A Non-BMI-Centric Approach To Diagnosing And Managing Obesity

Regular readers will be well aware of the limitations of applying BMI to obesity diagnosis – after all, BMI is a measure of size, not health. If there is one thing we have learnt, it is that good health is possible over a wide range of shapes and sizes and that using a static measure like BMI, will mean overdiagnosing obesity in people who have no relevant impairment in health and underdiagnosing obesity in people who would in fact stand to gain from obesity treatments.

As I have noted before, in a medical context, obesity should be defined as, “the presence of excess or abnormal fat tissue that impairs health“. In clinical practice this would mean asking the question (irrespective of BMI), “does this patient have a health impairment that is likely to get better with obesity treatment?” If yes, the patient most likely has obesity and should be offered obesity treatment. If not, the patient does not have obesity and will be unlikely to benefit from obesity treatment.

This approach would identify both the “high-BMI” individuals with medical issues likely to get better with obesity treatment, as well as the “normal-BMI” individual, who may stand to benefit from obesity management.

Not only will this “common-sense” approach to diagnosing obesity identify individuals over the entire BMI range, who would potentially benefit from obesity treatments, but will also help set specific targets for assessing the success of treatment.

Thus, if the presenting problem is hypertension (say in a patient with a BMI of 24 with clear signs of increased belly fat – but skinny arms and legs), then the goal of obesity treatment would be to lower blood pressure, rather than to simply reduce body weight. Similarly, a patient with a BMI of 24 with type 2 diabetes would likely benefit from obesity management in terms of better diabetes control. If, in these patients, effective obesity treatment (as measured by weight loss) does not lead to better hypertension or diabetes control, then their health issues are probably not related to their body fat, meaning that they probably don’t have obesity.

Thus, the obesity diagnosis and management algorithm would look something as follows:

Does the patient have any impairments in health likely to improve with obesity treatment? (list impairment/s)

if no, patient does not have obesity and does not need any obesity intervention (irrespective of BMI).

If yes, offer obesity treatment to see if this improves the health issues.

If health condition improves, continue obesity management.

If health condition does not improve, reconsider the obesity diagnosis (problem may be entirely unrelated to body fat). discontinue obesity treatment.

Basically, BMI, or for that matter body weight, does not have to enter the equation. The only thing that matters, is whether or not the presenting health problem actually responds to obesity treatment or not.

Fortunately, we have a long list of health issues that we know will improve with obesity treatment (e.g. hypertension, type 2 diabetes, sleep apnea, psoriasis, musculoskeletal pain, knee osteoarthritis, PCOS, etc. etc.), thus making clinical assessment of whether or not someone may have obesity relatively straightforward. Considering obesity treatments to better manage these conditions makes a lot of sense. On the other hand, treating obesity simply to lower body weight (without any measurable health benefits) does not.

Copenhagen, Denmark