Modeling - What Is Normal, Anyway?

One of my jobs is at an eating disorders partial hospitalization and intensive outpatient program. A treatment goal for all of our clients, across the spectrum of eating disorders, is to "normalize" eating behavior. One of the main interventions by which we work on this goal is supervised meals. As staff, we often eat with the clients - to make it less awkward, and to model "appropriate" or "normalized" eating.

Therefore, a perennial question among staff is: What is appropriate to eat with clients? What is normal?

It's comparatively easy to determine whether clients are eating what they should be eating. They are all on meal plans, so we just have to make sure 1) that they're eating the exchanges that are on their meal plan for a given meal, and 2) that they aren't following "abnormal" eating rituals. The latter includes "diet" or "light" foods, unusual food combinations, unusual patterns of cutting/slicing/biting/pacing themselves, and excessive condiments, or foods not considered appropriate to the meal in question (e.g., waffles for grains at lunch).

While there is some variation in what staff consider normal vs. abnormal eating behavior for clients, it becomes even less clear when we stop to think about our own eating - and whether we can eat that way in front of clients. It's complicated by the fact that 1) we aren't on meal plans, 2) we tend to be willing to eat a wider variety of foods, 3) we can eat certain foods (e.g., lean cuisine or diet coke) without it being part of a dangerous cognitive, emotional, and behavioral pattern, 4) there is natural variability in what food combinations people and cultures eat, and 5) many or most women in this culture have some disordered eating habits.

The most cautious response is to have staff follow the program's standard meal plan when eating with clients. However, while safe, this strategy isn't necessarily "normalized." For example, the standard meal plan only includes 1 grain with lunch, but I like to have a sandwich, with 2 pieces of bread. Trying to develop a less rigid policy becomes tricky, however. We definitely don't want to eat diet foods in front of clients - we don't want to portray "normal" eating as having to choose diet or low-fat options! But, if we don't let them have soda, should I avoid drinking my regular coke in front of them? Can I drink water for a meal, if they're drinking milk? I'm not going to measure out my foods like they do (that's not normal), but will it therefore look like I'm eating less? The possible pitfalls are numerous, both because of the ranger of staff's personal preferences, and because we have no way of knowing exactly how clients are perceiving our behavior.

Another dilemma that comes up is how to "normalize" social interactions during meals. We want to model "normal" mealtime conversation, while also guarding against dysfunctional group dynamics (i.e., competition among clients). As a result of the latter consideration, we don't let them talk about food, meal plans, or what anyone is (or isn't) eating. However, it's very "normal" to talk about food during a meal! It's sometimes hard for me to remember not to comment on something smelling good, or to mention a favorite food, or whatever. Because eating is anxiety-provoking for clients, it's common for staff to offer some form of distraction, for example trivia or other group games. This habit is also not particularly "normal" - most people engage in small talk or conversation during meals. Therefore, I prefer to try to model mealtime conversational patterns by facilitating conversation.

However, that puts me square in the middle of another therapeutic quagmire: small talk involves some amount of self-disclosure. It's considered "small" talk because it steers away from areas that may be vulnerable...but, we know as therapists that any bit of self-disclosure may have some greater significance to clients, or impinge in some way on the clinical work (via clients' assumptions, judgments, assumptions about our judgments, sense of themselves as similar to or different from us...etc.). In modeling "normal" mealtime interactions, I'm sharing things I would most likely not share in the course of typical treatment - where I'm from, where I went to college, what sports team I root for, etc. Nothing huge, but still, little details about myself that might color a client's later interactions with me. What is safe to share, and what isn't? In a group setting, with high client turnover, it's hard to know for sure.

As you can see, the simple question of "what is normal" is incredibly complex when asked about meals in a treatment context. What we eat, do, and say - about anything - has clinical significance. I therefore have no simple answers to guide a policy for myself and colleagues. Do you face similar dilemmas in your work? What do they center around? How do you navigate these kinds of challenges?