Jeroen Douwes

Massey University, Palmerston North City, Manawatu-Wanganui, New Zealand

Are you Jeroen Douwes?

Claim your profile

Publications (176)568.1 Total impact

  • The Journal of allergy and clinical immunology 08/2015; DOI:10.1016/j.jaci.2015.06.039 · 11.25 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: BACKGROUND: International collaborative cohorts the NINFEA and the ELF studies are mother-child cohorts that use the internet for recruitment and follow-up of their members. The cohorts investigated the association of early life exposures and a wide range of non-communicable diseases. OBJECTIVE: The objective is to report the research methodology, with emphasis on the advantages and limitations offered by an Internet-based design. These studies were conducted in Turin, Italy and Wellington, New Zealand. METHODS: The cohorts utilized various online/offline methods to recruit participants. Pregnant women who became aware volunteered, completed an online questionnaire, thus obtaining baseline information. RESULTS: The NINFEA study has recruited 7003 pregnant women, while the ELF study has recruited 2197 women. The cohorts targeted the whole country, utilizing a range of support processes to reduce the attrition rate of the participants. For the NINFEA and ELF cohorts, online participants were predominantly older (35% and 28.9%, respectively), highly educated (55.6% and 84.9%, respectively), and were in their final trimester of pregnancy (48.5% and 53.6%, respectively). CONCLUSIONS: Internet-based cohort epidemiological studies are feasible, however, it is clear that participants are self-selective samples, as is the case for many birth cohorts. Internet-based cohort studies are potentially cost-effective and novel methodology for conducting long-term epidemiology research. However, from our experience, participants tend to be self-selective. In marked time, if the cohorts are to form part of a larger research program they require further use and exploration to address biases and overcome limitations.
    05/2015; 4(2):e71. DOI:10.2196/resprot.3873.
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: In New Zealand, the burden of childhood obesity is greatest in Māori and Pacific children. In 687 infants from an internet-based birth cohort in New Zealand, we investigated ethnic differences in early life risk factors for later obesity, the degree to which these were explained by sociodemographic factors, and the extent to which ethnic differences in weight at age 3 months were explained by measured risk factors. The risk of having an obese mother was double in Māori and Pacific infants compared with NZ European infants (prevalence 24% and 14%, respectively; OR 2.23, 95% CI 1.23 to 4.04). Māori and Pacific infants had higher weights in the first week of life and at 3 months (mean difference 0.19 kg, 95% CI 0.01 to 0.38), and their mothers had higher scores on a 'snacks' dietary pattern and lower scores on 'healthy' and 'sweet' dietary patterns. These inequalities were not explained by maternal education, maternal age or area-based deprivation. No ethnic differences were observed for maternal pre-pregnancy physical activity, hypertension or diabetes in pregnancy, exclusive breastfeeding or early introduction of solid foods. Ethnic inequalities in infant weight at 3 months were not explained by sociodemographic variables, maternal pre-pregnancy body mass index or dietary pattern scores or by other measured risk factors. This study shows excess prevalence of early life risk factors for obesity in Māori and Pacific infants in New Zealand and suggests an urgent need for early interventions for these groups. Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to
    Journal of epidemiology and community health 03/2015; 69(6). DOI:10.1136/jech-2014-204464 · 3.29 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Recently the International Agency for Research on Cancer (IARC) Programme for the Evaluation of Carcinogenic Risks to Humans has been criticized for several of its evaluations, and also the approach used to perform these evaluations. Some critics have claimed that IARC Working Groups' failures to recognize study weaknesses and biases of Working Group members have led to inappropriate classification of a number of agents as carcinogenic to humans. The authors of this paper are scientists from various disciplines relevant to the identification and hazard evaluation of human carcinogens. We have examined here criticisms of the IARC classification process to determine the validity of these concerns. We review the history of IARC evaluations and describe how the IARC evaluations are performed. We conclude that these recent criticisms are unconvincing. The procedures employed by IARC to assemble Working Groups of scientists from the various discipline and the techniques followed to review the literature and perform hazard assessment of various agents provide a balanced evaluation and an appropriate indication of the weight of the evidence. Some disagreement by individual scientists to some evaluations is not evidence of process failure. The review process has been modified over time and will undoubtedly be altered in the future to improve the process. Any process can in theory be improved, and we would support continued review and improvement of the IARC processes. This does not mean, however, that the current procedures are flawed. The IARC Monographs have made, and continue to make, major contributions to the scientific underpinning for societal actions to improve the public's health.
    Environmental Health Perspectives 02/2015; DOI:10.1289/ehp.1409149 · 7.98 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Internet-based systems for epidemiological studies have advantages over traditional approaches as they can potentially recruit and monitor a wider range of individuals in a relatively inexpensive fashion. We studied the association between communication strategies used for recruitment (offline, online, face-to-face) and follow-up participation in nine Internet-based cohorts: the Influenzanet network of platforms for influenza surveillance which includes seven cohorts in seven different European countries, the Italian birth cohort Ninfea and the New Zealand birth cohort ELF. Follow-up participation varied from 43% to 89% depending on the cohort. Although there were heterogeneities among studies, participants who became aware of the study through an online communication campaign compared with those through traditional offline media seemed to have a lower follow-up participation in 8 out of 9 cohorts. There were no clear differences in participation between participants enrolled face-to-face and those enrolled through other offline strategies. An Internet-based campaign for Internet-based epidemiological studies seems to be less effective than an offline one in enrolling volunteers who keep participating in follow-up questionnaires. This suggests that even for Internet-based epidemiological studies an offline enrollment campaign would be helpful in order to achieve a higher participation proportion and limit the cohort attrition.
    PLoS ONE 12/2014; 9(12):e114925. DOI:10.1371/journal.pone.0114925 · 3.23 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Background Job insecurity has been identified as a risk factor for adverse health outcomes. Perceptions of job insecurity steeply increased during Europe's recent economic downturn, which commenced in 2008. The current study assessed whether job insecurity was associated with incident asthma in Germany during this period. Methods We used prospective data from the German Socio-Economic Panel for the period 2009–2011 (follow-up rate=77.5%, n=7031). Job insecurity was defined by respondents’ ratings of the probability of losing their job within the next 2 years and asthma as self-reports of physician-diagnosed asthma. Associations between job insecurity in 2009 (continuous z-scores or categorised variables) and incident asthma by 2011 were assessed using multivariable Poisson regression. Results The risk of asthma increased significantly by 24% with every one SD increase of the job insecurity variable. In dichotomised analyses, a probability of job loss of ≥50% (vs <50%) was associated with a 61% excess risk of asthma. A trichotomous categorisation of job insecurity confirmed this finding. Conclusions This study has shown, for the first time, that perceived job insecurity may increase the risk of new onset asthma. Further prospective studies may examine the generalisability of our findings and determine the underlying mechanisms.
    Journal of Epidemiology &amp Community Health 09/2014; 68(12). DOI:10.1136/jech-2014-204274 · 3.29 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Significant excess risks of lung cancer and haematologic neoplasms have been observed in slaughterhouse workers in eight New Zealand studies, and numerous studies conducted elsewhere. No specific causal agents have been identified, although a biological aetiology is suggested as the risk is highest in those areas where workers are exposed to live animals or to biological material containing animal urine, faeces or blood. This study aimed to assess the airborne bacterial microflora in the slaughterhouse environment in order to develop exposure categories for reanalysis of a meat workers' cohort.
    Occupational and Environmental Medicine 06/2014; 71 Suppl 1:A21. DOI:10.1136/oemed-2014-102362.65 · 3.23 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: To provide a detailed assessment of the extent and industrial distribution of workplace carcinogens in New Zealand, and to identify key industries and key carcinogens for which intervention would result in marked reductions in occupational cancer.
    Occupational and Environmental Medicine 06/2014; 71 Suppl 1:A47. DOI:10.1136/oemed-2014-102362.147 · 3.23 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: AIM:To estimate average infant daily intake of chlorinated persistent organic pollutants (POPs) through the consumption of breast milk in New Zealand.
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Published studies have shown that workers in animal slaughterhouses are at a higher risk of lung cancers as compared to the general population. No specific causal agents have been identified, and exposures to several chemicals have been examined and found to be unrelated. Evidence suggests a biological aetiology as the risk is highest for workers who are exposed to live animals or to biological material containing animal faeces, urine or blood. To investigate possible biological exposures in animal slaughterhouses, we used a metagenomic approach to characterise the profile of organisms present within an aerosol sample. An assessment of aerosol exposures for individual workers was achieved by the collection of personal samples that represent the inhalable fraction of dust/bioaerosol in workplace air in both cattle and sheep slaughterhouses. Two sets of nine personal aerosol samples were pooled for the cattle processing and sheep processing areas respectively, with a total of 332,677,346 sequence reads and 250,144,492 sequence reads of 85 bp in length produced for each. Eukaryotic genome sequence was found in both sampling locations, and bovine, ovine and human sequences were common. Sequences from WU polyomavirus and human papillomavirus 120 were detected in the metagenomic dataset from the cattle processing area, and these sequences were confirmed as being present in the original personal aerosol samples. This study presents the first metagenomic description of personal aerosol exposure and this methodology could be applied to a variety of environments. Also, the detection of two candidate viruses warrants further investigation in the setting of occupational exposures in animal slaughterhouses.
    PLoS ONE 08/2013; 8(8):e72226. DOI:10.1371/journal.pone.0072226 · 3.23 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Polybrominated diphenyl ethers (PBDEs) are present in many consumer goods. There is evidence that PBDEs are toxic to humans, particular young children. The purpose of this study was to assess indoor dust as an exposure source for PBDEs. Concentrations of 16 PBDEs were determined in dust samples from 33 households in New Zealand, and in breast milk samples from 33 mothers living in these households. Associations between dust and breast milk PBDE concentrations were assessed, and children's PBDE intake from breast milk and dust estimated. Influences of household and demographic factors on PBDE concentrations in dust were investigated. Indoor dust concentrations ranged from 0.1ng/g for BDE17 to 2500ng/g for BDE209. Breast milk concentrations were positively correlated (p<0.05) with mattress dust concentrations for BDE47, BDE153, BDE154, and BDE209 and with floor dust for BDE47, BDE183, BDE206, and BDE209. The correlation for BDE209 between dust and breast milk is a novel finding. PBDE concentrations in floor dust were lower from households with new carpets. The estimated children's daily intake of PBDEs from dust and breast milk was below U.S. EPA Reference Dose values. The study shows that dust is an important human exposure source for common PBDE formulations in New Zealand.
    Environment international 07/2013; 59C:255-261. DOI:10.1016/j.envint.2013.06.020 · 5.66 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Background and objectiveIncreased sputum neutrophilia has been observed in asthma, but also during normal ageing in asthmatics and non-asthmatics. It remains unclear what constitutes ‘normal’ neutrophil levels in different age groups. Methods We assessed the relationship between age and airway neutrophils of 194 asthmatics and 243 non-asthmatics (age range: 6–80 years). Regression analyses were used to assess this relationship adjusted for confounders including asthma status, atopy, gender, smoking and current use of inhaled corticosteroids (ICS). Age-corrected reference values for different age groups were determined using the 95th percentile of non-asthmatic participants. ResultsAge was positively associated with sputum neutrophils in both asthmatic and non-asthmatic adults (0.46% neutrophil increase/year (95% confidence interval (CI) 0.18, 0.73) and 0.44%/year (0.25, 0.64, respectively), but no association was found in the <20-year age category. Individuals with high sputum neutrophil counts (>95th percentile of non-asthmatic counts for any given age group) were significantly more likely to be asthmatic (odds ratio = 2.5; 95% CI: 1.3, 5.0), with the greatest effect observed in the older age group. Other factors that independently associated with increased sputum neutrophil levels included atopy in non-asthmatic adults, male gender and current use of ICS in asthmatic adults. Age-specific reference values for neutrophil percentage were under 20 years-76%, 20–40 years-62%, 40–60 years-63% and over 60 years-67%. Conclusions Airway neutrophilia is related to age in adults, with a neutrophilic asthma phenotype present in older adults. The use of appropriate age-specific reference values is recommended for future studies aimed at elucidating the role of neutrophils in asthma.
    Respirology 07/2013; 18(5). DOI:10.1111/resp.12079 · 3.50 Impact Factor
  • Neil Pearce · Jeroen Douwes
    [Show abstract] [Hide abstract]
    ABSTRACT: Epidemiology is currently undergoing changes in its underlying philosophy and approach, as a result of the rapid global changes which are transforming the world in which epidemiologists live and work. This necessitates a multidisciplinary "population approach" involving "multilevel thinking" about the determinants of disease. These issues are of relevance to the interface between human and animal epidemiology, which has received considerable attention in recent years, particularly as a result of the arrival of H1N1 influenza, and the increasingly obvious need for coordinated systems of surveillance for human and animal infectious diseases. However, the need for coordination between human and veterinary epidemiology is broader than that, and there is no need to restrict the "one world one health" concept to communicable disease. In the current paper we will therefore consider the interface between human and animal health for the study of non-communicable disease, particularly those involving occupational and environmental risk factors. These issues are illustrated with two examples: one involving environmental health (asthma); and one involving occupational health (cancer). We will also discuss the potential to use animal health data as indicators for human environmental health risks.
    Preventive Veterinary Medicine 06/2013; 111(3-4). DOI:10.1016/j.prevetmed.2013.05.010 · 2.51 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Breast milk samples of 39 first time mothers aged 20-30 were collected in 2007-2010 from rural and urban areas of New Zealand, following the fourth World Health Organization coordinated survey protocol. Samples were individually analysed for persistent organic pollutants (POPs) including dioxins and furans (PCDD/Fs), polychlorinated biphenyls (PCBs), organochlorine pesticides (OCPs) and polybrominated diphenyl ethers (PBDEs). The lipid adjusted concentrations of PCDD/Fs (mean toxic equivalent (TEQ): 3.54pg/g) and PCBs (mean TEQ 1.29pg/g) were low in comparison to those reported for other countries, and concentrations of dieldrin (10ng/g) and p,p'-DDE (379ng/g) and PBDEs were in the mid-range. Breast milk concentrations of PCDD/F-TEQ, PCB-TEQ, dieldrin and p,p'-DDE were significantly higher in rural compared to urban areas (+23%, 33%, 59%, and 44% respectively), while concentrations of several PBDEs and lindane were higher in urban areas. Concentrations of PCDD/Fs, PCBs and OCPs, but not PBDEs, increased with age, and higher body mass index was associated with lower concentrations of PCBs. Despite New Zealand's low body burdens of many chlorinated POPs in comparison to other countries, breast milk concentrations continued to decrease over time, with a decrease by half over the last 10years for PCDD/F-TEQ (-40%), PCB-TEQ (-54%) and OCPs -34 to -90%), indicating that regulatory measures continue to have beneficial effects. Continued monitoring is needed particularly for the brominated POPs for which little New Zealand specific data is available.
    Science of The Total Environment 05/2013; 458-460C:399-407. DOI:10.1016/j.scitotenv.2013.04.055 · 4.10 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Background. Asthma and depression are important public-health concerns worldwide. While some epidemiologic studies have shown asthma and wheezing to be associated with depression and anxiety, the patterns are unclear at the multinational level due to the lack of cross-study comparability. Our study examined the associations of self-reported asthma diagnosis and current wheezing with self-reported depression diagnosis and 30-day anxiety using an international survey. Methods. Using the 2002 World Health Survey, a standardized international survey conducted by the WHO, we estimated the associations between diagnosed asthma and current wheezing with diagnosed depression and 30-day anxiety via multiple logistic regressions for 54 countries worldwide. Results. Diagnosed depression and 30-day anxiety were associated with diagnosed asthma in 65% and 40% of the countries, respectively. Diagnosed depression and 30-day anxiety were associated with current wheezing in 83% and 82% of the countries, respectively. Conclusions. The association between asthma and depression was generally seen at the global level. These results indicated the importance of addressing the asthma-depression comorbidity as public-health and clinical management priorities, in order to improve the overall health of the countries.
    03/2013; 2013:929028. DOI:10.1155/2013/929028
  • [Show abstract] [Hide abstract]
    ABSTRACT: BACKGROUND: Airway inflammation is commonly assessed by sputum induction followed by a differential cell count (DCC) using light microscopy. This method is prone to intercounter variability and poor reproducibility. We aimed to develop a more objective method using flow cytometry (FCM). METHODS: Fifty-six sputum inductions were conducted in 41 adults (23 asthmatics). Sputum was processed, a cytospin prepared for DCC, and the remainder immunolabeled for FCM using CD45, CD14, and CD16-specific antibodies to distinguish major leukocyte populations. Aliquots of 15 samples were frozen at -80°C to assess the effects of cryostorage. DCC and FCM were compared, and viability of individual cell populations was determined by FCM. RESULTS: FCM and DCC, and fresh and frozen samples, were significantly correlated, R = 0.54-0.87; all P < 0.0001, and R = 0.57 to 1; P < 0.005, respectively. There was a significant neutrophil loss after cryostorage (from median 30.5-17.4% of total leukocytes; P < 0.0001). Cell viability was higher for lymphocytes compared to granulocytes or macrophages (P < 0.001). With the exception of the expected higher levels of eosinophils (P < 0.005), no significant difference in cell differentials or viability was observed between asthmatics and nonasthmatics using either DCC or FCM. CONCLUSIONS: FCM is a suitable means of assessing leukocyte populations in induced sputum. Sample storage at -80°C prior to FCM is feasible, but may be detrimental to neutrophils, although good correlations were still observed between fresh and frozen samples. Large differences in viability were found between individual cell populations suggesting that viability dye use may be necessary. © 2013 International Clinical Cytometry Society.
    Cytometry Part B Clinical Cytometry 03/2013; 84B(2). DOI:10.1002/cyto.b.21069 · 2.28 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Environmental levels of β-(1,3)-glucan, an inflammatory fungal cell wall component, have been suggested to be related to respiratory symptoms. However there is currently little data comparing β-(1,3)-glucan detection methods and/or results obtained in different laboratories. The aim of this study was to compare levels of β-(1,3)-glucans detected in household dust samples (n=40) using different extraction/detection methods (Limulus amebocyte assay (LAL), inhibition enzyme immunoassay (EIA) and sandwich EIA) in five different laboratories. Dust sample aliquots were sent to participating centres, extracted and analysed for β-(1,3)-glucan according to standard in-house procedures. Significant differences in the levels of β-(1,3)-glucan were observed between all laboratories geometric mean levels ranging from 15.42g/g to 47542g/g dust; p<0.0001) with the exception of those using a similar LAL method. The inhibition EIA used in laboratory D produced mean β-(1,3)-glucan measurements 80-100 times higher than the LAL assays, 4 times higher than the sandwich EIA in the same lab, 17.6 times those obtained with the EIA in lab E and 363 times those obtained in the EIA in laboratory C. Pearson’s correlations generally showed significant associations between methods and laboratories, particularly those using similar methodology (R ranging from 0.5 to 0.8; p<0.001), although some poor and even inverse correlations were observed. Bland-Altman analyses showed moderate to good agreement between most assays, although clear absolute differences were observed. In conclusion, although results obtained with different methods were significantly correlated and therefore comparable in relative terms, direct comparison of results between laboratories and assays may be inappropriate.
    02/2013; 15(2):405-411. DOI:10.1039/C2EM30749A
  • Collin Brooks · Neil Pearce · Jeroen Douwes
    [Show abstract] [Hide abstract]
    ABSTRACT: PURPOSE OF REVIEW: It has been hypothesized that increased cleanliness, reduced family size, and subsequent decreased microbial exposure could explain the increases in global asthma prevalence. This review considers the recent evidence for and against the 'hygiene hypothesis'. RECENT FINDINGS: Recent evidence does not provide unequivocal support for the hygiene hypothesis: the hygiene hypothesis specifically relates to atopic asthma, but some of the protective effects (e.g. farm exposures) appear to apply to both atopic and nonatopic asthma; asthma prevalence has begun to decline in some western countries, but there is little evidence that they have become less clean; Latin American countries with high infection rates have high asthma prevalence and the hygiene hypothesis relates to early-life exposures, but exposures throughout life may be important. SUMMARY: There is a considerable body of evidence which warrants scepticism about the hygiene hypothesis. However, these anomalies contradict the 'narrow' version of it in which microbial pressure early in life protects against atopic asthma by suppressing T-helper 2 immune responses. It is possible that a more general version of the hygiene hypothesis is still valid, but the aetiologic mechanisms involved are currently unclear.
    Current Opinion in Allergy and Clinical Immunology 10/2012; 13(1). DOI:10.1097/ACI.0b013e32835ad0d2 · 3.66 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The literature was reviewed to assess the relationship between the lipid adjusted concentration in human serum and breast milk (expressed as the serum/milk ratio) of a broad range of POPs in paired samples. Thirteen studies were identified, including seven studies that reported serum/milk ratios for polychlorinated dibenzo-dioxins and -furans (PCDD/Fs), ten for polychlorinated biphenyls (PCBs), five for polybrominated diphenyl ethers (PBDEs), and five for organochlorine pesticides (OCPs). Mean serum/milk ratios ranged between 0.7 and 25 depending on the compound and congener. For PCDD/Fs, PCBs and PBDEs, a clear trend of increasing mean serum/milk ratio by increasing molar volume, hydrophobicity and number of halogen substitutes was observed. The mean serum/milk ratios reported by the 13 studies summarized here will aid comparison between human POPs exposure studies using either serum or milk samples. More studies are needed to allow a valid comparison between data obtained from analysis of breast milk and serum samples for a broader range of POPs. Furthermore such studies may shed light on compound specific factors as well as other determinants that may affect the partitioning and partition kinetics of POPs between serum and breast milk.
    Chemosphere 08/2012; 89(8):911-8. DOI:10.1016/j.chemosphere.2012.06.049 · 3.50 Impact Factor
  • Neil Pearce · Jeroen Douwes
    [Show abstract] [Hide abstract]
    ABSTRACT: In recent decades there have been marked increases in asthma prevalence in Western countries. More recently, asthma prevalence has peaked, or even begun to decline, in Western countries, but many low and middle income countries are now beginning to experience increases in prevalence (although there is no evidence of increases in prevalence in India to date). "Established" risk factors for asthma cannot account for the global prevalence increases, or the international patterns that have been observed, or the recent declines in prevalence in some Western countries. It seems that as a result of the "package" of changes in the intrauterine and infant environment that are occurring with "Westernization", we are seeing an increased susceptibility to the development of asthma and/or allergy. There are a number of elements of this "package" including changes in maternal diet, increased fetal growth, smaller family size, reduced infant infections and increased use of antibiotics and paracetamol, and immunization, all of which have been (inconsistently) associated with an increased risk of childhood asthma, but none of which can alone explain the increases in prevalence. It is likely that the "package" is more than the sum of its parts, and that these social and environmental changes are all pushing the infants' immune systems towards an increased risk of asthma.
    The Indian Journal of Pediatrics 05/2012; 80(Suppl 1). DOI:10.1007/s12098-012-0764-2 · 0.92 Impact Factor

Publication Stats

6k Citations
568.10 Total Impact Points


  • 2003–2015
    • Massey University
      • Centre for Public Health Research
      Palmerston North City, Manawatu-Wanganui, New Zealand
    • Erasmus Universiteit Rotterdam
      Rotterdam, South Holland, Netherlands
  • 2001–2009
    • Universiteit Utrecht
      • Institute for Risk Assessment Sciences (IRAS)
      Utrecht, Provincie Utrecht, Netherlands
    • Columbia University
      • Department of Environmental Health Sciences
      New York City, NY, United States
  • 2006
    • The Chinese University of Hong Kong
      • Department of Paediatrics
      Hong Kong, Hong Kong
  • 2004
    • National Institute of Occupational Health (STAMI)
      • Department of Chemical and Biological Work Environment
      Kristiania (historical), Oslo County, Norway
  • 2000
    • Kuopio University Hospital
      • Department of Pulmonary Diseases
      Kuopio, Northern Savo, Finland
    • University of Iowa
      • College of Public Health
      Iowa City, Iowa, United States
  • 1995–2000
    • Wageningen University
      • Sub-department of Environmental Technology
      Wageningen, Gelderland, Netherlands
  • 1996
    • University of Amsterdam
      • Department of Molecular Cell Biology
      Amsterdamo, North Holland, Netherlands