Skip to content

Can We Trust Obesity Data?

October 22, 2013

Weight LossFat HealthFat ScienceFat NewsDiet TalkExercise

I read “Can We Trust Obesity Data” today and I think what they’re basically saying is that most people who are reporting their caloric intake and exercise routines to the National Health and Nutrition Examination Survey (NHANES) are liars.Calories in Calories Out 4The primary method used by NHANES to determine energy intake (calories eaten) is the 24-hour dietary recall interview.

These subjective data are then translated into nutrient food codes and assigned caloric values from food and nutrient databases, the researchers explained in the study published in PLoS One. But they found that the reported energy intake data on the majority of NHANES respondents (67.3% of women and 58.7% of men) were not physiologically credible.

In other words, these respondents had a physical activity level (the ratio of total energy expenditure and basal metabolic rate) of less than 1.35 — a level at which it is “highly unlikely that any normal, healthy, free-living person could habitually exist,” they wrote.

If these people who are interviewed by NHANES are dieting, they are going to be taking in fewer calories than are actually required for existence, since most diets restrict you anywhere from 800 calories a day (i.e., very low calorie diets) to 1,200-1,600 calories a day (i.e., low calorie diets), which is less extreme, but still less than what is required for maintenance, even if you do nothing but sit on your ass all day long.

If you’re dieting and exercising in order to lose weight, then when you report your restricted caloric intake plus the hours that you exercise plus all the activities you do in a normal day’s work and living your life, of course you’re going to have a physical activity rate that’s lower than “normal.” And, no, this is not an existence under which people can “habitually exist” (why do you think diets don’t work, hmmmm?).

But this type of regimen is exactly what doctors prescribe for their fat patients: restrict calories and add exercise, which, if followed, gives their fat patients a physical activity level (PAL) lower than what is needed to exist.

The expected average physical activity values for healthy men and women are around 1.55, with a range of greater than 1.35 to 2.40. Archer and colleagues found that the average values for all men and women were 1.31 and 1.19, respectively.

“These results suggest that without valid population-level data, speculations regarding the role of energy intake in the rise in the prevalence of obesity are without empirical support,” Archer told MedPage Today.

Valid population-level data? Apart from following people around all day and measuring yourself, what is more valid than interviewing people and having them tell you what they ate and how much movement they had in their lives for a specified period of time? Are they saying that people don’t/can’t accurately remember what they’ve done over the last 24 hours? Well, if they were asking me, I probably couldn’t remember everything I ate yesterday, or exactly what I did then either. Fibromyalgia’s brain fog and depression’s memory farts have totally fucked my short-term memory and haven’t helped my long-term memory all that much either.

NHANES has been around for 40 some years but even 60 years ago, researchers were questioning the use of self-reported energy intake, Andrew Brown, PhD, a post-doctoral trainee at the Nutritional Obesity Research Center at the University of Alabama at Birmingham, told MedPage Today.

“I agree with [Archer and colleagues] that using these data for population trends and making policy is not warranted,” Brown said. “They are mathematically biased — the numbers do not represent what they are meant to represent. The under-reporting in energy intake has nothing to do with actual energy intake.”

For 60 years, researchers have been calling people liars about their reports of what/how much they eat, but they still used the data, still applied for grants for studies knowing they were going to use that data (that they consider suspect and untrustworthy). Is is any wonder with that kind of attitude, their research could be considered flawed and their conclusions suspect?

Archer said he has received negative feedback from colleagues regarding the study. “There’s a lot of people dependent on the more than $1.3 billion federal dollars that support research using NHANES. The bureaucratic machinery and inertia has pushed this model forward even though we’ve known since the early 1990s that it is flawed,” he said.

And there’s the reason they’ve gone on using “flawed” data: “There’s a lot of people dependent on the more than $1.3 billion federal dollars that support research using NHANES.” Follow the money, follow the money.

The researchers also said public policy recommendations emphasizing healthy behaviors can influence participants to misreport their own unhealthy behaviors, making the data further unreliable.

Again with the “misreport their own unhealthy behaviors” BS. Misreport? Why not call a spade a spade and just say what you really mean — they’re lying.

Another potential problem is the flaws in the U.S. Department of Agriculture’s National Nutrient Databank System (NDBS), which is used for the post-hoc conversion of food and beverage consumption data into calories.

The database has not kept up with the exponential increase in food products into the American diet, particularly processed and restaurant foods — especially fast-food restaurants, Archer said in an interview.

Wait, you mean the NDBS isn’t keeping up with converting to calories the processed foods, restaurant food, and fast food? Couldn’t this account for some of the alleged “misreporting” if they are basing caloric rates on obsolete data?

“The nation’s major surveillance tool for studying the relationships between nutrition and health is not valid. It is time to stop spending tens of millions of health research dollars collecting invalid data and find more accurate measures,” he concluded.

Researchers have been studying the relationship between nutrition and health for decades, and still haven’t figured out a way to study how nutrition relates to health? Considering that every person is an individual with different likes, dislikes, and preferences, with different genes that influence what diseases they may have/get, different lifestyles, different socioeconomic statuses, etc, etc, etc, do you really think you’re ever going to find the answers to these questions? Personally, I think you could look for the next thousand years and you wouldn’t be any closer to a solution then than you are now. Too many variables to be controlled for will mess you up every time, and that’s what you have when you try to relate nutrition to health.

The only data you’re going to believe is what you can see with your own eyes, what you can measure yourself. I don’t know that there are very many people who would volunteer to be your guinea pigs in that scenario, and that’s the only way you’re going to get data that you’ll believe, since you’re so convinced that your interviewees aren’t being honest about how much/what they eat or how much movement they’re doing.

My conclusion from this is that no, no we can’t trust “obesity” data, not with the “unreliability” of the people who are reporting the data that researchers need in order to reach their conclusions, and that’s the problem with taking individual data, aggregating it, and then trying to apply your conclusions to a whole population of diverse people.

16 Comments leave one →
  1. Kala permalink
    October 22, 2013 2:23 pm

    I personally doubt that dieting could be skewing the numbers to that degree, it doesn’t seem realistic. I think the odds of surveying an individual for a study, and having them just happening to be on a highly restrictive diet that past 24 hours to be relatively low. I would be far more likely to believe that individuals were either too embarrassed to admit what they were actually eating (the selective under reporting mentioned in the cited article on the medpage article supports that claim), or were completely clueless about the actual serving sizes and were misreporting out of ignorance.

    Regrading reliable data. It depends on how it’s unreliable. Is the measurement error predictable? If it was highly variable and difficult to predict, then I could see the argument that perhaps the data shouldn’t be bothered with. But if measurement error is predictable (you can reliably model it with something, it seems like the believed trend is that people under report caloric intake and that the level of error increases with increased obesity), then I see no reason why you could not adjust for it in your mathematical model.

    I get the idea that you believe that people are under reporting their calories. That you believe that all these folks are really eating 1600-2500 calories a day as their normal diet, and it’s the models and the research that is flawed. Because ultimately what the research did was use some equations to estimate what a maintenance caloric intake would be for a person of whatever weight, and then looked at the discrepancy between that number and what they actually claimed they ate.

    Finally, regarding the last statement:

    Are you saying that you disagree with the central dogma of statistics? That one can estimate population parameters or their distribution based upon an aggregation and analysis of data from individuals? Is it all statistics, some types of statistical analysis? Do you believe that there’s no way to reasonably estimate population parameters and the relationship between variables using data from “a whole population of diverse people?”

  2. vesta44 permalink
    October 22, 2013 3:29 pm

    Kala – You doubt that dieting could be skewing the numbers? The article said that 67.3% of women and 58.7% of men ended up with a PAL that was considered to be physiologically unsustainable. So you find it difficult to believe that that many people could be dieting? Really? With the prevalence of diet ads, doctors telling their patients to “just lose weight” as the cure for all their ills, and the fact that the diet industry makes upwards of $60 billion a year, it’s difficult to believe that that many people could be dieting.
    I would bet that if you surveyed every doctor who treats fat patients and asked them how many of their patients are dieting, and how many of those patients are considered “compliant” – as judged by actual weight lost and maintained – most of those doctors would say most of their patients are “non-compliant”. In other words, because their patients aren’t losing weight or maintaining that loss, they’re lying about what they eat, how much they eat, when they eat, where they eat, and how much exercise they’re doing (been there done that with every doctor I’ve ever seen). So I find it very believable that those percentages of the people reporting to NHANES who have those “unsustainable” PAL numbers could be on diets that skew the numbers that low.
    As for whether I disagree with the central dogma of statistics – that one can estimate entire population parameters or their distribution based upon an aggregation and analysis of data from individuals, I’m not sure. The problem, I think, comes in when these studies are designed and funded. I’m more worried about where the funding is coming from, who is going to benefit from the results of the study, what agenda is it going to advance, and who is going to be hurt by those agendas.
    When it comes to studies on obesity and predicting how it’s going to affect, say, just the United States, I haven’t seen many that have a good track record for being accurate. Now, is that because the studies are designed to come to a particular conclusion, and a lot of the data gathered doesn’t really support that conclusion? Or are there other confounding factors that haven’t been taken into consideration? I don’t know because I don’t know enough about how studies are designed or who is chosen to participate. I can tell you, though, from the surveys about health that I’ve filled out online, that whoever is using my data is probably scratching their head and saying “This is impossible, these answers just don’t add up.” I think they would consider me a statistical anomaly, simply because these surveys can’t ask all the questions they need to in order to get a clear idea about the people answering them. Not to mention that the way some of those questions are phrased leaves a lot to be desired when trying to answer them. So maybe I’m saying that when it comes to obesity and statistics, yeah, I question the methodology used and its accuracy.

    • Kala permalink
      October 22, 2013 4:17 pm

      Yes, I absolutely think it’s unreasonable to assume that 2/3rds of all adults are currently dieting if I were to call up a somewhat representative sample of survey participants and ask them what they ate in the last 24 hours. I think error in reporting is far more likely an explanation. And yes, I still think all this despite the prevalence of diet ads and the amount of money the diet industry makes in a year. Exercise is recommended by doctors to nearly everyone regardless of size, and the fitness industry also makes a shit ton of money and has very pervasive advertising, so by your logic I should assume that more than half of all people in the USA are currently getting jacked. Yet the muscles of Americans are not swelling at some high rate, or are they and I haven’t noticed?

      I think you’re taking your own personal circumstances and the experiences of a very selective subset of people that you associate with in the FA community, and are extrapolating that out to the rest of the population, whether they be fat or not. You and your friends have dieted constantly for years and had WLS and done whatever else in past years. So thus in your mind, a lack of reporting error and a huge prevalence of dieting seems on point. This kind of assumption is a lot less realistic than using data taken from individuals in a highly organized observational study, and making conclusions about the population based upon that instead.

      I get the sense that you believe that mathematical models put every single person in the population on a single line, and that researchers feel that deviations are not possible. So you abhor the models and their conclusions, you feel that they could never be representative. There are people who take simple deterministic interpretations of those models without taking into account at all the random variation that exists in reality, and make exactly that conclusion. People who write about science in the media do this all day long. But anyone with any substantial understanding of the subject doesn’t do that.

  3. Rita permalink
    October 22, 2013 6:13 pm


    I work at a diabetes clinic as a nutritional counselor and I originally stumbled onto this site looking for a size acceptance forum for our patients. There doesn’t seem to be any for them. A small percentage of our patients have self-esteem issues related to their obesity so I was hoping to find them some sort of peer counseling to help them stay on their diet and avoid the foods that exacerbate their diabetes. Having diabetes can be very depressing and body image adds to that.

    As to the calorie thing, Most of our patients are not entirely honest about their caloric intake and that can be very dangerous. Nearly all of them under report. It’s due to their guilt and shame. We try to give them a graceful way out and nearly all of our male patients fess up and get with the program. It seems that men are far less body conscious than the women. Men tend to be more rational and less therapeutically resistant.than women.

    We give them simple formulas and eating plans that they can follow and the majority of them do well and get to a steady state but some are resistant and those are the ones who have the body image issues. That said, I am thinking that an online support group would be helpful to these people. Any suggestions.

    • October 23, 2013 10:49 am

      “…Men tend to be more rational and less therapeutically resistant.than women…”

      Wow. Just… wow.

      • Rita permalink
        October 23, 2013 12:49 pm

        It’s a common knowledge that when it comes to the Cluster B disorders women are therapeutically resistant. As you know, insulin is a hormone and in the case of Type 2 diabetes it’s the excessive fat that causes the insulin resistance but beyond that obesity causes hormonal imbalance which is more marked in females.

        Women are far more likely to have hormone imbalances and obesity exacerbates that more so in women than in men. Fat men and women have a different hormonal profile than lean men and women. when it come to most of the hormones. Body fat has a profound effect on all the hormone included CKK, the thyroid hormones, insulin and the sex hormones and that has an affect on behavior and emotions. Males are less adversely effected than females for a myriad on reasons. Also much depends on the distribution of fat i.e. visceral vs subcutaneous fat.

        I hope this answered your question but back to my question, can you recommend and online support group for fat folks?

        • October 23, 2013 2:17 pm

          Well, what with my fat ass and hormones so likely to be making me irrational and all, you probably wouldn’t want to place your trust in my unworthy recommendations. [rolleyes] Now, if you’ll excuse me, I need to be somewhere far, far away from you.

    • October 23, 2013 11:21 am

      How do you know they are under reporting? Just because they aren’t losing weight or is there something else that tells you that, like blood sugar readings?

      • Rita permalink
        October 23, 2013 1:11 pm

        Eventually they admit it. Males generally make no bones about it and they are looking for ways to control their hunger and since fluctuating insulin has a profound effect on hunger the aim is to get them off insulin.

        These VLC diets are a disaster. Putting anyone on extreme calorie restriction is medically unsound and IMO bad medicine. There is no one size fits all number for people. In the case of the super morbidly obese 1300 calories is not only cruel, it’s dangerous. Doctors do this sort of thing all the time with WLS patients knowing how dangerous is can be. WLS IMO should be avoided like the plague. This is not medical advice but I know some of the surgeons who do it and all the ones I know are jerks.

        Our aim is not weight loss per se but insulin and hunger control and that involves strict insulin and dietary monitoring. That includes monitoring blood insulin and sugar levels and glycemic loads in meals. The aim of that in part is to control hunger.

        • October 23, 2013 1:58 pm

          I really appreciate your answer, I really do.

          I have been SOOOOO many diets were I was accused of overeating/under reporting where I was doing neither that I was put off by your words and came out swinging, so I apologize. I was on a VCLD of about 450 cals a day and it worked for losing weight like the first month, but after that not so much, and it was emotionally draining.

          I think it’s a shame that real nutritional counseling isn’t available to people through insurance (at least never has been to me) until someone is diagnosed with someone with diabetes when I think a lot of people could benefit from counseling that doesn’t come from the perspective of weight loss.

          Try looking at the Fit Fatties forum. It’s more geared toward exercise, but no diet or weight loss talk is allowed. Maybe that’s what you’re looking for. It’s not supremely active, but it is very supportive of helping people with finding movement that supports their fitness goals.

          • Rita permalink
            October 23, 2013 2:53 pm

            The doctor who put you on that diet should lose his license.

            Nutrition does not have to be complicated. If it’s processed food don’t eat it. 70% of your calories from plant based carbohydrates. Get the rest from protein and healthy fats.

            The fit fat people are the ones with low levels of visceral fat. Unfortunately most obese people have too much visceral fat. Overall too much body weight takes its toll on the body especially the joints.

            Our patients who need to lose weight have the most success then they stick with the program. Starvation diets will not only fail but they are dangerous and can result in even more weight gain. The simple rule is for them not to eat less than their daily Basal Metabolic Rate or BMR. You can get a good idea of what your BMR is by multiplying your body weight by 10. Ex.. A person who weighs 130 pounds has a BMR of 1300 calories per day. A person with a body weight of 300 pounds has a BMR of 3000 cals per day. The more weight you gain the higher your BMI becomes. That’s why when you watch the fat people exploitation shows these people lose massive amounts of weight very rapidly but at what cost?

            • October 23, 2013 3:05 pm

              I approved your comments because they weren’t egregious, but they have grown increasingly out of line. I sent you an email to confirm that this was a valid account and it bounced back. You aren’t a nutrition counselor and you’re just spewing bullshit now. Your ignorance is tainting an otherwise valuable conversation. Visceral fat is even a questionable measure of fat and fit compared to the amount of fat in the liver. Furthermore, there’s no way to know exactly how much visceral fat you have without an expensive DEXA scan, rendering it a useless diagnostic tool.

              But the fact that you think a 130 pound person should maintain a 1300 calorie diet was the final straw for me. I don’t know who you are or what your qualifications really are, but I wouldn’t trust you with my dog’s healthcare.


              • October 23, 2013 4:14 pm

                Sorry I continued the conversation, she seemed mostly reasonable and well intentioned. I put on blinders to the warning signs, I’ll be more careful in the future..

                • October 23, 2013 5:13 pm

                  No need to apologize. If a comment is approved, then it has been vetted personally by me and I feel relatively confident that the commenter is being intellectually honest and fair. It was that last comment when I decided to verify email and confirmed my growing suspicions. It happens.


  4. October 23, 2013 7:12 am

    This reminds me of my frustrating experiences with the National Weight Control Registry. Over the years, they have routinely reported that registrants eat 1,300 cal/day, on average, but until recently they didn’t ask us what we think our calories are. (I told them 1,800, since they bothered to ask.) In the past, they deduced our calorie intake based on surveys that imagine plates that look like they were neatly assembled by USDA scientists from foods available in the 1960s, and they asked us to recall a whole year’s worth of eating and average servings. Now they have shortened the time frame we have to remember, but many foods we regularly eat are still omitted from the list of choices. It’s horribly flawed. Moreover, the old data is still used for historical and comparative studies (just as the NHANES data is regarded as a useful historical record). Yeesh.

    Now, because NWCR participants are presumed to be so virtuous, the historical research from NWCR land that deduced our calorie intake of 1,300/day was never used to say how we “under-reported” our food, but used to justify that people CAN live on these levels of calories, and fat people should just try it, doncha know. As you note, people cannot live for any length of time on 1,300 calories.

  5. October 27, 2013 11:01 am

    Yes nice escape for them, just call all the fat people liars.

    They are WRONG about metabolism.

    Guess they’ll have to lock all the fatties in a room and feed them through a hole in the wall—/sarc

    Funny for me, the blood sugars prove to the doctors I’m not a binge eater. Ironic isn’t it? I’m supposed to be laying on the floor after inhaling two boxes of ice cream or something with a blood sugar of 600 or something according to my body weight.

    It was 127 the other day.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: