Spatio-temporal patterning of small area low birth weight incidence and its correlates: a latent spatial structure...

Area of Science:

• * Public Health
• * Biostatistics
• * Epidemiology

Background:

• * Low birth weight (LBW) is a significant public health concern and a key indicator of population health.
• * Understanding the spatio-temporal patterns and influencing factors of LBW is crucial for targeted interventions.
• * Conventional modeling approaches often fail to capture the complex spatio-temporal dependencies in LBW data.

Purpose of the Study:

• * To analyze spatio-temporal variation in low birth weight (LBW) using flexible latent structure models.
• * To identify key risk factors and unobservable spatio-temporal effects associated with LBW.
• * To compare the performance of the proposed modeling approach against conventional space–time models.

Main Methods:

• * Employed flexible latent structure models for analyzing county-level LBW data in Georgia and South Carolina.
• * Incorporated analysis of known covariates and explored unobservable spatio-temporal components.
• * Evaluated model fit using goodness-of-fit metrics.

Main Results:

• * The proportion of the Black population was identified as a significant positive risk factor for LBW.
• * Higher income levels were found to be a negative risk factor, associated with lower LBW rates.
• * Two dominant residual temporal components were estimated, highlighting temporal dependencies.
• * The proposed latent structure models demonstrated a superior goodness-of-fit compared to conventional space–time models.

Conclusions:

• * Flexible latent structure models offer a more comprehensive approach to understanding LBW spatio-temporal dynamics.
• * Socioeconomic factors, specifically race and income, play a significant role in LBW.
• * The findings underscore the need for place-based and time-sensitive public health strategies to reduce LBW.