Is fluoridated water safe during pregnancy? Studies say yes

Analyzing nearly 11.5 million births across four decades, researchers use rigorous methods to show that fluoridated drinking water is not linked to lower birth weight.

Study: Community Water Fluoridation and Birth Outcomes. Image credit: Tatjana Meininger/Shutterstock.com

While fluoridated water is widely considered a public health benefit, its potential association with reduced birth weight has often raised concern. A recent study published online on JAMA Network Open explores this question in a large nationwide study.

Fluoridated water’s benefits and long-standing health concerns

The concept of community water fluoridation (CWF) is based on the observation that naturally higher water fluoride levels are associated with lower rates of dental caries at the community level. This large-scale public health experiment began in 1945, when Grand Rapids, Michigan, fluoridated its municipal water supply, with Muskegon, a nearby town, serving as the control.

The original experiment demonstrated its efficacy, with a 60 % reduction in pediatric dental caries in Grand Rapids by 1950. This triggered the widespread rollout of CWF, which covered 63 % of the US population by 2018, up from 3.3 % in 1951. The actual reduction in dental caries is likely to be somewhat lower, at 25 % to 35 %, probably because other fluoridated products are widely available.

Some adverse effects of fluoride are also well documented, such as fluorosis (teeth discoloration) with excessive intake. This is rare with CWF.

Other scientists have suggested that prenatal and early-life fluoride exposure could be associated with impaired cognitive development. Notably, observational studies have produced conflicting findings and have been subject to significant confounding. The current study does not directly evaluate neurodevelopmental outcomes but instead focuses on birth outcomes as indicators of early-life health.

The current study used birth weight as an alternative, complementary measure of the putative adverse effects of prenatal fluoride exposure. Birth weight is a reliable marker of overall infant and long-term health. Moreover, it rapidly and sensitively reflects prenatal exposures, reducing biases due to cumulative exposures or other unmeasured confounding factors operating after birth.

Birth weight is also a universally documented measure in the US, along with the mother’s county of residence, allowing accurate evaluation of community-level CWF exposure.

Observational studies have reported decreases in birth weight associated with higher fluoride levels in pregnant women. Maternal fluoride exposure raises fetal fluoride levels, potentially inducing oxidative stress. Fluoride may also alter maternal thyroid function or impact placental function. Thus, it may affect birth weight through multiple mechanisms, although these pathways were not directly examined in the current analysis.

Using county-level rollout to isolate fluoridation’s effects

The study used a cohort design spanning 1968 to 1988. The researchers chose an event-study analysis with a difference-in-differences (DID) approach. This technique tracks outcome differences in the intervention vs control groups at multiple time points before and after the specified event.

This statistical method is one of several that reduce confounding, demonstrating its value in assessing public health intervention outcomes, especially when experimental research is not feasible. The study design was made possible by the staggered initiation of CWF across US counties. Exposure rates were measured using data from the Centers for Disease Control and Prevention Fluoridation Census, which indicates the proportion of county residents who were supplied with fluoridated water.

Birth outcomes before and after the onset of fluoridation were compared within each county in the intervention group. Counties that had never implemented it or had not yet implemented it were in the control group. The researchers primarily evaluated the mean birth weight. Secondary outcomes included the rate of low birth weight, the mean gestational length, and the prematurity rate.

No detectable changes in birth weight after fluoridation

The study included 11,479,922 singleton births across 677 counties, with a mean birth weight of 3.34 kg and a mean gestational age of 39.5 weeks. CWF was implemented in 408 counties, comprising 60 % of the total.

The number of exposed counties increased steadily, covering 46 % of the population. However, 100 % fluoridation did not occur since not all water sources serving a given county were fluoridated at the same time, if ever. On average, CWF increased the proportion of county residents with access to fluoridated water by 32 percentage points.

Significantly, birth weights in the years preceding and after CWF rollout followed similar trends across exposed and control counties. Thus, the study found no evidence of an association between CWF and reduced birth weight.

The current study used a robust statistical technique that reduced the risk of confounding. These results support the interpretation that CWF has no detectable causal effect on birth weight, provided that the timing of CWF within a county is independent of other factors that drive birth weight.

Factors favoring this assumption include the staggered timing of CWF that was apparently systematically unrelated to other public health programs in the county. The study also uses comparisons within rather than between municipalities, yielding an intention-to-treat estimate at the community level rather than an estimate of individual fluoride exposure.

Additional evidence supporting internal validity includes the absence of differences in birth weight trends across counties, regardless of later CWF implementation, and multiple sensitivity analyses.

Limitations of the study include potential misclassification of exposure, as the researchers measured community access to fluoride rather than individual fluoride exposure based on biomarker use. It is possible that other environmental regulations were put into force, which improved water and air quality concomitantly with CWF, confounding the effect on infant health. However, these influences are limited by their national rather than water-district-level impact.

Further support comes from the various sensitivity analyses, which fail to confirm the hypothesis even when restricted to counties with more than 90 % fluoridation coverage, when model alternative exposures, or when accounting for changes in the overall composition of births. Even the most negative estimates discussed represent less than a 1 % change in mean birth weight, suggesting limited clinical significance.

Finally, other exposures, such as disinfectants or anti-corrosion agents added to the water supply, could impair the results of this analysis. However, the consistently null findings across all outcomes make this explanation unlikely.

Evidence supports fluoridation safety for birth outcomes

The study finds no evidence that community water fluoridation during pregnancy adversely affects birth weight, and adds to the body of evidence suggesting no detectable harm for this outcome. It also underlines the need for scientific rigor while framing safety studies to assess public health interventions. This is necessary to avoid drawing causal conclusions from less reliable observational study findings, which may be vulnerable to residual confounding and fade under more stringent quasi-experimental analysis.

Future research should attempt to provide more objective fluoride exposure measurements and capture non-CWF sources of fluoride intake, improving the generalizability of the results.

Download your PDF copy now!