Published on Fri Aug 13 2021

Exposure to pesticides and childhood leukemia risk: A systematic review and meta-analysis.

Maria A Karalexi, Christos F Tagkas, Georgios Markozannes, Xanthippi Tseretopoulou, Antonio F Hernández, Joachim Schüz, Thorhallur I Halldorsson, Theodora Psaltopoulou, Eleni Th Petridou, Ioanna Tzoulaki, Evangelia E Ntzani

Evidence remains inconclusive, and is inherently limited by heterogeneous exposure assessment and multiple statistical testing. We performed a literature search of peer-reviewed studies, published until January 2021, without language restrictions. We identified 55 eligible studies from over 30 countries assessing >200 different exposures of pesticides (n =160,924 participants)

1
4
9
Abstract

Despite the abundance of epidemiological evidence concerning the association between pesticide exposure and adverse health outcomes including acute childhood leukemia (AL), evidence remains inconclusive, and is inherently limited by heterogeneous exposure assessment and multiple statistical testing. We performed a literature search of peer-reviewed studies, published until January 2021, without language restrictions. Summary odds ratios (OR) and 95% confidence intervals (CI) were derived from stratified random-effects meta-analyses by type of exposure and outcome, exposed populations and window of exposure to address the large heterogeneity of existing literature. Heterogeneity and small-study effects were also assessed. We identified 55 eligible studies (n = 48 case-control and n = 7 cohorts) from over 30 countries assessing >200 different exposures of pesticides (n = 160,924 participants). The summary OR for maternal environmental exposure to pesticides (broad term) during pregnancy and AL was 1.88 (95%CI: 1.15-3.08), reaching 2.51 for acute lymphoblastic leukemia (ALL; 95%CI: 1.39-4.55). Analysis by pesticide subtype yielded an increased risk for maternal herbicide (OR: 1.41, 95%CI: 1.00-1.99) and insecticide (OR: 1.60, 95%CI: 1.11-2.29) exposure during pregnancy and AL without heterogeneity (p = 0.12-0.34). Meta-analyses of infant leukemia were only feasible for maternal exposure to pesticides during pregnancy. Higher magnitude risks were observed for maternal pesticide exposure and infant ALL (OR: 2.18, 95%CI: 1.44-3.29), and the highest for infant acute myeloid leukemia (OR: 3.42, 95%CI: 1.98-5.91). Overall, the associations were stronger for maternal exposure during pregnancy compared to childhood exposure. For occupational or mixed exposures, parental, and specifically paternal, pesticide exposure was significantly associated with increased risk of AL (OR