Restricting vs. Binge Eating: Surveyed Differences in Attitudes and Food Preferences

How do people with EDs feel about different ED behaviors? Also, how do food preferences in restriction relate to food preferences in binge eating?

Introduction

There is no single predictable set of attitudes for those with eating disorder behaviors, but a general understanding of ED attitudes is important in discussion, research, and clinical practice. To explain EDs, one should realize what motivates behaviors, how behaviors make one feel, and why behaviors may be resistant to change.

My previous survey found a range of attitudes associated with EDs — for example, they may be “a source of pride that one wishes would be recognized” or “a shameful secret that one tries to hide”— but did not apply attitude questions toward specific behaviors. Perhaps one feels successful in restriction and sick in binge eating, and a conflicted response to how one feels about their “ED” does not fully capture the difference.

Researchers may distinguish between different ED diagnoses, but this risks ignoring similarities across diagnoses. For example, bulimia nervosa, anorexia nervosa binge/purge subtype, and “other specified feeding/eating disorder” may all include individuals who restrict, purge, and binge; they are more separated by the frequency and duration of their behaviors (and weight classification) than the ways to explain and treat the behaviors.

This survey examines attitudes associated with ED behaviors, in hopes that the “mixed” nature of ED attitudes can be separated into more predictable patterns. Additionally, by separating respondents by their combinations of ED behaviors, this survey seeks to test claims about how food preferences in restriction relate to food preferences in binge eating.

Sample

To be eligible for the survey, one must be engaged in any of the following behaviors:

• restriction (here defined as deliberately eating a number of calories below the typical recommended minimum)
• purging (here defined as self-induced vomiting, laxative usage, or other compensatory methods to “eliminate” consumed food)
• binge eating (here defined with the provided DSM-5 criteria for binge eating)

The survey was advertised in online ED communities, much like the previous survey, and received 106 eligible responses. The distribution of behaviors is displayed below.

Number of respondents for each combination of ED behaviors

To summarize, a large majority of respondents engage in restriction, and both purging and binge eating have a small majority. The most common combination of behaviors is restrict+binge+purge. Many statistical tests in this report will compare two groups: the restrict+binge+purge group combined with the restrict+binge group, versus the restrict+purge group combined with the restrict-only group (in other words, binge vs. no binge among those who restrict).

Attitudes for Different ED Behaviors

For each reported ED behavior, respondents received a series of questions measuring attitudes toward that particular behavior, on a scale of 1 (“Strongly Disagree”) to 7 (“Strongly Agree”).

Historicist-Purposefulist Attitudes

A therapy approach that’s historicist-purposefulist, as coined by psychiatrist-writer Scott Alexander, claims disordered behaviors serve a purpose related to one’s past. Historicist-purposefulist therapy is a common form of ED treatment, but it may not be necessary; treatment centers that reject it, such as Mando Clinic and Kartini Clinic, report high patient success.

Regardless of whether it helps treat EDs, historicism-purposefulism is often helpful to explain one’s ED motivations. It seems that historicist-purposefulist explanations are most suitable for restriction and least suitable for binge eating, with purging in between.

Responses to the following questions contributed to the historicist-purposefulist score, from -24 to 24:

• +My personal past has contributed to [behavior].
• +I can identify what triggers me to [behavior].
• +[Behavior] serves a purpose for me
• +Norms of the society I live in contributed to [behavior].
• -I didn’t start [behavior] for a clear reason.
• -I cannot really explain what makes me [behavior].
• -Therapy for my emotional issues would not make me less inclined to [behavior].
• -It’s nobody’s fault that I have a [behavior] eating disorder; I just happen to be biologically susceptible.

Indeed, these claims are supported. Average attitudes toward restriction are more historicist-purposefulist than average attitudes toward binge eating (p = 0.00003), with purging apparently in between. Among those who binge and restrict, though, a higher historicist-purposefulist score for restriction is moderately correlated with a higher historicist-purposefulist score for binge eating (r = 0.3988, p = 0.0026).

There is no significant difference in historicist-purposefulist scores for restriction (p = 0.26) when comparing those who binge to those who do not binge; the same is true when comparing these groups’ historicist-purposefulist scores for purging (p = 0.41).

Involuntary vs. Voluntary Attitudes

Many EDs begin with a desire to feel “in control”, but over time, individuals with EDs often realize they are losing control over their eating behaviors; they cannot eat “normally” even if they try, and may have unwanted binge episodes that compel them to compensate. Still, they may insist that their weight loss is effortful and goal-driven, and that they are choosing not to recover simply because they do not want to “give up” yet.

In the previous survey, these contrasting attitudes were called compulsive vs. deliberate, but here, they have been renamed involuntary vs. voluntary. Responses to the following questions contributed to the involuntary score, from -24 to 24:

• +I wouldn’t know how to avoid [behavior] even if I tried
• +I engage in [behavior] even when I intend not to
• +My [behavior] habit controls me.
• +I wish people understood that [behavior] is not by choice.
• -I put a lot of effort into [behavior].
• -I don’t feel “urges” to [behavior]; I just decide to do it.
• -I feel like I’m in control when I [behavior].
• -I seek resources, tips, and/or new methods to engage in [behavior].

There is no doubt that lack of control is felt the strongest for binge eating (p < .00001). Purging also feels more involuntary than restriction (p = 0.00059). For restriction, higher historicist-purposefulist scores are associated with lower involuntary scores (r = -0.297, p = 0.028); apparently, one is more likely to believe restriction is done on purpose if it seems to be for a purpose.

There is no significant difference in the involuntary restriction scores between those who binge and those who do not binge (p = 0.30), nor in the involuntary purging scores between these groups (p = 0.40).

“Source of Pain” vs. “Coping Mechanism for Pain” Attitudes

In treatment, eating disorder behaviors are discussed as a “coping mechanism” for underlying distress. However, when one is so deep into an ED that it consumes one’s life, it becomes hard to see any connection between the ED and its original purpose; the ED becomes a source of pain in itself, something to cope with. The “coping mechanism” framing can sound especially out of touch when applied to binge eating, an egodystonic behavior that doesn’t feel as purposeful and voluntary as other behaviors.

Responses to the following questions contributed to the source-of-pain score, from -24 to 24:

• +My [behavior] habit creates more problems in my life.
• +[Behavior] brings me distress.
• +I wish I could live without [behavior].
• +[Behavior] feels like a form of self-harm.
• -My [behavior] habit serves as a coping mechanism.
• -I’d feel unpleasant emotions if I didn’t [behavior].
• -[Behavior] makes me feel better about myself.
• -If I had to give up [behavior], I’d need something to fill the void.

As predicted, binge eating is much more of a source of pain in one’s life than restriction (p < .00001), with purging in between.

Higher involuntary scores for restriction are associated with higher source-of-pain scores for restriction (r = 0.3474, p = 0.0094). For those who restrict without binges, restriction is much more of a source of pain (p < .00001). Even though restriction feels similarly voluntary in both groups, those who binge and restrict clearly view restriction as less of a hardship, probably because it is less present in their lives and/or they feel better about themselves when they do. This difference between the groups is not significant for purging, though (p = 0.29).

Food Preferences: Binge Eating vs. Restriction

Respondents who restrict were told, “For each of the following food groups, specify how strongly you would or would not like to include it during a day of RESTRICTION,” on a scale of 1 (“Very much would not like to include”) to 7 (“Very much would like to include”). With the same 1-to-7 scale, respondents who binge rated their preference for each food group during a binge episode.

By doing so, one could order the food groups from “more likely to be included in binges than restriction” to “more likely to be included in restriction than binges”: filter the dataset to only those who both restrict and binge, and for each food group, subtract one’s 1-to-7 restriction preference score from one’s 1-to-7 binge preference score, giving a new score (maximum 6, minimum -6), and then rank the food groups by their average score. The results are below:

The most common binge foods are calorie-dense and processed. This raises a question: are binge eating preferences biologically due to the foods themselves? Or, as “Intuitive Eating” dietitians claim, is it because calorie-dense processed foods are demonized in “diet culture,” creating a tempting “forbidden fruit” effect?

It’s clear that foods often eaten during restriction (e.g. vegetables, low-calorie fruits) rarely tempt one to binge, and the most common binge foods are typically not eaten during restriction. But this does not mean people with EDs binge on certain foods because they restrict them.

To test this, don’t compare one’s binge tendencies for restricted foods with binge tendencies for unrestricted foods. Compare whether those with restrict+binge(+purge) EDs who binge on a food are more restrictive of it than those with such EDs who don’t binge on it.

Put it this way: what do children find most boring to learn? Grammar rules, arithmetic, historical dates — basically, things they learn from school. What do children find most fascinating to learn? Probably things they don’t learn in school. Does that mean teaching something in school makes it boring? To find out, consider something that children may or may not learn in elementary school, like square dancing or art history. If those who learn it in school find it just as interesting as those who learn it outside of school, then the school setting probably isn’t why some topics bore children.

Similarly, consider foods that people with binge+restrict(+purge) EDs may or may not want to eat during restriction and see whether permission makes a difference. The table below shows the results of correlation tests — if the more a food is “off-limits” during restriction, the more one binges on it, the correlation coefficients (in the second column) should be close to -1. If the values are closer to 1, then eating a food during restriction is more associated with eating it during binges, and a value close to 0 indicates no correlation either way.

Turns out, binge eaters who allow a food when restricting are not less likely to want to binge on it than binge eaters who forbid that same food when restricting. For most food types, there is no significant difference; if anything, foods included in restriction might be more included in one’s binges, probably due to preference and availability. (The one exception is higher-fat animal-based protein sources among meat eaters.)

Next, let us examine whether a binge-eating problem, as a whole, is the result of extreme restriction, as IE would claim. If so, those whose EDs involve binge-eating would likely be more restrictive of foods than those with EDs who do not binge. Again, this view is not supported by the data.

The table above examines all the “binge foods” in the food preference questionnaire, excluding beverages and the foods that are much more favored in restriction than binge-eating. For virtually every food, those who would include it in binges seem just as (un)likely to include it in restriction as those who do not binge at all. If anything, binge eaters appear more likely to allow such foods, but most of these differences are statistically insignificant here.

The only statistically significant result is for sweetened diet products, consistent with findings above. Those with EDs who keep sweetened diet products may binge on them when cravings arise; those who do not eat them regularly would surely buy something cheaper, sweeter, and more energy-dense.

Overall, while binge foods and restricted foods appear to overlap, these findings suggest stronger restriction preferences don’t cause stronger binge food preferences.

Discussion

Describing people with EDs’ attitudes toward their “ED behaviors” misses key distinctions — different ED behaviors tend to be viewed differently. Restriction is viewed as more voluntary, less of a source of pain, and better explained by one’s past than binge eating, with purging in between. This suggests that different approaches are suited to address them, rather than generic treatment for avoiding “ED behaviors”; for example, the CBT exercise of examining “predisposing factors” and “perpetuating factors” may be more effective for restriction than binge eating, and “distress tolerance” urge-avoiding DBT skills may be more effective for binge eating than restriction.

Also, these findings weaken the claim that binge eating preferences are caused by restrictive attitudes toward those foods. Among those who both binge and restrict, one who avoids a food during restriction is not more likely to binge on it than one who allows that same food during restriction. Binge eating preferences are more likely due to the palatable, energy-dense properties of the foods themselves. If “Intuitive Eating” dietitians simply insist “diet culture” is the cause, patients may believe that permitting themselves to eat “forbidden foods” is the solution, leading to frustration if they continue to binge because the true cause is not addressed. It may be true that “restriction leads to bingeing,” as IE dietitians claim, but if so, the issue is likely restriction as a whole, not restriction of particular foods.

Thank you to all who participated! I wish you health and happiness.