Children of farmworkers or those residing in close proximity to crop production have an increased risk of exposure to pesticides (Coronado et al., 2011; Hyland and Laribi, 2017; Loewenherz et al., 1997; Lu et al., 2000). Exposures to neurotoxic pesticides in the prenatal period or early childhood may compromise healthy neurodevelopment (Bouchard et al., 2011; Burns et al., 2013; Eskenazi et al., 2007; Grandjean, 2006; Harari et al., 2010; Shelton et al., 2014; J. R. Suarez-Lopez et al., 2013). In a birth cohort in California, increasing levels of neurotoxic pesticides used within 1 km of maternal residence during pregnancy was associated with decreases in children’s cognitive function at 7-years (Gunier et al., 2017) and 10-years (Rowe et al., 2016) of age. For each standard deviation (SD) increase in agricultural use of total organophosphates (OPs) there was a 2-point decrease in Full-Scale IQ (Gunier et al., 2017). Three case-control studies of autism spectrum disorders (ASD) in California found that pregnant mothers had greater odds of having children later diagnosed with ASD if they lived near OP, pyrethroid organochlorine or other pesticide spray and application sites (Roberts et al., 2007; Carmichael et al., 2016; Ehrenstein et al., 2019).

The use of OPs and other classes of insecticides, fungicides and herbicides has been well documented throughout Pedro Moncayo County, Ecuador, a county with one of the highest concentrations of rose plantations on the continent and where flower production is a primary revenue driver (Grandjean et al., 2006; Handal et al., 2016; Harari, 2004; Suarez-Lopez et al., 2017a). Rose production is typically enclosed within greenhouses; however, pesticides may drift from the greenhouses through ventilation windows. Production of other flowers in open fields also exists, although in substantially smaller amounts. There is currently a need to understand not only the exposure patterns of children living near greenhouse agriculture, but also whether the neurobehavioral performance of such children may also be affected as a result of chronic pesticide drift from such crops. In the study of Secondary Exposures to Pesticides among Children and Adolescents (ESPINA: Estudio de la Exposición Secundaria a Plaguicidas en Niños y Adolescentes) based in Pedro Moncayo County, we observed that children living near flower crops had lower acetylcholinesterase (AChE) activity (reflecting greater exposure to cholinesterase inhibitor pesticides) and higher blood pressure compared to children living farther away (Suarez-Lopez et al., 2018). Because of these findings and our previously reported positive associations between AChE activity and neurobehavioral performance in boys (J.R. Suarez-Lopez et al., 2013) and transient alterations in neurobehavioral performance in relation to a peak pesticide spray season (Suarez-Lopez et al., 2017b), we hypothesized that children living near greenhouse floricultural production sites, assessed during the same time period with respect to harvest season, had worse neurobehavioral performance compared to children living farther away.

The mean age of children at the time of assessment was 6.6 years (SD = 1.6); 51% were male, 72% reported mestizo as their race, 21% indigenous, 2% other (white or black) and 5% had missing information (Table 1). of the 307 participants (50%) lived with a floricultural worker. The means of height-for-age z-score and hemoglobin concentration were −1.25 (SD: 0.96) and 12.7 mg/dL (SD: 1.2), respectively. On average, children in this sample lived 446 m from the edge of the nearest flower crop (SD: 344 m). In this sample of children, 6% lived within 50m, 11% lived within 100m, 27% lived within 200 m, and 61% lived within 500 m of a plantation’s edge. The average quantity of growing area within 100m of the residence was 292 m² (SD: 1406 m²).

The neurobehavioral subtest scores for subjects in our study sample were lower but with similar variability to those of the NEPSY-2 normative sample of U.S. children (data not shown)(J. R. Suarez-Lopez et al., 2013). The neurobehavioral domain scores of our participants ranged from 4 to 13, with mean (SD) scores of 6.6 (2.4) in Language, 8.8 (2.1) in Memory and Learning, 9.9 (3.3) in Sensorimotor, 9.6 (3.1) in Visuospatial and 8.5 (2.5) in the Attention and Inhibitory domain. The lowest domain scores were observed in the Language and Attention & Inhibitory Control domains. The percentages of children with scores of clinical concern for their age group were: 10% in the Memory domain, 12% in Visuospatial Processing, 15% in Attention & Inhibitory Control, 17% in Sensorimotor, and 33% in the Language domain (Korkman et al., 2007).

We observed that close residential proximity to floricultural crops was associated with poorer neurobehavioral outcomes in the domains of Attention & Inhibitory Control, Language and Memory & Learning. Associations were strongest among children living within 50m of a flower crop and present to some extent among children living between 51 and 100m. These findings were partially corroborated by sensitivity analyses using areas of floricultural crops near homes as a related construct of pesticide drift from flower crops.

Unlike short-lived biomarkers of exposure, proximity of a child’s home to agricultural crops may approximate the child’s ongoing and historical low dose exposure to pesticides through off-target drift or direct access to pesticide-treated areas. In the ESPINA study, we previously described positive associations between AChE activity (lower values reflect greater exposure to cholinesterase inhibitor pesticides) and the domains of Attention & Inhibitory Control, Memory & Learning, and borderline associations with the Language domain affecting boys more than girls (J. R. Suarez-Lopez et al., 2013). Alterations in the same domains were observed in the present study, which is consistent with previous findings. Epidemiologic studies provide increasing evidence that pesticide exposure during key developmental periods may be a risk factor for a range of neurocognitive deficits later in life, including attention deficit and hyperactivity disorder, autism spectrum disorder, developmental delay, slowed reaction time, and slowed motor control, poor verbal comprehension (Bouchard et al., 2011; Burns et al., 2013; Eaton et al., 2008; Eskenazi et al., 2008, 2007; Fenske et al., 2000; Grandjean, 2006; Grandjean et al., 2006; Handal et al., 2008; Harari et al., 2010; Horton et al., 2012; Kofman et al., 2006; London et al., 2012; Rauh and Margolis, 2016; Shelton et al., 2014; J. R. Suarez-Lopez et al., 2013)

People living closest to agricultural crops are at increased risk of pesticide exposure. In our analyses, children living within 100m of a flower crop, and especially within 50m, had lower neurobehavioral scores compared to children living farther than 500m. These findings suggest that the amount of pesticide drift from crops onto nearby homes can especially affect the neurobehavioral performance of children living within 100m. However, alterations in neurobehavioral performance may also be present at greater distances but the limited statistical power of our study to detect smaller differences precluded us from assessing this further. In previous analyses of the ESPINA study, we observed positive associations between residential distances to flower crops and AChE activity, with the lowest AChE levels observed among children living within 232m of a greenhouse floricultural crop (Suarez-Lopez et al., 2018). This supports the construct of residential distance to flower plantations as a pathway of exposure to pesticides Furthermore, we previously observed that children living closer to flower crops had higher systolic blood pressures, which indicates that additional physiologic processes may be affected among children living near pesticide spray sites (Suarez-Lopez et al., 2018).

Multiple investigations have studied the association between proximity of homes to agricultural crops and pesticide exposure (Coronado et al., 2011; Deziel et al., 2015; Loewenherz et al., 1997; Lu et al., 2000; Simcox et al., 1995; Ward et al., 2006). While maximal exposure attributable to pesticide drift, among these studies, varied from 60 to 750 meters, this collection of studies rather consistently indicates that homes residing closer to pesticide treated fields tend to have higher pesticide levels and that children residing closer to pesticide treated fields tend to reflect higher pesticide exposure levels using biomonitoring studies. In this study, exposure was modeled as distance to the nearest plantation in the primary analyses, based on the assumption that increased distance reflects lower exposures. An alternative measure, area of plantations within varying buffer areas, which is likely a better proxy for exposure to pesticide drift, was also explored. As expected, results showed consistent associations between these two related but different constructs of pesticide exposure, which strengthens our findings.

Several studies have utilized residential proximity to agriculture as a metric of exposure to pesticides when studying its associations with neurodevelopment (Butler-Dawson et al., 2016; Coker et al., 2017; González-Alzaga et al., 2015; Gunier et al., 2017; Handal et al., 2007; Roberts et al., 2007; Rowe et al., 2016; Shelton et al., 2014; von Ehrenstein et al., 2019). A number of these studies used data from California State’s Pesticide Use Reporting System, finding positive correlations between proximity of prenatal residence to areas of agricultural pesticide applications and neurodevelopmental outcomes in early childhood, specifically ASD (Roberts et al., 2007; Shelton et al., 2014). In our analyses the observed effect size in the logistics models were small, but the magnitude may have been attenuated by the non-linear dose-response relationship shown in Figure 3. Furthermore, the linear regression models indicated that a difference of 100m in residential proximity to floricultural crops is associated with a greater likelihood of the child scoring in the subclinical ranges for the Language and the Memory and Learning domains by 9% and 24% respectively. In the context of measurable outcomes, this is clinically significant in that identifying children with delayed development warrants early intervention by clinicians as well as educators. The expected distance of pesticide drift from flower crops to nearby homes was smaller in our study population compared to those of other studies likely because rose production is enclosed within greenhouses. Greenhouses in Pedro Moncayo County have air circulation vents or windows, which could allow the escape of fumigated pesticides during and after spraying. However, these analyses suggest that pesticide drift, even in this setting, could still be problematic in the context of pesticide exposure affecting the neurodevelopment of children living nearby. This body of evidence coupled with the growing number of studies describing neurobehavioral alterations associated with pesticide exposures (Bouchard et al., 2011; Burns et al., 2013; Eskenazi et al., 2007; Grandjean, 2006; Harari et al., 2010; Shelton et al., 2014; J. R. Suarez-Lopez et al., 2013) suggests that extending buffer zones or protective areas that separate the industry from the neighboring communities, could be an effective way to protect developing children (including during prenatal development) from the adverse effects of pesticide exposure.

The present study was subject to several limitations. Residential distance from treated floricultural crops was used as a proxy for childhood exposure to floricultural pesticides. Though a crude exposure assessment, the use of this exposure metric is supported by the existing literature and validated within our study population (Suarez-Lopez et al., 2018). Prevailing winds were not accounted for in the present analyses. This provides potential for non-differential misclassification of the amount of pesticide drift from plantations to homes and may have biased our findings towards the null (Gordis, 2014). Also, while the vast majority of the floricultural production in Pedro Moncayo County includes roses, which are grown inside greenhouses, a small amount of production of other flowers also occurs in non-enclosed fields typically located near the greenhouses. For this reason, it is plausible that some of the pesticide drift from crops, and hence the associations observed in this study, may be a result of both greenhouse and open field floricultural production. Nonetheless, residential proximity to crops is a useful construct of exposure as it is an indicator of chronic pesticide exposure, and provides practical information about the distances in which populations may have an increased risk of pesticide exposure and/or adverse health. While it does not allow us to determine which specific chemicals are influencing this association, it indirectly accounts for a mixture of the various agrochemicals used in floriculture. The floricultural industry in Pedro Moncayo frequently uses various pesticides including insecticides (organophosphates, neonicotinoids and pyrethroids), many classes of fungicides and to lesser extent, herbicides (Grandjean et al., 2006; Handal et al., 2016; Harari, 2004; Suarez-Lopez et al., 2017a). Many of the studies assessing neurodevelopment and pesticide exposure, including the ESPINA Study, are limited in that they study biomonitoring of few pesticides, even though it is unusual for one pesticide to be used without multiple others. It is plausible that pesticides or other neurotoxicant agrochemical exposures explain the neurobehavioral deficits seen among participants living near the flower crops. Determining the quantity and types of agrochemicals used over time and by location would improve precision but would be very difficult to ascertain in this agricultural setting. Another limitation related to exposure assessment is that we were not able to account for all potential routes of exposure to pesticides, including dietary intake. We did not have information on time-activity patterns, which would have provided better insight into participants’ outdoor exposures (Ring et al., 2019). There is also some uncertainty associated with using home location only to estimate exposure to environmental pesticides. Some children in the cohort went to school during the day, while younger children attended daycare or stayed home with a relative. Modeling exposure experienced across all daily activities and locations is beyond the scope of this current study; we choose to focus on exposures at the children’s home locations.

Another limitation of this study is that neurobehavioral outcomes were assessed at only one point in time, and thus we are unable to assess if the neurodevelopmental effects are permanent. Additionally, the NEPSY-2 is based on a US normative sample. Although this does not affect the internal validity of our findings, it is unclear how accurately the cut-off values for “low performance” apply to this study population.

This study has several strengths and thus contributes to our understanding of the effects of pesticide use in floricultural communities. This study is unique in that there was a wide distribution of participants’ residential distance to crops, and we had a considerable number of children living in close proximity (<100m) to flower crops, which allowed us to estimate effect sizes at short distances. Additionally, all participants were examined during a period of relatively homogeneous flower production and pesticide use. Children in this study were examined during a period of lower flower production (July-August) and pesticide use compared to other times of the year. In theory, this would reduce the off-target pesticide drift potential from crops, with resulting lower exposures to children living nearby. Considering that pesticide spray seasons may also have short-term neurobehavioral alterations in children (Ismail et al., 2017; Rohlman et al., 2016, 2014; Suarez-Lopez et al., 2017b), it is plausible that these observed associations would be stronger during peak exposure periods. Lastly the existing studies have focused on residential distance to agricultural open fields. To our knowledge, the present study is the first to characterize the associations of neurobehavior in relation to greenhouse agricultural production, which is generally though to result in reduced pesticide drift from crops. Many types of crops involve the use of greenhouses such as flowers, tomatoes, cucumber, a variety of herbs, lettuce, bell peppers, and eggplants. The present study findings may be applicable to such agricultural production.

Children living in close proximity to floricultural crops had poorer neurobehavioral performance in the domains of Attention & Inhibitory Control, Language and Memory & Learning. These findings indicate that pesticide drift from agricultural plantations may affect the neurobehavioral development of children living nearby. This association between residential distances to agricultural industries and health outcomes could be studied further to facilitate the development of policies and practices that protect agricultural populations from adverse effects of pesticide drift.