How Anorexia Damages Our Mental Capacities: Correlations

In the first part of this miniseries, I outlined the basics of what kinds of questions we can ask about anorexia and how the mind works. If we’re looking for evidence that anorexia nervosa goes hand in hand with impairment in cognitive functioning—the mind not working well—we don’t have to look very hard. More often than not, the two hang out together.

Evidence for correlations between cognitive impairment and anorexia

Let’s begin with executive functions, the collection of processes we rely on for cognitive control of behaviour, including attentional and inhibitory control and working memory. A systematic review of 64 eligible studies (including meta-analysis of 31) found that individuals with anorexia underperformed compared to healthy controls in tasks involving visual perception, decision-making, memory, social cognition, and set-shifting (moving between different tasks; Tomba et al., 2019).

A 2017 meta-analysis of 32 studies (Hirst et al., 2017) also found that executive functioning (including attention, planning, decision-making, monitoring, working memory, and inhibitory control) was impaired in anorexia, though less in anorexia than in bulimia. And a 2021 meta-analysis of 16 studies suggested that memory, working memory, and visuospatial abilities are affected, more strongly in older participants with anorexia (Stedal et al., 2021).

When it comes to general intelligence, verbal memory, and motor speed, a 2015 study (Kjaersdam Telléus et al., 2015) found that individuals with anorexia performed worse than healthy control participants.

As for emotion and social cognition, a systematic review of 32 studies found difficulties with regulation, expression, and awareness of emotions in anorexia (Tauro et al., 2021). The first meta-analysis I mentioned above (Tomba et al., 2019) also found evidence for deficits in emotion regulation. Diving into the subcomponents of empathy and theory of mind (our capacity to attribute mental states to other people), one study (Konstantakopoulos et al., 2020) found both diminished performance and inaccurate self-appraisal of empathic abilities in anorexia. Specifically, overall empathy and cognitive theory of mind were impaired, with cognitive deficits seeming to negatively affect the self-appraisal of empathy but not empathic skills per se.

Finally, there’s the big one: cognitive flexibility, or the ability to adapt cognitive strategies to a changing context. It’s often measured via set-shifting, as mentioned above: how readily we can switch back and forth between different activities or perspectives. A related measure is “central coherence," the ability to derive meaning from a collection of details—a literal measure of whether you can see the woods or just the trees.

It’s clear that both these measures of mental agility are strongly correlated with anorexia. As one paper puts it: “Anorexia nervosa (AN) has consistently been found to be associated with poor cognitive flexibility and central coherence” (Brockmeyer et al., 2022). Both are associated with poor quality of life, more directly when looking at self-report measures than at performance measures, and independently of depression. Poor central coherence and impaired ability to maintain a new attentional set correlate with both anxiety and quality of life (Hamatani et al., 2017). Correlations have also been found between amount of cognitive rigidity and amount of physical activity (in a pilot study by Lodovico & Gorwood, 2020). And the cognitive rigidity is linked to behavioural rigidity: not adjusting actions to environmental changes (Friederich & Herzog, 2011).

Stealing your capacity to be nimble in your thinking is a major component of what anorexia does to your life: "In AN, perseverative thinking and rigid behavior, including preoccupations with food, body shape, and compulsive exercise in the pursuit of thinness, are hallmarks of cognitive inflexibility. […] Cognitive inflexibility is thus fundamental to AN and is likely involved in the etiology of the disorder." (Dulawa, 2021)

We’ll come back to that kind of causal claim later in the series, but, for now, the research literature on cognitive flexibility aligns with what tend to be amongst the most miserable and tedious features of an anorexic way of life: rumination; rigid behavioural patterns; preoccupations with food, body, and exercise; and other forms of narrow, repetitive, and ritualistic thought and action.

Evidence against correlations between cognitive impairment and anorexia

Is there any counterevidence, finding no correlations between anorexia and impaired cognitive functioning, or indeed correlations between anorexia and improved cognitive functioning? Using tests of immediate memory, visuospatial/constructional, language, attention, and delayed memory, Rylander et al. (2020) report that “No baseline deficits in cognition were found in this sample of women with severe AN.” (They did, however, justify their use of a simple 12-item questionnaire, developed for use with older adults with cognitive decline, by saying that “our patient population would have had difficulty tolerating lengthier and comprehensive testing batteries given significant weakness and fatigue”—so I guess the unimpaired cognition lasts a good half-hour!)

Eating Disorders Essential Reads

When an Eating Disorder Isn't Just About Body Image

The Quiet Radicalism of Looking After Your Body

One study with a fairly young sample (average age 21) found impairment in information processing speed and planning in AN, but significantly better flexibility for the AN group than the healthy controls when adjustments were made for other cognitive functions and impulsivity, especially for a subgroup with less impairment overall (Cholet et al., 2021). But evidence of this kind is rare compared with the slew of evidence for impaired flexibility and other deficits.

One methodological factor to be aware of is that most of the studies I’m discussing involve performance measures that don’t tap into how individuals assess their own cognitive capacities. When self-perception is measured on standardized scales, there’s some evidence that it may not track performance particularly well.

One study finding impaired flexibility for people with anorexia versus controls found that self-reported cognitive flexibility (on a scale tapping agreement levels with statements like “I have many possible ways of behaving in any given situation”) was affected by other factors, including depression and insight, while test performance on a rule-changing task was affected by age (Lounes, Kahn, & Tchanturia, 2011). In part, this kind of disconnect just reflects the fact that if you measure different things, you’ll probably get different results.

But the potential gap between performance and self-assessment is worth bearing in mind when we consider the real-world relevance of the findings we’re discussing. Although insight into the damage done by an eating disorder may be high in many cases, sometimes persisting for many years without leading to the corresponding action, that doesn’t mean it has no biases or blindspots.

If we leave the lab and look at studies of how people actually experience their cognitive capacities in everyday contexts during illness, one study suggests an interesting twist on the strong pattern of the experimental results. Keeler et al. (2022) report that “Whilst respondents with AN appeared to perform remarkably well in their professional and educational lives, the cognitive difficulties were described as significantly impacting their ability to engage in life, particularly in the context of leisure and socialising.”

This gap between academic/professional achievement and the capacity to use your mind for things that aren’t work is pronounced in many anecdotal reports about anorexia. (I discuss the knotty relationships between anorexia and academic achievement from a personal perspective here.) To some extent, it’s as simple as: Sure, you can do most of what you need to, but not much of what you might otherwise have wanted to.

In part three we’ll look at what kinds of causal relationships might be hiding in amongst all these correlations.