Airborne Fungi and Bacteria: Background Levels in Office Buildings.

Introduction

Poor indoor air quality (IAQ) in office spaces is a growing concern among office workers. Investigation of health problems potentially related to IAQ poses new challenges for environmental health and safety professionals. A variety of volatile organic compounds (VOCs), fungal spores, bacteria, protozoa, pollens, and dust mites have been implicated as agents in IAQ-related health complaints (1). Symptoms associated with poor IAQ include irritation, headaches, fatigue, sinus congestion, allergic reactions, and asthma. These symptoms are general in nature and may be caused by factors other than poor IAQ. A building with a high prevalence of reported symptoms among occupants may be described as a "sick building"; a group of persons exhibiting these symptoms may be described as having "sick building syndrome" (2). Symptoms characteristic of sick building syndrome are often relieved shortly after the occupant leaves the building (3).

The number of buildings that may be labeled as sick buildings seems to have increased in the last two decades. It has been suggested that a major contribution to this increase is reduced provision of outside air in the interest of energy savings. Sundell summarized work on 210 office buildings in northern Sweden that seemed to demonstrate a "dose-response relationship between the outdoor airflow rate and the risk of having symptoms" (2). The problem is likely multifactorial. Menzies et al. and others have shown that simply increasing ventilation may not reduce complaints among building occupants (4-6). A laudable suggestion is proactive prevention of IAQ problems through a focus on HVAC system maintenance and operation, psychosocial factors, and handling of complaints (7). Measures of IAQ related to successful HVAC maintenance and operation would be a useful adjunct to such an approach.

Investigators commonly measure airborne culturable fungi and bacteria and total fungal spore concentrations along with carbon dioxide, carbon monoxide, and, increasingly, VOCs. Measurements of airborne fungi in office spaces have been used in several cases to confirm a problem with IAQ before successful remediation (8). Few data have been published, however, on background or baseline levels of airborne fungi or bacteria in office spaces. The U.S. Environmental Protection Agency (U.S. EPA) is completing its Building Assessment Survey and Evaluation (BASE), designed to address this data need (9).

The goal of the present study is to provide information on baseline IAQ consistent with the protocols for bioaerosols used in the U.S. EPA BASE study (10). On January 14, 1994, a major earthquake was centered near the campus of the California State University at Northridge. This earthquake presented an opportunity to study a variety of temporary office structures set up as a result of severe damage to existing buildings. Two large tentlike structures resembling Quonset huts, a doublewide trailer facility, and a newly constructed addition to the student union building (undamaged by the earthquake) were chosen for this IAQ investigation. One of the tentlike domes (called Dome A) housed the staff of the Admissions and Records Department. This structure was located on the north edge of campus near a trailer park (temporary classrooms) and an open area. During the half year of operation prior to this study, the campus Environmental Health and Safety Department had received a significant number of IAQ-related health complaints from the workers in Dome A. A comparable structure (Dome B) which housed the Graduate Studies Program and other administrators was located on the northwest edge of campus on a parking lot adjacent to a side street. The trailer (which housed a dean and her staff) was located in the center of a large paved parking lot near several other trailers on the east side of campus. The student union addition was located near the center of the campus.

Preliminary sampling began in the fall of 1995, and samples of indoor air and outdoor air were collected from each building location on five occasions throughout 1995-1996. An occupant questionnaire was administered at the beginning of the study to obtain measures of the magnitude and types of complaints potentially related to poor indoor air quality. The questionnaire confirmed that a probable sick building syndrome (SBS) situation existed in Dome A. Building surveys showed no visible signs of water damage, mold growth, or problems with air-handling units. A complete SBS investigation, with sampling for VOCs, particulates, endotoxins, and so forth, was not possible with available resources.

TABLE 1

Symptoms Reported by Structure Occupants (Percentage)

Symptoms                 Dome A   Dome B   Trailer   Student Union

Difficulty hearing        87.8     43.3     25.0           0.0
Eye irritation            85.4     53.3     12.5           0.0
Too hot                   82.9     66.7     37.5          16.7
Too cold                  80.5     83.3     25.0          50.0
Hard concentrating        78.0     46.7     12.5          16.7
Stress                    78.0     50.0     62.5          25.0
Sinus congestion          78.0     56.7     37.5          25.0
Headaches                 75.6     60.0     50.0          33.0
Sneezing                  73.2     40.0      0.0           8.3
Fatigue                   73.2     43.3     62.5           8.3
Odors                     73.2     13.3     12.5           0.0
Aching joints             48.8     30.0     25.0           8.3
Back pain                 43.9     16.7     25.0          25.0
Dry skin                  43.9     26.7     25.0          16.7
Itching                   34.1     13.3      0.0           0.0
Dizziness                 29.3      6.7      0.0          25.0
Nausea                    29.3      0.0      0.0           8.3
Twitching                 26.8      6.7     25.0           0.0
Chest tightness           26.8     13.3     12.5           0.0
Irritation                24.4     16.7      0.0           0.0
Heartburn                 17.1      6.7     12.5           0.0
Skin discoloration         4.9      0.0      0.0           0.0

Methods

Mesophilic Bacteria Concentrations

An Andersen N6 sampler was run at 28.3 liters per minute (L/min) for seven-minute sampling periods at three locations inside (front, middle and rear) and for five minutes at one outside location near the air-handling unit intake (except at the student union, where the outside sampling location was near the front entrance). On each sampling date, three replicate samples were collected consecutively at each of the locations. Sampling was performed in the late morning or the early afternoon during normal occupant activities. Sample flow rates were continuously monitored with a rotameter previously calibrated with a wet test meter. The Andersen N6 sampler was positioned on a ring stand at a height of 1.8 meters on a portable cart. The sampler was rinsed in 70 percent ethyl alcohol after testing of each location. Sterilized Andersen glass plates were filled with 27 milliliters (mL) of Trypticase[R] soy agar containing 0.4 grams per liter (g/L) of cycloheximide to prevent mold growth (10). After three days of incubation at 32 [degrees] C [+ or - ] 2 [degrees] C (in the dark), the plates were counted with a Quebec colony counter. Counts were adjusted for positive hole correction according to the table in the Andersen manual. Sterility controls were run for each sampling site.

Fungi Concentrations

Fungi were sampled with the Andersen N6, as above, with collection times of 5 minutes for inside locations and one minute for outside locations. Malt extract agar was used with 0.5 g/L of chloramphenicol to inhibit bacterial growth (11). The formulation was 20 g malt, 1 g peptone, 20 g dextrose, and 20 g agar per liter of purified water (10). The Andersen plates were counted and adjusted, as above, after incubation for five days at 25 [degrees] C [+ or - ] 3 [degrees] C under alternating fluorescent light/dark conditions. Morphology, texture, and color were used to categorize colonies into one of four groups:

1. Cladosporium,

2. nonsporulating,

3. Aspergillus and Penicillium, and

4. other genera.

Microscopic examination at 400X magnification was used for confirmation.

Evaluation of Microbiological Data

The bacteria and fungi colonies were counted, and airborne concentrations were reported as colony-forming units per cubic meter of air (CFU/[m.sup.3]). Fungi counts were also characterized by calculation of the relative prevalence of each fungal type. Statistical analysis (analysis of variance and means testing) was performed with the SAS System for Windows, Version 6.12 (SAS Institute, Cary, North Carolina).

Other Parameters

A portable IAQ monitor (Metrosonics AQ-502) was used for continuous carbon dioxide monitoring and for measuring carbon monoxide, relative humidity, and temperature.

TABLE 2

Types of Airborne Fungi (Percentage)

                           Dome A                 Dome B
Type                 Indoors    Outdoors    Indoors    Outdoors

Cladosporium            58         67          56         63
Nonsporulating          20         18          19         18
Penicillium             11          8          13         15
Other                   11          7          12          4

Results

SBS Case Study

Questionnaires were returned by 41 of 50 persons (82 percent) from Dome A, by 30 of 40 persons (75 percent) from Dome B, by eight of eight persons (100 percent) from the trailer, and by 12 of 40 persons (30 percent) in the student union building. Table 1 shows a high level of complaints in Dome A and some elevated numbers of complaints in the other temporary structures.

The average airborne fungi concentration in Dome A (geometric mean = 110 CFU/[m.sup.3], geometric standard deviation = 1.5) was higher than in Dome B (42 CFU/[m.sup.3], geometric standard deviation = 2.1). The geometric means of fungi in the trailer (100 CFU/[m.sup.3], geometric standard deviation = 1.7) and the student union (69 CFU/[m.sup.3], geometric standard deviation = 2.6) also were higher than those in Dome B. Multiple ANOVAs showed that Dome A and the trailer had similar fungi concentrations, that the student union had lower concentrations, and that Dome B had the lowest concentrations (p = .0001). Airborne fungi concentrations were not observed to be excessively high at any location. They ranged from 5 to 420 CFU/[m.sup.3] indoors and from 12 to 1,100 CFU/[m.sup.3] outdoors.

Likewise, for airborne bacteria concentrations, geometric means for Dome A (140 CFU/[m.sup.3], geometric standard deviation = 1.8), the student union (150 CFU/[m.sup.3], geometric standard deviation = 1.8), and the trailer (120 CFU/[m.sup.3], geometric standard deviation = 1.4) were similar to each other but greater than the geometric mean observed in Dome B (44 CFU/[m.sup.3], geometric standard deviation = 1.7) (p = .0001). No significant differences were seen among the geometric means for bacteria in Dome B, the trailer, and the student union. Sample airborne bacteria concentrations were also fairly low and ranged from 15 to 350 CFU/[m.sup.3] indoors and from 35 to 690 CFU/[m.sup.3] outdoors.

The ratios of the bacteria and fungi concentrations indoors to their concentrations outdoors provided evidence that indoor growth or amplification of bacteria or fungi was not occurring and thus not a likely cause of the SBS problem. The airborne bacteria were all lower indoors than outdoors. The ratios of indoor to outdoor fungi concentrations did not exceed 40 percent (35 percent for Dome A and 15 percent for Dome B). Reduction of numbers of organisms from the outdoor source air is commonly observed in properly functioning and maintained air-conditioned buildings (11,12). Nearly identical fungal diversity indoors and outdoors provided additional evidence for the absence of amplification (Table 2). Very similar distributions were observed for the trailer and student union.

Concentrations of carbon dioxide did not point to ventilation problems; levels were generally low (less than 800 parts per million). Ventilation systems were switched on and off erratically, making accurate estimation of the fresh air supply impossible; building occupants had access to air controllers and turned the systems on and off randomly. Relative humidity was acceptably low, with measured levels in the range of 18 to 40 percent.

Baseline Data and Variation

Variation in the numbers of airborne fungi and bacteria was high, and the data did not fit the normal distribution. Transforming the concentrations to logarithms did lead to normally distributed data in almost every case. The coefficients of variation of the transformed data ranged from 3.7 to 36.8; most were less than 25 percent (compared with greater than 50 percent for most nontransformed data). The data sets contained nine counts each (three locations per building times three replicates per location).

The geometric means of airborne fungi (from all five dates) at front, center, and rear locations are presented in Figure 1. The differences were not statistically significant except that the concentration at the front sampling location in the student union was higher than the concentrations at the center and rear locations (p = .0046). The student union was the only building with regular hallways that isolated air flow, and the front sampling location was near the main entrance where student traffic was heavy. For airborne bacteria, no statistical differences among the sampling locations were observed except at the center location in Dome A, where concentrations were statistically higher than at front and rear locations (p = .0053) [ILLUSTRATION FOR FIGURE 2 OMITTED].

Seasonal variation was not as great as was expected from changing moisture and temperature conditions outdoors. Outdoor airborne bacteria concentrations did not vary statistically among the five sampling dates (p = .4698). Seasonal indoor airborne bacteria concentrations did not vary statistically for the trailer (p = .0938) or for Dome B (p = .3299) [ILLUSTRATION FOR FIGURE 3 OMITTED]. The summer bacteria concentration in Dome A was statistically higher (p = .0307), and the February concentration in the student union was statistically lower (p = .0099). Differences among the other dates were not significant.

Outdoor fungi concentrations showed no significant differences throughout the year except for the student union low value in August (p = .0018). Fungi concentrations indoors [ILLUSTRATION FOR FIGURE 4 OMITTED] were more variable throughout the year: Dome A showed a statistically higher November concentration and a lower August concentration (p = .0001), Dome B showed statistically higher November and February concentrations (p = .0006), and the trailer showed a higher February concentration (p = .01). The student union showed a low August concentration (p = .0096) consistent with the observed low outdoor value.

Discussion

In terms of the occupant complaints, self-reporting of symptoms, although rapid and inexpensive, probably was not a good substitute for medical evaluations. Other studies have noted higher reporting of symptoms when occupants were aware of the SBS investigation (13). Psychosocial aspects of reporting were especially important in this situation, with post-earthquake stress and high job-related stress in Dome A as a result of unusual demands and lack of control, crowded and noisy conditions, poor lighting, and temperature swings. Together, these factors may explain many of the observed survey differences. This experience reaffirms the cautions expressed by others about the importance of establishing specific hypotheses for suspected chemical or biological etiologic agents based on medically indicated conditions (3,14,15).

The need for baseline data is connected to the larger issues of SBS or building-related illness (BRI) investigations and environmental prevention. Whether to sample air at all in SBS situations is presently a serious and controversial issue. Testing for airborne micro-organisms may not be necessary for finding amplifying sources of bioaerosols and can in fact prove misleading (by giving false negative findings), as Morey has repeatedly shown (16). In terms of prevention, a consensus is developing in the scientific community that

1. fungal growth should not occur to any significant degree in a properly designed and functioning building,

2. large amounts of visible fungal growth pose a potential health risk in the indoor environment and should be removed, and

3. stagnant water poses a potential health risk in the indoor environment.

Measures of success in achieving "acceptable" levels of indoor fungi or bacteria are needed (even toxic fungi are a normal part of outdoor and indoor environments). Such measures may involve culturable bacteria or fungi, total fungal spore counts, fungal VOC concentrations, or some other economical test. These measures would fit logically into the kind of proactive IAQ protocol recently proposed by Greene et al. (7). An appropriate analogy might be the current use of coliform bacteria to indicate acceptable potable water treatment.

The variation in fungi and bacteria concentrations observed in this study within buildings and throughout the year does not seem prohibitive of their possible use as IAQ indicators of building performance (at least in this part of southern California). More data are needed to determine the optimum numbers of sampling locations and replicates for a structure. More data from geographically disparate sites are also needed. Our data, the preliminary data from 16 U.S. EPA BASE study buildings from around the country (9), and the data from 23 complaint and noncomplaint office buildings in Sweden suggest that the mean bacteria and fungi concentrations are fairly consistent across a range of office buildings (17). These data are also within the range specified in suggested guidelines for indoor fungi concentrations (18). Additional comparisons of published baseline concentrations are in progress.

Conclusions

The statistically higher fungi and bacteria concentrations observed in the complaint dome (relative to the other dome) could have been misleading. Low indoor/outdoor ratios and similarities in fungal types provided evidence that factors other than bioaerosols were probably responsible for the observed SBS situation. Testing for additional parameters may have proved useful. The situation was fortuitously resolved when the staff in the problem dome were moved to better facilities shortly after the data collection phase of the study.

Baseline data from the several structures should prove useful in future efforts to assess prevention activities. The geometric means for airborne fungi were low and ranged from 42 to 110 CFU/[m.sup.3]. Geometric means for airborne bacteria were in a similar range of 44 to 150 CFU/[m.sup.3]. Differences in concentrations among different locations within structures were not generally significant. Likewise, seasonal variation of bacteria and fungi concentrations was quite low. The mean concentrations of airborne bacteria and fungi were similar to those reported in preliminary (1994) U.S. EPA BASE results from office buildings from around the country, as well as to those found in other office spaces in Sweden and elsewhere (9,17). Additional comparisons should be performed to demonstrate the feasibility of using culturable bacteria or fungi as performance guidelines for good IAQ management practices.

Acknowledgement: This research was supported by the California Urban Environmental Research and Education Center.

REFERENCES

1. Brown, K.S. (1996), "Sick Days at Work," Environmental Health Perspectives, 104:1032-1035.

2. Sundell, J. (1996), "What We Know and Don't Know About Sick Building Syndrome," ASHRAE Journal, 38:51-57.

3. Rafferty, P.J. (ed.) (1993), The Industrial Hygienist's Guide to Indoor Air Quality Investigations, Fairfax, Va.: American Industrial Hygiene Association.

4. Menzies, R., J-.P Farant, J. Hanley, E Nunes, and R. T. Tamblyn (1993), "The Effects of Varying Levels of Outdoor-Air Supply on the Symptoms of Sick Building Syndrome," New England Journal of Medicine, 328:821-827.

5. Hodgson, M.J.,J. Frohlinger, M. Karpf, S.A. Olenchock, W. Permar, C. Tidwell, and N.D. Traven (1991), "Symptoms and Microenvironmental Measures in Nonproblem Buildings," Journal of Occupational Medicine, 33:527-533.

6. Fanger, P.O., P. Bluyssen, G. Clausen, and J. Lauridsen (1988), "Air Pollution Sources in Offices and Assembly Halls," Energy Buildings, 12:7-19.

7. Greene, R.E., J.M. Casey, and P.L. Williams (1997), "A Proactive Approach for Managing Indoor Air Quality," Journal of Environmental Health, 60(11): 15-21.

8. Reynolds, S.J., C.E. McJilton, and A.J. Streifel (1990), "Elevated Airborne Concentrations of Fungi in Residential and Office Environments," American Industrial Hygiene Association Journal, 51:601-604.

9. Brightman, H.S., S.E. Womble, E.L. Ronca, and J.R. Girman (1998), "95 Baseline Information on Indoor Air Quality in Large Buildings BASE 95," <http://www.epa.gov/iaq/base/ref/html>.

10. Office of Research and Development and Office of Air and Radiation (1994), A Standardized EPA Protocol for Characterizing Indoor Air Quality in Large Office Buildings, Washington, D.C.: U.S. EPA.

11. Burge, H.A., K. Kreiss, P.R. Morey, J. Otten, K. Peterson, M. Chatigny, and J. Feeley (1987), "Guidelines for Assessment and Sampling of Saprophytic Bioaerosols in the Indoor Environment," Applied Industrial Hygiene, 2(5):R10-R16.

12. Holt, G.L. (1993), "Determination of Seasonal Indoor/Outdoor Fungal Ratios and Indoor Bacterial Levels in Non-Complaint Office Buildings," Ph.D. thesis, Birmingham Ala.: University of Alabama.

13. Mikatavage, M.A., K. Dillon, E. Funkhouser, R.K. Oestenstad, K.D. Reynolds, and V.E. Rose (1995), "Beyond Air Quality - Factors That Affect Prevalence Estimates of Sick Building Syndrome," American Industrial Hygiene Association Journal, 56:1141-1146.

14. Dillon, H.K., PA. Heinson, and J.D. Miller, eds. (1996), Field Guide for the Determination of Biological Contaminants in Environmental Samples, Fairfax, Va.: American Industrial Hygiene Association.

15. Bioaerosol Committee (1989), Guidelines for the Assessment of Bioaerosols in the Indoor Environment, Cincinnati, Ohio: American Conference of Governmental Industrial Hygienists.

16. Morey, P. (1994), "Studies on Fungi in Air-Conditioned Buildings in a Humid Climate," In E. Johanning and C.S. Yang, eds., Fungi and Bacteria in Indoor Air Environments: Health Effects, Detection and Remediation - Proceedings of the International Conference, Saratoga Springs, N.Y., October 6-7, 1994, Latham, N.Y.: Eastern New York Occupational Health Program.

17. Blomquist, G., and B. Andersson (1994), "Measurements of Micro-Organisms in Non-Industrial Indoor Environments in Northern Sweden," In R.A. Samson, O.C.G. Adan, B. Flannigan, M.E. Flannigan, E.S. Hoekstra, and A.P. Verhoeff, eds., Health Implications of Fungi in Indoor Environments: Air Quality Monographs, Volume 2, Amsterdam: Elsevier Publications.

18. Rao, C.Y., H.A. Burge, and J.C.S. Chang (1996), "Review of Quantitative Standards and Guidelines for Fungi in Indoor Air," Journal of the Air and Waste Management Association, 46:899-908.

Corresponding Author: John E. Schillinger, Ph.D., Associate Professor, Dept. of Health Sciences, California State Univ., Northridge, 18111 Nordhoff St., Northridge, CA 91330.