Keywords

1 Nutritional Approaches and Consideration of Abstinence

As discussed, FA is a valid construct, particularly as it relates to foods high in added sweeteners, fats, and refined ingredients (Chap. 11) [1, 2]. Many of those that believe in FA suggest that treatment should involve reducing intake of or abstaining completely from “addictive” foods (e.g., foods that prime the reward system, trigger craving, cause positive (from euphoric effects) and negative (from soothing effects) conditioning, have a more rapid effect on reward circuits , adversely affect appetite and satiety regulating homeostatic mechanisms, reduce the ability to feel full from any foods, etc.) [2,3,4,5,6,7]. These are also foods that tend to be high in fat or sugar, high in salt, and highly processed and easily/rapidly absorbed (Chap. 11).

If there is a rule of thumb for what abstinence might look like in FA, or what foods might best be avoided in a general way, it will be to reduce significantly or abstain from these more addictive foods (Chap. 11) . There are reasons to believe that completely eliminating these foods would be of benefit for health and recovery. In fact, some experts suggest that sugar is a toxin, and the World Health Organization recommends significantly limiting intake of sugar for all people for general well-being [8, 9], and reduction of sugary drink consumption is also associated with obesity prevention [10].

Examples of typically problematic foods (Chap. 11, Table 11.1) might include chocolate, ice cream, French fries, pizza, cookies, chips, cake, sugar beverages, and sugar cereals [2, 11,12,13]. A combination of sweet and fat has been proposed as being especially associated with addictive symptoms in humans [1, 14]. High sodium may be another important target [15]. Processed and refined foods are more palatable by design and generally wise to avoid across the board [16]. The use of the glycemic index , protein, and/or fat content in weight loss diets has produced varied and non-definitive results (Chap. 2), but it still might be wise to avoid those foods with high glycemic indexes [17,18,19,20], given what we know about the strong reinforcing effects of rapid blood glucose elevation on reward circuits (Chap. 8).

Some also or instead suggest focusing on significantly reducing or abstaining from individual trigger foods [2] or particular problem foods that the individual tends to overeat and crave [2, 3, 21, 22]. One related way to help people simplify their dietary approaches would be to help them choose a focus on between fat and sugar addiction [23]. Also relatedly, some researchers have proposed three of highly palatable foods , high fat and sodium , high fat and sugar, and high sugar and sodium (defined by percent calories coming from each of the categories) [15], and developed a “hyperpalatable (HP) food” numerical scoring system to determine whether a food fits or doesn’t fit into one of these three categories. This might be utilized and referenced to develop a diet to suggest to people with FA (e.g., avoid all foods that score highly in their particular trigger category), although it may not be a practical approach in the long run. The Yale Food Addiction Scale (YFAS) addressed food as a whole (not based on macronutrient content). A helpful questionnaire to develop next would be one to help patients identify which foods are most problematic for them.

A low carbohydrate ketogenic diet, which is essentially a one-size-fits-all abstinence-based approach, has been found to reduce reward, lower appetite, and increase satiety . In one small observational study, a low-carbohydrate ketogenic diet was initiated by three patients with obesity and comorbid binge eating and FA symptoms [18, 20, 24]. All patients tolerated the diet (macronutrient proportion: 10% carbohydrate, 30% protein, and 60% fat; at least 1200 calories/day) for 6–7 months without adverse events and reported significant reductions in binge eating episodes and FA symptoms including cravings and lack of control. Patients also lost 10–24% of their body weight and maintained gains in weight and binge eating up to 9–17 months after initiation [18, 20, 24].

In a study using a more individualized approach and focusing on food quantities, Vidmar and colleagues [2, 10, 21, 22] recently examined a FA model-based weight loss intervention phone app in a small group of obese adolescents which involved abstaining from “problem foods” (problem foods defined as foods they had cravings for or difficulty resisting). While weight loss was comparable to a standard weight loss intervention control group, the “abstinence” group had higher retention rates [10, 21, 22], and the app was more cost-effective (only two in-person visits and 12 contact hours total with mostly text and phone calls). A larger trial is now underway [2, 21, 22]. The food plan in this study involved staged withdrawal from problem foods (two problem foods at a time at 10-day intervals), then staged withdrawal from daytime (starting at a certain time of day, then progressively expanding the window) eating, and finally from excessive amounts of food consumed at meals reducing amounts in 2% increments.

12-step groups vary in their suggestions (abstaining from personal trigger foods and/or committing to 3–4 meals a day “with nothing in between” are commonly suggested in Overeaters Anonymous (OA) , whereas Food Addicts Anonymous encourages fixed food plans for all members). The primary goal of 12-step program engagement for substance use disorder (SUD) and “compulsive overeating” is abstinence . For food, what that means is often defined by the individual in collaboration with their sponsor [10, 25, 26]. Although these approaches have come under scrutiny because success rates have not been well documented and out of concerns that they might lead to ED development (Chaps. 4 and 5), they also haven’t been studied well [2]. Nevertheless, many individuals struggling with overeating report that applying an abstinence model helped them to control their eating [10, 25, 26]. If a more rigorous program is preferred by a patient, the provider should screen carefully for restrictive behavior prior to initiating and follow weights and disordered eating behaviors carefully [2, 27].

For patients to discover what foods are triggers for them, they could be instructed to keep a food journal. In this journal, they can list all foods they eat, paying attention to foods and patterns of eating that precede addictive eating or loss of control, and emotional responses to foods including improved mood, or increased cravings. This might help determine which of the approaches outlined above will be best suited for the individual patient.

2 Related Tips

2.1 Increase Satiety and Brain Health-Promoting Foods

In addition to avoiding the aforementioned highly palatable or triggering foods in particular, it’s important for patients with FA to not get excessively hungry and get adequate intake of brain health-promoting foods . Satiety -promoting foods (Chap. 11, Table 11.1) include protein, fiber, and whole grains [19]. Foods that are high in dietary protein and high in fiber increase satiety and encourage weight loss [7, 28,29,30]. Studies have reported that foods that are high in dietary protein tend to increase the perception of satiety are less liked than low-protein foods [7, 31] and produce an increase in sensory-specific satiety (meaning they don’t trigger immediate craving and overeating), compared to low-protein foods [7, 32]. Indeed, absorbed foods that might contain sucrose or fructose but that have high fiber contents (e.g., fruit) may not be prone to causing brain changes that lead to conditioning and binge use or overconsumption because fiber limits rapid absorption [9, 33, 34] (Chap. 9).

One commercial diet deserves mention here: https://www.drfuhrman.com/get-started/quick-start. This diet requires abstaining from many food groups and could be experienced as extreme for some patients. However, the website has a nice summary about foods to include that may have beneficial antioxidant effects and which might also promote satiety and brain recovery. It also suggests minimizing two macronutrients (carbohydrate, fat) and increasing various micronutrients . However, this approach has not been formally studied, and it is probably safer and less expensive to stick with nutritional recommendations that are less restrictive until we have more data to support the safety of these kinds of popular culture-based diets.

Furthermore, it is important to reverse nutrient deficiencies. Nutrient deficiencies can negatively impact mood (which could in turn increase craving for relief of negative mood) [35] and addictive behavior via dopaminergic mechanisms such as drug-seeking [35] and so will likely increase FA behavior too. Indeed, amino acids such as tryptophan, phenylalanine, and tyrosine are important for production of neurotransmitters such as serotonin, dopamine, and noradrenaline . Cofactors such as magnesium, zinc, chromium, selenium, folate, B12, lithium, and n-3 polyunsaturated fat are also important to replenish [35]. These nutrient deficiencies can be reversed with multivitamins but even more effective is to increase their intake through food that is rich in these components.

Antioxidants are also emphasized by some practitioners and experts [35, 36] (https://www.drfuhrman.com/get-started/quick-start). Recall that reactive oxygen radicals trigger inflammation, and this feeds into addictive behavior (Chap. 9). An antioxidant-rich diet reduces the experience of hunger and food intake suggesting that antioxidants may be able to reverse some of the deficits in the reward system that perpetuate obesity [36]. Antioxidant therapy has been shown to reverse impulsive behavior in general, as well [36]. An antioxidant and pro-oxidant food ratio of 2:3 per meal is the ideal nutritional ratio for good health and ideal weight in normal weight individuals, and a ratio of 3:4 is ideal for obese individuals because of their state of chronic oxidative stress, and inflammation is posited to be needed to promote recovery from obesity [36]. Additionally, N-acetyl cysteine (NAC) is an antioxidant that has some weak evidence for minimizing habit formation (e.g., conditioning) caused by L-Dopa, indicating that it might be able to help reduce compulsive use of HP foods and their power to usurp behavior and undermine weight loss attempts [37].

Finally, preliminary work supporting the potentially important role gut microbiota could play in food addiction via effects on brain reward circuitry (as we discussed in Chap. 10) has indicated probiotic supplementation might be useful. In one study probiotics administration significantly reduced weight, improved eating behavior, and decreased serum level of neuropeptide-Y compared to the placebo group [38].

2.2 Do Not Over-restrict Calorie Intake

Over-restriction is recognized by ED specialists worldwide as a major contributor to bingeing behavior and ultimately binge eating disorder (BED) and other EDs. In Alcoholics Anonymous one of the first things a “newcomer” learns is the acronym “HALT” suggesting that one should not get too “hungry, angry, lonely, or tired.” During recovery from all additions, it is imperative not to get excessively hungry. The same advice should hold true for recovery from FA [3]. Patients should be reminded that although the long-term goal may be weight loss , that abstinence is designed to reduce hedonic overeating, e.g., to reduce the eating behavior that goes above one’s homeostatic needs. Although weight loss will likely follow, and although in most people some calorie restriction may be safe, excessive starvation will undermine their attempts to get stable and may trigger bingeing [3, 10]. Promoting abstinence from certain foods as outlined above should not include restricting access to healthy foods [3, 10].

In support of these suggestions, recall how hunger and calorie restriction increases food cue reactivity and several other addictive processes in animal and human models (Chap. 9). Yo-yo dieting primes the reward system [39, 40]. Deprivation lowers the threshold for activation of reward pathways and increases the stress response, increasing sensitivity to both drugs of abuse and food as well as their conditioned cues, potentially increasing consumption, reinforcement, and future consumption of (due to conditioning) both drugs and food [35, 41,42,43].

That said, it’s important to mention the other side that several studies have shown how caloric restriction might actually improve brain health and particularly cognition [44]. Whether this applies to impulse control or in people with FA is not yet clear. Furthermore, rapid weight loss has not been found to predict worse outcomes , in some studies: one trial showed that larger initial weight loss during energy-restricting diets was associated with better long-term outcomes [10, 45], although cause and effect is unclear in this study because it was not prospective and randomized. Finally, there is growing evidence about the general health benefits of intermittent fasting [46] (also discussed in Chap. 2).

Likely clinical practice will involve individually tailored treatment and a bit of trial and error. If there is excessive bingeing, backing off on calorie restriction for a while might help someone get back on track, and then increasing restriction could be tried again. For whom more rapid initial weight loss is safe and for whom it is not will involve detailed assessment (Chap. 12) and careful following.

2.3 Realize that Craving Will Diminish with Time in Recovery

Craving is, of course, both a withdrawal symptom and a result of conditioning [47,48,49]. As previously discussed, across different substances (including food), the experience of craving and its cognitive and neural mechanisms are largely similar [10]. Indeed, withdrawal symptoms may increase in the short term, which results in greater conditioned cue reactivity and greater craving. For example, when trait chocolate cravers (who had normal weight) were instructed to refrain from eating chocolate-containing foods (but to maintain regular consumption of all other foods), they reported more intense chocolate craving after 2 weeks [10, 50]. As a result, many people with FA may experience, in the first few weeks of abstinence from a trigger food or a group of foods, strong desires to resume eating their trigger foods , and in the context of these cravings, their brains search for and latch on to any number of justifications to do so. These justifications can completely undermine what felt like a firm commitment, just days prior, and will take people off-course [2].

While short-term deprivation increases cravings for avoided foods, long-term restriction results in reduction of food cravings that can facilitate extinction of conditioned responses [2, 51]. When examining the effects of weight loss interventions in obesity, food cravings tend to decrease during energy-restricting diets over time [10, 52, 53]. And the decrease in cravings is selective for the types of food avoided: cravings for high-carbohydrate foods selectively decreased during a low-carbohydrate diet, while cravings for fatty foods decreased during a low-fat diet [10, 54].

Neuroimaging work in humans also indicates that cue reactivity can be extinguished over time and the ability of drug or emotional cues to trigger craving also diminishes. Before starting a diet, individuals with high YFAS scores had greater activation in reward regions of the brain in response to food cues. After maintaining a prescribed diet of 1600 kcal/day (50% carbohydrates, 30% fats, and 20% proteins) for 3 months, individuals with high YFAS scores compared to those with low YFAS scores had brains that were indistinguishable from one another (i.e., the differences between those with FA and without were no longer present) [55, 56]. Another study showed that activation to food cues reduced in medial prefrontal cortex and other cortical areas from before to after 12 weeks of a nutritional and behavioral program in which participants replaced high-calorie foods with high bulk low-calorie foods [57, 58]. Greater activation in the nucleus accumbens at study entry, indicating reward sensitivity, predicted less weight loss over the 12-week program [57, 58].

In alcohol use disorder (AUD) , as well, abstinence has been found to breed more abstinence and promote positive brain changes [59]. Abstinence from problem foods will likely ultimately do the same for people with FA. Interventional strategies that successfully reduce craving for and consumption of alcohol, tobacco, etc. can likely be applied to reduce craving for and consumption of food as well [10, 60, 61] to increase chances of success for getting more abstinence time (Chap. 14). With time in recovery, people can be reassured that things will get better and that they will experience the extinguishing of the food cue conditioning and reduced impulsivity and the process will naturally build on itself in a cyclical beneficial way [62]. People should know that although their cravings may increase initially, they will get better over time. It might be useful to suggest they consider the first month of abstinence as an experiment, to test the hypothesis that the cravings and dysphoria will likely pass. If it doesn’t pass after a certain period of time, they can, at that point, reassess their approach.

2.4 Abstinence Is Not Absolute: Avoid All-or-Nothing Thinking

Although abstention from problematic foods is ideal, the idea of complete abstinence from certain food items may not be accurate from a nutritional or neurochemical point of view in the same way that it is for substances of abuse. If one’s goal is to abstain completely from sugar, how to carry this out is not entirely clear, since fruit and vegetables also contain glucose, for example [10]. Moreover, it’s much easier to “slip” accidentally with food. Humans eat several times a day, and there is a lot of opportunity to eat a problem food without realizing it. For example, individuals who try to avoid eating sugar may still (inadvertently) consume some foods that contain sugar or, at least, other forms of carbohydrates [10].

Tailor-made hybrid models between inclusive and exclusive approaches have been useful according to some experts. These approaches usually require some trial-and-error and are best done under the supervision of an registered dietitian and a psychiatrist/psychotherapist who understands EDs, FA, SUDs, and the associations with other psychiatric diagnoses described herein [2].

In addition, it’s important to remember that individual differences exist and need to be taken into account in food plan development [21, 22]. Although fat, sugar, and highly processed foods are certainly the most likely culprits for most FA, it’s important to tailor food plans to the individual. Although, for some, complete abstinence might work best, it might also not be practical for others, especially in the long term , given the complexity of food and its intense cultural interweaving. Also, what is “addictive” to one person might not be a problem for another. The definition of abstinence in relation to food will likely not be the same from one person to the next.

2.5 Is It Better to Start More Extreme or Use a Graded Approach During Initiation?

Some approaches suggest making a rapid more extreme change in eating such as is seen with adopting a ketogenic diets [18, 20, 24], whereas others have been tested more graded approaches [21, 22, 63]. Until more research is done, it is not clear which approach is most useful and for which people. Relatedly, identifying the problem foods can be difficult. It might be wise to ask people to be overinclusive on what they abstain from and, once stabilized, slowly work back in foods they’re not sure about, until they start to lose stability around eating again. On the one hand, the world is full of food; it’s a challenge to abstain from certain foods, especially with food being so tied to socialization (even more so than liquor). On the other hand, abstinence might bring relief faster than a slow taper. The 12-step programs talk about surrendering to the program, which for people with FA might involve surrendering to a food plan. And when people achieve sobriety in the early days of recovery, they often report a “pink cloud” which reinforces future abstinence. The confidence and freedom from craving and obsession really feels good and might argue for a more rapid change initially.

In SUD treatment, either total abstinence or substance use reduction, a “harm-reduction” approach (decreasing substance use to a level that is non-problematic [64]), is considered reasonable treatment strategies. In SUD treatment, harm reduction might work for some people, but many people end up ultimately choosing abstinence because it’s so much easier.

2.6 Track Progress

Many weight management programs suggest tracking food intake or weight over time, which has been shown to improve outcomes. This might also be helpful in FA. However, in FA treatment, the goal is also to reduce the symptoms of FA (the sense of loss of control, craving, negative consequences of use, etc.). Remember the goals of behavior and dietary change are to enhance overall well-being and function. Patients might also consider tracking peace of mind , self-esteem, or personal sense of self-control to assess if what they’re doing with their diet and other self-care is helping or harming them over time. Thus, encouraging them to consider these as equally if not more important than weight loss or abstinence may be best, in the long run.

3 How to Incorporate FA Treatment into ED Treatment Programming

As we’ve mentioned there is great concern that the “abstinence model” might be taken too far by some individuals with ED or predisposed to ED and that FA-based nutritional approaches might increase bingeing, ED risk, resurgence of dormant ED, or worsening ED symptoms [2]. As a general rule, the ED treatment culture is not highly supportive of the FA model for these reasons . It has been argued that abstinence models may be ineffective or—as they may reinforce problematic dietary restriction—even be hazardous, particularly in individuals with bulimia and BED [10, 65]. As reviewed in previous chapters (Chaps. 3 and 5), whether restriction and/or dieting causes bingeing and obesity in all people and how to identify those more vulnerable to restriction induced ED or obesity is still not clear.

Current practice in cognitive behavior therapy (CBT) -based ED treatment programs aim to reduce dysfunctional dieting and restraint of any kind (including attempting to abstain from certain foods) in favor of emphasizing regular eating patterns with flexible and moderate food consumption and no forbidden foods [10] which arguably refutes the FA model [65, 66]. In fact, this specific nutritional aspect of CBT-based ED treatment may be of great benefit for many patients who binge eat. For example, a reduction in dietary restraint has been shown to moderate the increased effectiveness of CBT on binge eating in a sample of patients with bulimia [66, 67].

The truth is that an abstinence-based nutritional approach might be helpful for some and harmful for others [2], and at this point we don’t know for whom it is best to choose which. How to tailor nutritional recommendations for people with comorbid FA and ED may come down to three things. The first is the importance of doing a risk-benefit analysis. Potential harms of including an abstinence approach to eating are high for a patient with severe bulimia, but potential harms of not identifying and treating FA in someone with obesity is also high [2]. The second important issue is that providers shoud screen for restrictive patterns, which, if present, may indicate a greater risk of adverse outcomes from an abstinence-based approach. The assessment process is absolutely key (Chap. 12). Failure to consider restrictive eating patterns is an important criticism of FA that has led many ED professionals to reject the construct altogether [2]. The third is to recognize and identify impulsivity and cue reactivity as part of the eating pathology. Existing treatments sometimes fail to recognize impulsivity and susceptibility to environmental cues as key parts of the eating pathology [2, 68, 69]. If recognized, and if identified to be a bigger problem than, say, over-restriction, then an FA nutritional approach may prove more effective.

It will be very challenging, in residential settings especially, to implement divergent nutritional strategies where patients might compare food plans with one another [2], running the risk of exacerbating restrict-binge-purge patterns. How can one give conflicting messages to a patient one is treating, especially when a message of restriction might trigger other patients in the program, leading to potentially dangerous consequences? Some authors recommend that if an ED is present in addition to FA, clinicians should first provide more standard evidence-based treatments for the ED to see if the FA resolves [2, 70]. Indeed, studies suggest that FA most likely improves with ED treatment. For example, FA symptoms resolved when bulimia nervosa (BN) symptoms remitted [2, 70, 71] in a study using non-abstinence-based eating disorder treatment interventions. Also recall from Chap. 11 the animal studies that show that providing sugar and fat intermittently causes more addictive brain changes than continuous access animal models, which would imply that bingeing itself may be a cause of FA and a primary treatment target, avoiding restrictive approaches [72].

If a standard evidence-based ED treatment paradigm is ineffective, then an FA approach can be attempted next. When food alters dopamine circuitry, efforts by people with FA to moderate food intake while still eating “addictive” foods can make “intuitive eating” feel impossible, and the common ED paradigm which favors an “all foods fit” or “no bad foods” approach [2, 73] might trigger people with FA. In studies of OA, individuals struggling with overeating report that applying an abstinence model helped them to control their eating [10, 25, 26]. If an FA approach is deemed potentially useful, treatment of the FA might best be done in outpatient settings, or in settings designed for FA treatment, and progress can be supervised by clinicians and nutritionists experienced in EDs, SUDs, impulsivity treatment, and the FA concept [2] as well as medical personnel to watch for safety issues, in case dangerous bingeing or purging commences. Several residential and home-based intensive programs exist currently in the United States (SHiFT Recovery by Acorn – https://foodaddiction.com; COR – https://www.theretreat.org/programs/weekend-retreats/cor-retreats).

For obese patients with ED and FA, it may be wise to refer to an obesity clinic which utilizes behavioral weight loss therapy. A systematic review and meta-analysis found that structured and professionally run obesity treatments are associated with reduced ED prevalence, risk, and symptoms in children [2, 74].

Some would suggest that if some binging occurs following an attempt at abstinence should not lead to immediate cessation of the food plan. Saying that an abstinence approach is ineffective because people binge when they finally eat sweets is like saying that abstinence from alcohol is ineffective because those with AUD binge after taking the first drink [2, 11]. Failure to consider the possibility of increased craving and bingeing impulses during the first weeks of abstinence due to withdrawal is an important criticism of those that unilaterally say that if abstinence triggers binging, they should not try abstinence again [2, 10, 11].

It will also be important to distinguish between flexible and rigid restraint [2, 10, 75]. In some cases, restraint is related to a lower body weight, better weight regulation, and a better diet quality. In others, restraint predicts poor diet, overeating, and obesity [2, 27]. Relatedly, it is key to remember that abstinence doesn’t mean semi-starvation. Removing particular food groups from a person’s food plan should also be accompanied by an equal degree of emphasis on increasing healthy foods, adequate micronutrient, protein and fiber intake, and minimizing excessive calorie restriction and hunger.

Whether or not ED treatment providers accept the term “FA,” avoiding foods that trigger their own overconsumption to the best of their ability, deliberate inclusion of health-promoting foods, and many of the behavioral, lifestyle, and medication interventions that we will go over in Chap. 14 (which target impulsivity and habitual patterns of responding, depression and anxiety management, developing coping mechanisms, enhancing positive social connections, addressing cognitive distortions such as justifications, as well as neuromodulation , cognitive training, encouraging adequate sleep and exercise, and medication) can still be considered for some patients with ED, especially those with comorbid obesity [2, 3], and these interventions do not contradict standard ED treatment approaches (Chap. 14, Table 14.1).

4 What to Do with “Normal Weight” FA Patients?

Is it ever wise to suggest an abstinence-based approach in someone who doesn’t need to lose weight for health reasons (e.g., with a normal body mass index)? The jury is still out on this, but FA is associated with distress even if the person is of normal weight and is often related to feelings of loss of control. Perhaps assessing body image distortions could be helpful here, and if the goal is about weight maintenance rather than loss, and if the target symptom is loss of control rather than weight, then in combination with increasing healthy foods and eating adequate calories, and trialling an abstinence approach and some of the additional suggestions reviewed in the next section might be appropriate. More studies are needed, however, to confirm this.

5 Conclusion

Many dietary recommendations have been made regarding FA treatment; however none have been extensively studied. Given that many approaches appear to be useful for some, but maybe not all FA patients, it is important to individualize treatment. Some recommendations, which appear to make good nutritional sense such as avoiding or limiting HP foods, including more whole grains and fiber in one’s diet are probably safe to suggest to most FA patients. Treatment should be individualized, taking into account a patient’s comorbidities, especially noting the presence of EDs and how particular diets may affect those. It is also important to take individual preferences into account. More research in this area is needed to further our ability to target particular recommendations to individual patients.