Hacker News Reader

Comment by somesortofthing | Hacker News Reader

I feel like you're giving the people aggressively generalizing results from studies like these too much of the benefit of the doubt. The mice in this study were fed a diet that was more than 2/3 sucrose, compared to an average of ~20% in human diets. The level of attention given to these studies does far more to lower the level of discourse than the people who point out how exaggerated the model used to reach the study's conclusion is.

View on HN

This study, and others like it that use mice, would be a lot more interesting if they bucketed out some ranges for the test variable. For this one, I would have liked to see groups with 20, 40, 60, and 80% sugar diets.

>The mice in this study were fed a diet that was more than 2/3 sucrose, compared to an average of ~20% in human diets.

Honest question: do you know the reason why they do this?

It is not done by accident.

To maximize the possibility of producing a positive result for the hypothesis. And it's understandable that this is done given academia's incentive structure, and you could justify this study's existence as a foundation for further studies, but real damage is done to the discourse around these topics when people pick up on these results and generalize them across both input size and species lines. This study alone tells us virtually nothing useful about whether or not "High-sucrose diets contribute to brain angiopathy and higher brain dysfunctions" in humans, as the post's title implies.

>To maximize the possibility of producing a positive result for the hypothesis.

No. The reason is to approximate long-term exposure. There are obviously (big!) problems with that, but we don't have a better way of "speeding up time".

You really should try to understand things before you attack them with tired caricatures of legitimate arguments.

Reply on news.ycombinator.com