Fluoride is added to drinking water and salt for the prevention of dental caries (CDC, 2016). Other sources of fluoride include fluoridated dental products and supplements, certain foods that absorb naturally occurring fluoride, such as green and black tea, and foods that are sprayed with fluoride-containing pesticides (i.e., grapes; Nutrient Data Laboratory, 2015; Zohoori et al., 2013). Recent studies conducted in the United States (Abduweli Uyghurturk et al., 2020), Canada (Till et al., 2018), and Mexico (Thomas et al., 2016a,b; Castiblanco-Rubio et al., 2021) have reported positive associations between fluoride from dietary sources, including drinking water and salt, and urinary fluoride levels in pregnant women. Because of its ubiquity and its ability to pass through the placenta and blood-brain barrier to reach the fetal brain (Agency for Toxic Substances and Disease Registry, 2003), the safety of fluoride exposure in pregnancy has received much attention, both in endemic fluorosis areas (Jiménez et al., 2017) and communities that have fluoridation programs (Green et al., 2019; Bashash et al., 2017).

While it is not disputed that fluoride is a developmental neurotoxicant at high exposure levels, there are relatively few studies that have assessed fluoride’s potential neurotoxicity at levels found in fluoridated areas (i.e., 0.7 mg/L), particularly for pregnant women and young infants. In 2021, the National Toxicology Program (National Toxicology Program, 2020) conducted a systematic review on the impact of fluoride on neurodevelopmental outcomes. The NTP identified two high-quality prospective birth cohort studies that were conducted in Mexico City where salt is fluoridated at 250 ppm (Bashash et al., 2017) and in Canada where drinking water is fluoridated at 0.7 mg/L (Green et al., 2019). Both cohort studies found a 4-to 6-point lower Full-Scale IQ score in children per 1 mg/L increase in maternal urinary fluoride level; in the Canadian cohort the effect of maternal urinary fluoride on IQ was only found in boys while the effect of drinking water fluoride level on IQ was found in both boys and girls.

Most human developmental toxicology studies focus on global or composite test scores, such as Full-Scale IQ, which are derived from a diverse set of tasks (Kamphaus, 2019). While global outcomes are considered highly significant from a public health (Lanphear, 2015) and economic standpoint (Gould, 2009), a low composite score does not convey specific information about the child’s intellectual and cognitive profile. When there are strengths and weaknesses in a cognitive profile, the use of a composite score may have high sensitivity, but at the cost of low specificity (Fiorello et al., 2007). Partitioning Full-Scale IQ into domain specific intellectual abilities, such as verbal and nonverbal skills, may reveal particular cognitive domains that are more sensitive to neurotoxic exposures or may be differentially affected over time (Bellinger et al., 2016). For example, studies have demonstrated that prenatal and early-life lead exposure is more strongly associated with non-verbal intelligence compared to verbal intelligence between the ages of 2–7 (Bellinger et al., 1991; Desrochers-Couture et al., 2018; Dietrich et al., 1991, 1993; Factor-Litvak et al., 1999; Jusko et al., 2008; Wasserman et al., 1997). Similarly, prenatal and early-life fluoride exposure has been associated with greater deficits in non-verbal abilities than verbal abilities in preschool years (Till et al., 2020; Farmus et al., 2021; Cantoral et al., 2021); however, other studies examining early-life exposure to fluoride did not find observe this profile (Ibarluzea et al., 2021) or did not report verbal and non-verbal intelligence (Bashash et al., 2017). Whether non-verbal intelligence is associated with early life exposure to fluoride over the course of child development has not been examined.

In the present study, we examined the longitudinal and domain specific effects (i.e., verbal and nonverbal intelligence) of prenatal fluoride exposure on IQ in mother-child dyads from the Early Life Exposures in Mexico to Environmental Toxicants (ELEMENT) cohort. Since our prior publication on this cohort (Bashash et al., 2017), we received additional maternal urinary fluoride and creatinine data enabling us to examine children’s IQ at three separate time points (age 4, 5, and 6–12 years) and using a larger sample size at each time point relative to our prior work. We also examined the potential for sex-specific effects based on findings that boys may be more susceptible to prenatal exposure than girls (Cantoral et al., 2021; Comfort and Re, 2017; Green et al., 2019, 2020; Torres-Rojas and Jones, 2018).