In the last post, I found a 6th order polynomial relationship between health outcomes from County Health Rankings and COVID Mortality. The health outcomes ranking is a composite of length of life and quality of life rankings. A description of their model can be seen in the video above. First I will look at the length of life correlations with COVID mortality.
Length of Life
The graph above shows a strong 6th order polynomial association between the length of life rankings and COVID mortality. The linear association was also significant accounting for 45% of the variability but the polynomial equation accounts for 93%. Years of Potential Life Lost (YPLL) determines the length of life rankings. YPLL is the number of years lost of someone dies before age 75. For example, if someone dies at age 25 they have 50 years of potential life lost.
A similar pattern with a 6th order polynomial is seen in the above graph where YPLL replaced the length of life ranking. The y axis has been truncated at zero to better show the data points. The linear model explained 55% of the variability while the polynomial explained 91%.
Quality of Life
I next looked at the quality of life rankings and COVID mortality. The linear model was not significant accounting for 28% of the variability. The 6th order polynomial shows a different patter but better explains the model with 97% of the variability. Quality of life is a composite of four statistics: % in fair of poor health, # of poor physical health days, # of poor mental health days, and % of low birthweight babies.
It seems as though COVID mortality is being driven by length of life. Subsequently, the next step will be to see which quality of life statistic is a good predictor of COVID mortality.