Integrating data science for the analysis of early years health inequalities: a Birmingham, UK case study

Background

Early years health and development analysis is a key component of current government strategies, as it represents one of the main determinants for early intervention. A substantial body of research has applied data science methods—such as global cognitive scores, multilevel linear regression, stepwise linear regression, and logistic regression—to model risk factors for individual children. However, limited research has explored how socioeconomic status affects young children’s health and development, or how this correlation can be accurately measured.

Methods

In collaboration with Birmingham City Council’s Public Health, Education, and Skills directorates, Birmingham Community Healthcare NHS Foundation Trust, and the University of Nottingham’s School of Health Sciences, we conducted a comprehensive review and analysis of early interventions across Birmingham communities. The aim was to compare health and social deprivation indices to identify the data required to understand correlations between indicators from both domains, supporting evidence-based intervention planning. Using Pearson and Spearman correlations, factorial analysis was performed to identify deprivation variables most relevant to studying the relationship between socioeconomic deprivation and early health and developmental outcomes.

Results

The analysis revealed significant correlations between socioeconomic factors—such as income deprivation, universal credit, and unemployment claimants—and early developmental outcomes, including obesity and the proportion of children performing below expected levels in communication, fine motor, gross motor, personal-social, and problem-solving skills.

Conclusions

The findings highlight concerning public health patterns: a weak but notable correlation between socioeconomic indices and indicators of child and family poverty, linked to wider determinants influencing childhood obesity and poor school readiness at ages 2 to 2.5 years. These results underscore the need for stronger emphasis on understanding and addressing the relationship between local environmental conditions and public health outcomes.