the P. aeruginosa analysis because of the larger number of individuals excluded due to first recorded culture positive for Saureus (n = 1326). Hence, the study cohort consisted of 3196 patients. The demographic and clinical characteristics of this cohort were similar to those of the P. aeruginosa acquisition cohort (Table 1). A total of 1987 (62%) individuals acquired S. aureus during the study and 1209 (38%) remained S. aureus-negative. The overall S. aureus incidence was 26.7/1000 person-months (95% CI: 25.6–27.9) and the median age of S. aureus acquisition was 18 months. In contrast to the observed findings with P. aeruginosa, there was little variation in S. aureus incidence by season (26/1000 person-months in spring (95% CI: 24–28); 26/1000 person-months in summer (95% CI: 24–28); 29/1000 person-months in autumn (95% CI: 26–31); and 26/1000 person-months in winter (95% CI: 24–29)) (Fig. 4). Poisson regression models showed no difference in S. aureus incidence by season for the cohort as a whole or in each climate zone (Table 2).

image

Figure 4. Seasonal incidence rates for Staphylococcus aureus acquisition among young children with cystic fibrosis in the USA from 2003 to 2009. Whiskers represent 95% confidence intervals.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Transparency Declaration
  9. References

We observed significant seasonal variation in the rate of initial P. aeruginosa acquisition in a large US cohort of young children with CF over a 7-year period. Compared with winter, incidence of initial P. aeruginosa acquisition was significantly higher in summer and autumn and lower in spring. These seasonal differences in P. aeruginosa acquisition rates were seen in the continental and temperate climate zones but not in the dry zone, though our power to detect seasonal variation in the dry zone was limited by the small number of individuals residing in this zone. We also evaluated seasonal patterns in S. aureus acquisition as a comparison with an outcome that we hypothesized would not have a seasonal difference and found no seasonal variation in S. aureus incidence rates in the overall cohort or in any climate zone. These results strongly suggest that climatic factors that vary with season influence initial P. aeruginosa acquisition in CF patients.

To date, only two single centre studies have reported seasonal patterns of P. aeruginosa acquisition in children with CF, and to our knowledge no study has evaluated seasonal patterns of S. aureus acquisition in children with CF. In a retrospective medical record review of 300 Danish CF patients from 1965 to 1990, Johansen and Høiby [14] also reported seasonal variation in acquisition, with higher acquisition of P. aeruginosa from October to March compared with April–September. Although no quantitative results were presented, Farrell et al. [13] reported no differences in P. aeruginosa acquisition by season in 120 patients diagnosed through newborn screening in Wisconsin from 1985 to 1991, though their results may have been limited by small sample size.

Limited information is available regarding the seasonality of P. aeruginosa infection in other settings. Perencevich et al. [17] investigated the seasonality of bacterial infections (including hospital-acquired) among patients hospitalized at a single tertiary-care centre from 1998 to 2005 and found increased cases of P. aeruginosa and other gram-negative bacteria in summer months compared with winter. Further, they found that for a 10% increase in temperature there was a 17% increase (95% CI: 4–31%) in P. aeruginosa infections in warmer months (May–September) but not cooler months (October–April). Other studies investigated the temporality of P. aeruginosa infections in otitis externa [18-20] and keratitis [21] and found increased rates of infection in summer and autumn seasons.

Our finding of seasonal variation in P. aeruginosa acquisition rates suggests that meteorological conditions such as temperature, humidity or ambient air quality may be underlying risk factors for age at initial P. aeruginosa infection [1]. Such conditions could affect the density of P. aeruginosa in the environment of children with CF or could play an indirect role by influencing the activities of children during each season. Several investigators have found an association between P. aeruginosa prevalence and climatic factors, including warmer temperatures and higher humidity. Recently, Collaco et al. [22] investigated the role of environmental factors on CF lung disease in three distinct cohorts. Though they did not focus specifically on initial acquisition of P. aeruginosa, they showed that higher ambient temperature was associated with greater P. aeruginosa prevalence in both the USA and Australia. High-humidity environments have been linked to P. aeruginosa infections in non-CF infants [23], and rates of P. aeruginosa-related keratitis have been shown to vary by Köppen climate zone in Australia [24]. In addition, ambient air quality could be a risk factor for initial P. aeruginosa acquisition, as higher concentrations of particulate matter have been shown to adversely affect pulmonary exacerbation rates and lung function among CF patients [25].

Respiratory viruses have been shown to be important causative agents of pulmonary exacerbations in CF patients [26], and virally mediated damage to the respiratory epithelium could potentially predispose CF patients to bacterial airway infection [27]. Interestingly, we did not observe higher rates of P. aeruginosa acquisition in the winter, as might be expected if viral infections were the major risk factor for P. aeruginosa infection.

Risk factors in the environments of CF patients for initial P. aeruginosa infection remain poorly understood. With the recognition that initial P. aeruginosa acquisition is environmental in nature, recent investigations have focused on home [28] and macro-environmental factors [22] associated with P. aeruginosa infection in CF patients. In one small study, acquisition from the home environment was shown to account for at most 20% of all cases [29]. Other modes of acquisition including person-to-person transmission, nosocomial, and nebulizer equipment-related acquisition are rare. In a large prospective observational study, Rosenfeld, et al. [28] were unable to identify risk factors in the child's environment (e.g. hot tub exposure, day care, breastfeeding) associated with age at initial P. aeruginosa acquisition.

There are several limitations to the present investigation. First, the exact date of P. aeruginosa acquisition was unknown for individuals; rather, the date of acquisition was considered to be the date of positive culture. Second, in this observational study in young, generally pre-expectorating, children, c.90% of cultures were of upper respiratory tract samples (oropharyngeal swabs) rather than lower respiratory tract samples (sputum or bronchoalveolar lavage fluid). Oropharyngeal cultures are known to have moderate specificity and low sensitivity for lower airway P. aeruginosa [30]. Hence, it is possible that seasonal patterns of lower airway P. aeruginosa incidence differ from those observed in this study, though the portal of entry for lower airway infection is likely to be the upper airway. Third, we limited our analysis to initial P. aeruginosa acquisition; we did not evaluate the seasonal rates of chronic infection as we expected that the risk factors would differ. Finally, we limited our cohort to children diagnosed with CF before the age of 2 years and with a maximum age of 6 years (since the Registry only began collecting quarterly culture data in 2003 and data were only available through 2009); this could affect the generalizability of our results.

In summary, increased rates of initial P. aeruginosa acquisition were observed in young CF patients in summer and autumn compared with winter in this national study. Results of such analyses could inform recommendations regarding prevention strategies and clinical care, including reinforcing the importance of regular follow-up visits with cultures, particularly during higher-risk months and identifying high-risk populations that might benefit from more frequent monitoring. Similar approaches could identify environmental risk factors for other CF pathogens such as Burkholderia cepacia.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Transparency Declaration
  9. References

The authors thank Dr Bruce Marshall and the Cystic Fibrosis Foundation for providing the Patient Registry data. Portions of this work have been presented in abstract form at the 26th Annual North American Cystic Fibrosis Conference, Orlando, Florida, USA (11–13 October 2012).

Transparency Declaration

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Transparency Declaration
  9. References

Margaret Rosenfeld has received funds for consultancy from Genentech, and has received research funding from NIH, the Cystic Fibrosis Foundation and Vertex, Inc.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Transparency Declaration
  9. References