RESUMO
BACKGROUND: Application of pesticides in the vicinity of homes has caused concern regarding possible health effects in residents living nearby. However, the high spatiotemporal variation of pesticide levels and lack of knowledge regarding the contribution of exposure routes greatly complicates exposure assessment approaches. OBJECTIVE: The objective of this paper was to describe the study protocol of a large exposure survey in the Netherlands assessing pesticide exposure of residents living close (<250 m) to agricultural fields; to better understand possible routes of exposure; to develop an integrative exposure model for residential exposure; and to describe lessons learned. METHODS: We performed an observational study involving residents living in the vicinity of agricultural fields and residents living more than 500 m away from any agricultural fields (control subjects). Residential exposures were measured both during a pesticide use period after a specific application and during the nonuse period for 7 and 2 days, respectively. We collected environmental samples (outdoor and indoor air, dust, and garden and field soils) and personal samples (urine and hand wipes). We also collected data on spraying applications as well as on home characteristics, participants' demographics, and food habits via questionnaires and diaries. Environmental samples were analyzed for 46 prioritized pesticides. Urine samples were analyzed for biomarkers of a subset of 5 pesticides. Alongside the field study, and by taking spray events and environmental data into account, we developed a modeling framework to estimate environmental exposure of residents to pesticides. RESULTS: Our study was conducted between 2016 and 2019. We assessed 96 homes and 192 participants, including 7 growers and 28 control subjects. We followed 14 pesticide applications, applying 20 active ingredients. We collected 4416 samples: 1018 air, 445 dust (224 vacuumed floor, 221 doormat), 265 soil (238 garden, 27 fields), 2485 urine, 112 hand wipes, and 91 tank mixtures. CONCLUSIONS: To our knowledge, this is the first study on residents' exposure to pesticides addressing all major nondietary exposure sources and routes (air, soil, dust). Our protocol provides insights on used sampling techniques, the wealth of data collected, developed methods, modeling framework, and lessons learned. Resources and data are open for future collaborations on this important topic. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): RR1-10.2196/27883.
RESUMO
Phomopsins (PHO) are mycotoxins produced by the fungus Diaporthe toxica (also referred to as Phomopsis leptostromiformis). Lupin is the most important host crop for this fungus and PHO are suspected as cause of lupinosis, a deadly liver disease, in sheep. Lupin is currently in use to replace genetically modified soy in many food products available on the European market. However, a validated method for analysis of PHO is not available until now. In this work, a dilute-and-shoot LC-MS/MS-based method was developed for the quantitative determination and identification of phomopsin A (PHO-A) in lupin and lupin-containing food. The method involved extraction by a mixture of acetonitrile/water/acetic acid (80/20/1 v/v), dilution of the sample in water, and direct injection of the crude extract after centrifugation. The method was validated at 5 and 25 µg PHO-A kg(-1) product. The average recovery and RSD obtained were 79% and 9%, respectively. The LOQ (the lowest level for which adequate recovery and RSD were demonstrated) was 5 µg PHO-A kg(-1). Identification of PHO-A was based on retention time and two transitions (789 > 226 and 789 > 323). Using the average of solvent standards from the sequence as a reference, retention times were all within ± 0.03 min and ion ratios were within ± 12%, which is compliant with European Union requirements. The LOD (S/N = 3 for the least sensitive transition) was 1 µg PHO-A kg(-1) product. Forty-two samples of lupin and lupin-containing food products were collected in 2011-2012 from grocery stores and internet shops in the Netherlands and analysed. In none of the samples was PHO-A detected.