The study does not make me concerned that elevated Vitamin D level causes cancer. That is probably just a fluke of doing too many subanalysis tests. If you do 10 tests with a p=.05 you will have a 50% false positive rate even higher if you are looking for both a positive and negative change.
The interesting thing is that in the whole test and looking at all sub groups, it does not look like low Vitamin D levels is a statistically significant risk factor for cancer, even considering the seasonal factor in measurement.
Can’t find any more info on this study. They have a lock on it though it’s not a new publish.