I've spent some time living in Sweden and I think there's a number of important details here that are not sufficiently highlighted that can radically change how to interpret the results. Swedes have a tradition of eating candy once a week on Saturdays. It's incredibly common for them to not have candy all week, then go buy an assortment bag on Saturday, eat it all, and then not have any again for a week. We don't know how this affects health outcomes compared to say, having a bit of candy every day, the glycemic load profiles will be widely different and we don't know if that matters or not.

Fika, a coffee break with pastries, is a very strong cultural phenomenon. The pastries Swedes have are generally of very high quality and many of them also contain a considerable amount of fat, which as we know, helps buffer glycemic load. While it's mentioned in the study in passing, I think it should instead be one of the core focus points, it's hard to assess the effect of this as opposed to plain added sugar.

And lastly, Swedes are very outdoorsy people. On any day of the week, you can see people strolling on nature trails, of which there are plenty, and spending a lot of time active in general. I would not be surprised if a large part of the health outcomes we see in statistics can be attributed to variance in this.

So to summarize, while these results are interesting, I think there are too many hidden influences here that are not controlled for to really generalize these results in any way.