Mixed results from the diabetes prevention PREVIEW project
medwireNews: The PREVIEW project investigators have reported broadly positive results, but tempered by issues such as a high dropout rate.
The randomized trial section of the project involved patients with prediabetes and an elevated BMI and aimed to ascertain the optimal glycemic index (GI) and level of physical activity for the maintenance of weight loss and prevention of type 2 diabetes.
Just 4% of the 2326 included patients developed diabetes during the 3 years of follow-up, which was a much lower incidence than the investigators had anticipated. But PREVIEW had a high attrition rate, with only 43% of participants completing the full 3 years of the study, leaving the true diabetes incidence unclear.
However, Anne Raben (University of Copenhagen, Denmark), who presented these findings to delegates at the 54th EASD Annual Meeting in Berlin, Germany, said that they plan to contact all participants who dropped out to ascertain if they have since been diagnosed with diabetes.
The trial began with all participants put on a total meal replacement plan for the first 8 weeks, giving them 800 calories/day plus 100 g of low-calorie vegetables. They were aged an average of 52 years, two-thirds were women, and their average BMI was 35.4 kg/m2.
During these first 8 weeks, the average weight loss was 10.7 kg, with 1857 (79%) participants losing at least 8% of their starting bodyweight.
A total of 1854 participants moved into the weight maintenance phase of the trial, at which point they were randomly assigned to follow either a conventional diet, of 15% protein and 55% carbohydrate, or to a high-protein, low GI diet, of 25% protein and 45% carbohydrate. In addition, participants were additionally randomized within their dietary groups to undertake either 75 minutes/week of high-intensity exercise or 150 minutes/week of moderate-intensity exercise.
During the weight maintenance phase, the participants sustained their weight loss overall. However, there were no significant differences between the two diets in terms of participants’ risk for developing type 2 diabetes, and no differences between the four diet- and exercise-defined groups.
This may indicate that both dietary patterns were equally effective, but could be an effect of low statistical power, either because dropout was biased toward people who failed to maintain their weight loss and subsequently developed diabetes or because the effectiveness of the total meal replacement phase meant that PREVIEW became a victim of its own success.
Indeed, the researchers believe that the total meal replacement strategy was key to the trial’s effectiveness, because of the motivating effect of early and rapid weight loss.
“Many of us, after seeing what happened during those 8 weeks of fast weight loss, and the new psyche amongst our participants, we would recommend the total diet replacements,” said investigator Jennie Brand-Miller (University of Sydney, New South Wales, Australia), discussing the practical implications of the findings.
She stressed: “Probably one of the most important take-home messages from my talk is that we need to build confidence in the use of total meal replacements amongst health professionals.”
Also, although there was a significant difference in protein intake between the two dietary groups, based on participants’ urinary nitrate, the difference was not as large as the team had hoped to see, suggesting that incomplete adherence to the diets could also have contributed to the lack of difference in outcomes between the groups.
There were no differences between the groups in fat mass, fasting glucose, 2-hour glucose levels, glycated hemoglobin, and insulin use. Fasting glucose fell markedly early in the study but then increased again, whereas the other glycemia markers also fell but rose only slightly during follow-up; the latter was also true for insulin resistance.
The researchers also looked at participants’ behavioral characteristics, finding that those with improvements in insulin resistance over 3 years scored better for dietary restraint, physical activity, and vigor than those with no improvement, and worse for hunger and sleepiness.
And in a subgroup based at a single study site, they found that decreased insulin resistance was also related to reduced food reward-related brain activity, which was in turn associated with greater cognitive restraint, physical activity, and protein intake.
To support the randomized trial findings, the investigators analyzed five population studies, two from the Netherlands, one from Finland, one Canadian study, and one from New Zealand, incorporating a large age range and different dietary cultures and including around 170,000 people overall.
In this analysis, they found no significant associations between protein intake and diabetes incidence after accounting for confounders. This was also the case when considering animal and plant protein separately, and for prediabetes. There were some initial associations, but these disappeared after accounting for BMI, suggesting they were part of the same causal pathway.
However, when considering protein intake as grams per kg of bodyweight, increased intake was associated with reduced diabetes risk, and this was driven by intake of plant protein.
“I would have expected a result for protein per kg bodyweight, but perhaps more for animal sources,” said Edith Feskens (Wageningen University, the Netherlands), who presented the population study results. “But that’s up for discussion.”
medwireNews is an independent medical news service provided by Springer Healthcare. © 2018 Springer Healthcare part of the Springer Nature group