Investigators reported an association between routinely measured maternal white blood cell (WBC) counts in pregnancy and subsequent diagnosis of autism spectrum disorder (ASD) in offspring. The team used electronic health records from Clalit Health Services covering 303,234 singleton live births from 2011 through 2021. Within that cohort, 8,106 children (2.7%) received a clinical ASD diagnosis. In multivariable logistic regression models, elevated maternal total WBC counts—defined in this analysis as above the 97.5th percentile—were associated with higher odds of ASD in the child (adjusted odds ratio [aOR]=1.59; 95% CI, 1.43–1.75). The authors reported that these routine hematologic markers may serve as accessible signals of maternal immune activation in population-level risk assessment.
The study team observed that the association with elevated maternal total WBC counts was most pronounced during the 1st trimester. Reported estimates gave an aOR of 1.64 (95% CI, 1.45–1.85) for 1st-trimester elevations. The source did not provide additional mechanistic data linking trimester timing to developmental windows, nor did it detail the exact timing within the 1st trimester for individual samples. The paper confined its exposure definition to routine tests performed during the 1st and 2nd trimesters; tests outside those windows were not described in the abstract.
Data indicated similar positive associations for elevated counts of specific leukocyte subsets. The investigators reported comparable signals for neutrophils, lymphocytes, and monocytes alongside total WBC. The preprint did not report subset-specific adjusted odds ratios in the abstract text beyond stating that these subsets showed similar associations. The source therefore did not provide subset-by-subset effect sizes in the abstract available for this report.
To address potential familial confounding, the authors ran matched-sibling analyses that included 2,894 sibling pairs discordant for ASD. Many of the associations observed in the primary cohort analyses were attenuated in these sibling comparisons, which suggests that shared familial factors account for part of the observed relationship. The team also conducted chronicity analyses among multiparous mothers (n=72,102). In that analysis, mothers with persistently elevated WBC counts across multiple pregnancies had increased odds of offspring ASD (aOR=1.49; 95% CI, 1.08–2.01). Those chronicity results indicate that repeated maternal WBC elevation, rather than a single isolated high value, may carry additional risk signal.
The report described a retrospective, population-based cohort design using Clalit Health Services records. Exposures were maternal total and differential WBC counts obtained from routine prenatal testing during the 1st and 2nd trimesters. High and low exposures were operationalized as above the 97.5th percentile and below the 2.5th percentile, respectively. The primary outcome was clinical ASD diagnosis documented in the health system. The authors used multivariable logistic regression models as their primary inferential tool and supplemented those with sibling-matched analyses and chronicity analysis in multiparas to probe robustness and familial confounding.
The investigators stated that the dataset supporting these analyses is held by Clalit Health Services and is not publicly available because of licensing restrictions and patient confidentiality rules. The work received ethical approval from the Soroka University Medical Center ethics committee, as noted in the author declarations. The article is posted as a medRxiv preprint and has not been peer reviewed; the authors and medRxiv explicitly caution that the findings should not be used to guide clinical care until formal peer review and validation are completed.
The abstract itself notes several design features relevant to interpretation. The matched-sibling analyses attenuated many associations, indicating potential residual confounding by familial or shared environmental factors. The preprint did not present mechanistic data that would clarify how transient versus persistent leukocytosis might influence neurodevelopmental trajectories. The source also did not provide publicly accessible underlying data: access is restricted by Clalit Health Services. Finally, because this report is a preprint, peer review may alter interpretation, and the authors and platform caution against clinical application prior to validation. Key unanswered questions documented by the authors include the causal role (if any) of elevated maternal WBC in ASD pathogenesis, the clinical utility and specificity of routine WBC measures for individual risk prediction, and whether interventions that modify maternal inflammatory states would change offspring outcomes. These items remain to be addressed in further peer-reviewed analyses and prospective studies.