中国成人肥胖及糖前期/2型糖尿病患者基于食物的极低热量饮食计划体验研究
Experiences of a food-based very low-calorie diet programme among Chinese adults with obesity and prediabetes or type 2 diabetes: a COM-B-guided qualitative study
中国成人糖尿病患者人数持续增加,2021年达1.41亿,预计2045年增至1.74亿。极低热量饮食(VLCD)可改善2型糖尿病患者血糖控制,甚至实现糖尿病缓解。然而,饮食依从性差和长期效果不佳。本研究采用COM-B模型探讨中国成人肥胖及糖前期/2型糖尿病患者对基于食物的VLCD计划的体验,发现环境因素如家庭支持不足和文化差异影响饮食调整行为。
Introduction
The International Diabetes Federation (IDF) reported that over 537 million adults worldwide are currently living with diabetes, and this figure continues to rise. In China, one of the most severely affected countries, there were 141 million adults with diabetes in 2021, and this number is projected to reach 174 million by 20451. Additionally, the prevalence of prediabetes in China has reached 35.2%, indicating that a large potential population is at risk of developing diabetes in the future2. Type 2 diabetes mellitus, the dominant form of diabetes, accounts for more than 90% of all cases and is closely linked to unhealthy lifestyle choices, genetic predispositions, and increasing overweight/obesity rates3,4. T2DM can lead to a variety of acute and chronic complications, as well as psychological issues such as anxiety and depression, which significantly reduce patients’ quality of life. Furthermore, prediabetes is not only a predictor of future diabetes but also strongly associated with increased risks of cardiovascular disease, microvascular complications, cancer, dementia, and depression5.
Lifestyle intervention remains a cornerstone of T2DM management and plays a crucial role in preventing or delaying the progression from prediabetes to T2DM. The Da Qing Diabetes Prevention Outcome Study revealed that a 6-year lifestyle intervention programme could reduce the cumulative risk of diabetes progression by 39% in individuals with impaired glucose tolerance over the following 30 years6. Recent studies have increasingly highlighted the importance and benefits of enhanced lifestyle interventions in managing T2DM. VLCDs can achieve weight loss and improve biochemical marker levels in both patients with and without T2DM7,8. Furthermore, the VLCD has enabled some patients with obesity and T2DM to achieve normal blood glucose control and even achieve diabetes remission9,10,11.
Despite these benefits, adherence to intensive lifestyle interventions is challenging and long-term outcomes remain poor. Valabhji et al.12,13 investigated the effects of a total dietary replacement programme on T2DM remission and reported that participant engagement progressively declined, with only 36% of participants completing the programme. Similarly, Karmacharya et al.14 conducted an intervention using a low-calorie diet plan and reported that 35.7% of the participants were lost to follow-up within 12 months. Rai et al.15 reviewed the barriers and facilitators faced by South Asian populations with T2DM/prediabetes in Western countries during dietary adjustments and reported that a deep understanding of food can effectively promote dietary adjustment behaviours among populations with T2DM/prediabetes, whereas environmental differences can lead to social and cultural pressures, thereby challenging their adherence to dietary recommendations. For example, during weddings and festivals, hosts prepare many sweets, and if South Asian individuals with T2DM/prediabetes do not eat them, they may upset their hosts. Abel et al.16 conducted interviews with individuals with prediabetes with a BMI ≥ 25 kg/m² to explore motivations, facilitating factors, and challenges related to dietary changes under nurse-led guidance. The results indicated that uncontrollable environmental factors, such as widespread food advertising, a lack of family support, economic issues, the impact of other illnesses, and the ability to cope with challenges hinder the formation of healthy dietary habits. Whitfield et al.17 investigated the feasibility of VLCD for patients with prediabetic and early-stage T2DM in New Zealand. In addition to social pressures, food palatability, dietary monotony, and physical discomfort were identified as additional barriers to VLCD acceptance.
Moreover, the existing evidence largely originates from Western countries, where the VLCD often relies on dietary replacement products, which may not align well with Chinese dietary culture and habits. In contrast to Western culinary traditions that emphasize the natural flavors of ingredients, Chinese cuisine integrates diverse cooking techniques, skillfully utilizes seasonings, and prioritizes the harmonious integration of visual appeal, aroma, and taste, with communal dining being predominant. Influenced by climatic variations and religious-cultural factors, distinct dietary customs exist across regions. Concurrently, socioeconomic development has intensified work-life pressures, leading to increasingly prevalent unhealthy eating behaviours among Chinese residents, including irregular meal patterns, frequent takeaway/dining-out consumption, imbalanced meat-to-vegetable ratios, and binge eating, which collectively challenge the acceptability and feasibility of VLCD interventions domestically. Evidence regarding how individuals in China perceive and experience VLCD interventions, particularly food-based VLCD programmes, remains limited.
To inform the optimisation of intensive lifestyle management in this context, a theoretically grounded understanding of behavioural determinants is needed. The COM-B model, developed by Michie et al.18 in 2011, posits that behaviour is shaped by the dynamic interaction of capability, opportunity, and motivation. Within this framework, capability encompasses patients’ disease-management skills and psychological readiness; opportunity denotes external factors that facilitate health-promoting behaviours; and motivation comprises the affective responses and volitional processes underlying patients’ engagement in disease management. The COM-B model has been extensively applied to investigate determinants of individual health behaviours and serves as a theoretical foundation for designing and refining behavioural interventions19,20. We selected the COM-B model over alternative behaviour-change theories for four reasons. First, unlike unidimensional models, COM-B simultaneously captures capability, opportunity, and motivational determinants, which is essential for understanding complex dietary behaviours that require knowledge, environmental restructuring, and sustained motivation. Second, COM-B is the central component of the Behaviour Change Wheel (BCW), enabling direct translation of findings into intervention functions. Third, COM-B has been successfully applied to weight-management and diabetes self-management studies in diverse populations. Fourth, its explicit inclusion of ‘social opportunity’ is particularly suited to the Chinese context, where family dynamics and collective eating norms profoundly influence individual dietary choices.
Therefore, this study used a descriptive qualitative design guided by the COM-B model to explore the perceptions and experiences of Chinese adults with obesity and prediabetes or T2DM undergoing a food-based VLCD programme, with the aim of identifying key barriers to and facilitators of engagement in enhanced lifestyle management during and following the intervention.
Methods
Design
This study used a descriptive qualitative research approach to explore the experiences and perceptions of Chinese adults with obesity and prediabetes or T2DM who underwent a food-based VLCD programme. The study was informed by the COM-B model to identify barriers and facilitators to adherence to the VLCD and engagement in ongoing lifestyle management and to map the findings to COM-B domains. The study followed the Consolidated Criteria for Reporting Qualitative Research (COREQ)21.
Settings and participants
The study was conducted at Jiangsu Province Integrated Traditional Chinese and Western Medicine Hospital and Xiamen Traditional Chinese Medicine Hospital. These two hospitals were among the first institutions in China to implement inpatient VLCD therapy and had established standardized VLCD clinical pathways. All participants provided signed informed consent before the start of the study. The study was approved by the Ethics Committee of Jiangsu Province Integrated Traditional Chinese and Western Medicine Hospital (No. 2023-LWKY-051) and the Medical Ethics Committee of Xiamen Hospital of Traditional Chinese Medicine (No. 2024-K019-01).
Potential participants were purposively sampled from inpatients who had completed the VLCD programme between July and August 2024. Maximum variation purposive sampling was employed to ensure diversity across key demographic and clinical characteristics. The sampling matrix comprised sex (male/female), age (< 45 vs. ≥45 years), diagnosis (prediabetes/type 2 diabetes), and disease duration (< 6, 6–24, and > 24 months). Following each set of three interviews, the research team reviewed the matrix to identify underrepresented categories and purposively recruited additional participants to address these gaps.
Inclusion criteria and exclusion criteria
The inclusion criteria were as follows: aged 18–65 years, diagnosed with T2DM or prediabetes per the World Health Organisation (WHO) standards22; body mass index (BMI) between 24 kg/m² and 40.0 kg/m²; and the ability to comprehend the study and provide voluntary informed consent.
Participants were required to have a BMI between 24 and 40 kg/m². The lower threshold of 24 kg/m² aligns with the Chinese criteria for overweight (≥ 24 kg/m²) and obesity (≥ 28 kg/m²) defined by the Working Group on Obesity in China23, ensuring that participants exhibited clinically significant excess weight warranting intensive dietary intervention. The upper threshold of 40 kg/m² was set to exclude individuals with BMI > 40 kg/m² (or > 35 kg/m² with severe comorbidities), who are typically candidates for bariatric surgery rather than dietary intervention alone, thereby maintaining a homogeneous clinical indication for the VLCD.
The exclusion criteria included weight loss exceeding 6 kg in the past six months; current use of weight-loss medications or malnutrition; pregnancy; attempts to conceive, or childbearing potential without contraception; acute or severe chronic diabetes complications; history of gastrointestinal surgery; cardiovascular events within the last three months or hospitalization for unstable angina; coexisting malignant tumors; severe chronic infections; or significant respiratory diseases; and cognitive or communication impairments.
VLCD programme
All participants had completed a structured 9-day inpatient VLCD programme. The protocol included 1–2 days of a normal diet prior to hospitalisation, followed by baseline assessments and formal initiation of the intervention. The regimen comprised a 2-day buffer period at 600 kcal/day, a 5-day VLCD phase at 300 kcal/day, and a 2-day recovery period at 600 kcal/day. The macronutrient distribution was 55% carbohydrate, 25% fat, and 20% protein. Daily menus were designed by nutritionists and prepared by the hospital cafeteria according to a standardised set menu, delivered to patients at mealtimes. Individual modifications were permitted, however, for food allergies, religious restrictions, and strong aversions; for instance, participants with seafood allergies were offered poultry or legume alternatives of equivalent nutritional value. A registered dietitian reviewed each participant’s dietary restrictions within 24 h of enrollment and adjusted the menu accordingly.
The restrictive diet adhered to hypocaloric, oil-free, low-salt (< 5 g/day), and sugar-free principles, employing high dietary fibre, moderate protein, and low–glycemic-index staples for metabolic regulation. Breakfast comprised high-quality protein (milk, protein powder, or egg white) paired with low–glycemic-index whole grains. Lunch centred on abundant cold-dressed vegetables, supplemented with protein and minimal whole grains or low-sugar fruit. Dinner consisted of light vegetable soup or cold-dressed leafy greens with scant protein and staple food for strict caloric control. Cooking oil, sucrose, and sugar-containing condiments were strictly prohibited; monosodium glutamate, vinegar, ginger, scallion, and garlic were permitted only in moderation. To enhance cultural acceptability, the VLCD incorporated regionally familiar foods. At the northern site, staple foods included millet porridge and wheat-based noodles; at the southern site, rice congee and steamed fish were featured. Common Chinese cooking techniques (e.g., steaming, boiling, and cold-dressing) were used in place of Western-style methods (e.g., frying and roasting).
Following completion of the VLCD, patients received dietitian-led guidance on enhanced post-discharge dietary management and precautions. All participants were provided with: (i) a personalised transition meal plan (1,200 kcal/day, gradually increasing to 1,500 kcal/day over 4 weeks); (ii) a printed Chinese-language food-exchange handbook; (iii) a 30-minute one-to-one counselling session with a dietitian; and (iv) a follow-up appointment scheduled for 2 weeks post-discharge.
To ensure consistency, dietitians at both hospitals attended a 2-day training workshop prior to study commencement, and 10% of counselling sessions were audio-recorded and audited for protocol fidelity. VLCD programme and discharge education were standardised across both sites using a unified protocol manual developed by the research team.
Data collection
Data were collected through face-to-face, semi-structured, in-depth interviews conducted within 1–2 days after the participants completed the VLCD programme. The interviews were conducted by three postgraduate (Wang YL, Guo RJ, and Ge JM) nursing researchers who had received formal training in qualitative research methods.
Researchers proactively communicated with patients prior to the interviews, explained the purpose and significance of the interviews, obtained patient consent and cooperation, and recorded the entire process. The interviews were conducted in patient rooms or conference rooms and lasted between 20 and 40 min. The participants had the right to terminate the interview at any time, and the researchers were committed to the strictly confidential treatment of the interview content and the participants’ personal information. During the interviews, no participants refused to answer questions or exhibited emotional distress.
The interview guide was developed on the basis of the study objectives and the COM-B theoretical framework and was informed by a review of the relevant literature. The research team then drafted an initial interview outline, which was piloted with one eligible patient to identify areas for refinement. Data from this pilot interview were excluded from the formal analysis. After team discussions, the interview outline was finalized (Table 1). The interview questions focused on patients’ experiences during the VLCD process, their previous experiences with conventional lifestyle management, and their perspectives and intentions regarding continued lifestyle management following VLCD treatment.
Sampling and data collection continued until category and meaning saturation was achieved, defined as the point at which no new codes or categories emerged within the COM-B domains and additional interviews only repeated previously identified meaning units. Data saturation was reached incrementally. Following the 10th interview, no new COM-B categories emerged; following the 12th interview, no additional subthemes or meaningful variations were identified in the remaining two interviews. The remaining two interviews confirmed saturation and were included to ensure comprehensiveness. This process was monitored through weekly research team meetings, during which emerging themes were systematically compared with previously collected data. A saturation log was maintained, recording the date, participant characteristics, and the presence or absence of new themes for each interview.
Data analysis
All interviews were audio-recorded and transcribed verbatim within 24 h of completion by two researchers (Wang YL and Guo RJ). Personal identifiers were removed and replaced with numerical codes to ensure confidentiality.
Two researchers (Wang YL and Guo RJ) independently cross-checked all transcripts against audio recordings for accuracy, paying particular attention to medical terminology, paralinguistic markers (e.g., pauses, laughter), and non-verbal cues documented by the interviewers. Any discrepancies were resolved through repeated listening and discussion to reach consensus.
Within four weeks following each interview, all participants were provided with the interview transcript via email or WeChat for member checking. They were invited to review the transcript, confirm that it accurately reflected their experiences, and suggest any corrections. Of the 14 participants, 11 (79%) responded; no substantive corrections were requested, and three minor wording adjustments were incorporated into the final analysis.
Data were analysed using a COM-B–guided framework analysis. The analysis proceeded through five stages: (i) Familiarisation: Two researchers (Wang YL and Guo RJ) independently read all transcripts and listened to audio recordings to immerse themselves in the data. (ii) Identifying a thematic framework: A preliminary coding framework was developed based on the COM-B model (Capability, Opportunity, Motivation) and refined through inductive coding of three initial transcripts. (iii) Indexing: All transcripts were systematically coded against the framework using NVivo 12 software. Data that did not fit the initial framework were coded inductively and recorded in a separate ‘emerging themes’ memo. (iv) Charting: Coded data were extracted and arranged into thematic matrices, with rows representing participants and columns representing subthemes, allowing for cross-case comparison. (v) Mapping and interpretation: The research team reviewed the matrices to identify patterns, relationships between COM-B domains, and deviant cases. Representative quotations were selected to illustrate each category and to support interpretation.
The COM-B model encompasses six theoretical constructs: physical and psychological capability, physical and social opportunity, and reflective and automatic motivation. In this analysis, findings were structured under the three overarching domains—Capability, Opportunity, and Motivation—because the interview data naturally clustered at this level, and further subdivision would have produced themes with insufficient data support.
Quality control
Prior to data collection, three interviewers (Wang YL, Guo RJ, and Ge JM) completed a structured training programme comprising: (i) a 4-hour workshop on semi-structured interviewing techniques, facilitated by an experienced qualitative researcher (Li DD, with 10 years of qualitative research experience); (ii) a review of the interview guide and probing strategies; (iii) two pilot interviews with non-participant volunteers, followed by peer debriefing and feedback; and (iv) a discussion of reflexive journaling practices. To ensure consistency during data collection, interviewers convened weekly to discuss emerging themes, share field notes, and calibrate probing depth. All interviews were audio-recorded. Li DD randomly selected and independently reviewed 20% (n = 3) for quality assurance, providing feedback to interviewers to minimise drift.
To ensure coding consistency, two researchers (Wang YL and Guo RJ) independently coded the first three transcripts. Inter-coder agreement was assessed using percentage agreement and Cohen’s Kappa coefficient, yielding an initial agreement of 78% (Kappa = 0.71, indicating substantial agreement). Discrepancies were resolved through a 2-hour consensus meeting, during which the coding framework was refined to clarify ambiguous categories. The remaining 11 transcripts were then divided between the two coders, with each reviewing 20% of the other’s coding to detect drift. Any new coding discrepancies (n = 12 instances across all transcripts) were adjudicated by a third researcher (Li DD), and final consensus was achieved for all codes.
Reflexivity statement
The research team comprised three postgraduate nursing students (Wang YL, Guo RJ, and Ge JM) and two senior qualitative researchers (Li DD and Zhou Y). Wang YL and Guo RJ had 2–3 years of clinical nursing experience in endocrinology wards, which facilitated rapport-building with participants but also risked introducing therapeutic bias—that is, participants might provide socially desirable responses perceived as ‘good patient’ behaviour. To mitigate this, interviewers explicitly stated at the outset of each interview that they were present as researchers rather than clinicians, and that there were no ‘correct’ answers. Li DD, who had no clinical relationship with the participants, independently reviewed 20% of transcripts to identify potential interviewer-induced bias. No systematic bias was detected, although minor instances of therapeutic language (e.g., ‘that’s good’ in response to weight-loss reports) were noted and addressed in subsequent interviews through interviewer coaching. All researchers maintained reflexive journals throughout the study, documenting assumptions, emotional reactions, and analytical decisions. Key reflections were discussed in monthly team meetings and informed the iterative refinement of the coding framework.
Trustworthiness
Several strategies were used to ensure rigour and trustworthiness. Purposive sampling across two hospitals with established VLCD programmes enhanced the relevance and diversity of experiences. The interview guide was developed through a literature review, explicit mapping to the COM-B model, pilot testing, and team discussion. Credibility and dependability were supported by independent initial coding by two researchers, followed by consensus discussion and, when necessary, adjudication by a third researcher. A stable coding framework was applied consistently, with iterative data collection and analysis enabling ongoing assessment of category and meaning saturation within the COM-B domains. Confirmability was strengthened through reflexive consideration of researchers’ backgrounds and roles, and transferability was facilitated by detailed descriptions of the study context, the VLCD programme, and participant characteristics.
Results
Figure 1 presents the participant flow diagram. Between July and August 2024, 52 patients were screened for eligibility across both hospitals. Of these, 27 were excluded: 8 failed to meet the BMI criteria, 15 declined to participate (time constraints, n = 8; lack of interest, n = 5; other personal reasons, n = 2), and 4 were excluded due to comorbidities or cognitive impairment. Of the remaining 25 eligible patients approached, 21 consented to participate. Fourteen participants were interviewed, at which point data saturation was achieved.
Participant recruitment.
A total of 14 participants were interviewed, comprising 3 males and 11 females, with a mean (SD) age of 40.93 (9.36) years. Mean (SD) BMI was 29.52 (3.43) kg/m2 at admission and 28.14 (3.46) kg/m² at discharge. Most participants reported attempting lifestyle management and adjustments on the basis of healthcare providers’ advice before initiating VLCD treatment. The basic demographic and health information of the interview subjects is presented in Table 2.
Themes
A total of 8 subthemes were derived from the framework analysis, guided by the COM-B model, which reflects participants’ perceptions and experiences regarding the food-based VLCD programme in terms of ability, opportunity, and motivation. For further details, please refer to Table 3.
Capability
This theme encompassed factors related to participants’ physical and psychological capacity to engage in VLCD management, including their self-control and physical abilities.
Self-control
In this study, self-control refers to the participants’ ability to endure hunger, resist tempting foods during the programme period, and maintain long-term dietary management after the programme concludes.
Many patients reported experiencing hunger during the implementation of the food-based VLCD programme, but this hunger is generally manageable through coping strategies such as drinking water and diverting attention.
“Of course, I feel hungry; about two hours after eating, I start to feel hungry. I can feel the hunger, but I can only drink water.” (P8).
“I just zone out and forget about it.” (P12).
In addition to managing hunger, patients also needed to resist the allure of appetizing foods during the programme.
“I usually eat whatever is cooked, but now I have to choose what I can eat; I especially crave the things I can’t eat.” (P13).
The study also explored patients’ willingness to maintain dietary control after completing the programme. Some expressed concern over losing the health benefits they had gained, which motivated them to continue controlling their diet.
“Since I’ve managed to stick with it for so many days here, if I don’t continue, it might all be in vain.” (P6).
Conversely, others admitted that it was challenging to sustain dietary control in the long term.
“I can stick with it for a while but I don’t know how long it will last. I might feel a bit better and more comfortable and then lose that awareness of needing to continue.” (P5).
Limited physical ability
Some patients reported that they had previously managed their weight and health through exercise. However, physical limitations such as injuries or chronic conditions have restricted their ability to engage in high-intensity exercise, leading them to explore alternative strategies for weight and health management, including the adoption of VLCD.
“Why am I here? I could have lost weight through exercise; I wasn’t very overweight before, but my knees and back got injured, limiting my ability to exercise… If I were to exercise, it would definitely lead to pain every few days, which makes me irritable. So I thought maybe you have some other ways.” (P7).
Opportunity
This theme encompassed external factors that either facilitated or hindered patients’ adherence to VLCD and dietary management, including supportive physical environments, competing life priorities, and social support.
Supportive physical environment
A supportive physical environment can stimulate patients’ motivation by offering essential resources and encouragement, promoting better adherence to management plans. Many participants described that receiving dietary management in a hospital setting was more convenient and enabled higher adherence than managing it at home.
“I understand this; if I weren’t hospitalized, I would start eating randomly when I go back.” (P12).
“There are no difficulties; whatever the doctor arranges for me is what I do.” (P5).
In contrast, managing one’s diet at home or work often poses challenges. At home, patients have to consider family members’ preferences, and at work, they are limited to cafeteria meals.
“It’s definitely more challenging; when I’m not working, it’s manageable, but once I’m at work eating cafeteria food, it becomes impossible. I can only try to control myself and eat less because everyone eats the same there; it’s not possible for them to prepare a special meal just for me.” (P8).
Competing priorities
Another opportunity-related factor identified was competing life priorities, which often impedes patients’ ability to adhere to dietary management plans. Some participants stated that owing to the need to take care of their families and children or because of their busy work, they often find it difficult to maintain regular eating habits, which affects the implementation effect of the management programme.
“At that time, he suggested that I be hospitalized, but my child was studying then, and I didn’t have time. Therefore, during the summer vacation when I had time, I came over.” (P5).
“Actually, my biggest problem is work; my eating schedule is irregular. Basically due to irregular meals, I can’t guarantee it.” (P9).
Social support
Encouragement and support from family, friends, and healthcare providers to strengthen patients’ motivation to manage their diet are key facilitators. Some patients reported improved adherence to lifestyle changes after receiving emotional, practical, and financial support from family members.
“Absolutely! My partner bet with me that if I succeeded, he would give me 100,000 yuan. I want to lose weight, and he’s a bit overweight too. We both are a little overweight. When I lose weight, I want to tell him.” (P7).
Long-term supervision and guidance from healthcare providers also play crucial roles in helping patients maintain lifestyle changes.
“Here, we have nurses and doctors; if there are any issues, we will seek them out.” (P2).
“If it’s just one-sided from us, we won’t be able to stick with it; we need someone who can tell us how we can adjust this or that, or calculate how things should go. There needs to be a guiding form of interaction. In addition, then, at the end, give an evaluation or further suggestions, it should be a two-way process.” (P9).
Motivation
This theme captured the internal drivers that influence patients’ adherence to VLCD and lifestyle management, including their perceived threat of disease, feedback on treatment effectiveness, and adverse experiences during dietary restrictions.
Perceived threat of disease
Most patients perceived the health threats of diabetes and its complications, such as hypertension and hyperuricemia. They expressed concern about their overall health and were eager to improve their physical condition through dietary control, with some even hoping to lessen or eliminate the burden of the disease.
“I feel that I’m still young and hope to reverse it and reduce my medication.” (P1).
“I want to treat my diabetes; here, the diet is controlled, so I can’t eat whatever I want. I took this opportunity to lose some weight. It’s beneficial for my hypertension, for lowering blood sugar, and helps with my high uric acid levels. After discharge, I don’t want to take blood pressure or blood sugar medications anymore; I want to reduce my medication intake.” (P4).
Feedback on treatment effectiveness
Positive health improvements provided encouragement and a sense of achievement for patients, thereby motivating them to maintain regular eating patterns to sustain a healthy lifestyle.
“My blood sugar has improved significantly, and I’ve lost weight.” (P1).
“I feel much lighter; I’ve lost over 7 pounds, maybe 8.” (P11).
However, some patients expressed frustration over the reduced pace of weight loss in later stages.
“I lost weight quickly in the first three days, 5 pounds in three days, and I told all my friends… but then it stopped… The difficulty is that I’m not seeing the scale drop now.” (P12).
These findings suggest that while initial improvements can motivate patients, plateaus in progress may undermine their confidence and adherence.
Adverse experiences with dietary restrictions
Before initiating the VLCD, hunger was anticipated to be the primary challenge. However, during implementation, most patients reported that while they felt hungry, it was manageable. Rather than hunger, issues with food taste and quality posed more discomfort.
“This is my first time trying this dietary pattern, no oil, no salt. It feels hard to swallow sometimes. Sometimes I even feel like throwing up.” (P2).
In this study, patients’ daily meals were uniformly designed, prepared, and distributed, with adjustments made according to individual dietary restrictions. However, this approach still elicited feedback indicating a lack of necessary personalization. Many patients felt that their diet did not adequately cater to their individual preferences or needs, with some indicating that meals occasionally included foods they were unpalatable.
“The difficulty is that there are some dishes I don’t like, like loofah; I truly don’t like it, but it’s often included, haha. I don’t feel full; I think there should be more variety in the meals because some people truly don’t eat certain things.” (P11).
Inductive finding: a theme beyond the COM-B framework
Although the COM-B framework captured the majority of the data, an inductive theme—“temporal perception and rhythm adaptation”—was identified, reflecting participants’ subjective experiences of the VLCD timeframe (9-day inpatient programme followed by post-discharge maintenance). Some participants perceived the hospitalisation period as a “special interval suspended from normal life.” This perception of “temporal isolation” facilitated short-term adherence yet may undermine the sustainability of behavioural changes after discharge, as participants framed the VLCD as a “temporary project” rather than a “lifestyle transformation.” This theme did not map neatly onto any COM-B construct, suggesting that the temporal dimension should be incorporated into future research as an independent determinant of behavioural maintenance.
During data analysis, we employed a constant comparison approach to examine whether experiences varied across sites. Although the two study sites represented distinct regional dietary cultures, the consistency of themes across locations suggests that the identified barriers and facilitators may be generalizable to Chinese adults with obesity in diverse geographic settings. However, future studies should explicitly examine whether regional dietary preferences influence long-term adherence to food-based VLCD programmes.
Discussion
This study, grounded in the COM-B theoretical framework, investigated factors affecting enhanced lifestyle management among individuals with obesity and prediabetes or T2DM in China. Our findings identified a total of eight subthemes across the domains of capability, opportunity, and motivation, providing valuable evidence for designing targeted interventions and effective behaviour change strategies.
In the capability domain, two primary factors emerged as influential: physical limitations and self-control. However, these two factors do not act on behaviour in isolation; their influence is substantially moderated by opportunity- and motivation-related factors. Patients’ overall health status and the lasting impact of complications present significant barriers to maintaining a healthy lifestyle24,25. Lee et al.26 also pointed out that pre-existing health problems can restrict patients’ physical ability and motivation to exercise. Yet our study found that when physical limitations emerge, supportive options within the opportunity dimension—such as VLCD management—play a critical compensatory role. Some patients with limited exercise capacity found alternative behavioural pathways precisely by switching to VLCD. This practice suggests that lifestyle programmes best suited to individual circumstances (opportunity), combined with the willpower to sustain change (capability), enable patients to maintain long-term adherence.
Self-control is crucial for patients’ sustained adherence to management plans26. However, our study revealed that the depletion of self-control is strongly context-dependent. Patients’ difficulty in consistently resisting food cravings, or their tendency to abandon dietary plans when fatigued or inconvenienced, not only reflects the limits of individual self-regulation but also reveals that a lack of environmental opportunities—such as work pressure and social situations—directly consumes limited self-control resources. Therefore, in the process of intensive dietary interventions, simply emphasizing “firm beliefs” is insufficient. Intervention strategies should integrate the health belief model27 to foster positive beliefs, while simultaneously employing cognitive and behavioural approaches28—such as goal setting29 and self-monitoring30—to enhance capability. More importantly, it is essential to compensate for insufficient self-control by avoiding triggering environments and strengthening external supervision31.
Supportive relationships and environments play a significant role in initiating and sustaining lifestyle changes32, and their value extends beyond providing practical assistance to enabling capability and motivation33. Support from healthcare professionals, family, and friends constitutes a direct source of opportunity. Regular behavioural support from healthcare professionals assists patients in adhering to low-energy dietary interventions and maintaining weight loss34,35. In our study, many patients participated in the VLCD programme at the recommendation of family members or friends, with some families even establishing reward mechanisms—essentially using social support (opportunity) to compensate for deficiencies in individual self-control (capability). More notably, the expression by some patients of “not wanting to become a burden on their families” clearly illustrates how external support can be internalized into intrinsic motivation: family responsibility is transformed into a personal drive for change, thereby enhancing long-term adherence.
A supportive environment includes external supervision, structured support from lifestyle intervention programmes, a conducive setting for lifestyle change, and access to health-promoting options for lifestyle modifications32, also serve as critical catalysts for behavioural change. Our findings indicate that adherence is higher when management is delivered in hospital settings, highlighting the reinforcing role of structured environments (opportunity) in facilitating capability execution and sustaining motivation. However, not all patients have the conditions to receive management in hospital, and prolonged work and academic pressures can also disrupt dietary behaviours26. This suggests that the absence of opportunity—such as work-related stress—can impede behavioural change through dual pathways: inducing fatigue (which depletes capability) and generating anxiety (which undermines motivation). Therefore, future strategies must address the three dimensions of patients, healthcare professionals, and the healthcare system, with particular emphasis on reviewing work environments and policies26, so as to create a broader enabling platform for capability translation and motivation maintenance.
Motivation lies at the heart of the COM-B model18 and plays a central role in patients’ engagement with enhanced lifestyle management programmes36. In this study, weight loss and the hope of reversing diabetes were the primary motivations, and the sustainability of this motivation is closely intertwined with capability and opportunity. Research has confirmed that motivation is associated with health outcomes37,38: positive results enhance adherence, while poor outcomes undermine it. This reveals a critical positive feedback loop—early weight loss and improvements in blood glucose (as motivational outcomes) not only strengthen patients’ belief in the programme’s effectiveness, but also enhance their self‑efficacy (capability) and prompt them to make more active use of medical and family support (opportunity).
Furthermore, meal taste and personalization are important factors influencing adherence. Our study found that VLCD was easier to follow than expected and that hunger was generally acceptable39; however, poor palatability of the meal provisions emerged as a major barrier38. When dietary plans lack variety (insufficient opportunity), patients need to expend more self-control (capability) to sustain adherence, which leads to a decline in motivation. In contrast, incorporating individual preferences and cultural backgrounds (creating opportunity) can reduce the psychological resistance to behavioural change and lessen the over-reliance on self-control. Therefore, motivation is not merely the starting point of behaviour but also a product of the interplay between capability and opportunity. Enhancing the acceptability and sustainability of dietary plans requires not only reinforcing core motivation but also lowering the behavioural threshold through personalized meal plans, thereby achieving long‑term maintenance of motivation.
Although this study primarily sought to identify shared themes across diagnostic groups, exploratory subgroup comparisons revealed meaningful differences between the two cohorts. Patients with type 2 diabetes (n = 8) more frequently cited “perceived disease threat” and “treatment feedback,” consistently framing the VLCD as an essential tool for disease management and complication prevention. Their motivation was driven primarily by fear of prior hyperglycemic events or complications (e.g., retinopathy, neuropathy). These differences carry important implications for clinical practice: patients with type 2 diabetes may require more intensive motivational enhancement centred on complication risk education, whereas individuals with prediabetes may benefit more from interventions emphasising lifestyle optimisation and self-efficacy enhancement. However, given the small subgroup sample sizes, these findings should be regarded as exploratory and require validation in future studies with larger samples.
Strengths and limitations
Existing qualitative research on VLCD experiences has predominantly focused on Western populations using formula-based products26,28,31. The present study makes three key contributions to this literature. First, it evaluates a food-based VLCD programme that accommodates cultural eating practices, which may yield adherence patterns distinct from those associated with formula-based approaches. Second, it centres on Chinese adults—a population marked by distinctive dietary cultures, family-centred eating norms, and high diabetes prevalence—thereby bridging a critical gap in culturally specific evidence. Third, by operationalizing the COM-B model in this underexplored context, it uncovers culture-specific determinants of behaviour change, including family social support and traditional food preferences, thus broadening the framework’s relevance beyond Western settings.
However, several limitations should be noted. First, all study participants underwent VLCD therapy in a hospital setting. Given that VLCD treatment requires strict medical supervision, its implementation environment is inherently restrictive. This may limit the generalizability and transferability of the study findings to broader patient populations; in contrast, a less restrictive low-calorie diet can be administered in non-hospital settings. Second, although the study recorded changes in BMI, it did not incorporate long-term follow-up BMI data or biochemical markers, thereby limiting a comprehensive assessment of the sustained weight-loss effects of the intervention. Additionally, the study sample exhibited a sex imbalance, which may have impacted the representativeness of the results. Finally, although we inquired about patients’ willingness to continue implementing lifestyle management after the dietary restriction ended, unfortunately, no sustained follow-up investigations were conducted. This lack of follow-up prevents us from accurately understanding patients’ long-term adherence and behavioural changes.
Study implications
This study highlights multiple factors that influence lifestyle management among individuals with obesity and prediabetes or T2DM in China. The interview findings revealed that all three components of the COM-B model—namely, capability, opportunity, and motivation—play significant roles in patients’ lifestyle adjustments. Within capability, self-control and physical limitations were key influences; in opportunity, supportive physical environments, competing life priorities, and social support shaped adherence; and in motivation, perceived disease threat, positive treatment feedback, and adverse dietary experiences affected behavioural maintenance. Psychological capability, opportunity, and motivation appeared more prominently in participants’ narratives than did physical limitations.
These findings suggest that future interventions should strategically target these priority areas, facilitating the development of patient-aligned lifestyle management plans that enhance engagement. Drawing on the Behaviour Change Wheel (BCW) framework, future interventions may consider the following strategies to address the key barriers identified. To address insufficient self-control (psychological capability), training and enablement strategies combine inpatient cognitive-behavioural skills training with post-discharge mobile self-monitoring tools. For temptation from family and shared meals (social opportunity), environmental restructuring and enablement engage family members through concurrent nutrition education and the redesign of family-friendly, low-calorie menus. Negative eating experiences (automatic motivation) are mitigated through incentivization and restriction, increasing menu variety and introducing a “sensory satisfaction” metric. Finally, insufficient perceived disease threat (reflective motivation) is tackled through education and persuasion via personalised risk-communication tools for individuals with prediabetes, such as risk visualisation based on genetic and metabolic indicators.
At the hospital level, three measures are recommended. First, a dedicated VLCD kitchen should be established, staffed by chefs familiar with regional cuisines, to ensure cultural appropriateness. Second, a patient taste feedback mechanism should be introduced on days 3–5 of hospitalisation to enable timely menu adjustments. Third, a transitional recipe package containing locally available ingredients and simple recipes should be provided at discharge. At the policy level, food-based VLCD programmes should be incorporated into diabetes management clinical pathways via a China-localised implementation guideline, while their inclusion in the reimbursement scope of outpatient chronic disease management under medical insurance should be explored to improve accessibility.
Conclusions
This study examines the perceptions and experiences of individuals with prediabetes and T2DM during a VLCD intervention, analyzing factors influencing adherence. These findings suggest that future research should focus on developing targeted and sustainable lifestyle management programmes to address these identified factors, thereby enhancing dietary compliance and diabetes management outcomes.
来源:Nature 健康护理主题 · nature.com