Dust exposure and respiratory health among Tanzanian coffee factory workers
Doctoral thesis, Peer reviewed
MetadataShow full item record
Introduction: Exposure to organic dust may cause detrimental effects to the respiratory system of exposed workers. Organic dust is commonly contaminated with microbes and their derivatives such as bacteria and endotoxin, fungi, moulds and beta glucan. Few studies on exposure and health effects have been performed in primary coffee factories. The studies showed that processes in primary coffee factories cause emission of high dust levels. Work in coffee factories has been associated with respiratory health impairment. Coffee beans are of two main types; Arabica and Robusta. Before coffee is brought to the factory it is processed at the farm. At the farms the harvested Robusta coffee cherries are mostly dried under the sun (called dry pre-processing) while Arabica coffee cherries are depulped using water (called wet pre-process) and then dried as parchment coffee. At the end of harvest season, remaining Arabica coffee cherries are dried without being depulped (dry processed). The processes in primary coffee factories involves the pre-cleaning of coffee beans, the mechanical removal of the mesocarp and endocarp layers to get green coffee beans (GCB), the grading of the GCB, mixing them to produce a homogenous mixture by a process known as bulking and then packing in 60 kg bags. Loading and unloading of unprocessed coffee (parchment or coffee cherries) is done manually. For some processes, the machines are fed manually by production workers. Objective: The aim of this research was to determine dust exposure levels and to assess respiratory health of the production workers in Robusta and Arabica primary coffee factories. Materials and Methods: The research was conducted in Tanzania (in the Kilimanjaro and Kagera regions) in four primary coffee factories (factories; A, B, C and D) in three studies (in 2008 2009, and 2010). In the first study (2008) personal total dust (n=44) was sampled at a rate of 2 L/min from the breathing zone of the worker using side kick Casella pumps connected to closed-faced 25 mm conductive cassettes fitted with cellulose acetate filters. The samples were analysed gravimetrically. In addition, five samples were taken on glass fibre filters as pilot samples for analysis of endotoxin. We also assessed respiratory symptoms using an American Thoracic Society (ATS) standardized questionnaire among the production workers (n= 79) in comparison with a control group from a beverage factory (n=71). In the second and third studies personal total dust (n= 149) was sampled by same methods as in 2008, using closed-faced 37 mm plastic cassettes which were fitted with polycarbonate filters. Dust samples were gravimetrically measured and further analysed for endotoxin. Airway inflammation was assessed using NIOX MINO device in both studies. In the third study, lung function and respiratory symptoms were assessed among production workers (n=138) in four primary coffee factories and two control factories (n=120). A portable spirometer was used for lung function testing. Results: Personal total dust levels ranged from 0.25 to 36 mg/m³; geometric mean GM = 2.50mg/m³. Seventeen per cent of the samples were above the occupational exposure limit. There was a high correlation between total dust and endotoxin (r = 0.62). Endotoxin levels were higher when processing dry pre-processed coffee (mainly Robusta GM=10,800 EU/m³) than when processing wet pre-processed Arabica (GM=1,350 EU/m³). All endotoxin samples exceeded the health-based recommended value of 90 EU/m³. Using a mixed model analysis dry pre-processing was shown to increase the total dust and endotoxin levels by a factor of 2.5 and 7.2, respectively. Chronic respiratory symptoms were higher among coffee workers than controls. Having at least one asthma symptom was higher among the coffee workers compared to controls with odds ratio (OR) of 4.3 (95% CI; 1.9 – 9.9). Robusta coffee workers had higher prevalence of asthma symptoms (38%) than Arabica coffee workers (12%) (OR; 3.5, 95% CI; 1.4 – 9.0). Lung function parameters were not significantly different between coffee workers and controls. Nevertheless, in a linear regression model, controlling for age, height and type of coffee, there was a decrease in FEV1 and FEV1/FVC ratio related to an increase in cumulative total dust and endotoxin. In second study (2009) coffee workers had higher FENO levels than the controls, however, this was not found in the subsequent year 2010. Conclusion: This study revealed that work in coffee factory is associated with high dust and endotoxin exposure which may be associated with impairment of respiratory health. Processing dry pre-processed coffee increases the exposure levels significantly. Reduction of dust exposure is recommended.
Has partsPaper I: Sakwari G, Bråtveit M, Mamuya SHD & Moen BE. (2011) Dust exposure and chronic respiratory symptoms among coffee curing workers in Kilimanjaro: a cross sectional study. BMC Pulm Med.;11:54. The article is available at: http://hdl.handle.net/1956/7451
Paper II: Moen BE, Sakwari G, Mamuya SH, Kayumba AV, Larsson L, Pehrson C, Mashalla YJ & Bråtveit M. (2012) Respiratory Inflammation Among Workers Exposed to Airborne Dust With Endotoxins in a Coffee Curing Factory. J Occup Environ Med.;54(7):847-50. The article is not available in BORA due to publisher restrictions. The published version is available at: http://dx.doi.org/10.1097/JOM.0b013e318250ca24
Paper III: Sakwari G, Mamuya SHD, Bråtveit M, Larsson L, Pehrson C & Moen BE. (2012) Personal Exposure to Dust and Endotoxin in Robusta and Arabica Coffee Processing Factories in Tanzania. Ann Occup Hyg.;57(2):173-83. The article is available at: http://hdl.handle.net/1956/7452
Paper IV: Sakwari G, Mamuya SHD, Bråtveit M, & Moen BE. (2013) Respiratory symptoms, exhaled nitric oxide, and lung function among workers in Tanzanian coffee factories. J Occup Environ Med.;55(5):544-51. The article is not available in BORA due to publisher restrictions. The published version is available at: http://dx.doi.org/10.1097/JOM.0b013e318285f453