Serum Calcium and Cholesterol Are Associated in Mice

During the pandemic, our lab focused into several secondary data analyses. One of these took advantage of publically available data provided by Gary Churchill's laboratory at The Jackson Laboratories looking at what predicts variation in cholesterol in genetically diverse mice.

As part of his research into precision nutrition, Cody Cousineau was interested in this question and started by performing a machine-learning approach integrating data from over 800 mice utilizing a variety of physiological, morphological and clinical traits from male and female mice on both normal chow and high fat, high sucrose diets. Using a regression tree approach he input these 165 phenotypes and performed the analysis. The first two covariates identified were diet, and cholesterol both of which were expected.

The third predictor was calcium, which was somewhat surprising. We found that approximately 22% of the variance in total cholesterol was associated with calcium levels, a much larger proportion than we had anticipated. This relationship was true across both diets and sexes, suggesting an independent predictor that could be valuable for predicting and treating those with high cholesterol and increased cardiovascular risk

Check out the publication here and a short video describing this study below: