Abstract
Context
Eating habits and food craving are strongly correlated with weight status. It is currently not well understood how psychological and behavioral factors influence both weight loss and weight regain.
Objective
To examine the associations between psychological and behavioral predictors with weight changes and energy intake in a randomized controlled trial on weight loss.
Design and setting
The Prevention of Obesity Using Novel Dietary Strategies is a dietary intervention trial that examined the efficacy of 4 diets on weight loss over 2 years. Participants were 811 overweight (body mass index, 25-40.9 kg/m2; age, 30-70 years) otherwise healthy adults.
Results
Every 1-point increase in craving score for high-fat foods at baseline was associated with greater weight loss (-1.62 kg, P = .0004) and a decrease in energy intake (r = -0.10, P = .01) and fat intake (r = -0.16, P < .0001) during the weight loss period. In contrast, craving for carbohydrates/starches was associated with both less weight loss (P < .0001) and more weight regain (P = .04). Greater cognitive restraint of eating at baseline was associated with both less weight loss (0.23 kg, P < .0001) and more weight regain (0.14 kg, P = .0027), whereas greater disinhibition of eating was only associated with more weight regain (0.12 kg, P = .01).
Conclusions
Craving for high-fat foods is predictive of greater weight loss, whereas craving for carbohydrates is predictive of less weight loss. Cognitive restraint is predictive of less weight loss and more weight regain. Interventions targeting different psychological and behavioral factors can lead to greater success in weight loss.
Keywords: Weight Loss, Weight regain, Psychological factors, Behavioral factors, Diet, Predictive
Gradual weight gain over long period poses a great risk to health and increases the prevalence of obesity (1). Obesity is a major risk factor for the development of cardiovascular disease, type 2 diabetes, and several cancers (2). The prevention of gradual weight gain is therefore an effective strategy to prevent obesity and obesity-related diseases. Psychological and behavioral factors (i.e., food craving, cognitive restraint, and disinhibition of eating) influence body weight.
Food craving is a common psychological state defined as “an intense desire for a particular food (or type of food) that is difficult to resist” (3). On a weekly basis, both women and men report frequent experiences of food craving (4–6). Food craving has been associated with increased food intake and body weight. Previous evidence suggested that craving for specific foods (i.e., sweets, high-fat foods, carbohydrates/starches, fast-food fats) has been associated with an increase in intakes of the respective foods (7, 8). Higher frequency of food cravings has been positively associated with body mass index (BMI) (7). A meta-analysis including 3292 women and men demonstrated that visual food cue and the experience of craving significantly influence and contribute to eating behavior and weight gain independent of their BMI status (9).
Other psychological constructs of eating behaviors such as cognitive restraint and dietary disinhibition also affect food consumption and energy expenditure (10–12). High cognitive restraint has been suggested to be associated with intent to diet and controlled eating, but a number of studies suggest that restraint eaters tend to overeat after being exposed to a disinhibitor (i.e., consumption of forbidden food or breaking a dietary rule) (6, 13). Weingarten et al. report that when participants have been restrained to consume complex carbohydrate and protein for 3 days, there is an increase in craving for the restraint types of food (14). On the other hand, increased disinhibition of eating has been consistently associated with episodic overeating, binge eating, and adiposity (15, 16). Therefore, both dietary restraint and disinhibition demonstrate strong influences in dietary outcomes related to body weight (15, 17).
The relationship of food cravings and related psychological behavior with food consumption can be critical in the context of weight management. Few studies investigate whether the psychological and behavioral factors are predictive of weight changes, or whether same psychological and behavioral factors are predictive for both weight loss and weight regain during dietary interventions remains uninvestigated. The lack of targeted or tailored approaches that leverage the psychological constructs of eating behaviors on weight management may hamper the success of dietary and lifestyle modifications to combat the obesity epidemic.
Therefore, in the present study, we aimed to examine the associations between baseline psychological and behavioral predictors (i.e., food craving, cognitive restraints, and disinhibition) with weight changes and energy intake in a 2-year-long, randomized controlled trial on weight loss. We hypothesize that (i) craving for different types of foods have divergent associations with weight change and (ii) behavior predictors for weight loss and weight gain are indistinct.
Materials and Methods
Study design
The Prevention of Obesity Using Novel Dietary Strategies (POUNDS LOST) trial was a randomized clinical trial that examined the effects of 4 calorie-restricted, heart-healthy diets with varying macronutrient profiles on weight loss (18). Overweight or obese participants (n = 811), between age 30 and 70 years were randomly assigned to each diet group. Macronutrient profile of the 4 diets were: (i) low fat (20% of energy), average protein (15% of energy); (ii) low fat (20%), high protein (25%); (iii) high fat (40%), average protein (15%); and (iv) high fat (40%), high protein (25%). Participants were instructed to consume 150 mg or less of cholesterol per 1000 kcal, less than 8% of energy from saturated fat, and at least 20 g of fiber every day. Each participant received a tailored diet prescription based upon a 750 kcal/d energy deficit from his or her baseline energy needs. At year 2, 645 participants, or 80% of the study population, had completed the trial. Exclusion criteria were participants who were diabetic treated with oral medications or insulin, with serious gastrointestinal disease, alcohol or drug abuse, individuals in treatment for an eating disorder, unstable or recent onset of cardiovascular disease, or other serious illness. We also excluded those who were on weight-loss medications and other drugs that affect body weight such as some antipsychotic or antidepressant drugs, or corticosteroids; those with hypothyroidism defined by abnormal TSH; urinary microalbumin > 100 μg/g creatinine; or on unstable doses of medication for hyperlipidemia, hypertension; or with a psychiatric disorder. The goal for the study population was generalizability of the results to the population needing weight loss. Details in key findings from the POUNDS LOST study have published elsewhere (18). In brief, weight loss was similar across all 4 diet groups. Participants had the most weight loss during the first 6 months, and gradual weight regain from months 6 to year 1 and from year 1 to year 2. All participants gave written informed consent. The protocol was approved by the institutional review boards at Harvard T.H. Chan School of Public Health and the Pennington Biomedical Research Center in Baton Rouge, LA.
Assessment of body weight and BMI.
Body weight was assessed at baseline, 6, 12, and 24 months before breakfast. Height was measured at baseline. BMI was calculated as weight in kilograms divided by squared meter of height.
Dietary assessment.
At baseline, dietary intake was assessed using 5-day diet record. At 6 months and at year 2, we conducted telephone interviews in a random sample of 50% of study participants to assess dietary intake using 24-hour recall on 3 nonconsecutive days (18). Questionnaires on food craving, satiety, dietary restraint, disinhibition, and hunger were administered at baseline, 6 months, and year 2.
Food craving.
Measures of psychological and eating behavior factors were administered at baseline, 6, 12, and 24 months. The Food Craving Inventory-II is a 33-item, self-administrated measure designed to assess the frequency an individual experiences craving. Participants rated craving for particular food on a 5-point Likert scale ranging from 1 to 5, where 1 = never, 2 = rarely, 3 = sometimes, 4 = often, 5 = always (3). The Food Craving Inventory also has 5 subscales including high-fat foods (i.e., steak, fried fish, fried chicken), sweets (e.g., cake, cookies, chocolate, candy), carbohydrates/starches (i.e., rolls, baked potato, pasta, cereal), fast food fats (i.e., pizza, hamburgers, French fries, chips), and fruits/vegetables. We examined the association between total craving score, subscale score, and changes in body weight.
Ratings of hunger, craving, thoughts about food, and satiety.
We use the 100-mm visual analog scales to assess hunger, level of fullness after meals, thoughts about food, and food cravings (19, 20). Higher scores indicate greater hunger, fullness, more frequent thoughts about food, and greater cravings.
Dietary restraint, disinhibition, and hunger.
The Three Factor Eating Questionnaire consists of 51 yes/no questions to assess cognitive restraint, disinhibition (susceptibility to overeating), and perceived hunger (16).
Dietary restraint refers to the intent to diet, intent to monitor, and regulate food intake (21 items). The disinhibition scale measures uncontrolled eating in response to different stimuli, such as in the presence of palatable foods or emotional clues (16 items). Perceived hunger scale measures tendency to eat in response to the subjective sense of hunger (14 items). Higher scores indicate higher levels of cognitive restraint, disinhibition, and perceived hunger.
Statistics
We used multivariate linear regression to examine associations between psychological, behavioral factors at baseline, month 6, and month 12 and subsequent weight changes. The outcomes of interest were (i) weight loss at month 6 and (ii) weight regain from months 6 to year 1 and year 1 to year 2. The exposures were psychological and eating behavior factors including food craving, satiety, and restraint and disinhibition of eating.
In the present study, participants achieved the maximum weight loss of 6.2% (-5.75 kg) at month 6 and maintained an overall weight loss of 3.5% (-3.31 kg) after 2 years of intervention. Therefore, for analysis on weight change, the weight loss period was defined as months 0 to 6 and weight regain period as months 6 to 12 and months 12 to 24. For analysis during the weight loss period, the model was adjusted for age, gender, race/ethnicity, marital status, education level, income, BMI, body weight at baseline, diet group and interaction terms between diet groups and baseline psychological, behavioral factors. To further investigate the impact of insulin resistance on weight loss, we dichotomized participants to normal fasting insulin and high fasting insulin using 10.5 µIU/mL as the cutoff point (21) and adjusted for this categorical variable in the model. For analyses on weight regain periods (months 6-12 and months 12-24), we adjusted body weight at month 6 and month 12 in the same model, respectively.
Psychological and behavioral factors were examined as both continuous variables and categorical variables in quartiles. We used median values of each quartile calculate P for trend. Each subscale of the Food Craving Inventory (i.e., craving for high-fat foods, sweets, carbohydrates/starches, fast food fats, and fruits/vegetables) were mutually adjusted for each other to examine their independent contributions to weight changes. Spearman correlation was used to assess associations between psychological factors, behavioral factors, and changes in energy and macronutrient intake.
Baseline data are presented as the mean and SD. Percentage of weight changes were reported as means and 95% confidence interval. Analyses were conducted using SAS software, version 9.4.
Results
Participants characteristics
Participants in the overall POUNDS LOST study were predominantly white (81.8%), female (63.3%), and obese with an average BMI of 32.3 (Table 1). Females had less craving for high-fat foods, but greater craving for sweets and fruits/vegetables compared with males (P < .05 for all, Table 2). The study population consisted of 79% Caucasian, 16% black, 1% Asian, 4% Hispanic, and 1% other race or ethnic group. Among different races, Caucasians had the lowest craving score for high-fat foods, highest cognitive restraint, and lowest score for disinhibition and perceived hunger (P < .05 for all, Table 2). Participants had the most weight loss (-5.75 kg) during the first 6 months of intervention followed by a subsequent weight regain by 0.20 kg and 2.25 kg from months 6 to 12 and from 1 year to 2 years (Fig. 1). Being African American was associated with less weight loss (-4.90 kg) compared with being Caucasian (-7.01 kg, P = .0003) during the first 6 months; and race was not associated with weight regain from month 6 onward.
Table 1.
Baseline Characteristics and Scores of Food Craving, Hunger and Satiety and TFEQ (n = 811)
| All participants (N = 811) | Low-fat, Average-Protein Group (N = 204) | Low-fat, High-Protein Group (N = 202) | High-fat, Average-Protein Group (N = 204) | High-fat, High-Protein Group (N = 201) | |
|---|---|---|---|---|---|
| BMI | 32.7 ± 3.85 | 32.6 ± 3.7 | 32.6 ± 3.7 | 32.4 ± 3.9 | 33.2 ± 4.2 |
| Age, y | 51.9 ± 9.2 | 51 ± 9 | 50 ± 10 | 52 ± 9 | 51 ± 9 |
| Sex (n/%) | |||||
| Female | 515 (63.5) | 126 (62) | 135 (67) | 125 (61) | 129 (64) |
| Male | 296 (36.5) | 78 (38) | 67 (33) | 79 (39) | 72 (36) |
| Body weight | 93.0 ± 15.6 | 93.8 ± 15.8 | 91.5 ± 13.1 | 92.0 ± 16.9 | 94.3 ± 16.0 |
| Hypertension n (%) | 287 (35) | 70 (34) | 70 (35) | 67 (33) | 80 (40) |
| Food Craving Score | |||||
| High-fat foods | 2 ± 1 | 2 ± 1 | 2 ± 1 | 2 ± 1 | 2 ± 1 |
| Fast-foods fat | 2 ± 1 | 2 ± 1 | 2 ± 1 | 2 ± 1 | 2 ± 1 |
| Carbohydrate/starches | 2 ± 1 | 2 ± 1 | 2 ± 1 | 2 ± 1 | 2 ± 1 |
| Fruits/vegetables | 2 ± 1 | 2 ± 1 | 2 ± 1 | 2 ± 1 | 2 ± 1 |
| Sweets | 2 ± 1 | 2 ± 1 | 2 ± 1 | 2 ± 1 | 2 ± 1 |
| Hunger and Satiety Score | |||||
| Fullness | 63.3 ± 19.8 | 65 ± 20 | 62 ± 20 | 63 ± 20 | 63 ± 19 |
| Appetite | 65.7 ± 17.2 | 66 ± 18 | 64 ± 17 | 67 ± 16 | 65 ± 17 |
| Hungry | 43.1 ± 19.9 | 43 ± 20 | 43 ± 19 | 43 ± 21 | 42 ± 21 |
| Craving | 50.5 ± 25.9 | 51 ± 26 | 50 ± 26 | 50 ± 26 | 51 ± 27 |
| TFEQ | |||||
| Dietary restraint | 9.5 ± 4.3 | 10 ± 4 | 10 ± 4 | 9 ± 4 | 9 ± 4 |
| Disinhibition | 8.5 ± 3.4 | 9 ± 3 | 8 ± 3 | 8 ± 4 | 8 ± 3 |
| Perceived hunger | 5.6 ± 3.4 | 6 ± 4 | 5 ± 3 | 6 ± 4 | 5 ± 3 |
| Educational level | |||||
| High school or less | 76 (9) | 23 (11) | 15 (7) | 19 (9) | 19 (9) |
| Some college | 180 (22) | 47 (23) | 47 (23) | 45 (22) | 41 (20) |
| College graduate or beyond | 555 (68) | 134 (66) | 140 (69) | 140 (69) | 141 (70) |
| Marital status | |||||
| Married | 643 (79) | 132 (65) | 146 (72) | 144 (71) | 143 (71) |
| Divorced/separated | 127 (16) | 36 (18) | 24 (12) | 33 (16) | 29 (14) |
| Widowed | 5(1) | 8 (4) | 4 (2) | 4 (2) | 2 (1) |
| Never married | 29 (4) | 28 (14) | 28 (14) | 23 (11) | 27 (13) |
| Race | |||||
| White | 7 (1) | 159 (78) | 158 (78) | 165 (81) | 161 (80) |
| Black | 33 (16) | 33 (16) | 28 (14) | 33 (16) | |
| Asian | 365 (45) | 1 (<1) | 1 (<1) | 1 (<1) | 2 (<1) |
| Hispanic | 437 (54) | 8 (4) | 7 (3) | 9 (4) | 5 (2) |
| Other | 9 (1) | 3 (1) | 3 (1) | 1 (<1) | 0 |
| Income | |||||
| <$75,000 | 365 (45) | 97 (48) | 88 (44) | 86 (42) | 94 (47) |
| >$75,000 | 437 (54) | 105 (51) | 112 (55) | 116 (57) | 104 (52) |
| Total cholesterol (mg/dL) | 199 ± 38 | 203 ± 36 | 203 ± 37 | 204 ± 35 | 202 ± 37 |
| Glucose (mg/dL) | 92 ± 14 | 93 ± 12 | 92 ± 17 | 92 ± 12 | 92 ± 14 |
Values are mean ± SD.
Abbreviations: BMI, body mass index; TFEQ, The Three Factor Eating Questionnaire.
Table 2.
Baseline Scores on Food Craving, Hunger, and Satiety and TFEQ According to Gender and Race (n = 811)
| Gender | Race | ||||
|---|---|---|---|---|---|
| Food Craving | Female | Male | Black | Caucasian | Other |
| High-fat foods | 1.79 ± 0.60a | 1.97 ± 0.68 | 2.00 ± 0.68a | 1.82 ± 0.62 | 2.24 ± 0.58a |
| Carbohydrates/starches | 2.17 ± 0.68 | 2.13 ± 0.71 | 2.12 ± 0.65 | 2.15 ± 0.70 | 2.43 ± 0.52 |
| Fast-food fats | 2.29 ± 0.74a | 2.42 ± 0.78 | 2.23 ± 0.76 | 2.35 ± 0.76 | 2.54 ± 0.80 |
| Sweets | 2.49 ± 0.75a | 2.25 ± 0.78 | 2.28 ± 0.72 | 2.42 ± 0.77 | 2.54 ± 0.93 |
| Fruits/vegetables | 2.30 ± 0.81a | 2.12 ± 0.81 | 2.39 ± 0.78 | 2.19 ± 0.83 | 2.38 ± 0.88 |
| Hunger/satiety | |||||
| Hunger | 43.3 ± 20.5 | 42.9 ± 18.9 | 39.5 ± 21.9 | 43.8 ± 19.5 | 43.9 ± 19.7 |
| Fullness | 63.7 ± 20.6 | 62.7 ± 18.4 | 64.0 ± 21.6 | 63.2 ± 19.4 | 62.5 ± 24.0 |
| Appetite | 66.6 ± 17.2 | 63.9 ± 16.9 | 65.0 ± 18.9 | 65.8 ± 16.9 | 64.9 ± 13.4 |
| Craving | 49.1 ± 26.5 | 52.8 ± 24.7 | 51.8 ± 28.6 | 50.3 ± 25.4 | 48.3 ± 24.6 |
| TFEQ | |||||
| Dietary restraint | 10.4 ± 4.15a | 7.92 ± 3.97 | 11.1 ± 4.31a | 9.18 ± 4.19 | 9.47 ± 3.94 |
| Disinhibition | 8.76 ± 3.53a | 8.04 ± 3.24 | 6.95 ± 3.53a | 8.81 ± 3.35 | 8.18 ± 3.24 |
| Perceived hunger | 5.37 ± 3.34a | 5.91 ± 3.46 | 4.12 ± 2.86a | 5.86 ± 3.43 | 5.53 ± 2.94 |
Values are mean ± SD.
Abbreviations: TFEQ, The Three Factor Eating Questionnaire.
a P < .05.
Figure 1.
Changes in body weight (kg) from baseline to months 6, 12, and 24 and weight regain from months 6 to 12 and months 12 to 24 in response to study intervention over 2 years.
Food craving.
Craving score is predictive of weight change. The positive association can be interpreted as less weight loss given that participants had an overall weight loss throughout the intervention period. Every 1-point increase in overall craving score at baseline was associated with 1.04 kg less weight loss at month 6 (P = .004). Participants in the lowest quartile of craving score had the greatest weight loss (quartile 1 vs. quartile 4: -7.08 kg vs. -5.64 kg, P for trend = .0031). The association between baseline craving score and weight loss was modified by diet assignment at month 6 (20% fat, 25% protein: -6.50 kg; 40% fat, 25% protein: -6.11 kg; 20% fat, 15% protein: -6.03 kg; 40% fat, 15% protein: -5.89 kg; P for interaction .03).
Craving for different foods had divergent association with weight loss. Craving for high-fat foods at baseline was associated with greater weight loss at month 6 (-1.63 kg, P = .0006). In contrast, craving for carbohydrates/starches was associated with less weight loss (1.98 kg, P < .0001, Fig. 2A). The associations between overall craving, craving for high-fat foods, craving for carbohydrates/starches, and weight loss remained unchanged (P < .001 for all, data not shown) even after adjusting participants’ baseline insulin sensitivity in the model, a surrogate marker of insulin sensitivity (22). Interestingly, participants with insulin resistance (fasting insulin ≥ 10.5 µIU/mL) had greater craving for high-fat foods (P < .0001) and fast-food fats (P = .045). During the initial weight regain period (6-12 months), the associations between craving for high-fat foods and weight loss were no longer significant. Conversely, craving for carbohydrate/starches was associated with weight regain from month 6 to 1 year (0.54 kg, P = .04, Fig. 2B); craving for sweets was associated with weight regain from 1 year to 2 years (0.65 kg, P = .045, Fig. 2C).
Figure 2.
Association between psychological and behavioral factors and weight change.
Satiety.
More frequent thoughts about food (higher appetite score) at baseline was associated with a minimal increase in weight loss of -0.03 kg (P = .01) at month 6. During weight regain periods, satiety factors did not predict weight changes, with the exception of lower hunger score at 1 year was weakly associated with weight loss from year 1 to year 2 (Fig. 2C).
Dietary restraint, disinhibition, and hunger.
During the weight loss period (0-6 months), per 1 score-point increase in cognitive restraint was associated with of 0.23-kg less weight loss (P < .0001). Greater disinhibition of eating was associated with 0.12-kg more weight regain between 6 and 12 months (P = .001). Similarly, greater cognitive dietary restraint was associated with more weight regain (0.14 kg, P = .001) between 1 and 2 years during the trial.
Associations between psychological and eating behavioral factors and energy intake.
We next investigated whether the observed associations between psychological, behavioral factors at baseline and weight changes were attributable to differences in energy intake. During the weight loss period (months 0-6), high-fat food craving at baseline was correlated with a decrease in energy intake (r = -0.10, P = .01, Fig. 3A) and fat intake (r = -0.16, P < .0001, Fig. 3B). Carbohydrate/starches craving was weakly associated with a reduction in energy intake (r = -0.08, P = .04, Fig. 3A) and fat intake (r = -0.08, P = .05, Fig. 3B) during the same period. During the weight loss period, months 0 through 6, having higher appetite score at baseline was associated with a reduction in energy intake (r = -0.02, P = .014, Fig. 3C), whereas higher dietary restraint was associated with an increase in energy intake (r = 0.04, P < .001, Fig. 3D). The inverse association between craving for high-fat food and caloric intake and the positive correlation between dietary restraint and energy intake maintained at 2 years (data not shown). Surprisingly, we did not find any association between dietary disinhibition and energy or fat intake (data not shown).
Figure 3.
Association between baseline food craving score and changes in intakes of macronutrients and total calories during weight loss period (0-6 months).
Discussion
Our study expanded the literature by identifying unique psychological and eating behavior factors that predicted weight loss and subsequent weight regain during a randomized dietary intervention trial in 811 overweight and obese individuals. We found that craving for high-fat food was predictive of greater weight loss that was potentially attributed to a reduction in intakes of energy and fat among overweight and obese individuals in a weight loss program. Craving for carbohydrates/starches, on the other hand, was associated with both less weight loss and more weight regain. And this association was not explained by energy intake. Furthermore, dietary restraint was a strong predictor for less successful weight loss and more weight regain between years 1 and 2 probably through increased energy intake. Last, dietary disinhibition was a strong predictor only for weight regain. These results provide new evidence that psychological and eating behavior factors are important predictors of weight loss and weight maintenance during dietary interventions. Findings from our study can inform the design of more targeted approaches, such as reduce craving for foods that are high in carbohydrates/starches and enhance dietary restraint management skills. These factors are likely to optimize the outcomes of dietary modifications that lead to greater weight loss. Interventions to reduce disinhibition of eating through behavioral, psychological, and pharmacological interventions to prevent weight regain are likely to be fruitful.
Food cue-induced craving has been linked to increased consumption (9) and long-term weight gain in adults (23, 24). A recent published meta-analysis included 3292 participants conclude that food cue-induced craving is predictive of food consumption, body weight, and BMI, and is also associated with weight gain (9). Among 646 overweight and obese individuals (both female and male), higher frequencies of food cravings were positively associated with higher BMI, and cravings for specific food categories (high fat, sweets, carbohydrates/starches, and fast-food fats) were associated with increased consumption of craved food (7). In the present study, we found that craving for high-fat foods was associated with greater weight loss and craving for carbohydrates/starches was associated with less weight loss.
Carbohydrates and fats as energy sources have distinct pathways to the central nervous system to guide food choice (25). It was recently discovered that associations between energy density and reward cue are different for fat and carbohydrate (25). Energy density of fatty foods but not carbohydrate is positively associated with cognitive reward (25). Previous evidence suggested that obese individuals received greater cognitive rewarding from high-fat food when compared with lean individuals, which partially explained why people with obesity were more susceptible to consuming such foods (26, 27). Therefore, we hypothesize that among our participants, the overall healthy tailored diet prescription may disassociate the stimuli from high-fat foods thus hindering the rewarding feedback that leads to less consumption and greater weight loss. It is also plausible that during the intervention, individuals with craving for high-fat food may reduce consumption of these type of foods per dietary consultation, whereas, for craving for carbohydrate/starches, participants may substitute these types of foods with healthier carbohydrate options. Evidence from 1 recent study suggested that individuals were able to estimate energy density from fat more accurately compared with carbohydrate (25). This phenomenon may also contribute to unconscious consumption of carbohydrate, which may contribute to less weight loss and greater weight regain. We observed the association between craving and weight loss was modified by the diet group. Further research is warranted to explore the divergence in food cravings and how it is related to weight change.
Caloric restriction has a strong impact on body weight; and dietary restraint controls caloric intake, which is expected to achieve weight control effectively. Despite some short-term studies suggesting that higher dietary restraint was associated with greater weight loss (19, 28), long-term large-scale prospective studies suggested that cognitive restraint was not associated with weight loss (29–31) or BMI (32). A central hypothesis of dietary restraint theory is that the intent to diet may be disrupted or “disinhibited” by certain events such as dysphoric emotions, alcohol, or palatable foods (11, 33). This explains why individuals trying to restrict eating often lose control and overeat (13). Evidence from a prospective study demonstrated dietary restraint scores predicted an increased BMI over several years (34). Among overweight and obese people, diet restraint–induced weight loss was associated with physiological changes including decreased resting metabolic rate and reduced spontaneous physical activity (35, 36). Diet-induced weight loss was also associated with hormonal changes such as increase in concentration of ghrelin, and decrease in leptin and insulin sensitivity, which may promote counterregulatory eating and thus promote weight regain (10, 37, 38). Restrained eaters often respond to attractive food cues by eating more, thus leading to potential weight gain (39). However, Williamson and colleagues showed 2 decades ago that increased dietary restraint modulated the association of overeating with obesity in that some restraint eaters are able to maintain a normal weight or moderately overweight without overeating (11). The multidimensional construct of dietary restraint and its relationship to overeating and body weight has yet to be clearly defined.
Dietary disinhibition has previously been associated with loss of control of eating, overeating, and eating in response to emotional stress (40). An interaction between cognitive restraint and disinhibition has been described. Restraint behavior (i.e., dieting) involves constant refusing temptation from the stimuli (i.e., palatable foods) often creates a state of perceived deprivation. Eventually, when habitual dietary restraint is broken, individuals tend to overconsume in response to stimuli, which results in disinhibition (41). A high disinhibition score represents a risk factor for obesity and is consistently associated with high total energy intake and weight gain (42–44). In the present study, during the weight loss period, dietary restraint was predictive of less weight loss; during the weight regain period, dietary restraint and disinhibition were both predictive of more weight regain, indicating the relationship between cognitive restraint and body weight may change over time. We also observed that a higher hunger score was predictive of weight loss, suggesting that participants who had higher hunger score might withstand negative energy balance for a sustainable period to achieve weight loss (45). A recent report by Stinson et al. found that higher energy expenditure was associated with higher dietary disinhibition and greater susceptibility to hunger cues in men and women, potentially leading to uncontrolled overeating and weight gain (46). These findings provide a physiological basis and an energy-sensing mechanism for the psychological influences on human eating behavior.
The present study demonstrated the associations between food craving, eating habits, and weight changes in a randomized intervention trial. One of the strengths of our study is the use of repeated measurement on psychological, eating behavioral factor and its subsequent association with weight changes during both weight loss and regain periods over long periods. Our findings suggest that dietary intervention and consultation may help individuals curb their cravings of high-fat food and high-carbohydrate foods. Resisting intake of craved foods led to less consumption and greater weight loss in the context of a randomized trial. The repeated measures enable us to capture the changes of associations between psychological and behaviors factors after initial weight loss, thus providing evidence to further enhance weight maintenance in a weight management program. Limitations of the study include (i) results from the present study were derived from a clinical trial of weight in which the majority of study participants were women (64%) and Caucasian (80%), which may limit the generalization of these findings to general populations who are not on caloric restricted diets or those with different ethnicities; (ii) data on dietary intake was from self-report; previous evidence suggested that obese individuals tend to underreport food intake, which may contribute to unavoidable measurement error (47); and (iii) the current analysis did not take into account the genetic influences of weight loss and weight regain in response to dietary intervention. Future studies are needed to untangle these complex interactions. In the present study, 49% (n = 388) participants had fasting insulin ≥ 10.5 µIU/mL. The relationship between weight loss and food cravings was not modified by baseline insulin sensitivity, but this result needs to be interpreted with caution because we used fasting insulin as the surrogate marker of insulin sensitivity. Dynamic measures of postload insulin and glucose were not available in the POUNDS LOST trial to allow us use a more sensitive insulin sensitivity marker.
Conclusion
In the present study, we found that in a caloric restriction program with a nutritional education component, food craving and dietary restraint were the most likely factors that influence weight loss and weight regain. Participants were more likely to achieve weight loss through decreased energy and fat intake that was associated with craving of high-fat foods before dietary intervention. Dietary disinhibition is uniquely associated with weight regain. Future interventions that target these at-risk behaviors may lead to a more successful and sustainable weight loss. Better management of these psychological and behavioral factors may facilitate the prevention of gradual weight gain over time.
Acknowledgments
The authors thank all participants in the trial for their dedication and contribution to the research.
Financial Support: National Institutes of Health (NIH) U01HL073286, NIH R01DK115679, NIH/NIDDK 5R01DK097550.
Clinical Trial Registry: ClinicalTrials.gov number, NCT00072995)
Author Contributions: F.M.S. and J.T. designed the research; L.X., J.T., and D.H. performed the data analysis and interpretation and wrote the initial draft of the manuscript; C.M.C., L.Q., S.D.A., F.M.S., and J.T. critically revised the manuscript; G.B., D.W., S.D.A., and L.L. conducted the research; J.T. has primary responsibility for final content; and all authors read and approved the final manuscript and contributed to critically reviewing the manuscript.
Glossary
Abbreviations
- BMI
body mass index
- POUNDS LOST
Prevention of Obesity Using Novel Dietary Strategies
Additional Information
Disclosure Summary: None of the authors reported a conflict of interest related to the study.
Data Availability Statement: The datasets generated during and/or analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request. Data described in the manuscript, code book, and analytic code will be made available upon request application and approval.
References and Notes
- 1. Mozaffarian D, Hao T, Rimm EB, Willett WC, Hu FB. Changes in diet and lifestyle and long-term weight gain in women and men. N Engl J Med. 2011;364(25):2392–2404. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Zheng Y, Manson JE, Yuan C, et al. Associations of weight gain from early to middle adulthood with major health outcomes later in life. JAMA. 2017;318(3):255–269. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. White MA, Whisenhunt BL, Williamson DA, Greenway FL, Netemeyer RG. Development and validation of the food-craving inventory. Obes Res. 2002;10(2):107–114. [DOI] [PubMed] [Google Scholar]
- 4. Hill AJ, Heaton-Brown L. The experience of food craving: a prospective investigation in healthy women. J Psychosom Res. 1994;38(8):801–814. [DOI] [PubMed] [Google Scholar]
- 5. Pelchat ML, Schaefer S. Dietary monotony and food cravings in young and elderly adults. Physiol Behav. 2000;68(3):353–359. [DOI] [PubMed] [Google Scholar]
- 6. Weingarten HP, Elston D. Food cravings in a college population. Appetite. 1991;17(3):167–175. [DOI] [PubMed] [Google Scholar]
- 7. Chao A, Grilo CM, White MA, Sinha R. Food cravings, food intake, and weight status in a community-based sample. Eat Behav. 2014;15(3):478–482. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Martin CK, O’Neil PM, Tollefson G, Greenway FL, White MA. The association between food cravings and consumption of specific foods in a laboratory taste test. Appetite. 2008;51(2):324–326. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Boswell RG, Kober H. Food cue reactivity and craving predict eating and weight gain: a meta-analytic review. Obes Rev. 2016;17(2):159–177. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Sumithran P, Prendergast LA, Delbridge E, et al. Long-term persistence of hormonal adaptations to weight loss. N Engl J Med. 2011;365(17):1597–1604. [DOI] [PubMed] [Google Scholar]
- 11. Williamson DA, Lawson OJ, Brooks ER, et al. Association of body mass with dietary restraint and disinhibition. Appetite. 1995;25(1):31–41. [DOI] [PubMed] [Google Scholar]
- 12. Stinson EJ, Graham AL, Thearle MS, Gluck ME, Krakoff J, Piaggi P. Cognitive dietary restraint, disinhibition, and hunger are associated with 24-h energy expenditure. Int J Obes (Lond). 2019;43(7):1456–1465. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Lowe MR. The effects of dieting on eating behavior: a three-factor model. Psychol Bull. 1993;114(1):100–121. [DOI] [PubMed] [Google Scholar]
- 14. Coelho JS, Polivy J, Herman CP. Selective carbohydrate or protein restriction: effects on subsequent food intake and cravings. Appetite. 2006;47(3):352–360. [DOI] [PubMed] [Google Scholar]
- 15. Lawson OJ, Williamson DA, Champagne CM, et al. The association of body weight, dietary intake, and energy expenditure with dietary restraint and disinhibition. Obes Res. 1995;3(2):153–161. [DOI] [PubMed] [Google Scholar]
- 16. Stunkard AJ, Messick S. The three-factor eating questionnaire to measure dietary restraint, disinhibition and hunger. J Psychosom Res. 1985;29(1):71–83. [DOI] [PubMed] [Google Scholar]
- 17. Womble LG, Williamson DA, Martin CK, et al. Psychosocial variables associated with binge eating in obese males and females. Int J Eat Disord. 2001;30(2):217–221. [DOI] [PubMed] [Google Scholar]
- 18. Sacks FM, Bray GA, Carey VJ, et al. Comparison of weight-loss diets with different compositions of fat, protein, and carbohydrates. N Engl J Med. 2009;360(9):859–873. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Foster GD, Wadden TA, Peterson FJ, Letizia KA, Bartlett SJ, Conill AM. A controlled comparison of three very-low-calorie diets: effects on weight, body composition, and symptoms. Am J Clin Nutr. 1992;55(4):811–817. [DOI] [PubMed] [Google Scholar]
- 20. Wadden TA, Stunkard AJ, Day SC, Gould RA, Rubin CJ. Less food, less hunger: reports of appetite and symptoms in a controlled study of a protein-sparing modified fast. Int J Obes. 1987;11(3):239–249. [PubMed] [Google Scholar]
- 21. Hjorth MF, Ritz C, Blaak EE, et al. Pretreatment fasting plasma glucose and insulin modify dietary weight loss success: results from 3 randomized clinical trials. Am J Clin Nutr. 2017;106(2):499–505. [DOI] [PubMed] [Google Scholar]
- 22. Laakso M. How good a marker is insulin level for insulin resistance? Am J Epidemiol. 1993;137(9):959–965. [DOI] [PubMed] [Google Scholar]
- 23. Murdaugh DL, Cox JE, Cook EW 3rd, Weller RE. fMRI reactivity to high-calorie food pictures predicts short- and long-term outcome in a weight-loss program. Neuroimage. 2012;59(3):2709–2721. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Yokum S, Gearhardt AN, Harris JL, Brownell KD, Stice E. Individual differences in striatum activity to food commercials predict weight gain in adolescents. Obesity (Silver Spring). 2014;22(12):2544–2551. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. DiFeliceantonio AG, Coppin G, Rigoux L, et al. Supra-additive effects of combining fat and carbohydrate on food reward. Cell Metab. 2018;28(1):33–44.e3. [DOI] [PubMed] [Google Scholar]
- 26. McGloin AF, Livingstone MB, Greene LC, et al. Energy and fat intake in obese and lean children at varying risk of obesity. Int J Obes Relat Metab Disord. 2002;26(2):200–207. [DOI] [PubMed] [Google Scholar]
- 27. Rissanen A, Hakala P, Lissner L, Mattlar CE, Koskenvuo M, Rönnemaa T. Acquired preference especially for dietary fat and obesity: a study of weight-discordant monozygotic twin pairs. Int J Obes Relat Metab Disord. 2002;26(7):973–977. [DOI] [PubMed] [Google Scholar]
- 28. Teixeira PJ, Going SB, Houtkooper LB, et al. Exercise motivation, eating, and body image variables as predictors of weight control. Med Sci Sports Exerc. 2006;38(1):179–188. [DOI] [PubMed] [Google Scholar]
- 29. McGuire MT, Jeffery RW, French SA, Hannan PJ. The relationship between restraint and weight and weight-related behaviors among individuals in a community weight gain prevention trial. Int J Obes Relat Metab Disord. 2001;25(4):574–580. [DOI] [PubMed] [Google Scholar]
- 30. Provencher V, Drapeau V, Tremblay A, Després JP, Lemieux S. Eating behaviors and indexes of body composition in men and women from the Québec family study. Obes Res. 2003;11(6):783–792. [DOI] [PubMed] [Google Scholar]
- 31. Lowe MR, Doshi SD, Katterman SN, Feig EH. Dieting and restrained eating as prospective predictors of weight gain. Front Psychol. 2013;4:577. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Bellisle F, Clément K, Le Barzic M, Le Gall A, Guy-Grand B, Basdevant A. The Eating Inventory and body adiposity from leanness to massive obesity: a study of 2509 adults. Obes Res. 2004;12(12):2023–2030. [DOI] [PubMed] [Google Scholar]
- 33. Johnson F, Pratt M, Wardle J. Dietary restraint and self-regulation in eating behavior. Int J Obes (Lond). 2012;36(5):665–674. [DOI] [PubMed] [Google Scholar]
- 34. Hill AJ. Does dieting make you fat? Br J Nutr. 2004;92(Suppl 1):S15–S18. [DOI] [PubMed] [Google Scholar]
- 35. Heilbronn LK, de Jonge L, Frisard MI, et al. ; Pennington CALERIE Team Effect of 6-month calorie restriction on biomarkers of longevity, metabolic adaptation, and oxidative stress in overweight individuals: a randomized controlled trial. JAMA. 2006;295(13):1539–1548. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Martin CK, Heilbronn LK, de Jonge L, et al. Effect of calorie restriction on resting metabolic rate and spontaneous physical activity. Obesity (Silver Spring). 2007;15(12):2964–2973. [DOI] [PubMed] [Google Scholar]
- 37. Leidy HJ, Gardner JK, Frye BR, et al. Circulating ghrelin is sensitive to changes in body weight during a diet and exercise program in normal-weight young women. J Clin Endocrinol Metab. 2004;89(6):2659–2664. [DOI] [PubMed] [Google Scholar]
- 38. Mason C, Xiao L, Imayama I, et al. The effects of separate and combined dietary weight loss and exercise on fasting ghrelin concentrations in overweight and obese women: a randomized controlled trial. Clin Endocrinol (Oxf). 2015;82(3):369–376. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Polivy J, Herman CP. Restrained eating and food cues: recent findings and conclusions. Curr Obes Rep. 2017;6(1):79–85. [DOI] [PubMed] [Google Scholar]
- 40. Bryant EJ, King NA, Blundell JE. Disinhibition: its effects on appetite and weight regulation. Obes Rev. 2008;9(5):409–419. [DOI] [PubMed] [Google Scholar]
- 41. Lowe MR, Butryn ML. Hedonic hunger: a new dimension of appetite? Physiol Behav. 2007;91(4):432–439. [DOI] [PubMed] [Google Scholar]
- 42. Gallant AR, Tremblay A, Pérusse L, Bouchard C, Després JP, Drapeau V. The three-factor eating questionnaire and bmi in adolescents: results from the Québec family study. Br J Nutr. 2010;104(7):1074–1079. [DOI] [PubMed] [Google Scholar]
- 43. Hays NP, Roberts SB. Aspects of eating behaviors “disinhibition” and “restraint” are related to weight gain and BMI in women. Obesity (Silver Spring). 2008;16(1):52–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Zyriax BC, Wolf C, Schlüter A, Khattak AH, Westenhoefer J, Windler E. Association of cognitive dietary restraint and disinhibition with prediabetes–cross-sectional and longitudinal data of a feasibility study in German employees. Public Health Nutr. 2012;15(5):860–867. [DOI] [PubMed] [Google Scholar]
- 45. Batra P, Das SK, Salinardi T, et al. Eating behaviors as predictors of weight loss in a 6 month weight loss intervention. Obesity (Silver Spring). 2013;21(11):2256–2263. [DOI] [PubMed] [Google Scholar]
- 46. Stinson EJ, Graham AL, Thearle MS, Gluck ME, Krakoff J, Piaggi P. Cognitive dietary restraint, disinhibition, and hunger are associated with 24-h energy expenditure. Int J Obes (Lond). 2019;43(7):1456–1465. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Wehling H, Lusher J. People with a body mass index ⩾30 under-report their dietary intake: a systematic review. J Health Psychol. 2019;24(14):2042–2059. [DOI] [PubMed] [Google Scholar]