Exploring the Relationship between Fluorosis-Affected Areas in Chhindwara District and Regression Models: Bivariate and Multivariate Analysis of Water Fluoride Levels and Dental, Skeletal, and Non-Skeletal Fluorosis

<title>Abstract</title> <p>Fluorosis is a significant public health burden in fluoride-endemic tribal districts of central India, yet validated predictive models for fluorosis risk among tribal populations remain scarce. This primary cross-sectional epidemiological study, conducted by trained field investigators from ICMR–National Institute of Research in Tribal Health (ICMR-NIRTH), Jabalpur, assessed 60,595 individuals across 30 villages in Chhindwara District, Madhya Pradesh (2017–2019), using probability proportional to size (PPS) cluster sampling. Water fluoride levels (WFL) were categorised as &lt; 1.5, 1.5–3.0, and &gt; 3.0 ppm and measured by ion-selective electrode method. Significant dose-response relationships were observed between WFL and moderate dental fluorosis (χ²-trend = 73.64, p &lt; 0.0001), severe dental fluorosis (χ²-trend = 19.13, p &lt; 0.0001), knock knee in children (p = 0.008), and crippling fluorosis across both age cohorts (p &lt; 0.007). Multivariate regression identified pain in the stomach (β = 0.445, p = 0.001) and extreme weakness (β = 0.375, p = 0.003) as independent predictors of elevated WFL in adults, alongside farmer occupation (β = 0.243, p = 0.036). Childhood undernutrition worsened progressively with increasing WFL (χ²-trend = 21.49, p &lt; 0.0001). The regression-based prediction models developed provide field-applicable tools for fluorosis surveillance and support targeted defluoridation and nutritional interventions in tribal communities.</p>