From the New York Times
- is this all we have to work with re: their study? https://joovv.com/blogs/joovv-blog/the-ketogenic-diet-and-red-light-therapy
- no control group = SAD diet & no light therapy? control group still receives an intervention (keto diet)
- how much of the change in hormone levels can simply be attributed to cleaning up their diet?
- no way to blind participants to their diet. how would you blind red light therapy?
- results indicate "The preliminary results reflect approximately 50% of participants who completed the 12-week intervention." and looking at their results table, only 6 women completed the test so far, so the entire group is only n = 12? or has there been a bunch of study losses from participants who didn't complete?
- no summarized / aggregate data for control & experimental groups shown?
- not stratifying by confounders like age? age not shown for female participants.
- not taking menopausal status into consideration during analysis? or, only 3 post-menopausal & 3 pre-menopausal participants so far? small groups.
- as Dr Armory pointed out in the NYT article - how do hormone levels fluctuate throughout the day / months? how much variability is there between repeated measurements?
- estradiol increased for pre-menopausal participants but decreased for post-menopausal individuals? how does age of these two groups influence trends?
- not sure i'm understanding the difference between the two groups shown re: testosterone levels, both are labelled as control groups (?). only four participants in the first group and only 8 in the second group? small sample sizes.
- no confidence intervals shown but likely some overlap in groups, particularly re: estradiol measurements, since theres quite a bit of variability, especially so with estradiol (half saw increases, half saw decreases).
- obvious conflict of interest going on, and prob other issues I'm not aware of from my lack of subject matter knowledge on hormones, diet etc.
You've been pretty systematic in breaking down the limitations of this research. An RCT with 50% follow up is shocking. It's pretty amazing how in spite of these limitations, prominent media figures will run with it.
That's good timing...the Royal Society of Medicine are hosting a webinar with Ben Goldacre on Wednesday!