for rey writer

Health Risks Due to Coffee Dust

Marcus Oldenburg, MD; Cordula Bittner, MD; and Xaver Baur, MD

Objective: This study assessed current health risks due to occupational exposure to coffee dust. Methods: We performed a cross-sectional study in a coffee haulage company (n � 24), a coffee silo (n � 19), and a decaffeinating company (n � 17). Cross-shift and cross-week case histories of these employees as well as lung function values were recorded. During the handling of green coffee, measurements of airborne dust were conducted. Results: The employees in these workplaces were mainly affected by erythematous and rhino- conjunctival symptoms. They occurred especially in subjects exposed to a high dust load (> 10 mg of inhalable dust per cubic meter of air; n � 28) [Pearson �2 test, p � 0.020 and p � 0.023]. IgE antibodies to green coffee and castor beans were detected in 3 workers and 10 workers, respectively. The majority of them (two employees and six employees, respectively) had shown respiratory symptoms during the past 12 months. The preshift lung function values were below average but were not dependent on the level of the inhalable coffee dust exposure. Employees with a coffee dust load > 10 mg/m3 of air showed higher unspecific bronchial responsiveness more frequently than those with lower exposures. Conclusion: During the transshipment (especially during unloading) of green coffee, a high and clinically relevant exposure to irritative and sensitizing dust occurs. Therefore, efforts to reduce these dust exposures are generally recommended. (CHEST 2009; 136:536 –544)

Abbreviations: CI � confidence interval; FEV1/FVC% � FEV1/FVC percentage; kUA � kilounits; OR � odds ratio; PD20 � applied dose of methacholine inducing a 20% drop in FEV1; PM � particulate matter

C offee is a favorite beverage of millions of people.Only the two coffee species Arabica and Robusta are of economic importance. Coffee cherries are harvested up to twice a year and collected in jute bags. After emptying the bags, the cores, the so- called green coffee beans, are removed from the pulp of the coffee cherries. These beans are shipped worldwide, mainly in big bags inside 20-ton contain-

ers. During the transport and refinement of the green coffee in ports, exposures to irritative and sensitizing dust occur.1

A high frequency of asthmatic diseases among workers of a green coffee processing factory was already observed in the 1950s.1 It was also described that roasted coffee may be implicated in occupa- tional asthma due to heat-resistant allergens.2 In addition to coffee, castor beans were also regarded as a cause of asthma.3 Castor beans have a wide variety of uses. Among others, they are used in lamps, in laxatives, and as a fertilizer. In the countries of origin, the plantations of castor and coffee beans frequently border each other, and the harvested beans are transported by the same ship. Castor beans are thought to be an accidental contaminant of green coffee beans because some coffee bags may have been used previously for the transportation of castor beans or stored in conta- minated ship containers.3– 6

Nowadays, green coffee stored in big containers is imported by ships and transported by trucks of

Manuscript received August 11, 2008; revision accepted February 19, 2009. Affiliations: From the Institute for Occupational and Maritime Medicine (Drs. Oldenburg, Bittner, and Baur), University of Hamburg, Hamburg, Germany; and the Hamburg State Depart- ment for Social Affairs, Family, Health and Consumer Protection (Drs. Oldenburg, Bittner, and Bauer), Hamburg, Germany. © 2009 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (www.chestjournal.org/site/ misc/reprints.xhtml). Correspondence to: Marcus Oldenburg, MD, Department of Maritime Medicine, Institute for Occupational and Maritime Medicine (ZfAM), University of Hamburg, Hamburg State De- partment for Social Affairs, Family, Health and Consumer Protection, Germany, Seewartenstrasse 10, D-20459 Hamburg, Germany; e-mail: [email protected] DOI: 10.1378/chest.08-1965

CHEST Original Research COFFEE DUST EXPOSURE

536 Original Research

Downloaded From: http://journal.publications.chestnet.org/ by a California State University Northridge User on 04/24/2013

specialized haulage companies to coffee-processing factories (eg, coffee silos or decaffeinating compa- nies). The drivers of the haulage companies unload the coffee at tilting stations. In coffee silos, the coffee is roughly purged and placed in the silos for interim storage on call. In decaffeinating companies, the green coffee is impacted by steam after a rough cleaning and decaffeinated by chemical treatment. Afterward, the coffee is dried and polished. The aim of our study was to determine whether the exposure to coffee dust during loading and processing under the presently improved hygienic conditions (eg, use of effective exhaust systems during discharge) still represents a health hazard to employees.

Materials and Methods

We performed an occupational medical cross-sectional study in a haulage company mainly dealing with the transport of coffee, in a coffee silo, and in a decaffeinating company. Willing partici- pants were 24, 19, and 17 exclusively male subjects (82.8%, 100%, and 100%, respectively). All employees of the three investigated companies were personally informed about the study by the management and the company physician. Furthermore, a separate meeting was held to explain the objective of our investigation. The participation in this study was absolutely voluntary. All participants gave their informed consent, and the authors received institutional review board approval for the study.

We performed three investigations in each of the three companies: at the beginning of the shift on Monday morning, at the end of that shift, and on the following Friday noon. Each time the focus was on exposure-related symptoms, and a spirometry, including measurements of bronchial responsiveness by means of methacholine testing, was conducted. Preceding the investigation on Monday morning was a period of at least 1.5 days without exposure to coffee dust (median, 65.6 h [range, 38 to 744 h]).

Inhalative Dust Exposure During the Handling of Green Coffee

To assess the airborne level of alveolar or inhalable dust (particles with an aerodynamic diameter [particulate matter (PM)] �2.5 �m [PM2.5] and 10 �m [PM10]) during the handling of coffee, air measurements were performed using the sampling system of the Institute for Occupational Health and Safety, St. Augustin, Germany (Project-No. BIA 1061).7 Dust measure- ments were conducted with a glass fiber filter, type MN 85/90, by maintaining an airflow rate of 0.6 m3/h or 0.21 m3/h for the inhalable or alveolar dust fraction (according to EN 481), respec- tively.8 During five different working shifts, 12 personal and stationary gravimetric air measurements of the alveolar or inhal- able coffee dust were conducted over several hours.

The measurements were task related and therefore not completely identical with the prescribed procedure in order to obtain the shift-related mean value. In Germany, the general threshold limit value of dust is 3 mg/m3 for the alveolar fraction and 10 mg/m3 for the inhalable fraction. If these limits are maintained, no increased health risk is expected in the case of nonsensitizing agents (Technical Regulation for Hazardous Goods 900).

The employees at the tilting stations were presumably highly exposed to coffee dust. To evaluate the health effects of green

coffee, these workers (all employees of the haulage company and four employees of the coffee silo; n � 28) were allotted to the “higher dust-exposed group” (� 10 mg of inhalable dust per cubic meter of air) and the remaining ones to the “lower dust-exposed group” (� 10 mg/m3; n � 32).

Measurements of the Airborne Exposure to Germs

Air measurements of germs were performed in the decaffein- ating company to roughly assess the level of germs during the handling of organic raw coffee. Sedimentation plates were in- stalled in several company areas for 30 min and agar specifically incubated. To determine the total number of germs and the mold concentration, Caso and DG-18 agars were used.

Furthermore, the impaction method RCS Plus using the serial devices of the Biotest Company (St. Georgen, Germany) were applied for the quantitative determination of airborne germ concentrations. To measure the total germ count and the mold concentration, the TC-18 and DG-18 bands were applied with airflows of 100 L and 200 L, respectively. The colonies of bacteria and molds were counted, and the number of colony forming units were determined.

Questionnaire

The standardized interview included questions of demographic data as well as airway and skin symptoms. The questions regard- ing symptoms were mostly identical to those of the German National Health Interview and Examination Survey 1997/98 (BGS 98).9 Moreover, specific occupational parameters were recorded. In addition to the preshift questionnaire, we used a postshift and a postweek questionnaire focusing the subjects’ complaints in the course of time (see Online Data Supplement).

Asthma was defined in patients with current asthma medica- tion use and asthma attacks during the past 12 months. Further- more, work-exacerbated asthma was defined in cases of chest tightness, wheezing, cough, and shortness of breath during the exposure to coffee dust. Chronic bronchitis was diagnosed if coughing and phlegm occurred for a minimum of 3 months per year and for not less than 2 successive years.10

Serologic Parameters

The blood of 59 coffee workers was sampled for IgE determi- nation. Total IgE and specific IgE to green coffee beans (com- prising Coffea arabica, Coffea canephora/robusta, and Coffea liberica) and to castor beans were measured by fluorescent enzyme immunoassay (UniCAP; Phadia; Freiburg, Germany). To investigate immunologic cross-reactivity between coffee and cas- tor beans, 100 �L of the castor bean IgE-positive serum were incubated with 50 �L of green coffee solution (4 mg Robusta coffee protein per milliliter) overnight at 4°C. UniCAP sub- sequently measured the IgE binding to castor bean allergens. Specific IgE levels above the threshold of 0.35 kilounits (kUA) per liter of serum (CAP class 1)11 were interpreted as sensitization according to the recommendation of the manu- facturer.

Lung Function Analysis

During the three investigations just described, lung function measurements were conducted in a sitting position by using a nose clip. Lung function measurements with a portable spirom- eter (MasterScope 4.5; Jaeger; Wuerzburg, Germany) were performed by a trained technician in the company offices. Forced expiratory spirograms were performed for the whole study

www.chestjournal.org CHEST / 136 / 2 / AUGUST, 2009 537

Downloaded From: http://journal.publications.chestnet.org/ by a California State University Northridge User on 04/24/2013

population (n � 60) according to the recommendations of the American Thoracic Society.12,13 For assessment, the reference values by Brändli et al14,15 were used. They assumed14 an obstructive ventilation pattern when FEV1/FVC percentage (FEV1/FVC%) was lower than the limit for normal.

A methacholine challenge using the Pari Provocation Test (Pari Company; Starnberg, Germany) with spirometric measurements was conducted during each investigation.16 The applied dose of methacholine inducing a 20% drop in FEV1 (PD20) was calcu- lated by linear interpolation or extrapolation up to 2,000 �g of methacholine.16 Bronchial hyperresponsiveness was diagnosed when PD20 was � 300 �g of methacholine (inhaled cumulative dose).17 Acute cross-shift changes in pulmonary function (� lung function parameter) were expressed for each subject as a per- centage of the value before exposure.16

Statistical Analysis

Data were analyzed using statistical software (SPSS for Windows, version 13.0; SPSS GmbH Software; Munich, Ger- many). Continuous data were expressed as mean � SD and in case of nonnormal distribution as median (minimum to max- imum). For the nonparametric group of metric/ordinal values, the Mann-Whitney test and the Kruskal-Wallis test were per- formed. The Pearson �2 test determined the frequency differ- ences between the two groups. The provided p values were two sided, and an � value of � 0.05 was regarded as statistically significant.

To analyze exposure-response relationships between exposure and obstructive ventilation patterns, the logistic regression with or without adjusting for the potential confounding variable smoking status (nonsmoker, former smoker, and smoker) was conducted. Adjusted odds ratios (ORs) and 95% confidence intervals (CIs) were estimated.

Results

Dust Exposure During the Handling of Coffee

Haulage Company: We performed dust measure- ments in the ambient air of three representative tilting stations and found a high inhalable coffee dust

concentration during unloading (� 10 mg inhalable dust per cubic millimeter of air) [Table 1].

Coffee Silo: The inhalative dust load during dis- charge was equal to the exposure of truck drivers at tilting stations. High dust concentrations occurred during the operation of silo systems. The cleaning staff performed activities in different locations. The duration was similar to that in the two other working areas studied. Therefore, the average dust load during coffee unloading and system operation in the silo was calculated to estimate the exposure of these employees (Table 1).

Decaffeinating Company: The inhalative exposure during coffee discharge mostly performed by chain trough conveyors was distinctly lower than that of truck drivers and silo workers at the tilting stations. In the working area of system operators, foremen, and cleaning personnel, only low dust concentrations were measurable on account of the shielded con- veyor systems (Table 1).

The measured concentrations of germs and molds ranged from 114 to 1,279 cfu/m3 and from 108 to 325 cfu/m3, respectively. Measurements of the total number of bacteria in the atmosphere exclusively showed Gram-positive normal germs occurring in the environment (especially Bacillus spp). Measure- ments of molds revealed Cladosporium spp, As- pergillus niger, and Penicillium spp. The total mold concentrations in reference areas were distinctly higher than on system tracks exposed to coffee dust.

Demographic Data and Clinical Symptoms

The demographic data on employees of the haul- age company, silo, and decaffeinating company were

Table 1—Measuring Results of Personal Coffee Dust Concentrations in the Air During Different Activities in Investigated Companies

Variables

Dust Concentration, mg/m3 Employees

Alveolar Inhalable Haulage Company

(n � 24) Coffee Silo

(n � 19) Decaffeinating Company

(n � 17)

Coffee discharge At tilting station 1.38 11.65 (10.1–15.7) 24 4 3* By chain trough conveyors 3.80 (2.6–4.0)

System operation Coffee silo 0.85 (0.3–1.35) 7.05 (3.7–10.3) 7 Decaffeinating company 1.60 (1.3–1.9) 7

Activities in several areas (cleaning staff, foremen)

Coffee silo 1.12† 9.35† 8 Decaffeinating company 0.70 (� 0.2–1.1) 7

Data are presented as the median (minimum-maximum) or No. *Twenty percent of the coffee discharge at tilting stations and 80% by chain trough conveyors. †Mean value of dust concentrations during coffee discharge and system operation in silo.

538 Original Research

Downloaded From: http://journal.publications.chestnet.org/ by a California State University Northridge User on 04/24/2013

not significantly different (Table 2). On the basis of these anamnestic data, we placed the employees according to the exposure median in categories with current and cumulative exposure to low and high dust levels (Table 2). Compared to employees with lower dust exposure, an increase of symptoms, lung functional impairments, or allergologic findings was not observed in employees with anamnestically higher current or cumulative dust exposure.

The total group reported that Arabica coffee trig- gers symptoms twice as often as Robusta coffee (18 times vs 9 times). In particular, the coffee of Brazil-

ian origin was reported to elicit complaints. The participants exhibited mainly erythematous and rhi- noconjunctival symptoms during the past 12 months (Fig 1). Employees with currently higher dust expo- sure (� 10 mg/m3) complained more frequently than those with currently lower dust exposure about ery- thematous (42.9% vs 15.6%, respectively; Pearson �2

test, p � 0.020), conjunctival (53.6% vs 25.0%, respec- tively; p � 0.023), and sneezing (35.7% vs 25.0%, respec- tively; p � 0.366) symptoms during the past 12 months.

According to the questionnaire, symptoms of the lower airways in the sense of asthmatic complaints

Table 2—Demographic Data and Level of Anamnestically Recorded Coffee Dust Exposure in the Investigated Companies

Variables Haulage Company

(n � 24) Coffee Silo

(n � 19) Decaffeinating Company

(n � 17) p Value*

Participation rate, % 82.8 100 100 Age, yr 40.4 (7.7) 39.6 (10.5) 44.5 (10.3) 0.202 Smoking status, No. (% of company)

Nonsmokers 2 (8.3) 7 (36.8) 2 (11.8) Former smokers 6 (25.0) 4 (21.1) 6 (35.3) Smokers 16 (66.7) 8 (42.1) 9 (52.9)

Pack-years, No. 28.4 (17.0) 17.6 (9.1) 21.9 (11.3) 0.116 Period without coffee exposure before Monday

morning examination, h 77 (58–727) 63 (38–233) 66 (63–744) � 0.001†

Coffee dust exposure (according to anamnestical data)

Current, h/d 3 (1–8) 6 (0.5–8) 5 (0–8) 0.086 Cumulative, yr 10.7 (0.5–32) 11.8 (0.5–35) 17.3 (4–34) 0.074

Data are presented as mean (SD) or median (minimum-maximum) unless otherwise indicated. *Kruskal-Wallis test. †Indicates significant findings.

Figure 1. Prevalence (%) of symptoms during the past 12 mo among employees with coffee dust exposure � 10 mg/m3 and � 10 mg/m3 inhalable dust (n � 32 and n � 28, respectively); work-related symptoms are shown as hatched columns.

www.chestjournal.org CHEST / 136 / 2 / AUGUST, 2009 539

Downloaded From: http://journal.publications.chestnet.org/ by a California State University Northridge User on 04/24/2013

occurred more rarely, however. Because none of the investigated employees took any antiasthmatic med- ication or had suffered from an asthma attack during the past 12 months, no participant was regarded as having asthma. Thus, based on our questionnaire, there was no evidence for preemployment asthma in the study sample. During the period of exposure to coffee dust, work-exacerbated asthma (increase in chest tightness, wheezing, cough, and shortness of breath) developed in two subjects. Both subjects had a currently higher exposure to inhalable coffee dust. Furthermore, these two subjects showed IgE anti- bodies to green coffee and castor beans but no obstructive ventilation pattern; one employee exhib- ited bronchial hyperresponsiveness.

The diagnostic criteria of chronic bronchitis were fulfilled in three employees. All of them were non- smokers and had a currently lower coffee dust exposure. These three subjects had no obstructive ventilation pattern (risk group 0 according to the Global Initiative for Chronic Obstructive Lung Dis- ease criteria of COPD).18 Work-related erythema and conjunctivitis during the past 12 months were more often observed in employees exposed to � 10 mg/m3 than to � 10 mg/m3 (erythematous symptoms, 22.2% vs 9.4%, respectively [p � 0.172]; conjunctivitis, 42.9% vs 12.5%, respectively [p � 0.008]).

Postshift and postweek increases of work-related symptoms were mainly found in workers with a coffee dust exposure � 10 mg/m3 (erythematous symptoms, 21.4% vs 0%, respectively [p � 0.006]; conjunctivitis, 25.0% vs 6.3%, respectively [p � 0.090]; sneezing, 17.9% vs 15.6%, respectively [p � 0.631]; chest tight- ness, 14.8% vs 3.1%, respectively [p � 0.140]; wheez- ing, 7.1% vs 0%, respectively [p � 0.165]; cough, 17.9% vs 9.4%, respectively [p � 0.335]). The same exposure-dependent increase of work-related symp- toms was observed in the postweek examination.

Lung Function Parameters

Due to the respiratory complaints of five employ- ees, their methacholine challenge testing was not possible at the end of the week. The lung function comparison of the smoker group (n � 33) with that of the total group (n � 60) did not reveal noticeable differences. Therefore, we evaluated the lung function findings inde- pendent of the employees’ smoking status.

With regard to lung function parameters of the total group, the average percentage deviations from reference mean values by Brändli et al14 of FVC, FEV1, and FEV1/FVC% were 94.5%, 93.6%, and 96.0%, respectively, indicating slight lung function impairments. Seven employees showed an obstruc- tive ventilation pattern (FEV1/FVC � FEV1/FVC% lower limit of normal) before the Monday shift (two

workers and five workers with a dust exposure � 10 mg/m3 and � 10 mg/m3); three were smokers, and four were former smokers. The pack-year history of these seven people did not differ from those of the other smokers/former smokers (showing normal lung function) of the total group (mean, 25.6 pack-years vs 23.5 pack-years; Mann-Whitney test, p � 0.841). No significant association between obstructive lung func- tion impairment and dust exposure level (� 10 mg/m3

vs � 10 mg/m3 air) was found. Among the seven employees with obstructive ventilation patterns, five were assigned to risk group 1 and two to risk group 2 according to the Global Initiative for Chronic Obstructive Lung Disease criteria of COPD.18

On Monday before the shift, lung function was independent of the exposure level of inhalable coffee dust (Table 3). Results of the preshift methacholine challenge on Monday showed that workers with coffee dust exposure � 10 mg/m3 had bronchial hyperresponsiveness significantly more often (25.0% vs 6.3%). Postshift and postweek measurements also revealed bronchial hyperresponsiveness more fre- quently in workers with higher work-related dust concentrations. Therefore, in the three investigations we calculated a significantly lower effective PD20 for employees with a dust concentration � 10 mg/m3. Based on the anamnestic data about the period of coffee dust exposure (in years), no association be- tween cumulative dust exposure and lung function impairment was found (Fig 2).

The prevalence of postweek bronchial hyperre- sponsiveness (n � 10) was associated with dust ex- posure level (OR, 5.47; 95% CI, 1.04 to 28.74). After adjusting for smoking status (nonsmoker, former smoker, and smoker), this relationship remained significant (OR, 5.31; 95% CI, 1.01 to 28.01), and the smoking status did not reach statistical significance (OR, 1.33; 95% CI, 0.42 to 4.26). Thus, no major effect of smoking on bronchial hyperresponsiveness in employees was observed in this study.

In the total group (n � 60), the cross-shift course of lung function values revealed an average change in FEV1, FEV1/FVC%, and FVC of 0.8%, 2.2%, and 0.7%, respectively. The lung function com- parison before the shift with that at the end of the week showed a slightly stronger lung function im- pairment in the postweek measurement ( 2.5%, 2.1%, and 0.4%, respectively).

In the course of the week, the subjects with preshift bronchial hyperresponsiveness demon- strated work-related symptoms significantly more often than the rest of the group (chest tightness, 22.2% vs 6.0%, respectively [p � 0.014]; wheezing, 12.5% vs 4.2%, respectively [p � 0.05]; cough, 44.4% vs 7.8%, respectively [p � 0.003]; erythematous symptoms, 37.5% vs 4.2%, respectively [p � 0.002]; conjunctivitis,

540 Original Research

Downloaded From: http://journal.publications.chestnet.org/ by a California State University Northridge User on 04/24/2013

25.0% vs 20.8%, respectively [p � 0.05]; sneezing, 50.0% vs 16.7%, respectively [p � 0.033]). The seven subjects with obstructive ventilation patterns did not notice work-related airway symptoms. This may be due to a good adaptation to their lung function limitation.

Allergologic Findings

IgE antibodies to green coffee and castor beans were detectable in three workers and six workers

with an inhalable coffee dust exposure � 10 mg/m3

(Table 4). The three employees with IgE antibodies to coffee beans also demonstrated IgE antibodies to castor beans. Also, four subjects with a coffee dust load � 10 mg/m3 revealed IgE antibodies to castor beans. Elevated total IgE (� 100 kUA/L) occurred 20 times in the total group independent of the level of the subjects’ average coffee dust load.

Two workers and six workers with IgE antibodies to green coffee and castor beans, respectively, had respiratory symptoms during the past 12 months. In total, no association between specific/total IgE and work-related symptoms during the past 12 months was found. UniCAP inhibition assays with coffee extracts did not suppress the IgE binding to castor beans.

Discussion

This study shows that high exposures to inhalable dust (� 10 mg/m3) occur during the handling of green coffee. We observed great differences be- tween the haulage company, coffee silo, and decaf- feinating company. Concentrations of the alveolar

Figure 2. Baseline lung function of all employees independent of the number of years of cumulative coffee dust exposure.

Table 3—Lung Function Findings of Examined Employees (n � 60) Depending on Inhalable Coffee Dust Exposure

Variables

Coffee Dust Exposure (Inhalable Fraction)

p Value� 10 mg/m3 (n � 32) � 10 mg/m3 (n � 28)

Preshift measurement (Monday) FEV1, L 3.9 (0.7) 4.0 (0.6) 0.876* FEV1/FVC% 75.5 (6.7) 75.8 (6.0) 0.801* FVC, L 4.9 (0.7) 5.3 (0.8) 0.073* FEV1, % predicted 95.4 (13.4) 91.4 (10.8) 0.110* FEV1/FVC%, % predicted 96.2 (7.3) 95.9 (7.8) 0.988* FVC, % predicted 93.9 (8.8) 95.2 (12.8) 0.778* Positive BHR, No. (% of exposure

group) 2 (6.3) 7 (25.0) 0.042†‡

PD20, mg 2 (0–2) 1.5 (0–2) 0.013*† Postshift measurement (Monday)

Positive BHR, No. (% of exposure group)

4 (12.5) 7 (25.0) 0.212‡

PD20, mg 2 (0.1–2) 1.2 (0–2) 0.029*† Postweek measurement (Friday)§

Positive BHR, No. (% of exposure group)

2 (7.1) 8 (29.6) 0.031†‡

PD20, mg 2 (0–2) 2 (0–2) 0.020*† Postshift vs preshift findings, %

�FEV1 1.4 ( 15.7–18.1) 1.1 ( 11.9–37.6) 0.133* �FEV1/FVC% 2.1 ( 17.4–9.1) 2.8 ( 21.3–18.7) 0.859* �FVC 1.9 ( 10.3–27.5) 0.3 ( 25.1–59.1) 0.830*

Postweek§ vs preshift findings, % �FEV1 3.9 ( 100.0–18.9) 1.4 ( 11.9–41.0) 0.033*† �FEV1/FVC% 3.5 ( 100.0–18.4) 0 ( 14.0–19.2) 0.039*† �FVC 2.1 ( 100.0–10.0) 2.2 ( 22.1–38.5) 0.218*

Data are presented as mean (SD) or median (minimum-maximum). BHR � bronchial hyperresponsiveness. *Mann-Whitney test. †Indicates significant findings. ‡Pearson �2 test. §Five employees did not undergo bronchial hyperresponsiveness test due to respiratory complaints.

www.chestjournal.org CHEST / 136 / 2 / AUGUST, 2009 541

Downloaded From: http://journal.publications.chestnet.org/ by a California State University Northridge User on 04/24/2013

dust fraction � 3 mg/m3 were not detectable, how- ever. High exposures to inhalable dust were espe- cially found during the discharge of green coffee at tilting stations. The containers with green coffee were unloaded without protective equipment in a relatively short period of time (30 to 45 min). The alternative coffee discharge via chain trough con- veyor required more time (45 to 60 min) but resulted in a distinctly lower dust load due to the uniform and optimal velocity. The described transshipment of green coffee represents the current standard. An optimization of the coffee unloading process, such as the use of chain trough conveyors, is generally recommended. Engineering controls and personal respiratory equipment might be effective preventive measures. So far, they have not been realized, however.

The microbial ambient air measurements during the handling of coffee did not reveal an extraordinary concentration of bacteria or molds. Therefore, a noticeable influence on lung function caused by microorganism is unlikely.

Green coffee elicits sensitizing as well as irritative effects. According to the literature, coffee dust ex- posures primarily result in asthmatic symptoms and rhinoconjunctival complaints.19 –25 The subjects of our study with an exposure � 10 mg/m3 air showed significantly more frequent erythematous and rhino- conjunctival symptoms in the first examination as well as in the cross-shift and cross-week course. Considering our measurements of higher concentra- tions of the inhalable dust fraction in some investi- gated areas, these coffee workers are generally at a higher risk for irritant-induced symptoms of differ- ent organs such as erythema and conjunctivitis.

The predominance of upper airway symptoms in this study could be due to the high exposure to inhalable but not to the alveolar dust fraction. Our

questionnaire revealed manifest symptoms of the lower airways during coffee dust exposure in only two subjects. Both had a currently higher exposure to inhalable coffee dust and IgE antibodies to green coffee and castor beans. One employee showed bronchial hyperresponsiveness consistent with the beginning of work-exacerbated allergic asthma. In addition, one-third of the examined employees of both subgroups (� 10 mg/m3 and � 10 mg/m3) had increased total IgE levels, indicating an immunologic stimulation.

Our study group reported work-related symptoms to occur twice as often by Arabica than by Robusta coffee. This indicates that the Arabica species, re- garded as a high-grade and pure coffee, constitutes a higher health hazard than the qualitatively lower Robusta coffee. To our knowledge, this study de- scribes work-related symptoms due to different cof- fee species for the first time.

Lung function parameters of examined workers were, on the average, below reference mean values. Jones et al26 had already described lower mean FEV1 values in men exposed to green coffee dust. This is surprising because the working population is ex- pected to have supranormal values. In this study, however, no long-term effect of coffee dust exposure on lung function was observed. All three investiga- tions showed bronchial hyperresponsiveness to be significantly related to the dust exposure level; this relationship remained significant after adjusting for smoking status. Thus a major effect of smoking on the employees’ bronchial hyperresponsiveness is un- likely.

The reason why the spirometry improved slightly in the high exposure group at the end of the week cannot be explained definitively. This may indicate a work-organizational effect because in contrast to the employees with coffee dust exposure � 10 mg/m3,

Table 4 —Allergologic Findings of Examined Employees (n � 59) Depending on Fraction of Inhalable Coffee Dust Exposure*

Variables

Coffee Dust � 10 mg/m3 (n � 32) Coffee Dust � 10 mg/m3 (n � 27)

Exposure Group, No. (%) Respiratory Symptoms, No.† Exposure Group No. (%) Respiratory Symptoms, No.†

Total IgE � 100 kUA/L 21 (65.6) 9 18 (66.7) 8 � 100 kUA/L 11 (34.4) 5 9 (33.3) 5

Specific IgE Green coffee beans

CAP class � 1 32 (100) 14 24 (88.9) 11 CAP class � 1 0 0 3 (11.1) 2

Castor beans CAP class � 1 28 (87.5) 11 21 (77.8) 10 CAP class � 1 4 (12.5) 3 6 (22.2) 3

*One subject rejected blood sampling. †Sneezing, chest tightness, wheezing, and/or shortness of breath during the past 12 mo.

542 Original Research

Downloaded From: http://journal.publications.chestnet.org/ by a California State University Northridge User on 04/24/2013

those workers with high exposure (especially the truck drivers unloading the coffee at the tilting stations) had shorter assignments in these coffee dust areas on Friday and thereby the opportunity to compensate for their overtime worked.

Due to respiratory complaints, the methacholine challenge test with five subjects could not be per- formed on Friday. This indicates that the dust load had impaired their lung function so that the risk of bronchial hyperresponsiveness on Friday might have been underestimated. Zuskin et al27 showed that the inhalation of green coffee extracts can lead to an obstructive ventilation pattern in healthy subjects.

According to several, mostly older studies, the frequency of sensitization to green coffee beans is between 20% and 25% in coffee workers.4,5,27,28 Osterman et al6 found a level � 50%. In our study, however, specific IgE antibodies to green coffee beans were only detectable in three workers (5%). All of them were exposed to � 10 mg/m3. In recent years, coffee producers have started using new bur- lap or plastic bags and/or specific containers that considerably reduce the risk of castor bean contam- ination of green coffee beans.20 Furthermore, the use of dust exhaust systems during the coffee un- loading process in some of the investigated tilting stations improved the hygiene. This may explain the relatively low prevalence of sensitization to coffee dust in our study in comparison with former inves- tigations.

Our examination showed 10 subjects (17%) had IgE antibodies to castor beans. Four of them had a low coffee dust exposure (� 10 mg/m3). Inhibition test results of antibody-allergen reactions indicate that the identified castor bean sensitization cannot be explained by a cross-reactivity with coffee. The causative source of the castor bean allergens in imported green coffee has not yet been finally explained (castor beans were never transported or handled by the examined employees).6

A limitation of this study is its relatively small sample size, so it cannot be excluded that some findings might be accidental. The participation rate of the coffee dust exposed workers, however, was extraordinarily high and constituted a representative sample of each coffee company. Furthermore, in spite of the low number of examined employees, we observed statistically significant differences between workers with currently higher and lower coffee dust exposure. These differences not only covered bron- chial hyperresponsiveness but were consistently also true of work-exacerbated clinical symptoms among bronchial hyperresponsive subjects. Thus, the exposure-related bronchial hyperresponsiveness was associated with a higher frequency of work-related respiratory symptoms.

In summary, our results demonstrate a signifi- cantly higher prevalence of erythematous and con- junctival symptoms as well as bronchial hyperrespon- siveness among employees with a higher coffee dust exposure than in those with a lower one (� 10 mg/m3 vs � 10 mg/m3).

In addition to the known sensitizing potential, irritative effects of green coffee dust appear to be important pathogenetically. Thus, coffee bean dust still represents a health hazard to employees, so the improvement of hygienic conditions (ie, further op- timization of exhaust systems) is recommended.

Acknowledgment

Author contributions: Dr. Oldenburg was responsible for the study design, the statistical analysis as well as for the on-spot investigation. Dr. Bittner was engaged in the allergological investigations, including their interpretations. Dr. Baur was responsible for the study design and the interpretation of the employees’ lung function. Financial/nonfinancial disclosures: The authors have no sig- nificant conflicts of interest with any companies/organizations whose products or services may be discussed in this article. Other contributions: The authors wish to thank the volunteers and the management of the investigated coffee subcontracting firms and the haulage company. We also thank L. Barbinova, D. Johannsen, E. Nern, and B. Poschadel for the performed medical examinations (lung function and blood sampling). The authors also express their gratitude to the Pari Company in Starnberg, Germany, for making three Pari Provocation Test II(R) available for our cross-shift survey.

References 1 Figley KD, Rawling FF. Castor bean: an industrial hazard as

a contaminant of green coffee dust and used burlap bags. J Allergy 1950; 21:545–553

2 Lemière C, Malo JL, McCants M, et al. Occupational asthma caused by roasted coffee: immunologic evidence that roasted coffee contains the same antigens as green coffee, but at a lower concentration. J Allergy Clin Immunol 1996; 98:464 – 466

3 Patussi V, De Zotti R, Riva G, et al. Allergic manifestations due to castor beans: an undue risk for the dock workers handling green coffee beans. Med Lav 1990; 81:301–307

4 Romano C, Sulotto F, Piolatto G, et al. Factors related to the development of sensitization to green coffee and castor bean allergens among coffee workers. Clin Exp Allergy 1995; 25:643– 650

5 De Zotti R, Patussi V, Fiorito A, et al. Sensitization to green coffee bean (GCB) and castor bean (CB) allergens among dock workers. Int Arch Occup Environ Health 1988; 61:7–12

6 Osterman K, Zetterstrom O, Johansson SG. Coffee worker’s allergy. Allergy 1982; 37:313–322

7 Institute for Occupational Health and Safety. Project-No. BIA 1061: Personal dust sampler system for a flow rate of 10 l/min. September 30, 1999. Available at: http://www.hvbg.de/ e/bia/pro/pro1/pr1061.html. Accessed April 3, 2009

8 European Standard EN 481. Workplace atmospheres: size fraction definitions for measurement of airbone particles.

www.chestjournal.org CHEST / 136 / 2 / AUGUST, 2009 543

Downloaded From: http://journal.publications.chestnet.org/ by a California State University Northridge User on 04/24/2013

Available at: http://legacy.library.ucsf.edu:8080/i/e/m/iem52d00/ siem52d00.pdf. Accessed July 10, 2009

9 Thefeld W, Stolzenberg H, Bellach B-M. Bundesgesundheits- survey. Response, Zusammensetzung der Teilnehmer und Non- Responder-Analyse [German National Health Interview and Examination Survey: composition of participants, and analysis of nonrespondents]. Gesundheitswesen 1999; 2:57– 61

10 World Health Organization Report of an expert committee: definition and diagnosis of pulmonary disease with special reference to chronic bronchitis and emphysema. WHO Techn Rep Ser 1961; 213: 14 –19

11 Ewan PW, Coote D. Evaluation of a capsulated hydrophilic carrier polymer (the ImmunoCAP) for measurement of spe- cific IgE antibodies. Allergy 1990; 45:22–29

12 Hankinson JL, Viola JO. Dynamic BPTS correction factors for spirometric data. J Appl Physiol 1983; 44:1354 –1360

13 American Thoracic Society: standardization of spirometry. Am J Respir Crit Care Med 1995; 152:1107–1136

14 Brändli O, Schindler C, Leuenberger PH, et al. Estimated equations for 5th percentiles of lung function variables. Thorax 2000; 55:173–174

15 Brändli O, Schindler C, Künzli N, et al. Lung function in healthy never smoking adults: reference values and lower limits of normal of a Swiss population. Thorax 1996; 51:277–283

16 Oldenburg M, Latza U, Baur X. Exposure-response relation- ship between endotoxin exposure and lung function impair- ment in cotton textile workers. Int Arch Occup Environ Health 2007; 80:388 –395

17 Baur X, Huber H, Degens PO, et al. Relation between occupational asthma case history, bronchial methacholine challenge, and specific challenge test in patients with sus- pected occupational asthma. Am J Ind Med 1998; 33:114 –122

18 Pauwels RA, Buist AS, Calverley PM, et al. Global strategy for the diagnosis, management, and prevention of chronic ob- structive pulmonary disease: NHLBI/WHO Global Initiative for Chronic Obstructive Lung Disease (GOLD) Workshop summary. Am J Respir Crit Care Med 2001; 163:1256 –1276

19 Larese F, Fiorito A, Casasola F, et al. Sensitization to green coffee beans and work-related allergic symptoms in coffee workers. Am J Ind Med 1998; 34:623– 627

20 Thomas KE, Trigg CJ, Baxter PJ, et al. Factors relating to the development of respiratory symptoms in coffee process work- ers. Br J Ind Med 1991; 48:314 –322

21 Osterman K, Johannsson SGO, Zetterström O. Diagnostic tests in allergy to green coffee. Allergy 1985; 40:336 –343

22 Zuskin E, Valic F, Skuric Z. Respiratory function in coffee workers. Br J Ind Med 1979; 36:117–122

23 Karr RM, Lehrer SB, Butcher BT, et al. Coffee worker’s asthma: a clinical appraisal using the radioallergosorbent test. J Allergy Clin Immunol 1978; 62:143–148

24 Layton LL, Greene FC, Panzani R. Allergy to green coffee. Allergy Clin Immunol 1965; 36:84 –91

25 Zuskin E, Kanceljak B, Skuric Z, et al. Bronchial reactivity in green coffee exposure. Br J Ind Med 1985; 42:415– 420

26 Jones RN, Hughes JM, Lehrer SB, et al. Lung function consequences of exposure and hypersensitivity in workers who process green coffee beans. Am Rev Respir Dis 1982; 125:199 –202

27 Zuskin E, Valic F, Kanceljak B. Immunological and respira- tory changes in coffee workers. Thorax 1981; 36:9 –13

28 Lehrer SB, Karr RM, Salvaggio JE. Analysis of green coffee bean and castor bean allergens using RAST inhibition. Clin Allergy 1981; 11:357–366

544 Original Research

Downloaded From: http://journal.publications.chestnet.org/ by a California State University Northridge User on 04/24/2013