Misunderstanding the Data on Diet, Exercise and Mortality
ฝัง
- เผยแพร่เมื่อ 2 มิ.ย. 2024
- We’ve got another nutrition study making a splash in the media. This time we’ve got a combination of diet and exercise claims, complete with all our favorite things: Observational study, limited self-reports, and way too many factors to control for.
Related HCT episodes:
Nutrition Studies Are Terrible: • Nutrition Studies Are ...
Be sure to check out our podcast!
• Podcast
Other Healthcare Triage Links:
1. Support the channel on Patreon: vid.io/xqXr
2. Check out our Facebook page: goo.gl/LnOq5z
3. We still have merchandise available at www.hctmerch.com
4. Aaron's book "The Bad Food Bible: How and Why to Eat Sinfully" is available wherever books are sold, such as Amazon: amzn.to/2hGvhKw
Credits:
Aaron Carroll -- Writer
Meredith Danko - Social Media
Tiffany Doherty -- Writer and Script Editor
John Green -- Executive Producer
Stan Muller -- Director, Producer
Mark Olsen - Art Director, Producer
Vyacheslav Dumchev/iStock/Getty Images
Storyblocks/WeAre
CarlosGaw/Getty Images
alle12/Getty Images
luoman/Getty Images
GMVozd/Getty Images
SolStock/Getty Images
nazar_ab/Getty Images
gbh007/iStock/Getty Images Plus
dotshock/Storyblocks
Hangtime/Storyblocks
Glad to see the lack of control for class/wealth pointed out here, as that was my first question when I saw those studies going around. In many ways, the "healthy" diets that fitness and wellness shills push here in the US are too expensive (either in money or time) for poor or working class people. Unfortunately, health in the US has several paywalls.
Whenever I'm reading health news in my news app, I have conditioned myself to ignore the click-baity headlines and carefully read the full article. There's usually an 'a-ha' moment when I find the true (but often un-sexy) substance of the study buried in the article. It takes time and patience, but I want to be informed, not just dazzled - and I want to avoid passing on misinformation to others. Thanks to this channel for helping us all do the same. As Hank Green once said, "Separating the facts from the fertilizer."
I was a participant in a research trial where they would get me to fill an online survey, but then also call me randomly to ask my pain and excersize level from 6 months ago. Absolutely ridiculous to think anyone would remember at all and drastically increasing participant burden with annoying phone calls when I could tell the data was corrupted was so upsetting
There is no way people were remembering their diet accurately. I've been counting calories to lose weight and if I wait till dinner time to record lunch time calories, I've already forgotten half of it.
Snacks throughout the day are even worse, they are basically invisible to our brains.
Related note: i once was a participant in a nutritional diabetes risk study. They took a blood sample, urine sample, and asked me to fill out a survey. The survey was a list of probably 1000 foods and asked me how many servings of each I eat in a YEAR. There's absolutely no way anyone answered that survey correctly. At the end, they tried to refer me to eating disorder treatment because according to the survey I was only eating 800 calories a day. With common sense, I eat considerably more than 800 calories a day. It's just incredibly difficult to estimate how many bananas one eats per year, and incredibly boring to answer for over 1000 individual food items.
@@ThaliaIrwin That's the biggest flaw with epidemiology studies.
@@ThaliaIrwin How the hell did such a terribly designed study even get funding?!
@@SovietReunionYT Don't know, but I'll throw out the possibility that it would be because its costs were low.
Presuming the OP was accurate in describing it, it was nothing more than blood sample, urine sample and survey with clearly very, very little effort put into follow-up and follow-through. Compare what costs that would incur with a study that would give at least 'ballpark'-level accuracy. Perhaps something like...
- 6 months (26 weeks) total time
- Samples: daily for Week 1, then 2 samples for Week 2, then 1 sample per week for Weeks 3 & 4, then samples taken Week 8, 12, 19, and 26
* All samples taken with 10+ hours since last meal (eg: morning before breakfast)
== Total samples: 15
- Survey: daily diet log for one month (Weeks 1-4), weekly diet survey for the 3 days prior during Weeks 5-8, then one on Week 12, 19, 26
* All surveys purely report in by user (ie: doesn't show participants composite information like macronutrient or calorie totals)
== Total surveys: 35; reflecting 49 days over 26 weeks
The sample-taking costs will be somewhere around ~15x more and the effort to process the survey data somewhere around ~50x more. Plus there would be associated costs with participant drop-off, more likely need to reimburse participants for their daily/ near-daily effort, etc., etc., etc.
All of that and costs incurred versus 1-time sample, 1-time ultra-massive survey, minimal data-processing and follow-up.
... not to mention that any fish meal from the UK Biobank database is more likely to be fish&chips than your illustrated grilled salmon and asparagus 😅
I had salmon and asparagus last night! Lol.
The wrap against seed oils has blown up into urban legend with little evidence to support it. Watch Plant Chompers videos on oils and you’ll see what I mean.
Love the no BS approach this channel takes. 🤍
Lol plenty of BS on this channel though
@@Only1Jabstab Nah. It's just straight reporting on the results and methodology of peer-reviewed studies. Any inaccurate information comes from the studies themselves (which is inevitable as science progresses).
There's no over-editorializing, no wild jumping to conclusions - no BS.
@@Jesse__H if they are not doing their due-diligence and they continue following incorrect and misleading studies, they are complicit. Convenient that the science changes so often these days and everything is just forgiven because of it. Nevermind the millions of lives that have been lost, families destroyed, billions of dollars wasted, and countless businesses gone under.
But hey, you know, science and stuff.
People are too obsessed with a single magical answer to things.
The fact that tobacco use wasn't asked about/controlled for is weird.
That's why im forever weirded out by my country, illegalizing weed in ANY form (including for health purposes) yet forever makin tobacco (cigs) prices Extremely cheap.
@@homeland1128 Without even knowing your home country, I'll make the educated guess that the reason a lot of Tobacco corporation-funded lobbying.
@@FlyingDwarfman not surprised if that's what was happening cuz not only corruption cases sometimes are quite extreme here but also recently they're reducing sentences and bail money for corruption as part of the new laws, i hope some of that i mentioned are wrong but, welcome to indonesia i guess
I am subscribed but I haven't seen your videos in my feed for years now. Hank green mentioned this channel and I was reminded to check it out. You guy... click on the bell button. The algorithm sucks major time.
Thanks for this insight, I wouldn't have noticed the issues in taking the studies as hard evidence without looking much closer.
A very interesting discussion. Thanks for sharing.
Not to mention the project only followed people over 3 years. That's short even on the short timespan of a human life.
The methodology employed on this very limited data set leaves a lot to be desired. How can you even draw conclusions on lifespan on a 3 year follow up study?
I know a similar dataset was (or maybe is still being) gathered in a Rotterdam where participants were followed for at least over a decade and often from before birth (which takes the mother's decisions before and during pregnancy into account). Obviously people fell off because the intention was to follow people for their lifetime and measure a lot of things including diet and a whole bunch values from blood tests. I don't know if they're still doing that project and it would probably have even more confounders (but they actually note down a lot of them so you can control for them, as opposed to here). People are getting PhDs out of analyzing the gathered data so it's A LOT.
Thanks for this video. I heard someone else talk about this study, but it wasn't as informative as this one.
Fish gives me anaphalaxis. According to this study I would have died early no matter what. *eyeroll*
Can you do one on oral minoxidil for baldness please? NY times had a splashy article recently
This video brings up an interesting ethical principle of doing good. Particularly the discussion on diet and exercise decreasing mortality is a good headline that I imagine would help people to make some smarter decisions about their lifestyle. In this specific headline, the outcomes of publishing a statistically questionable study inspire people to make better decisions about their health. The problem I see is that many of the health-related studies we see in headlines and tabloids are founded on data acquired from research that lacks the necessary design to have appreciable, statistically significant results. Fad diets and weight loss pills are the first things that come to mind. My hope is that channels like this and others like it continue to review data and research methods and present them in an honest, easy-to-understand way. The general population doesn't have extensive education and practice evaluating statistical significance and research design so having access to insight from a medical doctor is invaluable when we live in a world where we are bombarded by so much information about the "next best thing". Keep it up, Dr. Carroll!
The purpose of news media is to sell copy (for print media) or sell ads (for digital media). "Informing the public" comes a very poor second
Thanks for this one. In my case diet really helped me to first, be committed,
second to not overeat or snack in between meals and most important it helped
to lose that belly fat faster and burn calories (while doing some moderate exercise)
Personally, I found a diet plan with universal diet guides; you can find them on google,
or any other diet guide that comes with lots of motivation and inspiration will work!
I heard this quote "there is freedom in commitment" and would really love to try to go on a new diet since the past ones didn't work 😅
Got it... People who make up stories about their healthy diets and exercise live longer…
Surgeon admiral Sam remains a much-acclaimed boss.
Indeed!!
Engagement
Yet another observational study.
It funny because people always complain that primary care doctors never know anything about nutrition or specific diets and its like.... because there is no good evidence on any of it! We all have common sense and a general idea of what is "good" and "bad" when it comes to food. That is as simple and as complex as you need to make it.
Nope. Applying "common sense" to nutrition is how you get horrible fad diets. Hard data is needed, and in the absence of any, the best course of action is to eat whatever you feel like eating. 99% of people never lose or gain weight long-term anyway. So diet should only take into account the actually known to be important things, like trans fats being horrible and easily digestible carbohydrates causing diabetes. Also take Omega3 supplements if you dont eat fish. Pretty much all other nutrition advice is unsubstantiated bullshit.
It’s frustrating
Media loves to use headlines like this as click bait, not to educate or to enlighten their audience. The New York Times especially loves these low quality studies and seldom speaks to the weaknesses each offer. I continually tell family to simply ignore all diet, nutrition and exercise studies for this very reason and instead just use common sense for all three.
Same results but less stress and less effort
So a vegetarian/vegan couldn't be a high diet because there's no fish?
the definitions for "low", "medium" and "high" quality screams inherent bias, which is a huge red flag on it's own, even disregarding all the flaws in data collection and statistical analysis...
I like how you said “Sam”, instead of how you usually say it, “Sam”. Nice. Variety is the spice of life!
Can you do an episode on accessing consistent and appropriate mental health care in the US, please?
My daughter has BPD, has aged off my insurance and has had more than 3 therapists and 3 psychiatrists in the last year because of insurance.
Because BPD is not hugely common, and because the reason she probably developed it (growing up hiding from my psychopathic mom when we made our break from her) it's difficult to find care that's really appropriate to her situation.
It would be great if she had a team that transcended insurance and would stick with her for a few years for continuity. It would also be great if there were online support groups for people with BPD, friends and family of people with BPD, and those recovering from psychopathic relationships.
I can find tons of 'interesting' research information on all of those subjects but none on how to get actual help with the situations.
Is anyone even researching mental health treatment accessibility?
I tell people all the time if weigh loss was as easy as hey fatty, put down the fork and go jog a mile and we're good to go- the fitness and diet industry would not be worth the billions it is.
Love how being vegan automatically exempts you from having a high quality diet 😂
If you make $13 an hour and rent is $1500, acquiring fresh fish or produce to eat is not a priority.
Flawed methods, in a health-science study!? 😮🙃
I couldn't understand shit what the f**k was that all about.
While the study quality wasn't fantastic there's still enough data to suggest that both a good quality diet and exercise increase lifespan and quality of life. And what i mean by a good quality diet is simply a diet with enough fiber, zinc, magnesium, omega 3 fats and a variety of polyphenols and phytonutrients and where you don't overeat calories chronically while also not eating too much saturated fat and apperently heme iron (wich is mostly in red meat) but i'm not gonna emphasise the heme iron that much because there's very little mechanistic data on how it harms you but a ton of good quality outcome data that suggest it probably does harm you in high enough doses.
That’s a tuuuruuubuuuul study. One has to wonder how it got through peer review when ye olde sniff test fails this bad
Before we criticize the media coverage (entirely driven by click bait) we should do a study about harm of such badly reported studies.
If adviced to "exersize and keep a quality diet" is this harmful or not? Granted, its not the right conclusion, but it probably does some good?
Lets ask "Do readers of improperly reported health studies in print and digital media have worse all cause mortality results, than readers of properly reported health studies?" And we take a controle group of people who dont read this at all. All cause mortality for 20 years, in ages 18-70, controlling for gender, income, education level, alcohol consumption and smoking habits. Readership recorded by user specific data for digital print, and yearly self report survey for print media.
I predict 0 significant difference between the two intervention groups and between intervention groups vs control group.
Suggesting in the discussion section that likely no significant groups inform their actual eating habits due to media information.
I predict statistically significant worse outcomes for the bad advice group, and possibly also for the good advice group, due to the psychological damage of frequently feeling shame over not being able to follow the advice that they're being told will make them healthier.
@@SovietReunionYT Aaaah that reminds me of a surgery study with prayer as intervention. The prayed-for group recovered worse and had more complications, dispite being randomized. Most likely due to the "performance pressure" patients experienced.
However - I would disagree with your prediction, because my study only sends a yearly survey and this isnt frequent enough to guild trip the study participants.
But to be fair, thats why the study needs a control group.
Hmmm... thinking about it, there probably is a study about how often one should survey a multi year group to optimize participation numbers and make significant measurements... What time if year should be send out the survey? Or do we also spread this out?
Sad that so much time and money is wasted on studies that are so inherently flawed. I’m no expert, but it’s easily apparent why this study has marginal value. Is this what researches do to justify their existence?
Honestly, the time and money was probably one researcher's afternoon. They used an existing data set.
Sure, the study wasn't perfect, but the good Doctor should admit that its recommendations are generally valid. As has been said many times "The perfect is the enemy of the good."
That's how the video ended! Maybe you commented before you got there
If I forecast the weather by flipping a coin and I get it right, I know you will have my back
The recommendations are only valid because we generally know them to be true beforehand. The study is so flawed that it adds very little if anything at all to that knowledge.
The study is about as useful as telling someone to use common sense.
You imply the study was good. It was clearly useless.