alt Hacker News> many studies of nutritional epidemiology that try to link dietary exposures to disease outcomes are founded on really dodgy data
I wonder if the data are always skewed in a particular direction. For example, do people typically underreport junk food and overreport salads? Or do they omit entire meals? Or snacks?
While far from being a potential silver bullet, I do wonder if continuous glucose monitoring could help with this. Your food log shows you didn't eat anything between noon and six in the evening but the glucose monitor shows a spike at 2PM? Your diet app could ask if you forgot to log something around that time. Maybe you want for a long walk aside while it was cold and that was the cause. Unless the question is asked, the tracking data for that time period will be questionable.