Pub. 7 2023 Issue 2

Why Health at Every Size Matters By Sara Walker, MD, MS This version of an apocryphal quote (attributed to many people, including William Osler and Carl Sandberg) was often repeated during my medical school education, including the start of our first lecture. The challenge, of course, is revealed in the rest of the quote — namely, that we don’t know which part it will be. Despite the quote’s repetition, we were rarely told the degree of confidence we could rely upon for each fact being shoved into our brains or how strong (or not) the evidence was behind the medicine. Much like this quote, society and the medical system start to believe statements if they are repeated enough. It can then be challenging to unseat long-held beliefs, even after evidence is published to the contrary. What is common knowledge gets accepted as truth without questioning the underlying evidence basis or even the potential for implicit bias driving its origin. Such as the much-vaunted BMI. The Body Mass Index (BMI) was created by a European actuary to determine the population statistics of Caucasian males. It was never intended for widespread dissemination nor clinical use.1 BMI was never validated on the overwhelming majority of the world’s population, i.e., females and nonCaucasians, and yet it is the defining characteristic of whether someone is “sick enough” to deserve treatment coverage for multiple diseases. For example, the DSM-5 mandates that a patient must be underweight in order to receive the diagnosis of “anorexia,” and BMI is how the severity is defined — not the presence of other factors such as bradycardia, hypokalemia or low bone density, but purely on the BMI. This bias/ distinction has also spawned the fatphobic diagnosis of “atypical” anorexia, in which a patient has all the characteristics of anorexia except for low BMI.2 This bias causes treatment initiation for “atypical” anorexia to be significantly delayed because of this distinction, despite the high mortality risk of eating disorders.3 In addition to the mental struggles a patient may face of not being “good enough” at having an eating disorder, insurance companies often deny coverage for things such as inpatient/ residential stays or even regular dietician visits, purely due to coding F50.89 (other specified feeding and eating disorders) rather than F50.01 (anorexia nervosa, restricting type). A patient can be seen in a family physician’s office for fatigue after having lost 40% of their previous body weight through severe restriction, be congratulated for losing weight and then be told to lose more weight to treat their fatigue — true story. Or the creation of obesity as a disease. In addition to the BMI being extrapolated from its original purpose, being considered “normal,” “overweight” or “obese” is purely based on a patient’s BMI. For example, the population map showed a stark change in 1998, when the normal/ overweight cutoff was lowered from 27.8 (men) or 27. 3 (women) to 25.4 The change was political, as studies actually showed that higher BMI was not associated with mortality until at least 40 was used as the cutoff.5,6 Ancel Keys, head of the Minnesota Starvation Experiment and also known food company spokesperson, was the one who advocated for BMI as the general measure of “fatness” without disclosing his conflicts of interest — after all, “50% of what you learn in medical school will be proven wrong.” Figure 1. Is this person healthy? Despite what appearances may suggest, her labs show K 3.1 and WBC 4.1 from daily caloric restriction of approximately 900 cals, with no purging behaviors. These behaviors led to constant fatigue even before starting residency and losing 200 lbs of force on leg press after her body cannibalized muscle for fuel. Less than six weeks after this photo, she began anorexia treatment the same week as PR’ing a half-marathon. She started treatment almost seven years after she started meeting the criteria for (atypical) anorexia after having lost nearly 40% of her body weight. Figure 2. With the change in BMI category definitions in 1998, the prevalence of obesity appears to have significantly increased between 1990 and 2000.41 23 |