Listing "risk factors" without quantifying them is useless waste of readers' time, but even then, "diet" is only one of eight listed, with three others being the obvious ones - alcohol, smoking, and low physical activity/obesity (arguably those should be two separate ones).

-

The chart you linked only talks about incidence ratio, and is more than adequately explained by improvements in access to tests, quality of tests, as well as improvements in healthcare in general, as people don't suffer and die today from what they did up to few decades ago - or anything else, really, since the world has been steadily improving across the board in every dimension.

In fact, non-linear effects of population growth alone could explain that chart: people talk more, including about colon cancer, so over time, more people in the population with access to testing would go test themselves after being made aware of the potential problem, biasing the sample.

Or, more fundamentally, the fact that medicine graduated from voodoo to proper science only around 100 years ago, would explain it too, because we're less than a century into doing proper studies about anything at all.