The fight against diabetes in Southern California is more complex than simply monitoring blood sugar levels. Emerging research highlights that lifestyle and social factors such as physical activity and food security play a crucial role in predicting who is at risk. But here's where it gets controversial... the interplay of these variables is more profound and region-specific than many health models have appreciated up to now.
In 2021, approximately 38.4 million Americans—around 11.6% of the population—were affected by diabetes. Notably, minority groups including American Indian and Alaskan Native, Black, Hispanic, and Asian American populations face a disproportionately higher burden of this disease. Given that Southern California has a large Hispanic community, primarily of Mexican descent, this region calls for tailored public health approaches that address its unique social landscape.
While previous studies have tied social and behavioral factors to diabetes, many relied on composite measures that often masked the influence of individual elements. The upheavals of the COVID-19 pandemic—disrupting daily routines and limiting access to healthcare—prompted researchers to revisit and refine these models with fresh data and methods.
This research, titled "Social and Behavioral Factors Associated With Diabetes in Southern California vs. the US," published in JAMA Network Open, aimed to identify specific social and behavioral clues that correlate with diagnosed diabetes. The study compared data from Southern California to national trends, providing crucial insights into regional differences.
The analysis encompassed data from 5,420 census tracts in Southern California, representing about 18.5 million adults, and 62,480 tracts nationwide, covering approximately 253 million adults. Data sources included CDC’s modeled estimates for 2024 at the census-tract level.
To tackle the complexity of factors involved, researchers employed an advanced machine learning technique called extreme gradient boosting. This approach considered 24 indicators—like rates of physical activity, food insecurity, health insurance coverage, and routine medical checkups—and modeled how these variables predicted diagnosed diabetes prevalence. The data was split into training and testing sets to ensure reliability, and explanations for each variable’s influence on the model were derived using a method called Shapley additive explanations.
Remarkably, the models explained over 95% of the variation in diabetes rates between different regions, both in Southern California and across the entire nation.
In Southern California, the most influential factors accounted for about 67% of the model’s predictive power. The key contributors included a lack of physical activity during leisure time (31%), irregular health checkups (14%), absence of health insurance (6%), and food insecurity (5%). Interestingly, binge drinking showed an inverse relationship: higher binge drinking was linked to a slight decrease in predicted diabetes prevalence, which might seem counterintuitive and warrants further exploration.
When comparing regional results to national data, some differences emerged. Nationally, additional significant factors influencing diabetes prevalence included higher obesity rates, the receipt of food assistance programs like food stamps, older age groups (65+), and racial or ethnic minority status—all elements that seemed less prominent or differently weighted in Southern California.
Overall, the average diagnosed diabetes rate was similar—about 11.3% in Southern California and roughly 11.5% nationwide—suggesting that regional behavioral and social factors significantly shape prevalence patterns. Understanding these nuances is critical for developing more effective, localized prevention and intervention strategies.
This research, thoughtfully conducted and carefully analyzed, underscores the importance of looking beyond traditional medical measures. If we recognize and address social determinants like physical activity and food security, we could make a real difference in reducing diabetes rates. But the question remains: how should we balance these social initiatives with medical interventions? Do you agree that public health initiatives focusing on social factors could be more impactful than solely promoting medical treatment? Share your thoughts—your perspective can help shape future health strategies.
