There is a lot of sound and fury about the BMI (body mass index). On the one hand, opinion leading doctors say we all should know our BMI and act accordingly. Other pundits point out that it's an inaccurate way to condem people. Both of these points of view arise from misuse of the BMI. The BMI is simply a way to account for the fact that your weight varies according to your height.

The BMI is also known as the Quetelet index after the Belgian astronomer, Mr. Adolphe Quetelet, who invented it in the early 19th century. He  measured the heights and weights of French and Scottish soldiers and found that their weights were proportional to the square of their heights. That means that if you plot a graph with the weights on one axis and the square of the heights on the other, you get a straight line. If you do this for weights and plain heights you get a curved line that's messier. We do this in meters for height and kilograms for weight since Mr Quetelet and most of the world works in the metric system. (for the defiantly contrary Americans, BMI would be weight in pounds X 705 divided by height in inches squared.)

In the olden days, life insurance companies, whose bottom lines are influenced by people who die, made charts of ideal body weight - "ideal" meaning least likely to die. They used "weight for height" and then there is a range of weights for every height. BMI charts are more compact and have replaced the original life indurance tabels for good, though it is important to note that BMI doesn't wipe out all the messiness of human variability. If you are a football player or very tall your BMI will skew. The life insurance charts had to account for this by another variable called body build. Then you have an even bigger chart with heights and a range of weights for each of 3 different body builds etc...

BMI didn't fix every thing but it is more useful for statistical purposes. For example, BMI makes it easier to make charts of fatness by age or sex or heart disease or over time. You have one number for "fatness" to plot against these other variables. We will make great use of BMI for these kinds of discussions.

So it was never meant to be a clinical tool for doctors in the office to assess an individual, unless those doctors want to make summary statistics of their patients. Don't worry about your BMI. Just look in the mirror. Does this dress make me look fat? In adults, whose heights don't change, doctors can just weigh you to see if you're getting fatter - as long as you're not in heart failure or dehydrated.... Percent of previous weight is a better way to judge changes from your doctor's perspective. But gaining or losing 10 pounds is a big deal if you started out at 100 (10% weight loss) but not if you were 300 pounds (3%).

The BMI routine is more complicated in children because it changes throughout childhood. Rather than reporting BMIs in children what we have to report is BMI percentiles for same aged boys or girls (different charts). Furthermore BMI charts aren't available for under 2 year olds. Babies are supposed to be laughing Buddhas. Five year olds have pot bellies and bird legs. Adolescents develop the sex shapes of adulthood. Back in the olden days, pediatricians, who always paid a lot of attention to growth and development, used to use height and weight growth charts. If somebody's height was in the 75th percentile and his weight was greater than the 99th percentile then we would know he was fat. The newfangled BMI is no better than that for a child. BMI is also slightly complicated in adult women. Remember men were used to find the best formula for determining BMI.

The modern era of the BMI was started by a paper (1) published in 1972 by Ancel Keys, a scientist who is famous for inventing the K-rations used for the soldiers during World War II and his starvation study on conscientious objectors to that war. He was also responsible for some of the earliest insights into diet and heart disease and was the first to popularize the Mediterranean diet. In that 1972 paper he said that the BMI was appropriate for population studies, and inappropriate for individual diagnosis. I don't know who it was who forgot that message and started this new BMI rage.

Can't Tell

But wait. There are reports that people can't tell if they are fat themselves or if their children are fat. Also doctors can't tell. That is why pediatricians are being encouraged to do BMIs on all their patients to see if they are fat. This seems ridiculous on the face of it. I don't buy it. But I am conflicted here. I am famous for saying that "obvious" things often turn out to be quite the contrary. (like obesity is caused by lazy gluttony or watching TV) So I am forced to spend your and my valuable time trying to find out how anyone could support the proposition that fat people might not realize they are fat.

Way back in 1987, Wright and Whitehead (2) reviewed the evidence and decided that some obese people were in denial. "As a result of negative attitudes toward fatness, stigmatization of overweight people has been consistently documented. In related studies, investigators have centered on self-perceptions of obese and overweight individuals. According to the data, the existence of body image disturbance among extremely overweight individuals is a common phenomenon....normal weight individuals (especially women) may also have difficulty in accurately assessing their body size."

In a report that comes from Cincinnati Children's Hospital(3), investigators asked 622 mothers with children 23 to 60 months of age -"Do you feel you are overweight right now?" and "I feel my child is- 1. very underweight 2. a little underweight 3. about the right weight 4. a little overweight 5. very overweight," among lots of other questions. They compared the answers with the heights and weights reported by the mothers of themselves, and the measured heights and weights of the children. They found that only 5% of the mothers who were overweight judged themselves incorrectly but 79%, of the mothers of overweight children said those children were not overweight, and this misperception was more common in mothers with less education. These authors say, "Childhood obesity prevention efforts are unlikely to be successful without a better understanding of how mothers perceive the problem of overweight in their preschool children." Why would mothers know they were fat but not know their children were? In the discussion these authors conjecture that maybe they did not want to label their children.

In another investigation of this, Galuska et al.,(4) from the Centers for Disease Control and Prevention, used data from the Third National Health and Nutrition Examination Survey (NHANES III, 1988-1994). They asked mothers of 5500 children aged 2 -11 years, "Do you consider [name of child] to be overweight, underweight or about the right weight?" Then they compared these responses with the measured BMIs. In the overweight children by BMI one third of the mothers reported their overweight child as "about the right weight." Getting it wrong was more common for sons, younger children, and not so obese children. Race didn't matter. These mothers must be much smarter than the mothers in Cincinnati and in other studies of this issue but they were still wrong a lot. Galusca et al. say that the failure to classify overweight children correctly by mothers "may reflect a failure of mothers to recognize overweight status of their child, a reluctance to admit that their child is overweight, or a lack of understanding of what 'overweight' means." The finding that mothers were 3 times more likely to accurately assess their overweight daughters than they were to accurately assess their overweight sons was disturbing. Maybe these mothers are not really blind.

To see what kids themselves say, using data from National Youth Risk behavior Surveys conducted between 1999 to 2007, Foti et al (5) asked 9 through 12th graders their heights and weights and whether they thought that was too big or too small. Sixteen percent overall thought they were too fat. Among non-overweight students, females were 3 times more likely than males to think that they were too fat, especially white girls (25%). This kind of incorrect perception decreased somewhat from 17.7% in 1999 to 14% in 2007. Among the overweight 68.5% overall, knew they were, especially again the girls (79.5%) and the white girls (85.2%) Over time those the perceptions didn't change. In a commentary in Contemporary Pediatrics of September 2010 p38, Dr. Michael Burke said, "Although I hesitate to stigmatize overweight children at school, I can see no way of solving childhood obesity unless its victims acknowledge their condition."

In Eckstein KC, et al.(6) parents of 223 children aged 2 to 17 were surveyed concerning their child's appearance and health. They then compared these responses with the children's' BMIs. Few parent's (36%) identified their overweight or at risk for overweight child correctly using words, but more (70%) selected a middle or heavier sketch from pictures. Parents of older children were more likely to judge correctly and to be worried if they perceived their child as less active than other children or recalled a doctor's concern. These authors conclude that "Identification of counseling strategies for pediatric practice that heighten parental awareness and concern about their child's overweight condition may be a needed first step in motivating families to participate in intervention programs that address child overweight."

One more example, in this exhausting but not exhaustive survey, is a study from McMaster Children's Hospital in Hamilton Ontario.(7) They collected 91 children aged 5 to 18 years who were visiting the Pediatric Gastroenterology Clinic for problems unrelated to obesity. Eight percent of these children were overweight and 15% were obese. They used verbal descriptions and pictures to ask the children, their parents and their doctors to characterize the children. Between 40 and 50% of the children and their parents underestimated the children's' body size by both methods. Thirty-three percent of the physicians also underestimated the overweight children's' body size. Even some obese children were not recognized. These authors conclude that, "This study demonstrates the need to further educate physicians to recognize obesity and overweight so that they can counsel children and their families."

So there you have it. Even though maybe we can tell that some people are obviously obese without doing a BMI calculation, let's accept that, for whatever reason, many of us can't or won't admit to weight problems, and that figuring out BMI, which puts a number on it, might be useful for talking to patients. The point of this as stated in all these studies is that people must face up to their problem before they can solve it.

I still have a problem with this. One problem with all these studies is that they presume BMI to be the gold standard. When measured against other determinates of body fat, BMI is more or less discrepant. For example, it has been shown that children with the same BMI show a marked variation in total body fat ranging from 8% to 22% as estimated from the isotope dilution method. (8) These authors concluded that non-invasive methods are presently not suited to assess the absolute amount of total body fat in 6-7 year old children. In another study of obese Korean children (9) the results of BMI showed 66.0% of participants as obese (boys 65.2%, girls 67.6%) compared to another index, called the obesity index, (defined as Actual weight - Classified by height standard weight/Classified by height standard weight×100), that measured 92.2% of participants as obese (boys 94.2%, girls 88.2%). A study in adolescents (10) found BMI to be specific, that is it is accurate when it says that someone is not obese, but not very sensitive, that means it is not very good for finding people who are indeed obese by triceps skinfold thickness and estimated percentage body fat as the criteria for adiposity. The predictive value of the BMI as an indicator of the risk of overweight relative to the triceps skinfold thickness varied among samples from 16.7% in Canadian girls to 80.0% in Asian girls. Another study of the validity of BMI in children and adolescents found that when using skin fold thickness as the standard, BMI was better than weight for height but when percentage body fat or total fat mass by DEXA scan was used as the standard, probably a better standard, then BMI was not better than weight for height in detecting under- or overweight.(11)

The other reason for measuring a child's BMI, besides to show him how fat he is, might be to monitor how that child is doing over time. Is he getting fatter or less fat? But a good normal BMI changes throughout childhood. To deal with this problem you could measure BMI z score, or BMI percentile. BMI z score and percentile are basically measuring where your BMI lies on the normal distribution of BMIs for your particular age and sex. But a study in an Italian kindergarten showed that it's even more complicated.(12) They concluded after a lot of mind boggling math that "Even though BMI z-score is optimal for assessing adiposity on a single occasion, it is not necessarily the best scale for measuring change in adiposity, as the within-child variability over time depends on the child's level of adiposity. Better alternatives are BMI itself or BMI%. Our results underscore the importance of using a relatively stable method to assess adiposity change when following children at risk of obesity."

My final and most important problem with the modern BMI imperative is that it blames the victim. Obese people, at least the few who do seem to know it, are not happy about it. Studies have shown that obese people and especially obese children are terribly persecuted by it. Shouldn't we have some effective treatment for those people who actually come to us for answers to their obesity problem before we impose this grief on the apparently oblivious? To say that the first step toward effective treatment of their predicament is to make them admit to it, makes it sound like that's at least part of the problem. If they only knew, they would be doing the right things to fix it. Maybe this cruelty on the part of doctors could be useless advice like the jobs old time doctors used to give when they really had nothing to offer, like ripping up sheets or boiling water, so people wouldn't feel powerless. Also measuring and reporting BMIs makes it look like we are paying attention and trying to do something.

One group of people who DO know they are obese are those who want surgery for it. Cincinnati Children's Hospital has a busy adolescent obesity surgery service. They reported in April of 2006 on 33 consecutive extremely obese adolescents asking for weight loss surgery.(13) High rates of quality of life problems were documented, way more than would be expected from the medical problems, like diabetes, that can accompany obesity. These poor souls also had a 30% rate of clinical depression by their own reports and 45% by their parents' reports. This 15% discrepancy must mean that some of the obese adolescents, who are not in denial about obesity, are in denial about depression since they looked more depressed to their parents than they admitted to. All of these problems are well known among the non-thin and should be part of the indication for this drastic surgical treatment. Adolescence is tough enough if you have a beautiful body. In the same issue of Pediatrics as this surgery study, there is another example of this liability.(14) Investigators, from the National Institutes of Health in Bethesda, Maryland, found 146 children aged 6 to 12 years who were at increased risk for adult obesity, which meant that they were overweight already (half of them) or their parents were overweight. These kids were examined and questioned about dieting, binge eating, disordered eating attitudes and depression. Then they were followed them for an average of 4 years. Binge eating and dieting predicted increases in body fat. Neither depressive symptoms nor disturbed eating attitudes were predictive. So disturbed eating ATTITUDES didn't hurt but disturbed eating BEHAVIORS (=binge eating and dieting) did. Things are already getting weird before adolescence. In the discussion of these findings they say, "..most likely, repeated dieting attempts before adolescence may reflect efforts by children and their parents to prevent the onset or worsening of obesity among children with unusually rapid, unremitting, weight gain. Multiple childhood dieting efforts would then be a marker for extreme susceptibility for weight gain." Are they proposing that dieting and binge eating could cause obesity, rather than the more reasonable conclusion that these behaviors are just futile in the people who have the greatest incentive?

Another one of the studies about the emotional liability of obesity in children is one from Australia.(15) Two thousand eight hundred and thirteen 11 year olds were asked to complete the Self-perception Profile for Children(16) and then their heights and weights were measured. They were trying to figure out why some domains of competence were impacted and not others in obese children, and why some obese children had low self esteem and not others? I'm not sure these investigators answered those questions very well but one thing that was mentioned in their discussion was the observation that in Mexico, where parents value child fatness as a sign of health, and food treats are socially important, obese school children from affluent families are indistinguishable from their normal weight peers in self-esteem and other psychological or social outcomes. So the social environment probably makes the most difference. The United States is more like Australia than Mexico in it's attitudes about obesity. Knowing your BMI is worse than useless.

1. Keys, Ancel; Fidanza, Flaminio; Karvonen, Martti J.; Kimura, Noboru; Taylor, Henry L. Indices of relative weight and obesity. Journal of Chronic Diseases 1972; 25 (6-7): 329-43.

2. Wright EJ, Whitehead TL.Perceptions of body size and obesity:a selected review of the literature. J Community Health. 1987 Summer-Fall;12(2-3):117-29.

3. Baughcum AE, Chamberlin LA, Deeks CM, Powers SW, Whitaker RC. Maternal perceptions of overweight preschool children. Pediatrics 2000;106(6):1380-6.

4. Maynard M, Galuska DA, Blanck HM, Serdula MK. Maternal perceptions of weight status of children. Pediatrics 2003; 111(5 Part 2):1226-1231.

5. Foti K et al.Trends in perceived overweight status among overweight and nonoverweight adolescents. Arch Pediatr Adolesc Med. 2010;164:636-42.

6. Eckstein KC, et al.Parents' perceptions of their child's weight and health Pediatrics; 117:681-90

7. Chaimovitz R, Issenman R, Moffat T, Persad R. Body perception: do parents, their children, and their children's physicians perceive body image differently? J Pediatr Gastroenterol Nutr. 2008 Jul;47(1):76-80.

8. L'abée C, Visser GH, Liem ET, Kok DE, Sauer PJ, Stolk RP. Comparison of methods to assess body fat in non-obese six to seven-year-old children. Clin Nutr. 2010 Jun;29(3):317-22. Epub 2009 Dec 29.

9. Yu OK, Rhee YK, Park TS, Cha YS. Comparisons of obesity assessments in over-weight elementary students using anthropometry, BIA, CT and DEXA. Nutr Res Pract. 2010 Apr;4(2):128-35. Epub 2010 Apr 28.

10. Robert M Malina and Peter T Katzmarzyk. Validity of the body mass index as an indicator of the risk and presence of overweight in adolescents. Am J Clin Nutr.1999;70:131S-136S.

11. Zugua M, et al. Validity of body mass index compared with other body-composition screening idexes for the assessment of body fatness in children and adolescents. Am J Clin Nutr 2002;75:978-85.

12. Cole TJ et al. What is the best measure of adiposity change in growing children:BMI, BMI %, BMI z-sore or BMI centile? European J Clin Nutrition 2005;59:419-25.

13. Zeller MH, Roehrig HR, Modi AC, Daniels SR, Inge TH. Health-related quality of life and depressive symptoms in adolescents with extreme obesity presenting for bariatric surgery. Pediatrics. 2006;117:1155-61.

14. Tanofsky-Kraff M, et al. A prospective study of psychological predictors of body fat gain among children at high risk for adult obesity. 2006;117:1203-9.

15. Franklin J et al. Obesity and risk of low self-esteem: A statewide survey of Australian children. Pediatrics. 2006;118:2481-7.

16. Hunter S. Manual for the Self-Perception Profile for Children. Denver, CO; University of Denver; 1985. Clin Nutr. 2010 Jun;29(3):317-22. Epub 2009 Dec 29.