Insulin Secretion Predicts Weight-Loss Impact of Diets

BOSTON, May 16 -- A simple test of insulin secretion may predict which obese patients would be better off with a low-glycemic load diet than a low-fat one, researchers here found.

Young, obese patients with high insulin levels on a glucose tolerance test had significantly more weight loss and body fat reduction with a low-glycemic load diet than a low-fat diet, said David S. Ludwig, M.D., Ph.D., of Children's Hospital, Boston, and colleagues.

In a randomized trial, participants with lower insulin secretion had similar results with both diets, they reported in the May 16 issue of the Journal of the American Medical Association.

The results may help explain inconsistent findings in other diet studies, the investigators suggested. Differences in individuals' insulin responsiveness to glucose may account for variations in efficacy, they explained.

"High postprandial insulin concentration has been postulated to decrease availability of metabolic fuels several hours after a meal, causing hunger and overeating," they wrote.

Their study included 73 adults ages 18 to 35 with a body mass index of 30 kg/m² or greater. They were randomized to six months on a low-glycemic load diet (40% carbohydrate and 35% fat) or a low-fat diet (55% carbohydrate and 20% fat).

At baseline, insulin secretion was measured by measuring serum insulin concentration 30 minutes after a 75-g dose of oral glucose.

Fifteen in the low-glycemic load diet group and 13 in the low-fat diet group had an insulin level above the median (57.5 ?IU/mL), which was considered high insulin secretion.

Diabetes was an exclusion criterion. Baseline characteristics were similar between diet groups, except low density lipoprotein cholesterol levels were higher in the low-fat diet group (126 versus 102 mg/dL, P=0.005).

Both groups attended 23 one-hour group nutrition education workshops and a one-hour private dietary counseling session followed by five half-hour motivational phone calls. Participants were instructed to use hunger and satiety cues rather than calorie counting to impose an energy deficit.

The low-glycemic load diet group was instructed to eat more nonstarchy vegetables, legumes, and temperate fruits (such as apples, pears, and peaches) and focus on healthy fat sources, such as nuts and oils. They were to limit intake of refined grains, starchy vegetables, fruit juices, and sweets.

The low-fat diet group was instructed to consume low-fat grains, vegetables, fruits, and legumes while limiting intake of added fats, sweets, and high-fat snacks.

Body weight, body fat percentage determined by dual-energy x-ray absorptiometry, and cardiovascular disease risk factors were assessed at six, 12, and 18 months.

The researchers found no difference in weight loss or body fat percentage change between diet groups overall (P=0.99 and P=0.81, respectively).

Nor was there a difference between diet groups for those with low insulin concentrations in weight loss (P=0.90) or percentage body fat change (P=0.11 at six months and P=0.56 at 18 months).

However, among those with high insulin concentrations, the low-glycemic load diet yielded more rapidly weight loss during the diet phase than the low fat diet (-1.0 versus -0.4 kg/month, P<0.001).

At 18 months, the low-glycemic load diet group with high insulin secretion also had greater overall weight loss than those on the low-fat diet (-5.8 versus -1.2 kg, P=0.004). Their body fat percentage was significantly lower as well at six months (?2.0 versus ?0.4, P=0.04) and at 18 months (?2.6 versus ?0.9, P=0.03).

The high insulin concentration patients assigned to the low-glycemic load diet showed no evidence of rebound weight gain during post-intervention follow-up.

Among the other findings, the researchers reported:

• Low density lipoprotein cholesterol improved more with the low-fat diet than the low-glycemic load diet at six months (?16.3 versus ?5.8 mg/dL, P=0.03) and at 18 months (?10.6 versus ?0.3, P=0.03).
• High density lipoprotein cholesterol improved more with the low-glycemic load diet at six months (1.6 versus ?4.4, P=0.002) though not at 18 months (?3.7 versus ?8.2, P=0.03).
• Triglycerides likewise improved more with the low-glycemic load diet than with the low-fat diet at six months (?21.2 versus ?4.0, P=0.02) but not significantly so at 18 months (?9.0 versus 2.0, P=0.18).
• Insulin secretion was not a significant factor in lipid levels, blood pressure, or fasting glucose or insulin levels.

"These findings are consistent with previous research demonstrating benefits of low-carbohydrate or low-glycemic load diets with regard to components of the metabolic syndrome," the researchers wrote.

The greater benefit of the low-fat diet on LDL cholesterol, which is not a component of the metabolic syndrome, may have been due to the baseline differences between groups or its lower saturated fat content.

Therefore, they speculated that a low-glycemic load diet in which saturated fat was kept low, such as by substituting vegetable for animal fat sources, would yield a benefit for all three cardiovascular risk factors rather than just two as was seen in the study.

The study was limited by self-reporting of dietary intake and use of tabulated glycemic index values.

However, "Because this study used dietary counseling rather than meals prepared in a metabolic kitchen, these findings should have direct relevance to the management of obesity in routine clinical practice," the researchers wrote.

Further study is needed both to look at other patient populations and to explain the relationship between diet and insulin secretion, they concluded.