Spatial distribution of SARS-CoV-2 incidence, social inequality, housing conditions, and density in South-Eastern France: keys for future epidemics

Introduction: Early in 2021, the SARS-CoV-2 incidence rate was higher in the East than in the West of the Alpes-Maritimes district in France. What was the impact of social deprivation, household overcrowding and population density per km² on this difference in incidence rate?

Methods: Cases were defined as persons with a first SARS-CoV-2 positive test detected between 04/01/2021 and 14/02/2021. We studied the « French Deprivation index » (FDep), rate of overcrowded households and population density/km². These indicators were compared between East and West and a Standard Incidence Ratio (SIR) and an Incidence Rate Ratio (IRR) were calculated for each indicator. The link between the incidence rate and the socio-economic variables per census blocks (IRIS) was analyzed with a GLM model. Lastly, a stepwise method was used to determine the East/West incidence thresholds for which an association was observed between the incidence rate and these three indicators.

Results: Among the 473 census blocks, 25,400 cases were geolocated among whom 23,867 not residing in nursing homes nor in long-term communal accommodation. Census blocks in the East included more overcrowded households (p = 0.009) and a higher population density (p < 0.001). In this area, the SARS-CoV-2 incidence was significantly higher in the most deprived census blocks (IRR = 1.614; 95%CI [1.530-1.703]), with a higher rate of overcrowded households (IRR = 1.583; 95%CI [1.508-1.663]) and higher population density (IRR = 1.062; 95%CI [1.023-1.102]). No difference was observed in the West. According to the GLM, in the East, the incidence rate was associated with the FDep index only, while no association was observed in the West. In the East, the association with FDep appeared for an incidence threshold of 210/100,000, while no threshold was identified in the West. Rates of overcrowded households were 310 vs. 370 and population density rates were 260 vs. 400 in the Eastern and Western areas, respectively.

Conclusion: Our results demonstrate the benefits of conducting a spatial analysis of socio-demographic and medical data. At the start of an emerging infectious agent-related epidemic, while surveillance is not yet operational, initial prevention measures could prioritize targeting populations according to their socio-demographic characteristics.