There are approximately 14 modifiable risk factors for dementia risk of which several are mentioned in the article. including socialisation, eating "healthy" etc. These are well-backed by research. However, little evidence was presented to support those facts .

A well presented short article highlighting the complexity of the situation and the requirement for further studies.

The evidence they present is shaky at best. Clear use of logical fallacies and personal attacks in their writing creates a very unbalance article. While the writing itself is clear, it is extremely misleading. Vaccines are one of the most effective public health interventions ever.

Overall actually a sensible article. There is a lot of evidence that diabetes (both type 1 and type 2) increase risk of dementias. There is lots of crossover between some of the mechanisms that cause damage in both diseases.

While cacao may contain polyphenols that are protective it would be very difficult to get these in sufficient quantity directly from chocolate. There is a small correlation detected in their study with dark chocolate consumption which is not found with milk chocolate. However correlation is not the same as causation. To be fair, the article points out that further randomised controlled trials are needed, which is true. This is exciting research but as always needs to be taken with a pinch of salt and validated in direct mechanistic trials. Please do not start eating 5 bars of dark chocolate a week and think it will decrease your risk of type 2 diabetes. Still the largest modifiable risk factor for diabetes is body fat. Eating multiple bars of chocolate per week makes reducing body fat a bit trickier, although technically still possible if other calorie intake is adjusted accordingly.