Exploring the Relationship Between Foodborne Outbreaks and Climate Change Using Social Media Data in the Years...

Area of Science:

• Environmental health
• Epidemiology
• Digital epidemiology

Background:

• Climate variability influences enteric infections, but research often overlooks gradual changes, focusing instead on extreme weather events.
• Traditional foodborne illness surveillance suffers from under-ascertainment, limiting the quantification of climate-sensitive disease burden.

Purpose of the Study:

• To evaluate symptom-related social media messages as a complementary signal for foodborne illness surveillance.
• To assess if combining social media data with climate anomalies improves characterization of climate-disease relationships.

Main Methods:

• Analysis of Twitter (X) messages from Australia and New Zealand (2016-2021) using symptom and food-exposure keywords.
• Natural Language Processing for data cleaning and symptom-ingestion term co-occurrence for relevance screening.
• Gaussian Mixture Models for spatial clustering, linked to climate anomalies and notified salmonellosis/campylobacteriosis counts via generalized additive models.

Main Results:

• Climate anomalies alone explained minimal variation in notified cases (R²=0.01-0.12).
• Social media message frequency alone showed modest explanatory power (R²=0.22-0.48).
• Models integrating message frequency and climate anomalies demonstrated substantially improved explanatory power (R²=0.87-0.97).

Conclusions:

• Social media-derived indicators, when combined with climate data, significantly enhance the monitoring of climate-sensitive foodborne illnesses.
• This integrated approach offers a more robust characterization of foodborne disease patterns compared to traditional methods alone.
• Further research should validate these findings across diverse platforms and datasets to assess generalizability.