Epidemiological data are not currently used in the risk assessment of chemical substances in a systematic and consistent manner. However, systematic reviews (SRs) could be useful for risk assessment as they appraise and synthesize the best epidemiological knowledge available.
To conduct a comprehensive literature search of SRs pertaining to pesticide exposure and various neurological outcomes, namely neurodevelopmental abnormalities, Parkinson’s disease (PD) and Alzheimer’s disease (AD), and to assess the potential contribution of SRs to the risk assessment process.
Search methods and selection criteria
Search was conducted in PubMed and Web of Science databases and articles were selected if the following inclusion criteria were met: being a SR, published until April 2015 and without language restrictions.
Data collection and analysis
For each neurological outcome, two review authors independently screened the search results for included studies. Data were extracted and summarized in two tables according to 16 criteria. Disagreements were resolved by discussion.
The total number of studies identified in the first search was 65, 304 and 108 for neurodevelopment, PD and AD, respectively. From them, 8, 10 and 2 met the defined inclusion criteria for those outcomes, respectively. Overall, results suggest that prenatal exposure to organophosphates is associated with neurodevelopmental disturbances in preschool and school children. In contrast, postnatal exposures failed to show a clear effect across cohort studies. Regarding PD, 6 SRs reported statistically significant combined effect size estimates, with OR/RR ranging between 1.28 and 1.94. As for AD, 2 out of the 8 original articles included in the SRs found significant associations, with OR of 2.39 and 4.35, although the quality of the data was rather low.
The critical appraisal of the SRs identified allowed for discussing the implications of SRs for risk assessment, along with the identification of gaps and limitations of current epidemiological studies that hinder their use for risk assessment. Recommendations are proposed to improve studies for this purpose. In particular, harmonized quantitative data (expressed in standardized units) would allow a better interpretation of results and would facilitate direct comparison of data across studies. Outcomes should be also harmonized for an accurate and reproducible measurement of adverse effects. Appropriate SRs and quantitative synthesis of the evidence should be performed regularly for a continuous update of the risk factors on health outcomes and to determine, if possible, dose–response curves for risk assessment.