it seems this got flagged for all sorts of reasons, but one significant observation is that they threw away data for people who eat more than 5000 kcal/day as "

it seems this got flagged for all sorts of reasons, but one significant observation is that they threw away data for people who eat more than 5000 kcal/day as "implausible". Those people are likely to be among the most unhealthy, and also the least likely to skip meals, moreover if skipping a meal caused you to go from 5500 to 4500, this would create a pretty significant effect in the data.