Three RCTs, two positive-quality and one neutral-quality, provide evidence that self-monitoring of food intake improves nutrition-related outcomes related to weight loss and compliance with renal diets. Three observational studies of neutral quality revealed that clients enrolled in cognitive behavioral weight-loss programs that were successful in losing weight were significantly more consistent with self-monitoring.

• Grade I means there is Good/Strong evidence supporting the statement;
• Grade II is Fair;
• Grade III is Limited/Weak;
• Grade IV is Expert Opinion Only;
• Grade V is Not Assignable.
• High (A) means we are very confident that the true effect lies close to that of the estimate of the effect;
• Moderate (B) means we are moderately confident in the effect estimate;
• Low (C) means our confidence in the effect estimate is limited;
• Very Low (D) means we have very little confidence in the effect estimate.
• Ungraded means a grade is not assignable.

Four RCTs, three positive-quality and one neutral-quality, assessed the efficacy of various types of meal replacement or structured meal plan strategies, as compared to self-selected diets in middle aged-adults and found the use of various types of meal replacements or structured meal plans helpful in achieving health and food behavior change in middle-aged adults. Additional research is needed to determine if benefits derived from temporary use of these behavioral strategies can be sustained over time.

• Grade I means there is Good/Strong evidence supporting the statement;
• Grade II is Fair;
• Grade III is Limited/Weak;
• Grade IV is Expert Opinion Only;
• Grade V is Not Assignable.
• High (A) means we are very confident that the true effect lies close to that of the estimate of the effect;
• Moderate (B) means we are moderately confident in the effect estimate;
• Low (C) means our confidence in the effect estimate is limited;
• Very Low (D) means we have very little confidence in the effect estimate.
• Ungraded means a grade is not assignable.

Two positive-quality (one RCT and one meta-analysis of seven RCTs) and one neutral-quality RCT found monetary rewards or reinforcement had no treatment effect.

• Grade I means there is Good/Strong evidence supporting the statement;
• Grade II is Fair;
• Grade III is Limited/Weak;
• Grade IV is Expert Opinion Only;
• Grade V is Not Assignable.
• High (A) means we are very confident that the true effect lies close to that of the estimate of the effect;
• Moderate (B) means we are moderately confident in the effect estimate;
• Low (C) means our confidence in the effect estimate is limited;
• Very Low (D) means we have very little confidence in the effect estimate.
• Ungraded means a grade is not assignable.

Two positive-quality RCTs, one in overweight and obese women and the other in post-menopausal women with diabetes, utilized interventions that incorporated problem-solving strategies. In both studies, use of problem-solving strategies resulted in improvements in key outcome measures, including maintenance of weight loss and in subjects with diabetes, was linked to improvements in fat consumption, self-efficacy and physical activity.

• Grade I means there is Good/Strong evidence supporting the statement;
• Grade II is Fair;
• Grade III is Limited/Weak;
• Grade IV is Expert Opinion Only;
• Grade V is Not Assignable.
• High (A) means we are very confident that the true effect lies close to that of the estimate of the effect;
• Moderate (B) means we are moderately confident in the effect estimate;
• Low (C) means our confidence in the effect estimate is limited;
• Very Low (D) means we have very little confidence in the effect estimate.
• Ungraded means a grade is not assignable.

One highly-intense lifestyle change study found social support was helpful and four traditional lifestyle change programs did not find it helpful. The definition of social support has evolved to include multiple dimensions of social support measured pre- and post-treatment. Two RCTs conducted in the 1990s manipulated social support and found no significant treatment effect. In an RCT published in 2006, multiple dimensions of social support were measured pre- and post-treatment and use of social resources was shown to mediate intervention effects on physical activity, fat consumption and HgA1C change. Additional studies are needed to measure impact of social support interventions on outcomes.

• Grade I means there is Good/Strong evidence supporting the statement;
• Grade II is Fair;
• Grade III is Limited/Weak;
• Grade IV is Expert Opinion Only;
• Grade V is Not Assignable.
• High (A) means we are very confident that the true effect lies close to that of the estimate of the effect;
• Moderate (B) means we are moderately confident in the effect estimate;
• Low (C) means our confidence in the effect estimate is limited;
• Very Low (D) means we have very little confidence in the effect estimate.
• Ungraded means a grade is not assignable.

One positive-quality RCT found a 30-minute motivational interviewing session, based on self-selected diabetic self-management goals, followed by three 10-minute phone calls at one, three and seven weeks, was significantly more effective than usual care in reducing dietary fat intake and increasing physical activity at one year in 100 adults with type 2 diabetes. A positive-quality RCT showed similar results regarding the value of clients' self-selected behavior change goals and demonstrated the effectiveness of goal-attainment training in realizing dietary improvements. One neutral-quality observational study found 422 clients with diabetes who used computer technology to self-select a behavior-change goal in an area of diet or exercise and received brief (eight to 10 minutes) counseling related to the goal, were successful in reducing fat intake two months later. Clients' active participation in selecting and setting goals led to the selection of a goal from the area that could use the most improvement and the goal that was most personally appropriate.

• Grade I means there is Good/Strong evidence supporting the statement;
• Grade II is Fair;
• Grade III is Limited/Weak;
• Grade IV is Expert Opinion Only;
• Grade V is Not Assignable.
• High (A) means we are very confident that the true effect lies close to that of the estimate of the effect;
• Moderate (B) means we are moderately confident in the effect estimate;
• Low (C) means our confidence in the effect estimate is limited;
• Very Low (D) means we have very little confidence in the effect estimate.
• Ungraded means a grade is not assignable.

One neutral-quality RCT assessed the additive effect of a cognitive restructuring component to a 10-week strictly behavioral weight-loss program in 63 middle-aged overweight subjects and found no significant difference between the treatment group and control group in any physiological, behavioral or cognitive measures at baseline, post-treatment and at three-month follow-up. Additional research is needed on the isolated effect of cognitive restructuring as part of a behavioral intervention on nutrition-related outcomes.

• Grade I means there is Good/Strong evidence supporting the statement;
• Grade II is Fair;
• Grade III is Limited/Weak;
• Grade IV is Expert Opinion Only;
• Grade V is Not Assignable.
• High (A) means we are very confident that the true effect lies close to that of the estimate of the effect;
• Moderate (B) means we are moderately confident in the effect estimate;
• Low (C) means our confidence in the effect estimate is limited;
• Very Low (D) means we have very little confidence in the effect estimate.
• Ungraded means a grade is not assignable.