Improved entropy-CRITIC population model based on temporal and spatial variability: Construction and application in...

Area of Science:

• Environmental Science
• Epidemiology
• Data Science

Background:

• Wastewater-based epidemiology (WBE) faces uncertainty due to population shifts.
• Existing population models have limitations in accuracy and reliability.
• Accurate population estimation is crucial for WBE data interpretation.

Purpose of the Study:

• To develop and validate a novel evaluation model (EC model) for precise population estimation in WBE.
• To assess the EC model's performance against traditional methods using Chinese city data.
• To apply the EC model for estimating lifestyle-related compound consumption and hypertension prevalence.

Main Methods:

• Constructed an EC model using an evaluation model framework and data characteristics.
• Aggregated weight coefficients from 16 Chinese cities into a national model.
• Validated the EC model against census population data and analyzed its variability, bias, and SRMSE.

Main Results:

• The EC model showed a strong correlation (r² = 0.98) with census data, indicating precise population dynamics.
• Low variability (RSD = 9.73%), minimal bias (-1.12%), and low SRMSE (14.75%) confirmed reliable population estimation.
• Northern China exhibited higher consumption of lifestyle compounds and greater hypertension prevalence (26.96%) than the South (16.01%).
• Hypertension positively correlated with alcohol (r=0.52) and nicotine (r=0.55) consumption.

Conclusions:

• The EC model offers a more reliable approach to population estimation in WBE compared to traditional models.
• The model successfully estimated lifestyle-related compound consumption and hypertension prevalence across China.
• Incorporating representative datasets and novel algorithms, the EC model enhances WBE strategy implementation reliability.