The study seems fine but the presentation is terrible.<p>There is no reason to believe there is even an association between diet and Type 1 diabetes. In the absence of evidence for even this basic fact, simply finding a correlation between a whole bunch of foods and Type 1 diabetes seems fraught with all sorts of issues beyond even the fact that the study design can simply find correlations and not causations any ways.<p>You can also see the biases that the news likes to present. The correlation is not causation proviso is barely mentioned, and most of the article shows it as a causal link and uses the speculation that I may be pesticides to justify this causal link, when it comes to bananas, oats and yoghurt.<p>However, the last couple of paragraphs which mention the correlation between staying up at night and Type 2 diabetes (a disease where lifestyle factors are known to have a causal effect), the fact that it’s a correlation is mentioned almost immediately and forms the bulk of the reporting.