Type 2 diabetes mellitus (T2DM) is preventable through lifestyle interventions focused on weight reduction via dietary and physical activity changes. However, maintaining weight loss long-term proves challenging for many. Understanding modifiable factors mediating successful weight loss maintenance is crucial. Psychological dimensions of eating behavior are potential mediators. These dimensions, including cognitive restraint (flexible and rigid), disinhibition, and susceptibility to hunger, influence food choices and intake. Previous research shows that increased cognitive restraint is associated with successful weight loss and maintenance, while reduced disinhibition and hunger also predict weight loss, at least short-term. Studies examining eating behavior dimensions in individuals with impaired glucose tolerance (IGT) are limited. The current study aimed to investigate how a lifestyle intervention to prevent T2DM affects eating behavior dimensions in individuals with IGT and whether early changes in these dimensions mediate the intervention's long-term effects on body weight.

Longitudinal weight management studies indicate that increased cognitive restraint is linked to successful weight loss in overweight or obese individuals and long-term weight maintenance. Increases in both rigid and flexible restraint have shown associations with weight loss, though the components may have divergent effects on health outcomes. Flexible restraint, in particular, has been linked to greater weight loss and better maintenance. Reduction in disinhibition and susceptibility to hunger have also been predictive of (at least short-term) weight loss. Studies on eating behavior dimensions among individuals with IGT are scarce. The DELIGHT trial showed that enhancing flexible control and decreasing disinhibition were beneficial for central adiposity and blood glucose control. A sub-study of the DPP indicated that increased dietary restraint predicted better long-term weight loss among individuals with IGT. Only one previous intervention study investigated eating behavior dimensions as mediators of weight loss and maintenance, finding that increased flexible restraint mediated greater 24-month weight loss in overweight/obese women. More evidence is needed regarding the role of eating behavior dimensions in long-term weight loss maintenance among those at risk of T2DM.

This sub-study utilized data from the Finnish Diabetes Prevention Study (DPS), a multicenter lifestyle intervention RCT. The current analysis included 38 men and 60 women with overweight and IGT from the Kuopio research center. Participants were randomized to an intensive, individualized lifestyle intervention group (n=51) or a control group (n=47). Body weight was measured annually for nine years. The Three Factor Eating Questionnaire (TFEQ), assessing cognitive restraint (total, flexible, and rigid), disinhibition, and susceptibility to hunger, was administered annually. The intervention involved seven face-to-face consultations with a nutritionist during the first year and subsequent sessions. The intervention focused on weight reduction (5% or more), dietary modification (reducing fat, increasing fiber), and moderate-intensity physical activity. The control group received general information on healthy lifestyle. Descriptive statistics, Pearson's chi-squared and correlation tests, Mann-Whitney U-test, independent samples t-test, and structural equation modeling (SEM) were used to analyze the data. Mediation analyses examined whether changes in eating behavior during the first year mediated the intervention's effect on 9-year weight change. Sex, age, baseline dietary restraint, and baseline weight were controlled for in the mediation analyses.

At 9 years, 43.6% of the intervention group achieved the 5% weight loss goal versus 23.5% in the control group (p=0.07). The intervention group experienced significantly greater weight loss than the control group during both the first year (-5.2 kg vs. -1.2 kg, p<0.001) and at 9 years (-3.0 kg vs. +0.1 kg, p=0.046). The intervention group showed significantly greater increases in total (p<0.001) and rigid (p=0.001) restraint of eating during the first year, with these differences persisting at 9 years (p=0.002 and p=0.004 respectively). Flexible restraint also increased significantly more in the intervention group during the first year (p=0.018) and showed a trend toward significance at 9 years (p=0.093). There were no significant differences between groups in changes in disinhibition or susceptibility to hunger. First-year changes in total, rigid, and flexible restraint were inversely correlated with 9-year weight change. Mediation analyses showed that first-year increases in total, flexible, and rigid restraint significantly mediated the intervention's effect on 9-year weight loss.

This study demonstrates the long-lasting effects of intensive, individually tailored lifestyle counseling on cognitive restraint of eating and body weight in middle-aged individuals with overweight and IGT. The significant mediation effects suggest that increasing cognitive restraint early in the intervention may be key to long-term weight loss maintenance. These findings align with some previous research showing associations between increased dietary restraint and successful weight loss in IGT populations. However, the concept of restrained eating is debated, with different measures potentially capturing different types of eaters. The TFEQ, used in this study, may identify individuals motivated by health and successful at consistently controlling food intake. The intervention's success may be linked to enhanced self-regulation skills, a concept closely related to restraint of eating. Goal setting, self-monitoring, problem-solving, and other self-regulation techniques were integral to the DPS intervention and may have contributed to the observed changes in eating behavior and weight.

Intensive, individually tailored lifestyle interventions have long-lasting positive effects on cognitive restraint and weight management in middle-aged adults with overweight and IGT. Early increases in cognitive restraint seem particularly important for long-term success. Future interventions should incorporate training in self-regulation skills, such as goal setting and self-monitoring, to enhance long-term weight loss maintenance and improve various health outcomes. Modern technology could expand the reach of these interventions.

The relatively small sample size and varied intervention duration are limitations. The study population comprised volunteers, potentially more health-conscious than the general population. The findings may not generalize to younger or older individuals. The mediation model assumes a unidirectional relationship between change in dietary restraint and change in body weight; however, a bidirectional relationship may exist.