Economist Lars Christensen analyzed the variation in COVID-related deaths between the US States in September and his analysis led him to believe that vitamin D is the principal factor that explains the disparity in deaths across states.
Christensen uses the African American population in each state as a proxy for vitamin D deficiency since as many as 80% of African-Americans suffer from vitamin D deficiency.
Uses the same metric, I used COVID data from NCDHHS (retrieved on November 6, 2020) and census data on the African-American population across North Carolina counties to see if Christensens hypothesis could similarily explain the variance in deaths within the state.
When analyzing the mortality rate and the share of the black population in each county in North Carolina, there is a moderate correlation between with pearson R value of 0.514
Running a simple regression, the percentage of total population that is African American is a signficiant explantory variable of county mortality rates. Graphically, we can see a linear correlation between the two.
I created a spatial analysis that shows that geographically shows this observed relationship. We see that many counties in north-east part of North Carolina, outside of Raleigh & Wake County, we have a particularily high level of mortality. These counties are one of the more rural parts of the state and so the high African American population in those counties best explains mortality rates in the region
In the western side of North Carolina, we generally see the lowest mortality rates which happen to coincide with a low population level of African Americans.
Christens vitamin D thesis seems to hold up at this more micro-level analysis. Health officials in North Carolina and nationwide should consider policies that address vitamin D deficiency in the fight against the pandemic.
Github: Covid in NC