> sunscreen has not been shown to cause any population wide reduction in skin cancer prevalence
Have you accounted for all variables? The moral hazard where people stay out in the sun longer when they're wearing sunscreen, the fact that testing shows there are manufacturers with poor quality control leading to significant differences between claimed SPF and actual SPF, have there been any changes in patterns of time spent outdoors over the years, and how about the fact that different cohorts have different rates and trends?
We know that aging increases the risk of cancer, we also know that repeated instances of UV skin damage make skin cancer more likely.
So, have a look at this graph (Figure 2).
https://www.stats.govt.nz/indicators/occurrence-of-melanoma
The rates in the 25-44 cohort, who grew up when health initiatives about "being sun smart" were introduced to school curricula and promoted in the media are lower and trending slowly downwards.
Whereas the rates for boomers with a lifetime of accumulated UV damage are trending upwards as aging leaves them vulnerable to cancer if all kind.
Endemic vitamin D deficiency is going to need a citation, I think. In my country at least, it's only been noted in children who have very high melanin already, and then dress in our local "tough gangsta" style with baseball caps and hoodies with the hood pulled up over the cap, thus drastically reducing the area of skin exposed.