Yes, while I like the article, it kind of distorts the evaluation in the original paper, where they don't only show the trend line but also it's 95% confidence interval (which very obviously significantly widens with age).