Drinking-water lithium linked to a higher prevalence of Wolff-Parkinson-White diagnoses in children

Higher lithium levels in drinking water tracked with a greater prevalence of Wolff-Parkinson-White diagnoses in childhood, raising the possibility that a common environmental exposure could influence development of cardiac accessory pathways before birth. The analysis focused on children under 18 in California and used county-level lithium measurements in drinking water against diagnosis rates captured through administrative codes. Because children are likely to live in the same county during fetal development, the design aimed to approximate in utero exposure rather than later-life contact alone.

The cohort was large: 2,329,090 patients with a mean age of 8.2 years, and 47.8% were female. Among them, 686 children, or 0.03%, carried a diagnosis of WPW. Counties with lithium concentrations at or above the 75th percentile showed 71% greater odds of WPW diagnosis, with an odds ratio of 1.71 and a 95% confidence interval of 1.44 to 2.03. Counties above the lithium health-based screening limit still showed a signal, with 29% greater odds of WPW, corresponding to an odds ratio of 1.29 and a 95% confidence interval of 1.05 to 1.58.

A single sentence captures the clinical tension here: a routine water contaminant aligned with a childhood arrhythmia phenotype that is usually regarded as rare and largely unexplained.

The biologic plausibility centers on the fetal origin of accessory pathways. Wolff-Parkinson-White reflects an atrioventricular conduction bridge that bypasses the normal AV node, enabling rapid reentrant tachycardias and, in uncommon cases, dangerous arrhythmia-related deterioration. The embryologic development of the annulus fibrosus and atrioventricular junction is tightly regulated, and anything that disturbs that process during gestation could conceivably alter pathway persistence. Lithium already has an established clinical association with congenital heart disease in the setting of maternal medicinal exposure, especially Ebstein’s anomaly, so environmental lithium exposure invites the same developmental question from a different route.

The findings did not stop at diagnosis prevalence. Validation analyses linked higher lithium areas with more SVT, syncope, palpitations, and catheter ablation procedures. Those endpoints fit the expected clinical footprint of accessory pathway disease: symptomatic tachycardia, episodic cerebral hypoperfusion, patient-reported palpitations, and definitive procedural treatment when symptoms or risk justify electrophysiology intervention. Seeing the same geographic exposure correlate with several downstream arrhythmia-related outcomes strengthens the internal consistency of the signal, even though the study remains ecological rather than individual-level.

The investigation extended beyond California. A statewide New York cohort showed similar results, giving the association broader geographic support and making it less likely that the California findings were a local anomaly. Cross-state replication matters here because county-level environmental studies are vulnerable to confounding by regional differences in healthcare access, coding practices, socioeconomic patterns, and referral behavior. Replication across separate populations does not eliminate those concerns, but it does make the pattern harder to dismiss.

The design also included positive and negative controls, as well as falsification endpoints, which is an important safeguard in ecological epidemiology. These checks help assess whether an observed relationship might simply reflect coding density or another geographic artifact rather than a true exposure-outcome link. Even so, the study cannot prove that lithium in water caused the observed diagnoses. County-level exposure estimates cannot measure the lithium dose received by any one pregnant person, and diagnostic codes cannot fully distinguish incident disease from prior recognition or differences in specialty access.

A second sentence deserves emphasis: the signal emerged in children, not adults.

That age restriction is clinically meaningful because it points toward fetal programming rather than cumulative lifetime exposure. Accessory pathways are congenital structures, so a maternal exposure window is biologically more relevant than postnatal exposure alone. The analysis therefore fits a developmental model in which low-level environmental lithium, if truly causal, could influence the persistence of conduction tissue that should normally regress. In practical terms, the question is not whether lithium acts like an antiarrhythmic drug in this setting; it is whether trace environmental exposure may intersect with cardiac morphogenesis at a vulnerable stage.

From a bedside perspective, the findings should not be overread as a reason to attribute an individual child’s WPW diagnosis to water exposure. The study does not provide patient-level pregnancy histories, direct measurement of maternal lithium intake, or pathway mapping from electrophysiology studies. Still, the association is strong enough to justify closer scrutiny of environmental exposures in congenital arrhythmia research, especially when a potentially modifiable factor is involved. A signal that persists above a health-based screening limit will attract public health attention because it bridges environmental monitoring and pediatric electrophysiology.

The most concrete next step is also the most demanding one: prospective work that measures maternal exposure, fetal timing, and clinical phenotype in the same individual. Until that happens, the present data remain an ecological association, but one with replication, internal validation, and a biologically credible developmental frame. The question now is how often environmental lithium exposure reaches the fetal heart during the narrow window when accessory pathways are being set in place, and whether that exposure can be reduced without unintended consequences for broader public health.


Reporting by Dr. Aishwarya N., MBBS for Medicine World Foundation, with additional data provided by Journal of Clinical Investigation.