Maintaining weight loss and associated risk reduction following lifestyle interventions remains a challenge, often due to a return to old habits. Understanding factors influencing the maintenance of behavioral changes is crucial for developing more effective treatments. Eating behavior is a key target in weight loss interventions. The Three Factor Eating Questionnaire (TFEQ) assesses cognitive restraint, disinhibition, and hunger. In shorter interventions, increased cognitive restraint alongside decreased disinhibition and hunger correlate with greater weight loss. However, less is known about factors predicting long-term maintenance of eating behavior changes. Stress, increasingly prevalent, is a potential factor. Stress activates the hypothalamic-pituitary-adrenal axis, leading to elevated cortisol and increased appetite, potentially hindering maintenance of low disinhibition and hunger. Maintaining behavioral change requires considerable cognitive resources and self-regulation, which may be diminished under stress. Perceived stress, a subjective appraisal of stressful situations, has been linked in cross-sectional studies to less cognitive restraint, but higher rigid restraint. High perceived stress has also been associated with overeating tendencies (disinhibition and hunger). This study uses data from the 3-year PREVIEW intervention to investigate whether perceived stress, measured after an active behavior change stage, predicts the maintenance of changes in eating behavior during a subsequent behavior maintenance stage, and whether frequently experienced stress during the maintenance stage is associated with eating behavior changes.

Numerous studies have demonstrated the efficacy of lifestyle interventions in treating obesity and reducing obesity-related morbidities. However, long-term weight loss maintenance remains a significant challenge, often attributed to a gradual relapse into previous unhealthy lifestyle habits. Eating behavior is a critical component of successful weight management, and several studies have shown associations between specific eating patterns and weight loss outcomes in short-term interventions (1 year or less). These studies consistently find that increased cognitive restraint, coupled with decreases in disinhibition and hunger, is associated with greater weight loss. One study even demonstrated a link between sustained decreases in uncontrolled eating and better long-term weight maintenance. However, knowledge regarding factors that predict the long-term maintenance of changes in eating behavior is limited. Stress, a common factor in modern society, has been identified as a potential factor influencing eating behavior and weight management. Studies have shown links between stress and alterations in eating patterns, suggesting a potential role in weight loss maintenance success or failure. The impact of perceived stress on the long-term maintenance of changes in eating behavior during lifestyle interventions remains unclear.

This secondary analysis used data from the PREVIEW intervention, a 3-year randomized controlled trial involving participants with overweight (BMI ≥ 25 kg/m²) and high risk of type 2 diabetes (n = 1311). The intervention comprised a 2-month low-energy diet phase followed by a 34-month weight maintenance phase. The first 6 months were considered the active behavior change stage, and the remainder the behavior maintenance stage. Eating behavior was measured using the Three Factor Eating Questionnaire (TFEQ), assessing flexible and rigid restraint, disinhibition, and hunger. Perceived stress was measured using the Perceived Stress Scale (PSS). Linear mixed effects models for repeated measurements analyzed the associations between stress and eating behavior. The study also examined the association between 3-year weight reduction success and changes in eating behavior. The analytical sample included 1311 participants who attended at least one visit after 6 months. Participants were excluded if they did not meet this criteria. The analysis regarding weight reduction success included 962 participants who completed the study. The study controlled for several factors including age, sex, intervention diet, and baseline eating behavior and BMI.

The study included 1311 participants (65% women), with a mean age of 54 years and baseline BMI of 34.3 kg/m². During the active behavior change stage (first 6 months), flexible and rigid restraint increased, while disinhibition and hunger decreased. Perceived stress at 6 months did not predict changes in eating behavior from month 6 to 36. However, higher perceived stress at 6 months was associated with lower flexible restraint and higher disinhibition and hunger from month 6 to 36 (all p < 0.001). Frequent high stress during the maintenance phase was associated with a greater decrease in flexible restraint (p = 0.026). The mean change in flexible restraint from 6 to 36 months was -1.1 (2.1) in participants with frequent stress and -0.7 (1.8) in participants without frequent stress (Cohen's d = 0.24 (0.04–0.43)). Frequent high stress was also associated with higher disinhibition from month 6 to 36 (beta estimate (95% CI) 1.00 (0.58–1.43), p < 0.001). Among completers (n = 962), successful weight reduction was associated with changes in all eating behavior dimensions. Participants with successful weight loss maintained higher flexible restraint, and showed greater decreases in disinhibition and hunger compared to those with partial or unsuccessful weight loss.

This study found that while perceived stress at a single time point did not predict long-term changes in eating behavior, frequent high stress during the maintenance phase was associated with difficulty maintaining improvements in flexible restraint. Higher initial stress levels were associated with lower flexible restraint and higher disinhibition and hunger throughout the maintenance phase. These findings align with the understanding that stress can impact appetite regulation and self-control, potentially hindering weight management efforts. The lack of a predictive effect of stress at a single time point suggests that capturing the longer-term impact of chronic stress may be more informative. The study highlights the importance of flexible restraint in weight management and suggests that interventions should focus on strategies to support this aspect of eating behavior, especially in participants experiencing high stress levels. The results also suggest that stress management techniques could enhance the effectiveness of weight loss interventions.

Perceived stress was associated with features of eating behavior that can hinder successful weight loss maintenance, particularly a decline in flexible restraint and increases in disinhibition and hunger. Future interventions should investigate whether incorporating stress reduction techniques improves treatment outcomes, especially for individuals experiencing high stress levels. Further research is needed to identify effective strategies for stress management within the context of weight loss interventions, and to better understand the interaction between stress and different aspects of eating behavior.

As a secondary observational analysis, the study could not rule out unmeasured confounding factors. Selection bias may be present due to the requirement of ≥8% weight loss for continued participation and a substantial dropout rate. Higher perceived stress was associated with higher likelihood of dropping out, potentially attenuating observed associations. The self-reported nature of eating behavior and stress measures introduces reporting bias. The scales used to measure flexible and rigid restraint had limitations regarding reliability and validity.