Archived

This content is available here for research, reference, and/or recordkeeping.

Date Available

4-21-2027

Year of Publication

2026

Document Type

Graduate Capstone Project

Degree Name

Master of Public Health (M.P.H.)

College

Public Health

Department/School/Program

Public Health

Faculty

Tianjun Lu

Committee Member

W. Jay Christian

Faculty

Yisi Liu

Abstract

Background

Pediatric brain tumors (PBT) are the most common solid tumors in children, with rising incidence and geographic variation in Kentucky. Ambient fine particulate matter (PM2.5) has been proposed as a potential environmental risk factor. This study evaluated the association between PM2.5 exposure and PBT incidence and examined spatial and temporal disease patterns.

Methods

A population-based ecological study was conducted using Kentucky Cancer Registry data for individuals aged 0–19 years diagnosed with PBT from 1995–2023. County-level PM2.5 exposure estimates were derived from high-resolution geospatial models. Associations were assessed using Poisson and zero-inflated Poisson, with and without adjustment for sociodemographic covariates. Correlation and spatial cluster analyses (SaTScan, QGIS) were performed.

Results

The overall observed association was inconsistent between PM2.5 and PBT incidence across all models (unadjusted IRR ≈ 0.97, 95% CI: 0.919–1.018; adjusted IRRs: 0.938–0.941, 95% CI: 0.838–1.050). Correlations were weak (Pearson r = -0.22; Spearman r = -0.18). Spatial analyses identified clusters of elevated incidences in north-central and central-eastern Kentucky.

Conclusions

PM2.5 exposure was not associated with PBT incidence and spatial patterns did not show a clear correspondence with PM2.5 levels. Limitations include the ecological design restricting individual-level inference, potential ecological fallacy, and county-level exposure that may not capture relevant exposure windows. Although these findings suggest no evidence of association, future studies with further stratified PBT subtypes, improved exposure assessment and individual-level data is needed.

Funding Information

This publication or project was supported by UK-CARES through Grant P30 ES026529. This publication or project was supported by the UNITE Research Priority Area at the University of Kentucky and the National Cancer Institute through the Cancer Center Support Grant P30CA177558. 

Download

Available for download on Wednesday, April 21, 2027

Contact Author

Share

COinS