Adiposity by Self-Report
If obesity were like ADHD
I often think about how the “invisibility” of mind — the inability of others to access our inner, subjective experiences through anything other than self-report and behavioral inferences — complicates the recognition and management of mental health problems. What if obesity were invisible in the same way as ADHD is? Such a comparison is hard to imagine and requires some suspension of disbelief but bear with me.
What if we couldn’t see adiposity, couldn’t measure it in terms of BMI, waist circumference, or body fat percentage, and had to rely on something indirect like self-perception and other-perception of being obese? This clinical construct of self/other-perceived obesity would not be identical to our current conception of obesity, a chronic state of excess adiposity that impairs health. It would be heavily mediated by subjective perception and shaky behavioral inference. The clinical criteria would invoke items like feeling heavy or weighed down; feeling full or bloated out of proportion; short of breath or fatigued on exertion that others manage; a persistent sense of occupying too much space; eating without hunger or grazing; difficulty stopping once started; eating in response to affect; preoccupation with the next meal; avoiding activities because of perceived size, etc, etc. People would obsess over whether your parents thought you were fat as a child or you’ve only become fat as an adult, and if that “counts.” Who gets believed in the perceived obesity world? The conspicuous eater, the person who visibly labors up stairs. Who gets believed in ADHD? People resembling the hyperactive-boy template. It would not even be clear to many that perceived obesity can be or should be reconceptualized in terms of adiposity because so many clinical features, such as feeling heavy or preoccupation with food or feeling fatigued on exertion are not specific to excess body fat.
Obesity is a clinical problem and weight loss for obesity is the treatment. But we also understand very well that even people in the (non-obese) overweight and “normal” weight range desire weight loss. They desire it for sociocultural reasons, for psychological reasons, for physical fitness reasons. Weight loss in the absence of clinically excessive adiposity is considered enhancement rather than treatment per se, but the boundaries are fuzzy. And we also know that some people desire weight loss for “pathological” ends, to fuel their eating disorders, for example. We can adjudicate these cases because we can measure body fat percentage, or more commonly, we have anthropometric surrogate measures like BMI. But what if you couldn’t? How would you reliably, unproblematically distinguish weight loss for pathological, enhancement, and treatment goals if you had no way to measure body fat percentage or use surrogate measures like BMI? In such a world, it’s easy to imagine that individuals with eating disorders would self-diagnose as obese and would want access to GLP-1 as a treatment for obesity. We would have entire communities dedicated to explaining how anorexia is a poorly recognized form of obesity and if only we didn’t gate-keep access to GLP-1 and believed patients when they told us they are fat.
Such a state of affairs would also lead many to be skeptical of the existence of obesity itself. It’s all a sham! It’s all enhancement and self-destruction! People need to live with obesity. Diet and exercise if you want. Focus on social determinants of health, don’t make it a problem inside a person.
Studies conducted in samples inadvertently enriched for excess adiposity would show tremendous benefits of weight loss treatments across a wide range of measures, including quality of life and mortality. The default medical stance would be that obesity is under-treated, rightly so. And yet, anorexics claiming to be obese and starving themselves with GLP-1s would also be a well-recognized phenomenon, leaving many people confused and vexed, generating endless commentary on how the diagnosis of obesity had been stretched too far, had lost all validity, had been gentrified. And much of this confusion would be dispelled by a definitional physical substrate (adiposity) or a physical proxy measure (BMI).
This is the situation we find ourselves in with ADHD and stimulant treatment. ADHD is as real a clinical problem as obesity, but our epistemic access to it is indirect and mediated via self-report and behavioral inference. Unlike obesity, we do not (cannot?) define it in terms of something we can objectively measure. We do not, e.g., define it in terms of attentional impairments on neuropsychological testing. Perhaps ADHD is as muddled and heterogeneous as “perceived obesity” in my thought experiment which can’t be strictly defined in terms of adiposity either. ADHD is a construct that seems to encompass motivational and task initiation difficulties, executive function difficulties, and impulse control difficulties. And stimulants are attractive to people for a lot of reasons. They offer significant benefits to many people who meet diagnostic criteria for ADHD, but they are also desirable for people who don’t technically have ADHD but have some other problem. People who are struggling to manage motivation, boredom, energy, focus, sleep, etc. Individuals with obsessive and perfectionistic tendencies can worry about having attentional impairments without actually having any attentional impairments. Similar to weight loss treatments being used in the service of an eating disorder, stimulants can be used in a harmful way. Addiction and abuse is one example, but also situations where people are desperately trying to maintain an unsustainably active or productive lifestyle as well as people who are in a stimulant-fueled mania or psychosis. Fitting for our analogy: stimulants are also actively sought and misused for appetite suppression by people with eating disorders.
Nearly everything people cite as evidence against ADHD — contested thresholds, shifts in thresholds and prevalence, the demand for treatment, the simultaneous visibility of medication misuse, swinging clinical opinions between suspicion and enthusiasm — follows from the epistemic situation alone, and would appear just as reliably around a condition few otherwise doubt. (This is a defense of ADHD’s reality only in the negative sense of removing some bad arguments against it.)
Physical measures and shared external yardsticks don’t resolve all the problems of clinical diagnosis. Obesity demonstrates that. Thresholds can still be contested. Surrogates can be highly imperfect. We still have to appeal to evaluative judgments like “excessive” and “dysfunctional.” The appropriate response — intervening on the person vs intervening on the environment — can still be debated. A physical substrate doesn’t tell us what the causes are. And effective treatments can end up targeting processes tangential to causes with the potential for uses and misuses. But at least a public anchor allows us to adjudicate, however imperfectly, between interventions being employed for treatment, enhancement, and pathological ends.
See also:
Psychiatry at the Margins is a reader-supported publication. Subscribe here.
Two textures worth adding.
The enhancement branch is already running at scale. Undergraduates on many campuses have learned the ADHD self-report script well enough to walk out with Ritalin specifically for exam season. It's not a boundary case. It's how the diagnostic infrastructure actually functions when the incentive gradient points one way and self-report is the only door in.
The second one I'm less sure about, and it's the one I've been circling in my own writing on layered honesty about diagnosis. The obesity substrate you use as the stable anchor is itself institutionally indexed. In a food-scarce environment, efficient fat storage was adaptive, not pathological. What changed wasn't the substrate. It was the world the substrate landed in.
If that's right, the same move might apply to ADHD. Traits that read as attentional pathology in an environment demanding sustained focus on sedentary abstract tasks weren't necessarily pathological in the environments those traits evolved in. The industrial revolution didn't just make our epistemic access indirect. It made "attentional pathology" a category the substrate is being read through.