PREVIEW is a 3-year, randomized, multicenter study including up to 2,500 prediabetes participants designed to test the main hypothesis that a high-protein-low glycemic index diet will be superior to the recommended diet for weight maintenance and prevention of type 2 diabetes (T2D) and its complications in individuals with impaired fasting glucose and/or impaired glucose tolerance. The data from PREVIEW was also used for combined population studies, for analyses on dietary restraint, physical activity, sleep, circadian rhythm, and stress, as well as a more detailed investigation on the brain and liver.

• PREVIEW randomized subjects to one of four intervention groups:
  • High protein (25% energy), moderate carbohydrate (<50% energy), low glycemic index (<50) diet with high intensity physical activity (75 min/week).
  • High protein (25% energy), moderate carbohydrate (<50% energy), low glycemic index (<50) diet with moderate intensity physical activity (150 min/week).
  • Moderate protein (15% energy), moderate carbohydrate (55% energy), medium glycemic index (>56) diet with high intensity physical activity (75 min/week).
  • Moderate protein (15% energy), moderate carbohydrate (55% energy), medium glycemic index (>56) diet with moderate intensity physical activity (150 min/week).
• Subjects were followed for 36 months.
• Predicted incidence of diabetes was 15.8% in the moderate protein diet and 10.5% in the high protein diet.

Patient population

• 5,472 adults were screened.
• 2,326 adults were eligible.
• Two-thirds were women.
• Mean age 52 years.
• Mean BMI 35.4 kg/m².
• During the lead-in period of two months, patients were prescribed a low-calorie diet, during which a mean weight loss of 10.7 kg was achieved.
• 79% of all eligible subjects achieved ³8% weight loss on a low-calorie diet.

PREVIEW diet/exercise/behavioral intervention

• There were no significant differences in developing T2D between the two types of dietary intervention groups (high-protein vs. low-protein).
• There were no significant differences in developing T2D among all four intervention groups.
• The cumulative incidence rate of developing T2D was 4%.
• There were no significant differences between the four groups in fasting glucose, insulin levels, 2-hour glucose, HbA_1c levels, or HOMA-IR during the course of the study.
• An initial loss of body weight with a high-protein-low glycemic index diet was seen that might be related to an overall healthier lifestyle.

PREVIEW population studies

• The aim of this analysis was to substantiate the findings of the intervention study using data from population-based studies (Lifeline cohort, n=119,843; NQplus, n=1,584; Young Finns study, n=1969; Quebec family study, n=751; New Zealand Adult Nutrition Survey, n=3,347).
• There appeared to be an inverse relation between protein intake and the incidence of diabetes, even after adjustment for waist circumference; additional study is needed on this relation considering an animal or plant source of protein.
• There was a positive association between glycemic index and glycemic load and the risk of diabetes in women, but not in men.
• There was no clear evidence for an additive effect of glycemic index and protein for risk of diabetes, but an additive effect was seen for weight change.

Dietary restraint, physical activity, sleep, circadian rhythm, and stress

• This analysis used data from PREVIEW to look at behavioral characteristics and possible associations with insulin sensitivity and BMI.
• No differences were found between intervention groups in behavioral scores or circadian rhythm.
• Similarly, there were no differences in dietary restraint, physical activity, or sleep.
• Differences were observed in HOMA-IR and BMI: both were primarily associated with cognitive restraint and physical activity, and both were secondarily associated with stress and mood disturbance.
• Stratification for gender showed associations of cognitive restraint and disinhibition with HOMA-IR, and between stress and BMI, exclusively in women.
• In men only, there were associations between physical activity and HOMA-IR, and between BMI and sleep disturbance.

Spotlights on brain and liver

• This analysis had the objective of identifying possible changes in brain responsiveness of food-related brain areas to visual food cues and to identify possible changes in body fat deposition and intrahepatic lipid content, both in relation to diet and insulin resistance.
• Functional MRI showed that brain reactivity was positively correlated with insulin resistance, percent body fat, and body weight.
• Reduction in brain reactivity was related to increased protein intake and an increase in cognitive control of eating behavior.
• For the liver sub-studies, PREVIEW intervention was related to reductions in intrahepatic lipid content, as well as visceral and subcutaneous adipose tissue.

• The PREVIEW study found that were no significant differences in developing T2D considering a high vs. moderate intake of protein, and that there were fewer T2D cases than expected after 3 years of intervention: only 62 subjects of the 2,326 participants developed T2D by the end of the study, and the probability of not developing diabetes was 96% vs. the projected 85-90%.
• Several gender-specific differences were found that warrant further study.
• Changes in brain reactivity correlated with insulin resistance, percent body fat, and body weight, and may be related to cognitive control of eating behavior.
• PREVIEW intervention may lead to reduced intrahepatic lipid content, as well as visceral and subcutaneous adipose tissue.