As shown in Fig. 2, a period of caloric restriction causes BERs to eat significantly more food then when sated. This increase consists of greater chow intake, which is typical of rats hungry from metabolic deficit (25, 31). BEPs, too, eat proportionally more chow than when sated and so appear to respond normally to hunger. Also, because hungry BEPs do not eat more total kilocalories than hungry BERs, it can be implied that they also have normal satiety. However, as also shown in Fig. 2, the amount of total kilocalories that BEPs eat under restricted conditions matches the amount of calories they take in under sated conditions. This behavior is clearly abnormal especially given the behavioral indices of normal hunger and satiety cues in these rats. The responses just described are observed in the 4th hour of feeding but can all be observed as early as after the 1st hour of feeding (5). Clinical binge eating also appears to be unaffected by hunger and satiety cues. Indeed, hunger is one of the weakest triggers of binge eating and binges are diagnostically defined as occurring in the absence of hunger (1, 32–34). In humans, a more potent trigger is stress (34–38).

As seen in Fig. 3, at 2 h following foot shock, stressed BERs consume less food than when not stressed. This is the expected and normal response of rats to laboratory stressors (26, 27). However, stressed BEPs fail to display this normal hypophagia and in fact appear to be completely unaffected by shock. By 4 h, stress causes BERs to remain hypophagic. Notably, they are eating less PF, not less chow. By this time BEPs appear somewhat affected by the stress, but their decrease in intake is due to forsaking the healthy chow over PF. Their PF intake remains abnormally elevated. By 24 h, the intake of each group normalizes to match their counterparts’ intake under nonstressed conditions. The behavior of BEPs under stress resembles that of human binge eating in that stress triggers overeating versus undereating and is associated with increased consumption of PFs (34–37). In human binge eating, stress may actually make PFs more rewarding (38). Just how rewarding BEP rats find PF can be seen by how much punishment they are willing to tolerate for it.

As shown in Fig. 4, BEPs make significantly more M&M® retrievals than BERs. This difference reaches significance at shock levels of 0.25 mA and higher. At 0.40 mA and higher, only one BER versus 8 BEP rats braved shock for M&M’s®. Only BEP rats continue to cross at 0.60 mAs (Fig. 4). As also seen in Fig. 4, the retrieved M&M’s® are consumed as evidenced by the 2-fold greater kcal intake of M&M’s® by BEPs versus BERs across all shock levels (16). In sum, the BEP rats’ willingness to tolerate increasing pain and anxiety associated with foot shock models the addictive-like nature of human binge-eating where motivation to binge-eat persists despite the mounting psychological and physical consequences directly associated with this behavior (1, 39–41).

When BEP and BER rats are fed a high-fat diet for 2 weeks, exactly half of the BERs and half of the BEPs develop obesity while the other half of each BEP and BER group resist weight gain. The obese-prone rats gain approximately 8.3% of initial body weight vs. a 1.9% gain by the obese-resistant rats (p  <  0.01) (5). This is due to the obese-prone rats’ failure to decrease their normal volume of food when forced to eat the more calorie-dense high-fat diet. Importantly, BEPs are as likely to do this as BERs. Each group is also as likely to voluntarily restrict the amount of the high-fat diet which results in maintaining normal weight.

At the end, the DIO protocol yields four subgroups that can be used to investigate possible biological differences between bulimia nervosa, BED, non-binge-eating obesity, and healthy controls (see Table 1). For example, some individuals with bulimia or BED resist obesity through compensatory behaviors, including limiting caloric intake (1, 42). Similarly, the obese-resistant BEP rats reduce their volume of high-fat intake, thereby reducing total kilocalories consumed to maintain normal body weight. Not all human motivations to restrict caloric intake can be modeled in rats, but there may be a common physiology between bulimia nervosa patients and BEP-obese-resistant rats that enables them to reduce caloric intake amidst PF, a physiology possibly compromised in BEP-obese-prone rats and obese individuals with BED.

As is typical of the BEP/BER model in the author’s hands, Klump and colleagues found that roughly one-third of an initial group of thirty rats could be clearly classified as BEPs and one-third as BERs based on their significant difference in amount of PF intake (17). Of major importance is that when they observed PF intake patterns across time, they found that the onset of the BEP phenotype does not appear until mid-late puberty (P39-P58). This seminal change in PF intake occurred as chow intake during chow-only days, and body weights, remained the same for both groups. Only PF intake differed. The emergence of PF binge eating shortly after puberty was replicated in a separate squad of N  =  36 rats that were exposed to more frequent feeding tests during puberty (three times per week vs. one time per week) (17). The results are a compelling parallel to the age of onset for eating disorders, which is after puberty (1, 2, 43). Results from the OVX study revealed that OVX caused an expected increase in general food intake and body weight across all hormone-depleted rats, regardless of BEP or BER status. However, the OVXed BEP rats still continued to eat significantly more PF than the OVXed BER rats (21). The fact that BEP/BER patterns remained stable despite removal of estradiol indicates that other—yet unknown—signals activated at puberty are needed to express binge eating. While these results were at first surprising given that the BEP pattern appears after and not before puberty, they are consistent with the fact that a significant number of men, and not only women, develop BEDs (2).