Nondietary psychological app program leads to sustained weight loss due to trained physiological satiety perception

Abstract

Obese people are mostly unable to maintain successful weight loss after the end of a dietary change. One reason is that conventional weight reduction concepts neglect physiological hunger and satiety perception, leading to a relapse to previous eating habits on the long run. We examined the long-term efficacy of a psychological smartphone weight loss program, which avoids any dietary instructions and aims at relearning of satiety perception. Parameters of body weight alterations and psychological features, for example, satiety perception, food cravings, and emotional eating, were explored in a nonrandomized experimental study comprising 75 obese participants. Measurements occurred at baseline, two times during program application, as well as at 6- and 12-month follow-up. Participants displayed significant weight loss during the entire study period (p = .029) and showed an improved body composition at the 6-month follow-up (p = .018). These effects were associated with increased satiety perception, as well as reduced food cravings, and emotional eating habits. Notably, all improvements in measured parameters significantly sustained between the end of the program and the 12-month follow-up (p < .005 for all). Psychological relearning of satiety perception may outclass dietary approaches in terms of long-term efficiency.

INTRODUCTION

Despite a variety of therapeutic approaches, the long-term effectiveness of widespread weight loss programs is extremely limited (Canuto et al., 2021; Jauch-Chara & Oltmanns, 2014; Kheniser et al., 2021). After ending the therapy, most people gain weight again (Amigo & Fernandez, 2007; Ge et al., 2020; Korkeila et al., 1999). At most one fifth of overweight people can maintain a weight loss of 10% of their original body weight for at least 1 year (Kraschnewski et al., 2010; Wing & Phelan, 2005). One of the reasons for the long-term failure of conventional energy-restrictive diets is the neglect of physiological hunger and satiety perception. Diet plans give instructions on the choice of food, the number of calories, and the exact time when to eat. In contrast, physiological food intake control regulated by hunger, satiety, and specific taste preferences is mostly not considered, which inevitably leads to a return to previous eating habits and in the end: body weight regain. Thereby, dieting and restrictive eating can promote overeating (Birch et al., 2003) and disturbances in food intake control (Wardle & Beales, 1988). Even at an early age, overweight and obese children show increased scores for eating in the absence of hunger (Kral et al., 2012; Lansigan et al., 2015), which is associated with their parents' restrictive child feeding practices (Fisher & Birch, 2002). In line, findings of an intervention study demonstrated an impaired brain but not peripheral (i.e. gut hormonal) satiety response after a 24-week weight loss program in obese 9- to 11-year-olds, which may promote future weight regain (Roth et al., 2022). Against this background, the concept of intuitive eating, which is defined as eating in response to physiological hunger and satiety cues (Tribole & Resch, 1996; Tylka, 2006), gains importance. Numerous studies revealed negative correlations between intuitive eating and body weight (Denny et al., 2013; Gray et al., 2012; Herbert et al., 2013; Owens et al., 2023). In line, Bacon et al. (2005) demonstrated in a randomized 6-month intervention study comprising weekly group meetings followed by six monthly aftercare sessions that an intuitive eating approach enables chronic dieters to maintain long-term body weight in a 2-year follow-up. Moreover, the intervention improved the awareness of body signals linked to hunger and satiety perception and reduced weight loss attempts by dieting. In contrast, diet group participants initially lost body weight and improved susceptibility to hunger perception after 1 year but finally regained weight at 2-year follow-up (Bacon et al., 2005). An additional psychological factor underlying chronic overeating is the emotional eating, that is, eating to regulate feelings irrespective of any hunger perception (Bennett & Latner, 2022; Mai-Lippold et al., 2024; Robinson et al., 2021). In order to regulate emotions, people may eat for reasons such as stress, frustration, or boredom in the absence of hunger or appetite. Psychological approaches may be helpful in this context. For instance, a combined mindfulness-prolonged chewing intervention for 8 weeks led to maintained weight loss, increased satiety awareness, less food craving, and decreased emotional as well as external eating in obese participants (Schnepper et al., 2019). This highlights the need to take psychological factors such as the lack of eating rituals and widespread emotional eating into account, which both correlate with body mass index (BMI) (Bénard et al., 2018; Wansink & van Kleef, 2014) and are generally not considered in popular diets either. It is known that trusting physiologic hunger and satiety cues decreases as the BMI increases (Denny et al., 2013). In the context of weight loss diets, restrictive eating is a risk factor for emotional eating (van Strien, 2018) and is directly linked to weight regain due to dietary disinhibition after dietary lapses (Herman & Mack, 1975). Against this background, we investigated the effectiveness of a behavioral therapeutic smartphone-app for weight loss, which explicitly abstains from diet and exercise instructions. Instead, the program focuses on relearning the perception of hunger and satiety and food-independent emotion regulation. Over a study period of 1 year, body weight, BMI, body composition, satiety perception, emotional eating, and food craving were examined.

METHODS

Study design

Our prospective interventional study was performed on self-selected obese participants (BMI > 30 kg/m²) who were of legal age (≥18 years) and had a smartphone (iOS or Android) available to use the smartphone app program. Participants were recruited via advertisement, mailing lists, and placards. Exclusion criteria were internal, neurological, or psychiatric diseases, which prohibited alterations in dietary habits, known internet dependency, and body weight-influencing medication. Each participant gave written informed consent and received an expense allowance for each participation day regardless if they lost weight or ceased from participation. The study has been carried out in accordance with the Declaration of Helsinki and was approved by the ethic committee of the University of Luebeck. The study comprised a period of 1 year, during which participants were examined at five time points. At the beginning of the study, individuals received free access to the app program (NUPP, Oakwood & Son UG [limited liability], Luebeck, Germany). They were instructed to use the app throughout the 3-month therapy cycle. Examination dates (±3 days each) were directly before starting the program, after 1 month of usage, at the end of the program application 3 months thereafter, and at 6- and 12-month follow-up. In the morning of each examination date, participants arrived at the human laboratories after an overnight fast.

Sample description

A total of 75 Caucasian, German participants were enrolled in the study, of which 41 (55%) were female. At study onset, volunteers displayed an average age of 50.09 ± 1.56 years, an initial body weight of 106.84 ± 2.08 kg, and a BMI of 35.44 ± 0.47 kg/m². With regard to overweight classification by the World Health Organization (WHO, 2000), more than half of the study participants (53%) were assigned to obesity class I (BMI 30.00–34.99), about one third (32%) to obesity class II (35.00–39.99), and 15% of the subjects had a BMI ≥ 40 kg/m² (obesity class III). After 1 month of program application, 71 individuals came to the second examination date. At the end of the behavioral program after 3 months, a group of 66 subjects participated in the examination again. At 6-month follow-up, still 62 returned for study examination, and at 12-month follow-up, 33 individuals completed the entire study period. A flow chart of the study sample and drop-out rates is shown in Figure 1. Although the drop-out rate by nature increased continuously with each examination date, only 13 subjects (17%) stopped participating in the study during the first 6 months. The drop in participant rate 1 year after study onset may be explained by the fact that the last examination date has been appended in the course of the ongoing study to gain additional information on long-term efficacy. A sample description of the main variables as well as a comparison of completers versus drop-out participants at baseline measurements are showed in Table 1. Comparing completers of the study with the drop-out group at baseline, we did not find any significant differences in body weight (p = .114, t-test) or BMI (p = .056, t-test).

FIGURE 1

Participant sample during the 12-month study.

TABLE 1. Baseline sample characteristics for the whole sample, completers of the study in all measurements, and the drop-outs.

		Total		Completers		Drop-out
Sample size (n)	Study sample	75		33		42
	Male	34		16		18
	Female	41		17		24

		M	SEM^a	M	SEM^a	M	SEM^a	p-value
Age	Year	50.1	1.6	53.8	1.9	47.2	2.3	.034*
Body height	Meter	1.7	0	1.7	0	1.7	0	.656
Body weight	kg	106.8	2.1	103.1	2.8	109.8	3	.114
BMI	kg/m^2	35.4	0.5	34.4	0.6	36.2	0.7	.056
Waist circumference	cm	111.2	1.6	109.6	2.1	112.5	2.2	.353
Hip circumference	cm	119.3	1.1	116.7	1.4	121.4	1.6	.035*

• Note: Comparisons of completers versus drop-outs are presented.
• ^a Standard error of the mean.
• * p < .05.

The psychological weight loss program

The tested behavioral therapeutic program differs from conventional weight loss programs in several points: Relearning of the individual physiological perception of hunger and satiety is at the center of the concept, thereby aiming at the permanent reduction of meal portion sizes and snacking frequency without predetermined instructions of specific food choice. The program completely dispenses with dietary instructions, calorie counting, and exercise programs. Moreover, coping strategies to practice food-independent emotion regulation are integral part of the concept. The program comprises a 3-month application cycle consisting of three pillars named Rule, Emotion, and Daily. The Rule module contains weekly behavioral instructions for 3 months (in total 12 Rules), which build on each other and are to be maintained over the long term (e.g. conscious chewing and focusing taste perception). Behavioral instructions aim at generating a cognitive awareness of appetite-controlled food intake and restoring physiological perception and differentiation between hunger and satiety. The Emotion module is an interactive question-and-answer tool, in which psychological support can be worked on up to two times a week (e.g. stress coping). The main goal of this pillar is to generate awareness for emotional eating and to develop alternative strategies for the regulation of emotions. The Daily module consists of daily push notifications, which are informative or convey behavioral support. In addition, the program includes a gamification element to activate the users' reward system and distract them in case of a crisis such as food cravings.

Data acquisition

Anthropometric measurements

Body weight was determined by a digital body scale (Kern MPB 300K100, Kern & Sohn GmbH, Balingen, Germany), and BMI was calculated with body height thereafter. Body composition was measured by bioelectrical impedance analysis (BIA, Nutribox, Data Input GmbH, Poecking, Germany). The percentage of fat mass refers to body fat mass in relation to body weight and the percentage of muscle mass describes the proportion of body cell mass in body weight. Fat-to-muscle-mass-ratio was calculated from these measures.

Questionnaires

Rule compliance

Several questions addressed the compliance to each of the 12 Rules during the last 4 weeks as main behavior instructions of the program (e.g. “In the last four weeks, did you eat three meals every day?”; “… chew your food long and consciously?”; “deliberately fill smaller portions on your plate?”; “… consciously pay attention to the taste of your meal?”). Participants were asked to estimate the application frequency on a scale of 1 (never) to 6 (always). Values were averaged for each Rule and overall compliance was calculated from the average of all Rules.

Hunger and satiety

In this questionnaire, seven statements on the perception of hunger, satiety, as well as food cravings were listed (e.g. “I can feel exactly when I am satiated”; “I feel when I am hungry”; “When I am satiated, I stop eating right away.” “When I have cravings, I eat something right away”). The frequency of these items was assessed on a scale of 1 (never) to 6 (very often). An exploratory factor analysis revealed the two factors satiety perception (Cronbach's α = .759) and food cravings (Cronbach's α = .688). We performed a principal component analysis (PCA) on seven items using varimax rotation. The Kaiser–Meyer–Olkin measure (KMO) indicates the suitability of the data for a factor analysis KMO = .728 and Bartlett's test of sphericity was significant (p < .001) indicating that correlations between items were sufficiently high. Only factors with eigenvalues ≥1 were considered. Two factors with eigenvalues exceeding 1 accounted for 58.79% of the total variance (please see the Rotated Factor Matrix in Appendix A).

Emotional eating

Emotional eating was assessed using a modified version of the subscale “emotional eating” from the German version of the Dutch Eating Behavior Questionnaire (DEBQ) (Nagl et al., 2016). In the applied version, nine items on emotional eating behavior were rated on a 6-point Likert scale (e.g. “I have a desire to eat when I am depressed or discouraged”; “I have a desire to eat when I am feeling lonely”; “I have a desire to eat when I am cross”; “I get the desire to eat when I am anxious”). In our modification, we added two items (“I have a desire to eat when I am excited” and “I have a desire to eat when I am afraid”) and removed three original items, which were not applicable. The reliability of this analysis yielded a Cronbach's α of .916, which is consistent with the results of previous studies (Nagl et al., 2016) and reflects a very good internal consistency.

Sample size and statistical analyses

For power calculation (R Foundation, version 3.3.2), we assumed a moderate average weight loss of 5 kg with a standard deviation of 10 kg within 6 months. For a power of .90 and a significance level of .05, this calculation resulted in a minimum sample size of 44 subjects. To compensate for a high drop-out rate of typically 50% in weight loss studies (Moroshko et al., 2011), 31 additional subjects were recruited. Statistical analysis was based on intention to treat including participants who completed baseline measurements and got access to the app program. Missing data were not imputed. The Shapiro–Wilk test was conducted to ensure normal distribution in factor levels of all measured parameters. Despite satiety perception after 3 months of program use, factor levels in all parameters were not significant and met the assumption of normal distribution for analyses of variance (ANOVA). Statistical analyses (using Superior Performing Software System, version 27, IBM SPSS, Armonk, NY, USA) were based on ANOVA for repeated measurements including the factor time. In case of violation of the sphericity conjecture, we corrected according to Greenhouse–Geisser procedure. Pairwise mean differences between time points were tested using Fisher's LSD. For pairwise comparisons of single time points or groups, paired or unpaired Student's t-test were used, respectively. To describe the relationship between Rule compliance and weight loss, simple linear regressions at 3-, 6-, and 12-month follow-up were applied. Furthermore, for measures of the strength between two continuous variables, Pearson correlation coefficients were calculated. All statistical tests were performed at the two-sided .05 level.

RESULTS

Descriptive details of the main measures, for example, body weight, BMI, Rule compliance, satiety perception, food cravings, and emotional eating for all time points of measurements are shown in Table 2.

TABLE 2. Outcome measures after 1 and 3-month program use and at 6- and 12-month follow-up.

	n	Minimum	Maximum	M	SEM^a	SD	p-value
Body weight in kg							p = .029*
Baseline	75	77.9	178.0	106.9	2.1	18.0
1 month	71	77.5	177.0	104.7	2.0	16.8
3 months	66	76.8	181.1	104.5	2.2	17.7
6-month follow-up	62	76.6	180.0	104.3	2.3	17.9
12-month follow-up	33	79.0	140.4	100.7	2.6	14.7
Body mass index (kg/m^2)							p = .036*
Baseline	75	29.8	47.8	35.4	0.5	4.0
1 month	71	29.5	44.6	34.8	0.4	3.7
3 months	66	29.3	45.7	34.9	0.5	4.1
6-month follow-up	62	29.5	47.4	34.6	0.5	3.9
12-month follow-up	33	29.7	40.4	33.7	0.5	3.1
Rule compliance							p < .001***
Baseline	75	2.3	5.1	3.7	0.1	0.6
1 month	70	2.8	5.1	4.2	0.1	0.5
3 months	63	2.9	5.6	4.5	0.1	0.6
6-month follow-up	61	2.8	5.4	4.5	0.1	0.6
12-month follow-up	33	3.5	5.9	4.5	0.1	0.5
Satiety perception							p = .001**
Baseline	74	1.8	5.4	3.53	0.1	0.8
1 month	67	2.2	5.4	3.9	0.1	0.8
3 months	66	2.2	5.8	4.2	0.1	0.7
6-month follow-up	61	2.2	5.6	4.0	0.1	0.8
12-month follow-up	30	2.6	5.4	4.1	0.2	0.8
Food cravings							p = .001**
Baseline	74	1.0	6.0	3.74	0.1	1.2
1 month	67	1.0	5.0	3.3	0.1	1.0
3 months	66	1.0	5.5	3.1	0.1	1.0
6-month follow-up	62	1.0	5.0	3.1	0.2	1.2
12-month follow-up	30	1.0	4.5	3.0	0.2	1.0
Emotional eating							p = .003**
Baseline	73	1.1	5.1	3.1	0.1	1.1
1 month	66	1.2	5.4	2.9	0.1	1.0
3 months	66	1.0	5.1	2.7	0.1	1.1
6-month follow-up	62	1.0	5.4	2.7	0.1	1.1
12-month follow-up	29	1.3	4.6	2.7	0.2	1.1

• Note: All scales 1 (never) to 6 (always/very often). p-values for the overall effect including all time points of the study.
• ^a Standard error of mean.
• * p < .05,
• ** p < .01, and
• *** p < .001.

Rule compliance, satiety perception, food cravings, and emotional eating

Analyses along all time points revealed a significant increase in Rule compliance over the entire study period (F[3, 86.04] = 25.23, p < .001; Figure 2 and Table 2). Adherence to Rules increased continuously during the 3 months of active app use (M_Diff = −0.71, 95% CI [−0.94, −0.48], p < .001). Thereafter, it persisted at this elevated level until the end of measurements (p = .930, repeated measures ANOVA); that is, participants consistently followed the Rules even after program application had ended (Figure 2). Rule compliance correlated negatively with body weight across all measured time points (r[306] = −154, p = .008), which strongly indicates that adherence to the app Rules is associated with successful weight loss. There were no differences in Rule compliance between individuals with different obesity grades (p = .518, two-way repeated measures ANOVA). In terms of satiety perception, we observed a highly significant increase over the entire study period (F[2.95, 76.78] = 6.14, p = .001; Figure 2 and Table 2). Post hoc pairwise comparisons revealed a significantly increased scoring of satiety perception after 3 (M_Diff = −0.44, 95% CI [−0.74, −0.15], p = .004), 6 (M_Diff = −0.48, 95% CI [−0.87, −0.10], p = .016), and 12 months (M_Diff = −0.57, 95% CI [−0.88, −0.26], p = .001) compared with baseline measurements. Furthermore, including all time points, there was a significant correlation between baseline-adjusted body weight change with satiety perception (r[297] = −.165, p = .004) and food cravings (r[298] = .155, p = .007), but not emotional eating (r = .070, p = .230). Reporting of food cravings' frequency distinctly decreased during the entire measurement period (F[2.82, 75.99] = 5.90, p = .001; Figure 2 and Table 2). In accordance with satiety perception, post hoc pairwise comparisons demonstrated significantly lowered cravings scoring after 1 (M_Diff = 0.50, 95% CI [0.05, 0.95], p = .032), 3 (M_Diff = 0.57, 95% CI [0.07, 1.07], p = .027), 6 (M_Diff = 0.77, 95% CI [0.29, 1.25], p = .003), and 12 months (M_Diff = 0.79, 95% CI [0.37, 1.20], p < .001) as compared with baseline measurement. Emotional eating also decreased in the course of the study (F[4, 100] = 4.37, p = .003; Figure 2 and Table 2). Respective scores were significantly reduced after 3 months of app use (post hoc pairwise comparisons: M_Diff = 0.39, 95% CI [0.08, 0.70], p = .015), as well as after 6 (M_Diff = 0.38, 95% CI [0.04, 0.72], p = .029), and 12 months as compared with baseline (M_Diff = 0.49, 95% CI [0.17, 0.81], p = .004). In order to test for a connection between Rule compliance and improvements in the three parameters satiety perception, food cravings, and emotional eating, we performed respective correlation analyses. Statistical evaluation including all time points evidenced a significant correlation between Rule compliance and the three parameters (satiety perception: r[292] = .468, p < .001, food cravings: r[293] = −.292, p < .001, emotional eating: r[290] = −.224, p < .001). It may therefore be assumed that Rule compliance is associated with improvements in these measures.

FIGURE 2

Questionnaire scores of Rule compliance and psychological aspects trained by the app. Mean values ± SEM of Rule compliance (scale never [1] to always [6]), satiety perception, food cravings, and emotional eating (all scales never [1] to very often [6]) during the 12-month study. Data were analyzed using repeated measurements ANOVA. **p < .01, ***p < .001.

Body weight, BMI, and body composition

Analyses of the data sample revealed a body weight reduction ranging from 0.1 kg to a maximum of 8 kg (M = 2.54 ± 0.32 kg) after 3 months of app use. At the 12-month follow-up, weight loss ranged from 0.2 kg up to 15 kg (M = 5.28 ± 0.92 kg) as compared with baseline measurements. Analysis of variance including all time points revealed a significant effect of time on body weight over the entire 12-month period (F[1.76, 54.49] = 3.98, p = .029). Differential post hoc pairwise analyses between baseline measurements of body weight and single time points revealed a significant continued body weight reduction between baseline and 1 (M_Diff = 0.73, 95% CI [0.17,1.28], p = .012), 3 (M_Diff = 1.11, 95% CI [0.10, 2.13], p = .033). 6 (M_Diff = 1.28, 95% CI [0.10, 2.45], p = .034), and 12 months (M_Diff = 2.50, 95% CI [0.61, 4.39], p = .011; Figure 3a). Remarkably, body weight remained at the reduced level and did not rebound after end of the intervention (p = .144) as mostly observed after calorie-reduced diets. In analogy to body weight, we found a significant time effect of the weight loss program on BMI (F[1.81, 56.17] = 3.67, p = .036) during the entire study period (Figure 3b). Differential post hoc comparisons of respective single time points with baseline measurement showed a significantly reduced BMI after 1 (M_Diff = 0.24, 95% CI [0.06, 0.42], p = .012), 3 (M_Diff = 0.35, 95% CI [0.02, 0.68], p = .037), 6 (M_Diff = 0.41, 95% CI [0.01, 0.81], p = .043), and 12 months (M_Diff = 0.79, 95% CI [0.18, 1.39], p = .01; Figure 3b). BMI remained at the decreased level between 3 and 12 months (p = .159). Taking a closer look at weight loss effects, we found that after 3 months of app use 59.1% of the users (n = 39) lost body weight while 49.1% maintained or gained weight (n = 27). In order to explain the weight loss failure rate, we analyzed to which extent study participants followed the Rules in the app program. We found that weight loss failure could be explained by decreased efforts to follow the app Rules as we revealed a significant negative correlation between Rule compliance and body weight across all measurement time points. In particular those who showed distinct weight loss (“program responders”) were associated with high Rule compliance scores, while “nonresponders” were correlated with low Rule compliance after 3 months of app use (r[62] = −.365, p = .003). Moreover, “responders” displayed a significant ascending development in terms of Rule compliance after 3 months of app usage as compared with “nonresponders” (t[61] = 3.14, 95% CI [0.15, 0.68], p = .003). In 12-month follow-up measurements, 63.6% (n = 21) of users have lost weight 1 year after starting the app, while 12 participants showed no further weight reduction. In order to examine the effectiveness of higher Rule compliance in predicting body weight reduction after 3, 6, and 12 months, we performed a linear regression analysis. Results indicated that Rule compliance explained 13% of the variation in body weight change after 3 months of app use (F[1, 61] = 9.118, p = .004; Table 3). In analogy, Rule adherence had a significant promoting impact on the extent of weight reduction after 6 months (F[1, 59] = 4.980, p = .029; Table 3) and explained 8% of the variation in body weight change. The described impact of Rule compliance on body weight reduction was not seen after 12 months (p = .793; Table 3), possibly due to the elevated drop-out rate and minimized sample size of n = 32. Analyzing body composition with paired t-tests, data revealed some alterations in response to the intervention. The percentage of fat mass significantly declined from baseline to 6 months thereafter (t[61] = 3.31, p = .002; 37.88% vs. 36.64%). This was also the case by trend between baseline and after 12 months (p = .060; 36.55% vs. 35.55%). The percentage of muscle mass increased by strong trend between baseline and 6 months thereafter (p = .052; 31.23% vs. 31.84%). Accordingly, we observed a significantly decreased fat-to-muscle ratio 6 months after study onset as compared with baseline (t[61] = 2.44, p = .018; 1.29 vs. 1.23).

FIGURE 3

Body weight and body mass index (BMI) alterations upon app usage. Post hoc pairwise comparisons between baseline and each time point (mean values ± SEM) of body weight (a) and BMI (b) measurements comprising volunteers having participated in all five examinations during the study. All marked comparisons *p < .05, n = 32.

TABLE 3. Linear regression analyses for rule compliance predicting body weight reduction.

Predictor variable	B	Standard error B	Beta	p-value	Regression results
Intercept	7.447
Rule compliance program end	−1.832	0.607	−.361	.004**	R^2 = .130 F(1, 61) = 9.118
Intercept	6.717
Rule compliance 6-month follow-up	−1.700	0.762	−.279	.029*	R^2 = .078 F(1, 59) = 4.980
Intercept	−0.335
Rule compliance 12-month follow-up	−0.456	1.727	−.047	.793	R^2 = .002 F(1, 31) = 0.070

• * p < .05, and
• ** p < .01.

DISCUSSION

Our data demonstrate that a nondietary psychological weight loss program leads to distinct reductions in body weight as well as BMI and improves body composition. Strikingly, in contrast to prevailing conventional weight loss programs, which mostly display body weight regain after ending of the intervention, the psychological app program even causes persistent reductions of body weight for at least 9 months after program cessation. Due to the intentional avoidance of predetermined restrictions in food choice, the behavioral therapeutic program prevents and reduces food cravings and fosters satiety perception leading to slow but sustained weight loss without the commonly known struggle between preferred and prescribed foodstuff. As demonstrated, behavioral therapeutic app support moreover abolishes emotional eating habits. Looking at previous studies, a meta-analysis of randomized trials published in 2020 compared several dietary programs (3 months of duration) containing structured advice for daily macronutrients/food (e.g. WeightWatchers) or calorie intake/low carbohydrate (e.g. Atkins and Ornish) and quantified the effectiveness on body weight after program ending, and at 6- and 12-month follow-up. Results showed that all examined types of diet were associated with distinct reductions in body weight as compared with free choice nutrition after 3 months, an effect that diminished at 12-month follow-up among all 14 popular listed diets (Ge et al., 2020). This leads to conclude that the most challenging difficulty fighting obesity is to prevent body weight regain in the long term. Dieters who gain back more weight than they lost are rather the norm than an unlucky exception (Mann et al., 2007). Consequently, multiple repeated attempts to lose body weight through dieting exacerbate the obesity problem and lead to a further increase in body weight driven by biological mechanisms (Busetto et al., 2021). Among others, the latter comprise persistent metabolic slowing 6 years after body weight reduction (Fothergill et al., 2016) or gut hormone dysbalance in conjunction with persistently increased hunger scores despite weight regain 52 weeks after body weight reduction (Sumithran et al., 2011). In fact, weight loss strategies such as eating very little, consuming formula diets, and skipping meals are strongly associated with the largest weight regain over a 10-year period (Neumark-Sztainer et al., 2012). These results have impressively been demonstrated in the Growing up today study comprising 16,000 participants in a 3-year follow-up (Field et al., 2003) and 1902 individuals over a 10-year period (Neumark-Sztainer et al., 2012). As the authors revealed, dieting and unhealthy strict weight control behaviors are the strongest predictors for weight regain (Neumark-Sztainer et al., 2012). In conclusion, and in line with Benton and Young (2017), it can be assumed that counting calories and decreasing calorie intake by specific prescribed meals only have a limited short-term effect on body weight loss. So alternative strategies addressing the psychological aspects of overeating and weight loss are urgently required. One of these alternative approaches may represent a comprehensive community-based behavioral weight loss program, which included 20 sessions of behavioral therapeutic group treatment under psychologist guidance over a 6-month period (Latner et al., 2013). The purpose of these group meetings focused on behavioral change strategies, that is, actively controlling and slowing down eating velocity and improving satiety awareness, as well as moderate physical activity of 150 min per week. After the 6-month treatment period, all participants continued self-supported weekly group meetings and respective body weight alterations were monitored 6 and 18 months after treatment finish. Similar to our data, the adherence to treatment was high, body weight reduction occurred in the course of the treatment period, and achievements were maintained for further 6 and 12 months thereafter. However, group therapy programs require comprehensive psychological staff and time expense and—in contrast to an app intervention—focus on calorie intake control as well as food intake monitoring via eating protocols. Both are strategies, which are deliberately avoided in the psychological app tested in our study in order to generally draw off the attention from calories and rather concentrate on physiological satiety perception. Moreover, the app program does not categorize foods into healthy and unhealthy and, as a consequence, a restriction of “allowed” foods eventually leading to craving for “forbidden” foodstuff. This is in line with an unexpected previous outcome of a 22-month behavioral weight-loss program among 182 obese adults, demonstrating that a recommended reduction in “problem food consumption,” that is, high-sugar or high-fat processed foods, distinctly prevented any weight loss (Gordon et al., 2020). Looking at the specific causes underlying the success of our tested app program, analyses showed that the effect of body weight reduction mainly depends on the adherence to the basic Rules provided by the app. In contrast to participants with a low Rule compliance, volunteers with a high adherence successfully reduced body weight. Rule compliance predicted body weight alterations after 3 months of app use as well as at 6-month follow-up. In line, we observed a distinct fat mass reduction and simultaneous muscle mass increase mainly during the first 6 months. Thereafter, both parameters persisted at this improved level. Another cause underlying the successful and sustained weight loss through the app use is the fact that the surveyed Rule compliance with the app concept was constantly high throughout the 12-month study period. As every therapeutic intervention, adherence to a weight loss program is a compelling requirement to succeed (Jacobs et al., 2017; Turner-McGrievy et al., 2019). Thurner-McGrievy et al. compared seven methods of tracking adherence to mobile dietary self-monitoring and found that self-monitoring declined over time. Moreover, all examined methods of self-monitoring exhibited fewer than half of all study participants still tracking after 10 weeks (Turner-McGrievy et al., 2019). In fact, an app specifically developed to maintain previous weight loss, which was based on conventional weight loss strategies (tracking weight, food intake, and physical activity) did not succeed (Brindal et al., 2019). Right to the contrary, after 6 months, more than half of the participants regained up to 2% of their initial body weight prior to weight loss and were probably in the middle of the yo-yo effect. Another crucial factor to achieve sustained weight loss is the individual's awareness to feel satiated and avoid food cravings, which both generally prevent overeating. Therefore, one main goal of the tested app program is to elevate satiety perception and reduce food cravings. It is known that altered satiety perception and obesity are tightly related (Hellstrom, 2013; Wolnerhanssen et al., 2017), and it has previously been demonstrated that the BMI increases as trusting physiological hunger and satiety signals decrease (Denny et al., 2013). Therefore, the app program implements mindful food intake, that is, without disruptive media use, reducing eating velocity, recreating awareness for the food intake procedure, and following the individual's own instinct as reference cues for hunger and satiety. However, this approach is not entirely new. Mindful-based interventions are part of binge-eating disorder therapy (Kristeller et al., 2014) but are gaining more and more interest also in the treatment of obesity. High scores in mindful eating are associated with lower BMI (Kes & Can Cicek, 2021) and meta-analyses showed that mindful training has a significant positive effect on weight reduction as compared with nonintervention and active treatment controls (Carriere et al., 2018; Fuentes Artiles et al., 2019). Carriere et al. (2018) examined 18 studies using mindfulness-based interventions among 1160 participants and demonstrated that body weight reduction has been mediated by alterations in eating behavior. Hence, participants in mindful-based interventions showed continued weight loss at follow-up, while participants in diet and exercise programs slightly gained weight. Moreover, participants in the mindfulness training of 5 months reported significantly reduced levels in reward-driven eating at the 6-month postintervention examination, which, in turn, induced weight loss at the 12-month follow-up (Mason et al., 2016). However, in our study, not exclusively the perception of physiological satiety signals is addressed by the app program. It is known that emotional eating is strongly associated with body weight status and is more substantial in former and current dieters than in individuals without a history of dieting (Peneau et al., 2013). Food intake, as a response to emotions such as boredom, sadness, or happiness, which are decoupled from hunger perception, is called “emotional eating.” Emotional states can have major effects on eating behavior. Most findings indicate overeating during negative emotional states in healthy obese individuals as compared with normal (Reichenberger et al., 2021) and underweight individuals (Geliebter & Aversa, 2003). In the context of weight loss diets, restrictive eating is associated with emotional regulation difficulties (Haynos et al., 2018), which is a risk factor for emotional eating (van Strien, 2018). Our data show a significant decline in emotional eating behavior during the program cycle period, an effect that persisted until follow-up measurements and is certainly part of the successful strategy in sustained weight loss. Looking at our data, a plateau in terms of Rule compliance, satiety perception, food cravings, and emotional eating becomes evident after 3 months of the study. In order to find an explanation for this phenomenon, we presume some kind of a ceiling effect. Continuous learning all of the 12 Rules accompanied by regular reminders to keep applying them initially leads to the observed alterations during the 3 months of app use. Thereafter, the program ends and no further new contents are conveyed to participants. At this point, all learned cues are apparently kept in mind (in particular because the app can be furthermore used) but—of course—due to the cease of new messages, all parameters remain at the same level as at the end of the program. This potentially successful internalizing of the program content, specifically the Rules, can be speculated to be the underlying cause of the ongoing weight loss even after the program itself had ended. We could imagine that continuing the program over a longer period may even boost successful weight loss upon using the app in the future. However, some aspects of our study may limit the interpretation of the data. One point is the lack of a nontreated obese control group for comparison in our within-subject design. However, a suitable control group (matched for gender, age, etc.) can hardly be implemented in such a study as potential participants must compellingly be obese and refrain from any active attempt to lose weight by conventional weight loss programs; that is, it is required that therapeutic treatment is refused to people who are in need of it. This would have been highly questionable in terms of ethical standards. A consideration for further studies could be the comparison of our psychological approach with a conventional calorie-reduced diet program as a control group. On the other hand (as specified in the introduction), numerous previous data already demonstrate that—in contrast to our psychological app program—conventional weight loss interventions lead to rapid weight regain after ending. Another limitation may be the drop-out rate at the 12-month follow-up as we cannot exclude that some participants (i.e. those who failed to lose or maintain body weight) dropped out disproportionally often. Moreover, we cannot exclude a possible social-desirability bias in the self-reports. To avoid such unconscious and conscious social bias as far as possible, we ensured a high degree of anonymity during the interviews and confidentiality with received answers. In order to further amplify the point of the drop-out rate, self-motivation is a further aspect to be addressed here. In this context, one could consider that the compliance of our self-selected study participants may be higher than for patients who receive a prescription from a doctor to use the app, that is, in a clinical setting. Indeed, self-motivation is crucial to succeed in any therapeutic approach, which is based on behavioral changes. This is true in terms of all weight loss concepts, which require a fundamental contribution by the individual itself. Our participants—as most obese people—desired to lose weight. We think that our psychological program does not differ from other approaches in this regard, that is, a prescription from the doctor does not help if the patient is not motivated to implement new habits. However, our participants were requested to stick to the program at least for 1 year, and it seems likely that they did so because they successfully lost weight without all the privation (e.g. forbidden or nonpalatable foodstuffs, hunger periods, joyless exercising) they were used to by conventional concepts before.

CONCLUSION

Our data demonstrate that a smartphone-based behavioral therapeutic weight loss program may lead to a sustained body weight loss in obese participants. In contrast to conventional therapeutic approaches such as calorie-reduced dieting, the app program may avoid weight regain at least for 9 months after the end of the intervention. This effect was associated with improved satiety perception, reduced food cravings, and less emotional eating. Our results demonstrate that psychological aspects underlying overeating must compellingly be addressed in weight loss concepts to succeed on the long run.

ACKNOWLEDGMENTS

Anonymized participant data (e.g. study protocol, statistical analysis, informed consent, and study report) underlying the results reported in the article can be shared with researchers presenting a methodologically sound proposal 6 to 24 months after publication of the article.

CONFLICT OF INTEREST STATEMENT

All authors report no conflict of interest.

ETHICS STATEMENT

The Ethics Committee of Lübeck approved the present study.

APPENDIX A: EXPLORATORY FACTOR ANALYSIS

	Factor
	Satiety	Food cravings
Item 01: I can feel exactly when I'm full	.823	−.125
Item 03: When I'm full, I stop eating immediately	.817	−.105
Item 07: When I'm not hungry, I eat less	.683	−.155
Item 06: I always eat up everything	−.666	.044
Item 02: I can feel when I'm hungry	.472	−.245
Item 05: When I have food cravings, I eat something immediately	−.063	.877
Item 04: I regularly have food cravings	−.229	.836

• Note. Rotated factor matrix. Extraction method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization. Rotation converged in 3 iterations.

Open Research

DATA AVAILABILITY STATEMENT

Anonymized participant data (i.e., study protocol, statistical analysis, informed consent, study report) underlying the results reported in the article can be shared with researchers presenting a methodologically sound proposal 6 to 24 months after publication of the article.