Airborne Exposures and Ambulatory Peak Expiratory Flow in Drywall Finishers

• Study Population
• Ambulatory Peak-Flow Monitoring
• Industrial Hygiene Evaluation
  
  
  • Personal sampling for total and respirable airborne dust
  • Instantaneous sampling
  • Area sampling for total and respirable airborne dust
    
    
• Statistical Analyses

Results

• Questionnaire Results
• Technical Discussion of Industrial Hygiene Results
• Personal air-sampling results
  
  
  • Task-specific instantaneous sample results
  • Area air-sampling results
  • Particle sizing
  • Ambulatory peak-flow monitoring results

Discussion
Conclusions and Recommendations
References

Appendixes
A. Personal air sampling results
B. Individual task-specific instantaneous sampling results
C. Area air-sampling results
D. Particle size distribution during personal sampling
E. Individual maximum peak expiratory flow and dust exposure

Tables

1. Summary of 8-hour time-weighted-average personal air-sampling results by task for site 1
2. Summary of 8-hour time-weighted-average personal air-sampling results by task for site 2
3. Personal air-sampling results and average ratio of respirable to total dust, by site and task
4. Task-specific instantaneous sample results at site 1
5. Task-specific instantaneous sample results at site 2
6. Comparison of site-specific personal data to task-specific instantaneous sample results
7. Linear regression model to predict daily peak-flow variability

Figures
A-1. Site 1 personal air sampling results
A-2. Site 2 personal air sampling results
B-1. Task-specific instantaneous sampling results
D-1. Particle size distribution during joint-compound application
D-2. Particle size distribution during sanding
E-1- E-10. Maximum peak expiratory flow and 8-hour personal time-weighted average total dust exposure, by date and subject, 26

Several studies have evaluated airborne exposures of drywall finishers, revealing a range of exposures that included high levels of total and respirable dust (Schneider and Susie 1993; Ken Mead, NIOSH, personal communication, June 1996 ). (Respirable dust is dust small enough to go deep into the lungs.) The high dust levels, coupled with concern among drywall workers about possible adverse health effects of the exposures, indicated a need for further exposure studies of drywall workers.

The prospective cohort pilot study described here sought to determine the relationship of airborne dust exposures, job task, and expiratory peak-flow measurements in a small group of drywall finishers.

The study had three main goals. First, the authors characterized total and respirable dust exposures -- concentrations and identities -- during a typical workweek for the drywall finishers. With this information, the authors intended to assess when dust-control measures are most needed in drywall finishing.

Second, the authors sought to evaluate possible acute health effects of inhalation of the dust generated during drywall finishing -- whether exposures during drywall finishing tasks lead to acute decrements or increased variability in lung function. This was done by using a series of peak expiratory-flow rate measurements obtained throughout the study week. Serial measurement of peak flow rates has been shown to be a sensitive indicator of bronchial reactivity, possibly associated with irritation of the mucosal lining and coughing and wheezing (Hetzel and Clark 1980).

Until recently, however, when the National Institute for Occupational Safety and Health (NIOSH) developed a portable electronic spirometer, it was difficult to get repeated reliable measures of expiratory flow rate for construction workers, who regularly move about the worksite. (When someone blows into a hose on the spirometer, it records the force and volume of air coming out of the lungs.)

So, a third goal of the study was to evaluate use of the new spirometer for research in the construction workplace.

The management of Precision Walls, Inc., one of the largest commercial drywall companies in North Carolina allowed employees to participate in this study. Participants were each paid $150 in return for collecting information even while at home.

In July 1996, 10 drywall finishers representing 2 work groups enrolled in our study. One work group, numbering 4, worked in a 4-story building at the preliminary stages of drywall finishing (site 1). The other work group, including 5 workers, was performing the final stages of drywall finishing in a 2-story building (site 2). We enrolled the supervisor of these two groups as a tenth subject.

After giving informed consent, participants completed a health questionnaire written in English and Spanish. To gather information about current respiratory symptoms, questions adapted from the American Thoracic Society Standardized Respiratory Questionnaire (Ferris 1978) and the International Union Against Tuberculosis bronchial symptoms questionnaire (Abramson 1991). Work-related respiratory symptoms were identified by asking subjects whether they had ever been bothered at work by wheezing, cough, chest tightness, or shortness of breath. Additionally, the authors collected medical, smoking, and occupational history.

Participants were asked not to alter their work routine, finishing materials, or home activities.

Ambulatory Peak-Flow Monitoring

At the beginning of the study, a NIOSH-certified respiratory technician measured each participant's lung function using a stationary flow spirometer. The technician also taught participants to use the portable belt spirometer. Subjects were asked to collect 3 consecutive measurements of peak expiratory flow every 2 hours while awake on workdays and non-workdays for one week. An alarm was set on the belt spirometer to ring every 2 hours after the first time it was used each day, reminding the subject to take the measurements. The instrument recorded measurements electronically and provided feedback to the subject, flashing a green light if the measurement was satisfactory.

Industrial Hygiene Evaluation

The industrial hygiene evaluation was conducted at both construction sites throughout the workweek of the study. Drywall finishers worked from 7 a.m. until 3 p.m., Monday through Friday, with an hour for lunch.

Airborne dust concentrations were obtained by personal, area, and instantaneous sampling. The air-sampling methods were based on Methods 0500 and 0600 for total and respirable dust, not otherwise regulated, as described in the NIOSH Manual of Analytical Methods (Eller 1994).

Personal sampling for total and respirable airborne dust. All subjects, except the supervisor (who did no finishing work during the study), wore personal breathing-zone samplers for total dust throughout each 8-hour work shift. The sampling method used a flow rate of 2 liters per minute and filter cassettes containing tarred 37 millimeter, 5 micrometer(µm) PVC filters, because excessive filter loading was not expected (when the filter would be too full to produce an accurate measure). Also, 2 subjects at each site wore personal breathing-zone samplers for respirable dust using a 10 millimeter cyclone and a filter cassette containing a tarred 5 µm PVC membrane. The flow rate was set to 2.2 liters per minute for a 4 µm cut-point. Personal 8-hour time-weighted average exposures for total and respirable dust were determined by weight.

Instantaneous sampling. The authors used a Grimm 1105 dust monitor and aerosol spectrometer for instantaneous dust sampling. The sampling was intended to characterize each phase of drywall finishing and to identify high-exposure tasks. Tasks were distinguished by equipment or tools used. Measurements were collected over 1-to-15-minute intervals.

The Grimm 1105 dust monitor and aerosol spectrometer measured particle concentration (0.001 to 100 mg/m³) and size (0.5 to 20 µm). The Grimm monitor reported results according to particle size distributions known as the inhalable, thoracic, and alveolic fractions. (Inhalable dust particles are the largest ones and can be breathed in through the nose.) These fractions coincide with the American Conference of Governmental Industrial Hygienists (ACGIH) particle-size selective sampling criteria for airborne particulate matter (American Conference of Government Industrial Hygienists 1997).

Area sampling for total and respirable airborne dust. On each workday, area sampling was used to measure total and respirable portions of airborne dust using the methods described above for personal airborne dust sampling. Selection of an area to sample proved difficult because each day the drywall finishers worked throughout their building site. This sampler thus was not near drywall-finishing work during most of the workday.

Statistical Analyses

Prior to analysis, names were removed from the questionnaires, spirometry, and personal airborne-dust samples, and each individual's data were assigned a blind identifier allowing linkage of his information. Information from each questionnaire and from personal airborne-dust collections were entered into a personal computer-based data manager; peak expiratory flow measurements were downloaded into the data manager.

The authors calculated reliability of peak expiratory flow measured at each time point by determining whether the second-highest reading was more than 90% of the highest peak-expiratory flow measurement (see Enright and others 1995). All analyses used only the maximum peak flow measured at each time. The authors plotted maximum peak expiratory flow and total airborne dust by time for each subject across all days. These plots were assessed for trends suggesting possible effects of learning (becoming familiar with the device), day of the week, and dust exposure on peak flow.

Daily peak-flow variability was determined for each day for each participant. Considering only the maximum peak flow measured each time, the daily variability was calculated as:

daily maximum peak flow minus daily minimum peak flow
	daily maximum peak flow.

The authors did not obtain 8-hour personal time-weighted average respirable dust measurements for each subject for each workday. However, the authors collected personal total-dust concentrations for each workday and recorded the work activity for each subject for all workdays. Thus, the average ratio of respirable to total dust exposure measured for a task by site was used to estimate the missing personal respirable exposures.

For statistical analyses, SUDAAN software was used, in order to adjust for clustering in the data because of multiple measurements by each individual. SUDAAN was used to perform Cochran-Mantel-Haenszel tests of the association between daily peak-flow variability and possible explanatory variables, while adjusting for the effects of smoking. For these tests, daily peak-flow variability was categorized as >=20% or <20% (value associated with asthma) and the tests were repeated for daily peak-flow variability>=8% or <8%. The possible explanatory variables included workday vs. non-workday, prevalent respiratory symptoms, work-related respiratory symptoms, and ethnicity. Each of the other possible explanatory variables (age, height, FEV₁/FVC, respirable dust, total dust, and years in the drywall trade) was categorized by its average value.¹

The authors developed a linear regression model to estimate the contribution to daily peak-flow variability of the effects of airborne exposure as measured by 8-hour personal time-weighted average total dust exposure, smoking, prevalent respiratory symptoms, and age. ²A second regression model replaced total dust exposure with 8-hour personal time-weighted average respirable dust exposure. Dust exposure and age were continuous variables. Smoking and prevalent respiratory symptoms were categorized as present and absent (smoker vs. nonsmoker; had symptoms vs. no symptoms).

In addition, the authors developed a logistic regression model assessing daily peak-flow variability as a categorical variable (one model using>=20% and a second model using >=8%). The logistic regression model used the same 2 sets of independent variables as were used in the linear regression models to produce a total of 4 logistic regression models.

The 10 study subjects included 7 Hispanics, 2 Native Americans, and one Caucasian. Their ages ranged from 25 to 46 years old (average age 35) and their drywall work experience varied from 5 months to 18 years (average time working in the trade was 7 years). Seven had never smoked and 3 were current smokers (11, 22, and 27 pack-years). Two smokers had a history of physician-diagnosed sinusitis and one nonsmoker had been diagnosed as having allergies and childhood asthma. None had asthma as an adult.

Six subjects reported prevalent respiratory symptoms. Three indicated they had been awakened at night by an attack of coughing in the last 12 months. Three others reported nighttime cough and additional lower-respiratory symptoms. Five study participants reported work-related coughing; three of those reported additional work-related lower-respiratory symptoms, such as shortness of breath or wheezing.

Technical Discussion of Industrial Hygiene Results

Personal air-sampling results. A summary of the personal air sampling for the 2 construction sites is presented in tables 1 and 2. For the individual personal air-sampling results, see appendix A.

The average ratio of total dust to respirable dust for each site was similar for a given task (table 3). These data suggest that sanding creates a substantial exposure to total airborne dust; however, much of the airborne dust is not small enough to be respirable.

Task-specific instantaneous sample results. The authors measured exposures during 1 to 15 minutes of mud application, pole sanding, and hand sanding (tables 4 and 5). (Pole sanding is using sandpaper on a pole to sand hard-to-reach areas.) Each sample is the average value over one minute. A weighted average value was calculated for all the sanding data by adding the product of each one-minute average value and the number of samples for each value used to determine the average and then dividing by the total number of samples:

add for all avg. values: (1-min. avg. value times the number of samples in that set)

total number of samples.

Personal total and respirable dust data were compared to task-specific instantaneous sample results (table 6). The Grimm 1105 dust monitor yielded higher results than did the personal sampling, possibly because of difficulty positioning the Grimm monitor close to each worker's breathing zone. In addition, the Grimm samples likely represent instantaneous concentrations during actual work whereas the personal monitors integrate exposure over the entire work day. Personal monitors thus include periods of low to moderate work activity.

Area air-sampling results. Total dust concentrations ranged from less than 0.06 to 2.22 mg/m³ (see appendix C). The relevance of these measurements is uncertain. Because drywall finishers worked throughout each building site and moved around a lot, it was difficult to select appropriate areas to obtain representative samples.

Particle sizing. The study used measurements collected with the Grimm 1105 monitor to create a histogram of the particle-size distribution, based on percentage of the inhalable mass during joint compound application and during sanding. The percentage (by mass) of particles in 8 particle-size categories -- 0.5-1 µm, 1-2, 2-3.5, 3.5-5, 5-7.5, 7.5-10, 10-15, >15 -- was combined from both sites (figs. D-1 and D-2). During joint-compound application, the greatest fraction of the inhalable mass consisted of particles with diameters between 5 and 7.5 micrometers. During sanding, particles greater than 15 micrometers were about 50% of the inhalable mass.

Ambulatory peak-flow monitoring results. Participants measured peak flow an average of 7 times during each study day (range 3 - 8.8 times per day). Subjects performed an average of 3.2 maneuvers each time they used the monitor. Ninety percent of the peak expiratory-flow measurements were reliable, based on criteria previously described.

Figures E-1 through E-10 (appendix E) plot maximum peak-expiratory flow and 8-hour personal time-weighted average total-dust exposure by time for each study day for each subject. Visual assessment by multiple observers revealed no consistent pattern of an effect of total-dust exposure or day of the week on peak flow.

Daily peak flow varied by 0.2 to 40%, with an average of 11.3% (standard deviation 7.8%). Cochran-Mantel-Haenszel summary odds ratios showed that smokers were 24.94 times more likely than nonsmokers to have daily peak-flow variability >=20% (95% CI 2.82 - 220.82). Those who had worked more than 7 years in the drywall trade were 15.33 times more likely than those who had a shorter work history to record daily peak-flow variability >=20% (95% C. I. 1.77 - 133.04). Also, those who reported work-related respiratory symptoms were 9.5 times more likely than subjects who did not report work-related respiratory symptoms to record daily peak-flow variability >=20% (95% C. I. 1.10 - 81.95). Years working in the drywall trade and work-related respiratory symptoms could not be evaluated based on smoking in the Cochran-Mantel-Haenszel analysis, because of small numbers. Age, height, prevalent respiratory symptoms (without regard to work), baseline lung function, ethnicity, and non-workdays were not significantly associated with daily peak-flow variability greater than 20%, when smoking was considered.

Smoking was not significantly correlated with daily peak-flow variability >=8%. The crude odds of measuring daily peak-expiratory flow >=8% were 4.66 times higher (95% CI 1.67- 13.02) among subjects who reported prevalent respiratory symptoms compared to those without symptoms.

Years working in the drywall trade, work-related respiratory symptoms, age, height, baseline lung function, ethnicity, and non-workdays were not significantly associated with daily peak-flow variability >=8%.

When daily peak expiratory-flow variability was considered as a continuous variable, linear regression modeling (table 7) showed that workers who already had respiratory symptoms were much more likely to have a higher daily peak-flow variability. Eight-hour personal time-weighted average total dust, smoking, and age did not significantly predict a higher daily peak expiratory flow variability. A second linear regression model substituting 8-hour personal time-weighted average respirable dust for total dust showed the same results.

Logistic regression modeling of daily peak-flow variability <=20% proved unreliable, because of the small sample size. Only 9 (12%) of 79 study days yielded daily peak-flow variability <=20%. Logistic regression modeling showed that 8-hour personal time-weighted average total dust exposure, respirable dust exposure, smoking, prevalent respiratory symptoms, baseline lung function, and age did not significantly predict daily peak-flow variability <=8%.

Personal average dust exposures during sanding were at least 10 times higher than exposures associated with mud application using a premixed paste that consisted mainly of water and limestone. Task-based instantaneous measurements showed the same relationship.

Instantaneous readings were made during pole sanding at site 1 (but not at site 2) and showed results higher than exposures during hand sanding. This finding differs from findings in an experimental NIOSH study, where hand sanding without dust controls was almost 1.7 times as dusty as pole sanding without controls (Mead, Fishbach, and Kovein 1995). The inconsistency may be explained by difficulties positioning the Grimm dust monitor near the breathing zone for hand sanding when finishers wore stilts in this pilot study. Thus, the results for hand sanding reported here may underestimate true exposures.

At both sites for all tasks, the air sampling results for respirable dust were well below the ACGIH threshold limit value of 3 mg/m³(the OSHA PEL is 5 mg/m³). At both sites, the crystalline silica concentrations were well below regulated values (see tables A-1 and A-2).

The Grimm dust monitor yielded higher results than did the 8-hour personal dust sampling (table 6) -- except for the higher Grimm task-specific instantaneous measurements during mud application at site 1. The discrepancy may be explained by difficulty positioning the Grimm monitor near the breathing zone. The Grimm measurements may also reflect airborne dust generated by workers engaged in a variety of activities adjacent to the study subjects.

Area sampling provided little useful information, because the drywall finishers moved quickly from area to area. Thus, a prolonged area sample did not reflect the workers' airborne exposure throughout the day see appendix C.

The study shows that ambulatory peak expiratory-flow monitoring every 2 hours is feasible in the construction workplace. Most of the participants in the study fully complied with instructions, including using the device at home and during the weekend. The ambulatory peak-flow monitor proved reliable and easy to use. None of the participants had questions or difficulties using the device. Additionally, because measurements were recorded electronically, possible inaccuracies and fabrication of values were eliminated (see Quirce and others 1995).

The study used daily peak-flow variability <=20% as a level that may be abnormally high, indicating asthma (Hetzel and Clark 1980). However, limitations to using such a threshold have been described (Higgins and others 1989). The study used the threshold of <=8% as a more sensitive maker of variability, because healthy individuals without asthma have been reported to have a daily variability in peak expiratory-flow rate of as much as 8% using measurements made 4 times a day (Quackenboss and others 1991). The more-sensitive threshold loses specificity and the small sample size may have limited the authors' ability to detect significant associations with this outcome.

Cochran-Mantel-Haenszel testing showed that smoking, work-related lower respiratory symptoms and working as a drywall finisher longer than 7 years were significantly associated with daily peak-flow variability <=20%. These results must be interpreted with caution, because only 75% of the cells for each test contained more than 5 subjects. The absence of significant associations between these factors and daily peak-flow variability 8% may be because of the loss of specificity using 8% as a threshold.

The linear regression model indicated that having current respiratory symptoms increased daily peak-flow variability by 8% compared to individuals who had no respiratory symptoms, taking into account total dust exposure, smoking, and age. This finding is consistent with the expectation that symptomatic individuals record higher peak-flow variability than asymptomatic individuals. The small sample size likely restricted the study's ability to detect other possible risk factors for increased peak-flow variability.

Although it would have been preferable to study one larger work group, at the time of this study the contractor was working on sites requiring work groups of only 4 or 5 drywall finishers. Studying two groups at different stages of the work, however, showed that airborne dust measurements may differ according to the phase of construction, as well as by task.

Still, personal-sampling measures of airborne total-dust exposures during drywall finishing were elevated, indicating that control measures during sanding are needed. The proportion of respirable dust was much higher during mud application (up to 37% of total dust) compared to sanding (up to 10.9%). Nevertheless, total dust measurements during mud application were low, yielding low respirable fractions. The study subjects used a premixed paste for mud application and these results may be lower than might be found when drywall finishers mix joint compound from powder.

The study identified no acute drops in peak flow related to amount of dust exposure or day of the week. Current lower-respiratory symptoms increased daily peak-flow variability, after controlling for smoking, total dust exposure, and age. Additionally, the data suggest that working as a drywall finisher longer than 7 years and experiencing work-related lower respiratory symptoms may increase risk for daily peak-flow variability 20%, a level associated with asthma.

It is clearly feasible for construction workers to take a series of measurements while awake -- every 2 hours-- using a portable spirometer.

The Grimm dust monitor proved to be an inefficient way to collect exposure information. Additionally, area dust-sampling measurements do not appear to be useful in evaluating potential respiratory effects during drywall finishing.

A larger study simultaneously measuring serial peak flow and dust exposures should be considered to further explore this relationship. Any future study should use 8-hour average personal-dust measurements to assess airborne exposures.

American Conference of Governmental Industrial Hygienists. 1991. Particulates not otherwise classified (PNOC). In: Documentation of the Threshold Limit Values and Biological Exposure Indices, Sixth Edition, Volume II. Cincinnati, Ohio, 1166-67.

1997. TLV's and PEL's, Appendix D: Particle Size-Selective Sampling Criteria for Airborne Particulate Matter, American Conference of Governmental Industrial Hygienists. Cincinnati, Ohio.

Eller, P.M., ed., NIOSH Manual of Analytical Methods, Fourth Edition. 1994. Cincinnati, Ohio: U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control and Prevention. (Particulates not otherwise regulated, Total, Method # 0500; Particulates not otherwise regulated, Respirable, Method # 0600). DHHS (NIOSH) Publication 94-113.

Enright PL, and others. 1995. Ambulatory monitoring of peak expiratory flow, reproducibility and quality control. Chest, 107(3): 657-61. Ferris BG. 1978. Epidemiology standardization project. American Review of Respiratory Diseases, 118:1-120.

Hetzel MR, Clark TJH, 1980. Comparison of normal and asthmatic circadian rhythms in peak expiratory flow rate. Thorax, 35:732-38.

Higgins BG, and others. 1989. The distribution of peak expiratory flow variability in a population sample. American Review of Respiratory Diseases, 140:1368-72.

Mead KR, Fishbach TJ and Kovein RJ. 1995. In depth survey report: A laboratory comparison of conventional drywall sanding techniques versus commercially available controls. Cincinnati: US Department of Health and Human Services, Public Health Service, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health.

Quackenboss JJ, and others. 1991. The normal range of diurnal changes in peak expiratory flow rates. Relationship to symptoms and respiratory disease. American Review of Respiratory Diseases, 143:323-30.

Quirce S, and others. 1995. Peak expiratory flow monitoring is not a reliable method for establishing the diagnosis of occupational asthma. American Journal of Respiratory and Critical Care Medicine, 152:1100-1102.

Schneider, Scott, and Pam Susie. 1993. Final Report: An Investigation of Health Hazards on a New Construction Project. Washington, D.C.: CPWR – Center for Construction Research and Training.