To the Editor: Streptococcus pneumoniae is a leading cause of bacteremia, meningitis, and pneumonia in older persons and causes the highest mortality in this age group.1 Asymptomatic pharyngeal colonization precedes respiratory or systemic disease and plays an important role in horizontal spread in the community. The highest colonization rates have been documented in children.2 Colonization is seldom reported in older subjects.

A prospective epidemiological study was performed to establish the prevalence, risk factors, and dynamics of pneumococcal colonization in community-dwelling (n=109) and institutionalized (296 nursing home residents and 98 hospitalized) older persons using bacteriological culture and molecular diagnostic (polymerase chain reaction (PCR)) techniques on nasopharyngeal swabs (NPSs).

Two NPSs, one from each nares, were obtained per participant using Dacron polyester-tipped swabs on an aluminium shaft. The first NPS was plated directly on a selective agar containing polymyxin B and incubated for 48 hours. The second NPS was wringed in Todd-Hewitt (TH) broth, preincubated for 18 to 24 hours, plated on a blood agar, and then processed like the first NPS. When identified, pneumococci were serotyped.

The overall pneumococcal colonization rate was 4.2% (21/503) (5.5% (6/109) in the community, 4.1% (12/296) in nursing homes, and 3.1% (3/98) in the hospital, P=.69) (Table 1). Fourteen of 21 colonizing pneumococci were recovered by direct plating on a selective agar and seven of 21 after enrichment in TH broth.

Table 1. Subject Characteristics and Pneumococcal Colonization Rate
CharacteristicTotal N=503Community n=109Nursing Home n=296Hospital n=98 P-Value
  1. SD=standard deviation.

Age, mean ± SD80.3 ± 10.066.2 ± 4.584.3 ± 7.483.8 ± 6.4.001
Comorbidities, mean ± SD2.5 ± 1.81.0 ± 1.12.7 ± 1.43.6 ± 2.4.001
Antibiotic use within the previous 3 months, n (%)139 (27.6)8 (7.3)93 (31.4)39 (39.8).001
Hospitalization within the previous year, n (%)162 (32.2)11 (10.1)102 (34.5)51 (52).001
Influenza vaccine within the previous year, n (%)376 (73)43 (39.4)258 (87.2)66 (67.3).001
Pneumococcal vaccine, n (%)187 (37.2)14 (12.8)148 (50)25 (25.5).001
Katz score, mean ± SD14.7 ± 7.28.0 ± 0.216.4 ± 7.117.6 ± 6.8.001
Pneumococcal colonization, n (%)21 (4.2)6 (5.5)12 (4.1)3 (3.1).69

No significant differences in age and sex distribution, presence of comorbidities, vaccination status, hospitalization or antibiotic use history, or functionality (Katz score) were evident between colonized and noncolonized subjects.

Ten colonized and 30 noncolonized nursing home residents were reswabbed after 1, 2, 4, 8, and 12 weeks. During these 3 months after the primary sampling, pneumococcal carriage occurred in five of the 10 subjects colonized at first sampling (50.0%), compared with eight of the 30 (26.7%) subjects not colonized at first sampling (P=.25). Colonization could be detected only once (2/13), intermittently (4/13), or for longer periods (1/13 during 2 weeks, 6 during ≥1 month). The pneumococci recovered per resident during primary sampling and follow-up belonged to the same serotype.

In a subset of nursing home residents (n=199), deoxyribonucleic acid was extracted from the NPSs and amplified using real-time fluorescence PCR with a lytA gene (encoding for autolysin) probe with a high sensitivity and specificity for S. pneumoniae as descibed previously.3 The pneumococcal colonization rate detected using PCR was 2.0% (4/199) in this group. When PCR was taken as the criterion standard, the bacteriological culture technique had a sensitivity, specificity, and positive and negative predictive values for detecting nasopharyngeal colonization of 50%, 100%, 100%, and 99%, respectively.

This is the first study that specifically describes prevalence and dynamics of pneumococcal colonization in older subjects in a nonoutbreak setting. The overall prevalence of nasopharyngeal pneumococcal colonization was low (4.2%). Higher colonization rates in older subjects have been described in families with small children (4.6%), during respiratory tract infections (6.5–8.7%), and during outbreaks of pneumococcal disease in nursing homes (23%).4–6

During the 3-month follow-up, more than one-quarter of nursing home residents carried a pneumococcus. It is likely that family visits with young children and healthcare personnel with young children, acting as intermediary carriers, can transmit pneumococci to institutionalized older adults.

No specific risk factor for pneumococcal carriage in older adults could be identified in this study. Neither colonization nor its risk factors by themselves but rather immunosenescence (reduced and deficient immunological response) and comorbid illness (cardiopulmonary disease, smoking, alcoholism) contributes to the high susceptibility of elderly persons to pneumonia and invasive pneumococcal disease.7,8

The broth enrichment technique augmented the recovery of colonizing pneumococci by 33%, although the combination of direct inoculation and broth enrichment in the recovery of colonizing pneumococci is likely to underestimate the true pneumococcal carriage rate. In adults, a supplementary oropharyngeal swab can yield 30% more carriers.9 Compared with PCR with a lytA probe, the bacteriological culture technique resulted in a good specificity (100%) but low sensitivity (50%) for detecting pneumococci. These results suggest that the pneumococcal colonization rate detected using lytA-PCR is twice that of the culture technique that relies on the recovery of viable pneumococci.

In conclusion, the prevalence of nasopharyngeal carriage of pneumococci in older persons is low, although over a period of 3 months, pneumococcal colonization was a frequent event. The origin, dynamics, risk factors, and optimal detection method of pneumococcal colonization in older persons need further study.


We would like to acknowledge Marc Van Ranst, virologist of the department of microbiology of the University Hospitals Leuven, Belgium, for his expert advise on the application of molecular diagnostics by PCR on the nasopharyngeal samples.

The results of this study were partially presented at the 18th European Congress of Clinical Microbiology and Infectious Diseases, Barcelona, Spain, April 19 to 22, 2008 (abstract P1708).

Conflict of Interest: The editor in chief has reviewed the conflict of interest checklist provided by the authors and has determined that the authors have no financial or any other kind of personal conflicts with this paper.

This study was partially funded by a clinical PhD grant from the Research Foundation Flanders (FWO-Vlaanderen) granted to Johan Flamaing.

Author Contributions: All authors participated in the design, analysis and interpretation of the data, and preparation of the manuscript. J. Flamaing was responsible for acquisition of subjects, NPSs, and laboratory work.

Sponsor's Role: The Research Foundation Flanders (FWO-Vlaanderen) had no role in the design, methods, subject recruitment, data collection, analysis, or preparation of the letter.