Introduction

The obesity epidemic is a global public health issue. In 2016, approximately 1.9 billion adults were people with overweight, and over 650 million adults were people with obesity [1]. According to the World Health Organization (WHO), approximately 60% of European adults were people with overweight or obesity as of 2022 [2]. Severe obesity increases the risk of various diseases, including adverse cardiovascular events, non-alcoholic fatty liver disease, and cancer, thereby endangering the lives of the population with obesity [3,4,5]. However, research indicates that even when using the same weight loss strategy, there is significant individual variation in the degree of weight loss among people with obesity. This often results in a cycle of losing weight and regaining it, complicating long-term weight maintenance [6]. Weight maintenance is defined as the intentional loss of weight and subsequent maintenance of that weight for at least six months [7]. Weight cycling is a cyclical cycle of weight loss (≥ 5 kg) and weight regain, affecting 20–30% of adults [7, 8]. Several studies have found that weight cycling increases weight, the metabolic burden and chances of develo** chronic disease of people with obesity [9, 10].

Most current medical weight loss interventions combine lifestyle changes, such as diet and exercise, with medication and surgery being recommended for those with severe obesity and co-morbidities [10]. Energy restriction and changes in macronutrient distribution are common dietary interventions, and exercise interventions alternate between aerobic and resistance exercise training. However, adhering to these exercise and healthy eating plans is challenging for most participants, especially those who have experienced weight cycling. After multiple weight regains, physiologically in terms of body composition changes and metabolic alterations and psychologically in terms of internal drive [11].

Depression, anxiety, and other forms of distress are referred to as psychological distress [12, 13]. Previous studies have established a strong association between psychological issues such as anxiety and depression with weight cycling or weight regain [14, 15]. Large-sample statistical analyses have revealed that populations with weight cycling were more likely to experience depressive symptoms than non-weight cyclers and that there was a link between frequent weight cycling and depression [16]. In addition, participants with bipolar disorder and a history of weight cycling, particularly those who often regained lost weight or struggled to sustain their weight loss, experienced more frequent relapses of symptoms such as mania and depression [17].. A subsequent study discovered that adults with baseline anxiety and depression exhibited lower levels of weight maintenance over two years, were more likely to gain weight, and that depression severity had a dose-response effect with significant weight gain [18,19,20]. In contrast, some remission of major depression and dysthymia was linked to improved weight control and loss. It indicated a clear relationship between psychiatric and psychological distress, weight cycling, and weight maintenance. Mental and psychological distress have been identified as essential factors in the development of certain diseases. Psychological stress has an impact on irritable bowel syndrome. Severe mental illness and cardiovascular disease have some root causes in common, and those who have a mental illness are more likely to develop cardiovascular disease [21, 22]. Currently, studies on obesity-related diseases suggests that mental stress may influence changes in eating behavior and metabolic disorders among diseases associated with obesity. According to an increasing number of studies, independent mental diseases are closely related to the occurrence of metabolic diseases [23]. We hypothesize that reducing psychological distress in people with obesity and a history of weight cycling will lead to better weight-loss outcomes.

Eating behavior can also be influenced by psychological distress. Previous cross-sectional studies have discovered a link between psychological distress and emotional eating, binge eating, and external eating. Emotional eating behaviors, in general, are mediators of the effect of psychological distress on health-related quality of life [24]. Psychological distress regulation is strongly linked to problematic eating behavior in adolescents [25]. The indirect effect of psychological distress on emotional eating via emotional dysregulation was found to be significant in mediation analyses [26]. Furthermore, 25% of adults eligible for weight loss surgery exhibit food addiction and binge eating behavior, which is linked to a high prevalence of psychological distress, including depression and anxiety disorders [27]. It is important to note that eating behavior has been linked to both psychological distress and weight loss levels. Previous research has found that changes in eating habits for weight loss are associated with better outcomes [28]. The presence and severity of eating behaviors may play a role in mediating the relationship between psychological distress and weight loss levels. However, most prior studies have primarily focused on individual factors like depression, anxiety, attitudes towards healthy eating, or exercise habits, only examining their isolated impact on weight maintenance [29].

There have been few studies that look at the link between psychological distress, eating habits, and weight loss maintenance. We proposed a hypothesis that explores a sequential pathway involving psychological status, eating behavior, and weight control. This hypothesis is based on theoretical assumptions derived from existing research in the field, suggesting a complex interplay between these factors in individuals experiencing weight cycling. Our study aims to empirically test this pathway to better understand the mechanisms underlying successful weight management. This sequence is speculated to influence the effectiveness of weight loss in populations experiencing weight cycling. We used a cohort of weight loss clinic participants in this study. This study looked at the relationship between psychological distress, appetite, eating behavior, and weight loss outcomes in outpatients who had a history of weight cycling. It also examined whether the mediation effect of eating behavior between psychological status and body mass index (BMI) change found in previous studies can be replicated and whether this mediation effect is dependent on appetite. Using a structural equation model (SEM) for path analysis, we assessed the mediation and moderated mediation between psychological status, eating behavior, and weight loss. To the best of our knowledge, this is the first study to investigate the relationship between psychological distress and weight control in a cohort of participants with obesity and a history of weight cycling, as well as to mediate the role of eating behavior. This evidence will inform future improvements in psychological interventions for participants with a history of obesity and weight cycling (people experiencing weight loss difficulties), as well as provide a theoretical foundation for advancements in multi-disciplinary team (MDT) weight loss intervention strategies aimed at promoting physical and mental health.

Method

Participants

The Population with Weight Cycle Obesity Study at Peking Union Medical College Hospital was used in this investigator-initiated cohort study. The goal was to investigate the mechanisms and interventions that explain the differences in weight maintenance outcomes in this population. The study included outpatients recruited in the Clinical Nutrition Department of Peking Union Medical College Hospital in April 2022. Participants had purposefully lost and regained weight before attending the clinic. Weight loss prior to the visit occurred on its own and was not part of the study’s weight loss program. Weight cycling was defined as one or more episodes of weight loss in the previous two years, followed by weight recovery of 5% or more of the baseline weight before weight loss. The inclusion criteria were: aged 18 to 50 years; 28 ≤ BMI < 35 kg/m2; ability to comply with the weight loss intervention; and a history of weight cycling. The exclusion criteria were: weight loss of more than 10% over the past year; taking medication that affects weight, glucose and lipid metabolism, blood pressure, lipids, and hormones currently or over the past six months; currently pregnant or breastfeeding women; a history of severe cardiovascular, gastrointestinal, psychiatric, and cancer diseases; with abnormal liver and kidney function; and inability to complete the study well for various reasons. The researchers examined the causes of weight loss in the participants, excluding weight loss caused by various diseases (see exclusion criteria for details) and including only weight loss caused by conscious weight loss (e.g., dieting or increased exercise) or other lifestyle intervention programs for weight loss (non-drug and surgical treatments) that were begun and completed before the participants entered the study. People with obesity and a history of weight cycling comprised 153 (53 men and 100 women) participants who completed the three-month outpatient intervention program and six-month follow-up. Pregnancy, withdrawal, missed visits, and the sudden onset of illnesses that made participation in the study impossible were all reasons for non-completion. The Human Ethics Committee of PUMC Hospital approved the study (No. ZS-3067), and all participants provided written informed consent. The study was carried out following the principles of the Helsinki Declaration. It was published in Clinical Trials (NCT05311462).

Measurement

Participants’ height and body weight were measured using standard methods in light clothing and without shoes by a mechanical column scale (seca, China) and a bioelectrical impedance body composition analyzer (seca, China), respectively. The specialist prescribed a routine intervention program in the outpatient clinic for these participants. The diet was high-protein and energy-restricted (20–25 kcal/kg/d, of which protein was 1.5 kg/d). The exercise program included 40 min of aerobic exercise and 20 min of resistance exercise daily.

The program required six months of program adherence and daily weight recording, with weekly professional follow-up to ensure compliance. Following the completion of the intervention program, the specialist advised participants to maintain a healthy diet (following the Chinese Dietary Guidelines 2023) and check in with the online community every week. Six months of follow-up had elapsed when the article was finished, and the study had been running for nine months.

At baseline, at the end of the three-month intervention program, and after six months of follow-up, we measured participants’ height and weight. Body mass index (BMI) was calculated as BMI = weight (kg) / height (m)2. ΔBMI1 was the difference between BMI 1 (baseline) and BMI 2 (three months), while ΔBMI2 was the difference between BMI 1 and BMI 3 (six months). Weight loss was calculated as follow-up BMI (kg/m2)– baseline BMI (kg/m2). A questionnaire collected basic information such as name, gender, age, smoking, drinking, exercise, sleep habits, and medical histories. The questionnaire also asked all participants if they had ever tried to lose weight, how many times they had, and how much they had lost. Weight cycling was defined as having more than one weight loss episode in the previous two years and gaining more than 5% of the pre-weight loss baseline weight.

Scales

All participants independently completed self-report questionnaires under the supervision of investigators, including sociodemographic information (e.g., age, gender, BMI, ethnicity) and the following scales, which were returned to the researchers immediately:

Symptom-Checklist-K-9(SCL-K-9) [30]

The SCL-K-9 is a short, unidimensional version of the SCL-90-R Symptom Checklist 90 Revised (SCL-90-R) [31], a self-report questionnaire that assesses general psychiatric symptoms and psychological status and is recommended as a validated screening scale for psychological assessment by the American Society for Metabolic and Bariatric Surgery. It consists of nine SCL-90-R items (#24, #28, #31, #34, #43, #57, #58, #75, #77) and best represents (i.e., has the highest total item correlation) all of the original SCL-90-R subscales. The SCL-K-9 has been used among participants with overweight/obesity [32, 33]. Even in the absence of significant binge-eating symptoms, it has the same discriminatory validity as the SCL-90. Higher mean values indicate more severe symptoms. It suggests that the SCL-K-9 is appropriate for this study to investigate weight control correlates in participants with obesity and a history of weight cycling. SCL-K-9 was used in this study to assess psychological distress. The Cronbach’s α for this scale in the current study was 0.82 and 0.85 at baseline and three months.

Dutch Eating Behavior Questionnaire (DEBQ) [34]

DEBQ was assessed using the Chinese version, which has high reliability and validity [35,36,37]. It is a self-report questionnaire used to assess adults’ eating habits. It consists of 33 items divided into three subscales: restrained eating (ten items), emotional eating (13 items), and external eating (ten items). Restrained eating is a deliberate effort to limit your food intake to control your weight. Emotional eating is eating in response to a negative emotion like stress or loneliness. External eating is eating in response to a food attraction, such as sight or smell [38]. The subscales of emotional eating and extrinsic eating were scored on a 5-point Likert scale (from 1 “never” to 5 “often”), with higher scores indicating a higher degree of emotional eating and a greater tendency to extrinsic eating. The abstinence subscales (#1, #2, #7, #8, #13, #14, #21, #22, #28, #29) were scored inversely on the 5-point Likert scale, from 1 “always” to 5 “never.” In this study, we used “reverse scoring,” meaning a higher score indicates a greater desire for the tasty food. DEBQ achieved high reliability in data collected at various time points. Cronbach’s αvalues for data collected at baseline and three months were 0.90 and 0.91, respectively.

Power of Food Scales (PFS) [39]

PFS assesses and measures appetite and the psychological impact of living in a food-rich environment. It has three subscales representing different food proximity levels of: food available (thoughts about food generally; e.g., “I find myself thinking about food even when I’m not physically hungry”), food present (a draw toward food directly available; e.g., “If I see or smell a food I like, I get a powerful urge to eat them”), and food tasted (desire and liking of food at first taste; e.g., “When I eat delicious food I focus a lot on how good it tastes”). The scale has been used with good internal consistency among the Chinese, with a Cronbach’s α of 0.92 [33]. These are the advantages of this study.

We acknowledge that the current research has limitations. The data in this study come from a single-center cohort of participants with obesity and weight cycling. Because this study was conducted on young and middle-aged populations, the current study recognizes that the results may not apply to the general population. Furthermore, as with most large-scale studies, our findings depended on the reliability of each participant’s self-reported questionnaires, potentially subject to recall bias. Finally, latent variables were not introduced because the primary focus of our study was on relationships between observed variables. Although latent variables can provide a more in-depth understanding and interpretation of concepts not directly observed in the model, it was beyond the scope of our study. The lack of latent variables is acknowledged as a study limitation, and future research may consider a more thorough exploration of latent factors influencing the study model.

Weight loss was evident during the intervention (short-term), but maintaining weight is a multi-faceted physical-psychological-social effort. This study focuses on psychological distress during outpatient interventions, given that short-term psychological interventions are more achievable within treatment protocols and improving the psychological state during the follow-up period [52]. We should continue to follow up with participants in the future to obtain longer-term follow-up data, and more multi-center studies with large cohorts and longer follow-ups could be conducted.

Conclusion

This study discovered that psychological distress was related to weight control levels in weight-cycling populations. More emphasis should be placed on psychological distress in this population. It also discovered that food behavior moderates the relationship between psychological health and weight control. Therefore, in the program of weight loss intervention, measures should be considered to reduce the psychological distress of people with a history of weight cycling to prevent the occurrence of poor weight control.