Prevalence and Duration of Asymptomatic Clostridium difficile Carriage among Healthy Subjects in Pittsburgh, Pennsylvania

Abstract

Previous studies suggested that 7 to 15% of healthy adults are colonized with toxigenic Clostridium difficile. To investigate the epidemiology, genetic diversity, and duration of C. difficile colonization in asymptomatic persons, we recruited healthy adults from the general population in Allegheny County, Pennsylvania. Participants provided epidemiological and dietary intake data and submitted stool specimens. The presence of C. difficile in stool specimens was determined by anaerobic culture. Stool specimens yielding C. difficile underwent nucleic acid testing of the tcdA gene segment with a commercial assay; tcdC genotyping was performed on C. difficile isolates. Subjects positive for C. difficile by toxigenic anaerobic culture were asked to submit additional specimens. One hundred six (81%) of 130 subjects submitted specimens, and 7 (6.6%) of those subjects were colonized with C. difficile. Seven distinct tcdC genotypes were observed among the 7 C. difficile-colonized individuals, including tcdC genotype 20, which has been found in uncooked ground pork in this region. Two (33%) out of 6 C. difficile-colonized subjects who submitted additional specimens tested positive for identical C. difficile strains on successive occasions, 1 month apart. The prevalence of C. difficile carriage in this healthy cohort is concordant with prior estimates. C. difficile-colonized individuals may be important reservoirs for C. difficile and may falsely test positive for infections due to C. difficile when evaluated for community-acquired diarrhea caused by other enteric pathogens.

INTRODUCTION

Clostridium difficile infections (CDIs) have shown dramatic increases in incidence and morbidity during the past decade (1). While C. difficile is well established as a health care-associated pathogen, incidence estimates for community-acquired CDIs vary between 6 and 30% (2–4). The source of C. difficile in community-acquired cases is unclear. Prevalence estimates for asymptomatic C. difficile colonization range from 7 to 15% among healthy, non-health care workers outside the United States to 1 to 4% among health care workers in the United States and abroad (5–9). Individuals colonized with C. difficile could represent a potential reservoir of strains imported into hospitals, as well as a source of false-positive clinical test results for CDIs among patients with community-acquired diarrheal illness, especially norovirus and Clostridium perfringens. We studied the prevalence, genotypic distribution, and duration of C. difficile carriage among healthy adults in Allegheny County, Pennsylvania, and evaluated potential risk factors for asymptomatic colonization. We also examined the relative abundance of C. difficile and the sensitivity of a commercial nucleic acid test for detection of C. difficile in the stool of healthy colonized individuals.

MATERIALS AND METHODS

Healthy adult residents (≥18 years of age) in Allegheny County, Pennsylvania, were recruited for participation via print and electronic advertisements in September 2012 through April 2013. Individuals who reported chronic constipation or diarrhea, the presence of an ostomy or a history of colon resection, a history of CDI, recent or anticipated hospitalization or surgery, or employment involving direct patient contact in a health care facility were excluded. Eligible participants provided epidemiological data via a written questionnaire and then recorded and categorized all dietary intake for 10 days using an online food diary; after completion of the food diary, subjects submitted a stool specimen. Specimens were shipped from the subjects to the laboratory, where they were stored at room temperature and inoculated for incubation within 7 days after collection. Aliquots of the stool specimens were frozen at −80°C. Subjects whose stool specimens yielded C. difficile on at least one occasion were referred to as C. difficile colonized and were asked to submit additional epidemiological and dietary data, along with additional stool specimens on a monthly basis.

Stool specimens (approximately 0.01 to 0.1 g, using a 10-μl inoculation loop) were cultured by broth enrichment using cycloserine-cefoxitin-mannitol broth with taurocholate and lysozyme (CCMB-TAL) (Anaerobe Systems, Morgan Hill, CA), as described previously (10). PCR amplification and sequencing of tcdC were used to infer the toxigenicity of isolates. C. difficile isolates that were tcdC negative were confirmed as nontoxigenic by lok1/lok3 PCR assays (11). In addition, stool specimens that were positive for C. difficile in culture underwent loop-mediated isothermal amplification of tcdA (illumigene C. difficile; Meridian Bioscience, Cincinnati, OH), according to the manufacturer's instructions. Quantitative cultures to determine C. difficile stool density were performed using serial 10-fold dilutions of approximately 0.1 g stool diluted in sterile deionized water; anaerobic culture, in triplicate, of 100 μl of each dilution was performed on cycloserine-cefoxitin-mannitol agar (CCMA) for 72 h. C. difficile isolates were typed using tcdC genotyping (12). The tcdC genotypes were assigned according to the PubMLST database (http://pubmlst.org/cdifficile). Ethical approval was granted by the University of Pittsburgh institutional review board.

Epidemiological and dietary data submitted by the study participants were analyzed to identify risk factors for C. difficile carriage. Differences in the prevalence of categorical variables were assessed using Fisher's exact test, and differences in the distribution of continuous variables were assessed using the Wilcoxon test. All odds ratios (ORs) and 95% confidence intervals (CIs) for putative C. difficile exposure risks were calculated using exact logistic regression, and results are reported as either odds ratios or median unbiased estimates, as appropriate. Analyses were conducted with SAS software (version 9.3; SAS Institute, Cary, NC).

RESULTS

One hundred six (81%) of 130 enrolled participants submitted stool specimens. Sixty-nine participants (65%) were female, 65 (61%) were <35 years of age, 71 (67%) identified their race as white, 57 (54%) had attained a college degree or higher, and 100 (94%) were self-reported omnivores. Seven (6.6%) of 106 participants were colonized with toxigenic C. difficile. There was no recovery of nontoxigenic C. difficile. The median ages of colonized and noncolonized individuals were 24 years (range, 19 to 37 years) and 30 years (range, 18 to 71 years), respectively (P = 0.25). There was no association between C. difficile colonization and race (P = 0.26), ethnicity (P = 0.64), consumption of raw beef (P = 0.41) or seafood (P = 0.65), or exposure to a physician's office, emergency department, or urgent care clinic (P = 0.88), antibiotics (P = 0.36), children <4 years of age (P = 0.56), or individuals known to have CDIs (P = 1.0). Participants who reported exposure to pets (median unbiased OR estimate, 0.14 [95% CI, 0.00 to 0.74]; P = 0.05), exposure to a dentist's office (OR, 0.56; P = 0.12), or public restroom use of ≥7 times/week (OR, 0.47 [95% CI, 0.16 to 1.37]; P = 0.18) exhibited less-prevalent C. difficile colonization, but these associations did not reach statistical significance. The most frequently owned pets for non-C. difficile-colonized participants were mammalian species (40/43 participants [93%]); no C. difficile-colonized participants reported pet ownership. Two (17%) of 12 participants who reported antibiotic use in the 90 days prior to participation had C. difficile colonization, compared to 5 (5.3%) of 94 participants with no antibiotic use, but this observed difference was not significant (OR, 3.5 [95% CI, 0.30 to 25.2]; P = 0.36).

The microbiological details for the 7 C. difficile-colonized participants are presented in Table 1. Six colonized individuals (86%) submitted >1 stool specimen. Stool specimens from 2 individuals (29%) yielded C. difficile on 2 successive occasions, 1 month apart. The median density of C. difficile in stool specimens of colonized participants was 3.0 × 10⁴ CFU/g. The quantity of C. difficile in 2 stool specimens was <1 × 10¹ CFU/g. The illumigene assay detected 3/9 (33%) stool specimens that were positive in broth enrichment cultures, each of which exhibited ≥3.0 × 10⁴ CFU/g stool. A total of 7 tcdC genotypes and 7 multilocus sequence types (MLSTs) were identified among colonized participants, including tcdC 3 (a genotype associated with the epidemic PCR ribotype 001 lineage) and tcdC 20 (a genotype associated with PCR ribotype 078, which has been detected in food animals). Genotype tcdC 1 (associated with the epidemic PCR ribotype 027) was not observed in this study.

TABLE 1.

Stool testing characteristics of 7 healthy subjects who tested positive for C. difficile in Pittsburgh, Pennsylvania

Participant	Visit no.	C. difficile culture resulta	Quantitative culture result (CFU/g stool)	Nucleic acid test result	tcdC genotype	MLST
1	1	Positive	2.7 × 103	Negative	5	58
	2	Negative
	3	Negative
2	1	Positive	<101b	Negative	20	11
	2	Negative
3	1	Positive	8.7 × 103	Negative	19	8
	2	Positive	4.9 × 104	Positive	19	8
4	1	Positive	3.0 × 104	Positive	14	53
	2	Negative
	3	Negative
5	1	Positive	<101b	Negative	53	129
6	1	Positive	8.0 × 104	Negative	3	3
	2	Negative
7	1	Positive	1.1 × 103	Negative	10	110
	2	Positive	1.6 × 106	Positive	10	110

^aIn broth enrichment culture.

^bC. difficile did not grow with direct plating of 0.03 g stool, i.e., the quantity tested with three of the initial 10-fold dilutions (see Materials and Methods).

DISCUSSION

The prevalence of C. difficile colonization in this small healthy cohort is concordant with contemporary values derived from Japanese populations and is higher than estimates from the United States, many of which were reported prior to the early 2000s (5, 7–9). This difference may result from the well-documented increase in rates of health care-associated CDIs in the United States since 2000, resulting in spillover into outpatient populations (2). We think, however, that it is more likely that the higher prevalence of C. difficile colonization observed in our cohort and in prior Japanese studies, in comparison with earlier U.S. estimates, results principally from the incorporation of essential C. difficile spore germinants (sodium taurocholate and lysozyme) in the culture medium used, in contrast to the cycloserine-cefoxitin-fructose agar (CCFA) used in prior U.S. studies, which is now known to be substantially less sensitive (13). Our data support a potential role for asymptomatic carriers in the introduction of organisms into hospitals and the transmission to susceptible community dwellers.

It is notable that the mean stool density of C. difficile among our healthy participants was 3.0 × 10⁴ CFU/g, whereas the reported value for C. difficile stool density in symptomatic CDI patients was reported to be 4.0 × 10⁶ CFU/g (14). C. difficile-colonized individuals with a low colonic density of organisms may not participate in the CDI transmission cycle; this is unlikely, however, because the infectious doses for CDIs in animal models are low (15) and C. difficile-colonized patients within hospitals are well documented to be responsible for contamination of hospital rooms and transmission to new patients (16–18). It is also notable that the majority (67%) of C. difficile-colonized participants who submitted >1 stool specimen tested negative for C. difficile 30 days after submission of their positive samples. Further study is needed to quantify the duration of C. difficile colonization in asymptomatic persons.

Of interest was the detection of tcdC genotype 20 (inferred ribotype 078) C. difficile carriage in one participant. In a recent study, our laboratory determined that 2% of uncooked ground pork sausage products were contaminated with this C. difficile lineage in the Pittsburgh, Pennsylvania, area (19). However, the current study was underpowered to identify risk factors that were significantly associated with food consumption or other participant characteristics, including the interesting lower prevalence of C. difficile colonization observed among individuals with pet ownership or frequent public restroom use.

This investigation underscores the need for caution when interpreting the results of commercial nucleic acid-based C. difficile detection assays. Three (33%) of 9 stool specimens from healthy C. difficile-colonized individuals were positive by a commonly used commercial nucleic acid test for C. difficile. These data reinforce the notion that nucleic acid-based C. difficile assays lack specificity and may overdiagnose CDIs in lower-risk populations when additional clinical criteria are not considered or when testing for other enteric pathogens is not performed (20, 21).

It is also notable that the nucleic acid test used in this study was insufficiently sensitive to detect toxigenic C. difficile in 67% of the stool specimens submitted by healthy colonized individuals that tested positive for C. difficile in broth enrichment culture. This finding should provide a note of caution to investigators seeking to screen healthy donors for fecal microbiota therapy (FMT) as a treatment for CDIs without using toxigenic anaerobic cultures. This finding probably reflects the limit of detection for the illumigene assay, which is not optimized for organism densities lower than those typically encountered in CDI patients (∼10⁶ CFU/g stool). Furthermore, nucleic acid-based assays may be insufficiently sensitive to identify asymptomatic, C. difficile-colonized, hospital inpatients, the active surveillance of whom may have a role in strategies for the reduction of hospital-acquired CDIs (17).

The principal limitation of this investigation is the small sample size. A larger sample is necessary to identify risk factors for C. difficile colonization among healthy individuals. Because we performed broth enrichment cultures with only 0.01 to 0.1 g of stool, our culture method might have missed subjects with low C. difficile colonization densities (i.e., <10 to 100 CFU/g), but this culture method was recently shown to be the most sensitive method currently in use (13). Nonetheless, it is conceivable that our estimates of C. difficile colonization prevalence would be higher if larger quantities of stool were sampled.

In conclusion, the prevalence of asymptomatic C. difficile colonization in this cohort was 6.7%, a value that is concordant with estimates from studies performed outside the United States. Nucleic acid testing for CDIs is not as sensitive as anaerobic toxigenic cultures to identify carriers, but false-positive test results for CDIs could result from testing of carriers with other causes of diarrheal illness.