Welcome to Debate Club! Please be aware that this is a space for respectful debate, and that your ideas will be challenged here. Please remember to critique the argument, not the author.

How do we judge a healthy weight range? BMI is no longer valid?

1810121314

Replies

  • MarkusDarwath
    MarkusDarwath Posts: 393 Member
    robininfl wrote: »
    This isn't true. Under 20 or over 25 BMI risk does increase, particularly for men. This is a meta-study with a total of over 10 million participants in 239 studies of weight and risk of death, "all cause mortality", so correlation no proven causation, but a large chunk of people studied.

    http://www.thelancet.com/journals/lancet/article/PIIS0140-6736(16)30175-1/abstract

    Fair enough. I'm not going to try nitpicking details about quality of the base studies or the aggregation of studies across varying population groups. Given that aggregation studies tend to balance out flaws in component studies, one would have to basically re-do the meta-study to assess the validity of their conclusions, and that could well go either way.
    Again, I want to say that I support aesthetic goals and do not believe you have any sort of obligation to get to the healthiest possible weight for your height.

    I'm pretty sure 15% body fat will be close to optimum healthy weight, regardless of how much muscle I add to it. :)
    What surprised me in this study was the increased death rate for those between 18.5 and 20 BMI, not the increase in overweight and obese. I would have thought that carrying less weight was healthier until you got to that 18.5, but that isn't borne out in the big study.

    I would imagine that's 1. difference in resilience in the face of accident or disease, and 2. people in that group being more physically active and prone to risk taking.
  • ouryve
    ouryve Posts: 572 Member
    edited August 2016
    Hornsby wrote: »
    Here's a read out from a scan I had done in March to give you an idea of what you will see.

    Hmmmm....you weighed 90.1 pounds on the left side, and 88.9 pounds on the right side: did you find yourself tilting to the port side, or walking in circles? :D

    ;)

    Though not actually that off. I'm quite lopsided due to accumulated injuries. Gammy right foot meant that my right shin was almost an inch smaller than my left, a while back, though I seem to have stabilised it,vfinally, after many years of embarrassing falls. Conversely, developed a massive right forearm compared with left, due to a combination of knitting style and arthritic left elbow!
    robininfl wrote: »
    robininfl wrote: »
    BMI is just built off insurance risk tables though. Health risk increases over (or under) a given mass for a given height. Saying that Most People can't get a medical benefit from getting under a 25 BMI is nonsensical. All BMI is saying, literally (in the original sense of the word) is that Most People do get a health benefit from getting under a 25 BMI. What it doesn't say is which people are part of that "Most People" group.

    Statistically speaking, most people -don't- get a particular benefit from being under 25BMI. The actual cutoff for increased risk is 30 for those without other health risk indicators, or 27-28 for those who do have other indicators (such as high blood pressure, glucose or cholesterol.)

    The "overweight" category has no meaning by itself.

    This isn't true. Under 20 or over 25 BMI risk does increase, particularly for men. This is a meta-study with a total of over 10 million participants in 239 studies of weight and risk of death, "all cause mortality", so correlation no proven causation, but a large chunk of people studied.

    http://www.thelancet.com/journals/lancet/article/PIIS0140-6736(16)30175-1/abstract

    "Mortality was lowest in the
    BMI range from 20·0 kg/m² to less than 25·0 kg/m²,
    and was significantly increased just below this BMI
    range and in the overweight range just above it."

    Again, I want to say that I support aesthetic goals and do not believe you have any sort of obligation to get to the healthiest possible weight for your height. All of us make compromises in some direction, nobody is doing 100% of the "right things" to be healthy. When you have information, though, you can make an informed choice about it. Overweight does increase health risk.

    What surprised me in this study was the increased death rate for those between 18.5 and 20 BMI, not the increase in overweight and obese. I would have thought that carrying less weight was healthier until you got to that 18.5, but that isn't borne out in the big study.

    Good luck to you in getting to your personal goals.

    I think one of the more overlooked problems at lower BMI, particularly for women, is a sharp drop in bone density, which can severely affect mobility and overall health in later life. I suspect that correlation and causation may become muddied, with this issue. Though. After all, many skinny people get that way eating shite, too. if you live on tabs, instant coffee and toast, then you may have a BMI of 20, but you are likely to be malnourished.
  • concrete_daisies
    concrete_daisies Posts: 44 Member
    ninerbuff wrote: »
    Getting an idea of what your lean body mass is (even if it's off by say 3%), you can get an ideal body weight if you know it and use this formula.

    Your lean body mass divided by (1 minus the body fat percentage you want to be)

    So say you're 120lbs lean body mass and want to be 20% body fat. Then apply the formula.

    120/(1-.20)
    120/.80= 150lbs

    So 150lbs would be the target weight (approximate) to be at 20% body fat.

    A.C.E. Certified Personal and Group Fitness Trainer
    IDEA Fitness member
    Kickboxing Certified Instructor
    Been in fitness for 30 years and have studied kinesiology and nutrition

    9285851.png

    This is misleading, generally when you lose weight you lose some lean body mass as well as fat. The exact proportion will depend on how you are eating, whether you are doing weightlifting, and how quickly you drop the weight. But you are never going to be able to keep ALL of the lean body mass.
  • Packerjohn
    Packerjohn Posts: 4,855 Member
    tomteboda wrote: »
    2efii6w4ts0z.jpg
    Also I've been dying to share this. Because statistical distributions do not mean what people seem to think they mean. A LOT of people fall outside 1 standard deviation of the mean. A fair number fall outside two standard deviations. And in a world with 7.4 billion people, 14 million people fall outside 3 standard deviations. That's a lot of people!

    14M is .2% of 7.4B or 2 out of 1000 people. 14M is a lot of people but pretty insignificant when taken as a part of the world's population.
  • tomteboda
    tomteboda Posts: 2,171 Member
    edited August 2016
    Packerjohn wrote: »
    14M is .2% of 7.4B or 2 out of 1000 people. 14M is a lot of people but pretty insignificant when taken as a part of the world's population.

    It's approximately the population of Norway and Sweden combined. Yet people meet folks from those countries all the time! It's utterly amazing.

    Also, 3 standard deviations would be 0.3% or 22.2 million... An even greater number! I changed my math when I was confused by my own graphic /facepalm.

  • tomteboda
    tomteboda Posts: 2,171 Member
    @stevencloser this thread is not a random population sample.
  • Packerjohn
    Packerjohn Posts: 4,855 Member
    tomteboda wrote: »
    Packerjohn wrote: »
    14M is .2% of 7.4B or 2 out of 1000 people. 14M is a lot of people but pretty insignificant when taken as a part of the world's population.

    It's approximately the population of Norway and Sweden combined. Yet people meet folks from those countries all the time! It's utterly amazing.

    Also, 3 standard deviations would be 0.3% or 22.2 million... An even greater number! I changed my math when I was confused by my own graphic /facepalm.

    Maybe in your world people meet residents of Norway and Sweden, how about the billions of people in China, India South America and Africa (outside of those in the higher end areas of some of the major cities)?

    3 out 1000 as outliers is not a big deal.
  • tomteboda
    tomteboda Posts: 2,171 Member
    @packerjohn you are missing the point entirely.
  • Packerjohn
    Packerjohn Posts: 4,855 Member
    edited August 2016
    tomteboda wrote: »
    @packerjohn you are missing the point entirely.

    I would disagree and suggest you are missing the point. BMI is a very reasonable starting point for the vast majority of the population in determining a healthy weight range. Understand there are outliers, but they represent a small number compared to the total population, hence the term outlier.

    A 275 pound 5'5" individual wants to lose weight. Don't you think the BMI calculation will provide them with a reasonable range to shoot for in the vast majority of cases? If you don't think it is what, that is widely available at low cost, would be better? Do you think it would yield a significantly different result than BMI for most people?
  • tomteboda
    tomteboda Posts: 2,171 Member
    I think, enabling the thinking of "Oh I must be one of those outliers" by making it sound more common than it is is not good for the health of the population.

    Did anyone even read my last paragraph?
  • Dove0804
    Dove0804 Posts: 213 Member
    edited August 2016
    tomteboda wrote: »
    I'm not rejecting its use, but noting that it has some very serious limitations and it should absolutely NOT be a gold standard on its own by which indivdiuals are judged to be healthy or unhealthy (much less a moral judgment point).

    Yes, I did read your last paragraph. I followed along and for the most part agreed with your entire post, I'm just pointing out that again- this is exactly the point many of us are making in this thread. Nobody here said it should be a gold standard or wasn't without its serious limitations. You make lots of good statements.
    tomteboda wrote: »
    2efii6w4ts0z.jpg
    Also I've been dying to share this. Because statistical distributions do not mean what people seem to think they mean. A LOT of people fall outside 1 standard deviation of the mean. A fair number fall outside two standard deviations. And in a world with 7.4 billion people, 14 million people fall outside 3 standard deviations. That's a lot of people!

    This is where I lost you a bit. I don't think anyone here was discussing statistical distributions in the sense you are. Otherwise, I wouldn't be using term "outlier"- as outliers don't technically exist when we're talking distribution curves.
    What are you representing with that normal distribution curve? While the basic principles of distribution curves and standard deviations are not going to change, the subject matter and what you're applying it to certainly does. You're using the entire world's population as a data set. Were you talking about what weight/BMI/BF%/whatever the world population currently is? In that case I think we've established that being too many standard deviations away from the mean is not a good thing, and there's also the assumption that the average is healthy to begin with. Are we talking about BMI's ability to predict health in a population where everyone is healthy (employing Central Limit Theorem)? In that case saying BMI does not apply to 14-20 million people is being misleading because you were assuming an entirely healthy population to begin with (along with a ton of other assumptions). Statistics are great but they can't be used in a vacuum without labels and definition- the clinical application is incredibly essential to consider as well.
    Your discussion of statistical distributions isn't what bothered me- it's the blanket-statement message you were sending at the end there without defining what you were showing, which makes me inclined to agree with this:
    I think, enabling the thinking of "Oh I must be one of those outliers" by making it sound more common than it is is not good for the health of the population.
  • MarkusDarwath
    MarkusDarwath Posts: 393 Member
    Packerjohn wrote: »

    I would disagree and suggest you are missing the point. BMI is a very reasonable starting point for the vast majority of the population in determining a healthy weight range. Understand there are outliers, but they represent a small number compared to the total population, hence the term outlier.

    Unless that "vast majority" are only those within one standard deviation of mean, in which case your "vast majority" is just over 68%. Yes, that is still a sizeable majority, but if the metric is less than reliable for 32% of the populace then it absolutely should not be used as a standard go-to.
    No method should become a default for assessing individual health unless it is reasonably reliable for at least 95% of people, preferably more.
    A 275 pound 5'5" individual wants to lose weight. Don't you think the BMI calculation will provide them with a reasonable range to shoot for in the vast majority of cases? If you don't think it is what, that is widely available at low cost, would be better? Do you think it would yield a significantly different result than BMI for most people?

    At best, a BMI calculation would tell that person they are likely at high risk for metabolic issues. Chances are at that size, they already know this.
    Simply looking at their height on a BMI chart and saying, "you should weigh this much" is not terribly helpful if there's no reference to body composition. Even if the weight on the chart is appropriate for that individual, setting such a distant goal for the severely obese is likely to be a motivation killer.

    Any means of actually measuring body fat, even if it's not the most exacting, will give a good indication of where the person is starting from and can be broken down into a range of shorter and longer term goals that are specific to that individual and their particular health needs.

  • This content has been removed.
  • Packerjohn
    Packerjohn Posts: 4,855 Member
    Packerjohn wrote: »

    I would disagree and suggest you are missing the point. BMI is a very reasonable starting point for the vast majority of the population in determining a healthy weight range. Understand there are outliers, but they represent a small number compared to the total population, hence the term outlier.

    Unless that "vast majority" are only those within one standard deviation of mean, in which case your "vast majority" is just over 68%. Yes, that is still a sizeable majority, but if the metric is less than reliable for 32% of the populace then it absolutely should not be used as a standard go-to.
    No method should become a default for assessing individual health unless it is reasonably reliable for at least 95% of people, preferably more.
    A 275 pound 5'5" individual wants to lose weight. Don't you think the BMI calculation will provide them with a reasonable range to shoot for in the vast majority of cases? If you don't think it is what, that is widely available at low cost, would be better? Do you think it would yield a significantly different result than BMI for most people?

    At best, a BMI calculation would tell that person they are likely at high risk for metabolic issues. Chances are at that size, they already know this.
    Simply looking at their height on a BMI chart and saying, "you should weigh this much" is not terribly helpful if there's no reference to body composition. Even if the weight on the chart is appropriate for that individual, setting such a distant goal for the severely obese is likely to be a motivation killer.

    Any means of actually measuring body fat, even if it's not the most exacting, will give a good indication of where the person is starting from and can be broken down into a range of shorter and longer term goals that are specific to that individual and their particular health needs.

    Unless that person is unusually muscular a bodyfat% goal or BMI goal will result in a goal in pounds that is very similar

    As far as goal setting, depending on the individual a health care provider could give a series of step goals to get to the final to prevent giving up/burnout.
  • MarkusDarwath
    MarkusDarwath Posts: 393 Member
    RLester67 wrote: »

    The article you linked does not support the idea that BIA is less accurate than BMI, except that in some cases it doesn't accurately show trends in fat loss. BMI, as several have pointed out, is not even a measure of body fat, so it really can't even be compared as a means of determining body composition.
  • MarkusDarwath
    MarkusDarwath Posts: 393 Member
    Packerjohn wrote: »
    Unless that person is unusually muscular a bodyfat% goal or BMI goal will result in a goal in pounds that is very similar

    I am not unusually muscular, yet I have mathematically demonstrated more than once that a better-than-average BF goal of 15% still puts me well above the very lower limit of "overweight" for my height according to BMI.
    I do have broad shoulders and a disproportionately long torso (that latter is a factor I had overlooked), but if that is enough in itself to qualify me as "unusually muscular" then there is a lot greater percentage of such folks in the general population than BMI defenders are admitting to.





  • MarkusDarwath
    MarkusDarwath Posts: 393 Member
    You are basing your claim on measurements taken by a notoriously inaccurate device and assuming your muscle mass will be the same when you reach goal.
    I'm not exactly convinced.

    And BMI defenders base their claims on appeal to authority. That's not exactly convincing either.

  • tomteboda
    tomteboda Posts: 2,171 Member
    Let me clarify then. The BMI vs health is charted on a normal probability distribution. The confidence intervals in most correlation studies suggest only 1 standard dviation is taken into account, sometimes 2 and never 3 because bmi is being used as a proxy for visceral fat. So a lot of folks get labeled as "at risk" by bmi who aren't carrying dangerous levels of visceral fast. A pretty large number get "healthy" labels so carry visceral fat because of the BMI proxy as well. However, BMI is cheap and statistically not much different in predictive results than body fat analysis, largely because most analysis methods also cannot distinguish between visceral and subcutaneous adiposity.

    Ultimately, a person's metabolic health as tracked by blood sugar, cholesterol, blood pressure, heart strength (stress test) and blood chemistry are far more important for an individual. Weight gives a crude estimate of body fat that can be tracked over time on an individual.

    Personally, I think the most important thing is functional ability. Can you do the things you want to do? Is your weight impacting your life personally, if you're fully honest? We don't all have to be marathon runners or heavy lifters or have the physique of a professional athletes to be healthy.

    I've tried very hard to stay away from the personal in this discussion , so accusing me of being an "enabler" stings. Yet I will confirm that as I understand it, the science behind dropping "healthy" to a BMI of 25 from 27.5 was spurious at best. The health outcomes of individuals between 25 - 27.5 ON AVERAGE are better than those between 18.5 - 20.

    The statistics just don't bear out the antipathy towards overweight status that several posters have demonstrated. In particular, one person with a very low BMI attacked even people within the normal range for being too fat, but statistically, by the study cited, that person is at greater risk than others between 25-30! Yet no sane person is going to suggest we raise the minimum acceptable BMI, and drawing any conclusion about that person's health would be the venture of fools.

    There has been a lot of creep about aesthetics in this thread, which is terrible when discussing health policy because aesthetics change by time and culture.
  • TR0berts
    TR0berts Posts: 7,739 Member
    I love when people acknowledge that their info source is full of bunk. Except, of course, the one bit that they agree with.

    Cherry picking, at its finest.
  • Packerjohn
    Packerjohn Posts: 4,855 Member
    Packerjohn wrote: »
    Unless that person is unusually muscular a bodyfat% goal or BMI goal will result in a goal in pounds that is very similar

    I am not unusually muscular, yet I have mathematically demonstrated more than once that a better-than-average BF goal of 15% still puts me well above the very lower limit of "overweight" for my height according to BMI.
    I do have broad shoulders and a disproportionately long torso (that latter is a factor I had overlooked), but if that is enough in itself to qualify me as "unusually muscular" then there is a lot greater percentage of such folks in the general population than BMI defenders are admitting to.





    Have you been lifting weights for a long time? If not, the ideal of a man in his mid 40' getting to 15% bodyfat at your height and 220 pounds most likely will not happen.

    15% BF is in the range for college baseball players, who have most likely been playing other sports recently in hs and are usually on a structured lifting program. Not to mention naturally high testosterone levels. Given your current size you may want to get your levels checked. It will be hard enough trying to get to 15% bodyfat at the weight you're looking at let alone with low levels

    http://www.ncaa.org/health-and-safety/sport-science-institute/body-composition-what-are-athletes-made

    Again, best of luck but you've picked some pretty challenging goals
This discussion has been closed.