Diagnosis of anorexia nervosa (AN) and bulimia nervosa (BN), while including such physiological data as weight and the reproductive status of the individual, are primarily based on questionnaires and interviews that rely on self-report of both body-related concerns and eating-related behaviors. While some key components of eating disorders are psychological and thus introspective in nature, reliance on self-report for the assessment of eating-related behaviors and nutritional status lacks the objectivity that a physiologically based measure could provide. The development of a more physiologically informed diagnosis for AN and BN would provide a more objective means of diagnosing these disorders, provide a sound physiological basis for diagnosing subclinical disorders and could also aid in monitoring the effectiveness of treatments for these disorders. Empirically supported, physiologically based methods for diagnosing AN and BN are reviewed herein as well as promising physiological measures that may potentially be used in the diagnosis of AN and BN.
I don't think any decent physician will discount a person's personal experiences; those do have meaning in treatment. But when you're confused or unaware of what's really going on, it helps to have some objectivity.
This would also show things like the level of starvation and malnutrition, which, to me, play just as much a role in eating disorders as weight and the presence of menses.