When my younger sister first told me she wanted to become a vegetarian, I was worried. My biggest fear was that she would, like I did, develop an eating disorder. In high-school, I didn’t eat meat for roughly 14 months, and though I can’t be sure now of what my reasons were at the time, in retrospect, I do think in large part it was just a convenient way to avoid yet another food group. It was a legitimate reason to restrict my intake.
But is there any evidence that this behaviour (becoming vegetarian as a convenient way to restrict intake) is common among individuals with eating disorders? What is the relationship between dietary restraint, eating disorder symptoms, and vegetarianism? Is vegetarianism a risk factor for developing an eating disorder or do eating disorders lead many to adopt a vegetarian diet as a socially acceptable excuse to avoid eating specific foods? And, is there a difference between vegetarians that do not engage in dietary restraint and do not display eating disorder symptomatology, and those that do?
Thus far it appears that the research in this area is both contradictory and filled with methodological problems:
- Researchers have found inconsistent results when evaluating the level of dietary restraint between omnivores and “vegetarians” (I’ll get to why I put this in quotes). Some found higher levels of restraint in vegetarians (Barr et al., 1994; Gilbody et al., 1999; Trautman et al., 2008; Worsley & Skryzpeic, 1998), others found the opposite (Curtis & Cormer, 2006; Janelle & Barr, 1995) and yet others found no differences at all (Barr & Broughton, 2000; Fisak et al., 2006; Larsson et al., 2002). So, there is no real consensus.
- What’s more, a lot of these studies lumped vegetarians (ovo-, lacto-, and ovolacto-vegetarians) into one group, while others lumped semi-vegetarians (pesco-, pollo-, and pescopollo-vegetarians) and ovo-, lacto-, and ovolacto-vegetarians into one group. (You can read about what these mean here.)
- Finally, higher restraint scores in the vegetarian groups may not mean increased disordered eating behaviour but might just be a by-product of the question types. Of course they restrict their intake in a sense (they are cutting out a food group!) but it may not be maladaptive.
What about direct studies on whether or not vegetarianism is a precursor to eating disorders? Two studies evaluated this question and found that roughly half of the individuals in treatment for anorexia nervosa were “vegetarian” (in one study that just meant they didn’t eat red meat). Interestingly, the vast majority (~95%) began to adhere to a “vegetarian” diet only after the onset of their ED (Kadambari et al., 1986; O’Connor et al., 1987).
Wanting to improve on the previous studies, C. Alix Timko and colleagues set out to figure out whether there are differences between vegans, vegetarians, semi-vegetarians, and omnivores in attitudes towards food and health, and eating disorder symptomatology (study #1). They also wanted to further categorize the differences between omnivores and semi-vegetarians (study #2). I’ll write about study #1 here and cover study #2 in my next post.
They recruited participants through university psychology departments, flyers, and the internet. All participants were at least 18 years old. Table 1 shows the demographic information of each subgroup (%ages calculated from the total sample size in each group).
The participants then had to complete a whole bunch of questionnaires to evaluate their eating patterns, self-esteem, dieting behaviours, emotional status (depression, anxiety, stress) and so on.
They also asked the participants what their reasons were for starting and continuing the vegetarian diet:
If you look at the table above, you’ll see that semi-vegetarians were more likely than vegetarians and vegans to start and continue the diet for primarily health, as opposed to ethical reasons (they had to pick one main reason).
EATING ATTITUDES TEST
Although there were no statistically significant differences on the EAT (Eating Attitudes Test)-26 test, which examines disordered eating, the numbers approached significance semi-vegetarians scored higher than vegetarians, followed by vegans and omnivores (semi-vegetarians>vegetarians>vegans>omnivores).
DRIVE FOR THINNESS
Interestingly, vegans and omnivores had higher scores than vegetarians and semi-vegetarians (vegans, omnivores>semi-vegetarians, vegetarians).
Semi-vegetarians had higher levels of restraint than omnivores and vegans (semi-vegetarians>omnivores, vegans). Vegetarians were in the middle and did not differ from any of the groups.
There were no differences in depression, anxiety or stress between the groups. And individuals reported similar levels of physical activity.
When the authors compared the differences between the aforementioned test scores between those who chose the vegetarian diet primarily for weight control reasons to those who did not, they did not find any differences in the levels of disordered eating.
The conclusions the authors drew from this data was that although the samples were all rather healthy, the semi-vegetarian group did exhibit more disordered eating patterns and disordered eating attitudes than any of the other groups. (But, don’t forget about the unexpected finding that vegans and omnivores had higher levels of the drive for thinness than the semi-vegetarian and vegetarian groups.)
The semi-vegetarians had a more disordered eating pattern as evidenced by higher levels of restraint, external eating, hedonic hunger, and avoidance of food cues. Although not significantly different, semi-vegetarians had higher EAT-26 total and diet scores than the other groups. In contrast, vegans and omnivores did have higher (and statistically equivalent) scores on the EDI-DFT [Eating Disorder Inventory, Drive for Thinness Subscale] than either vegetarians or semi-vegetarians.
All in all, this study suggests that previous findings which indicated that vegetarians engage in more disordered eating patterns might have been due to the fact that vegetarians and semi-vegetarians were grouped together. Timko and colleagues hypothesized that those findings might have been due to the relatively high proportion of semi-vegetarians included in the vegetarian groups. In this study, if participants said they were vegetarian but it later turned out they just cut out red meat, they would be put into the “omnivores” group, whereas in some other previous studies, these red-meat avoiding participants would be grouped under “vegetarians.” (Misleading, don’t you think?)
The authors do warn that these findings are preliminary and more research needs to be done to replicate these findings. This is particularly so because there were many comparisons made in this study (between each group and every other group for every test and test-subscale) that the chances of type I error (a false positive) are significant (Read about the problem with multiple comparisons here, or here). So, it will be important to replicate this study with planned comparisons. (I’m not a statistician, but if this confuses you, do ask me, I can explain it in simpler terms.)
I don’t know much about the proportion of females to males among vegetarians, but I do wonder if the researchers in this study analysed the questionnaire results from males and females separately, would they find anything interesting or significantly different between male and female responses? And is that difference, if it exists, due to a real difference in behaviours or attitudes, or a by-product of questions being interpreted differently by different genders?
The vegan and vegetarian groups were overwhelmingly composed of females in this study. Indeed, overall 3 out of every 4 participants were females. So, analysing the genders separately here would be difficult (the male groups would be too small), but I do think it is something to consider.
As mentioned above, this paper had two studies. I’ve reported on the first one in this post and I’ll report on the second one, which compared eating-related attitudes and behaviours between semi-vegetarians and omnivores in more detail, in the next post.
Oh yes, and getting back to my sister, she’s still a vegetarian (it is been a few years now), for primarily ethical and environmental reasons. And thankfully, she doesn’t have an eating disorder or any disordered eating behaviours (as far as I know, anyway).