- Click here for explanation of classification scheme.

To test the hypothesis that consumption of High-glycemic index (GI) foods induces a sequence of hormonal changes that lead to decreased availability of metabolic fuels, excessive hunger, and overeating in obese adolescent males. 

1. Pubertal boys (mean age 15.7±1.4 (SD) years
2. Obese--defined as >120% of ideal body weight (National Center for Health Statistics [NCHS] percentiles, 1979)
3. Healthy (assessed by physical examination and laboratory analysis--complete blood count, electrolytes, liver function tests, glycosylated hemoglobin, thyroid stimulating hormone and urinalysis)

 

Participants were exluded from study if they were:

1. Nonobese adolescent boys
2. Non-healthy

Recruitment --13 participants recruited to participated in study

Design--cross-over randomized study consisting of three separate 24-hour admissions (Children's Hospital, Boston) separated by a 1- to 2-week washout-period

Blinding used (if applicable)--no

Intervention (if applicable)

Study protocol:

12 obese adolescent males were randomly assigned to receive different test meals (Low-, Medium-, or High-GI meals) at 3, 24-hour admissions to the General Clinical Reseach Center (GCRC), Children's Hospital, Boston. There was a 1- to 2-week washout period between admissions.

Each admission was conducted in an identical manner. For each admission:

• After admission at 6 pm, subjects received a standard Low-GI dinner (40% energy from carbohydrate, 30% from protein, 30% from fat; total energy 18.5% of predicted resting metabolic rate (RMR) calculated from body weight at preadmission physical examination).  A bedtime snack was also provided (same macronutrient breakdown as dinner; energy content 5% of predicted RMR).
• At 6:45 am the following morning, after an intravenous line was placed and baseline blood samples were obtained and hunger scale measurements were taken ("not at all hungry" to "extremely hungry"). Participants received test meals at breakfast that had a Low-, Medium-, or High-GI to be consumed completely in 20 minutes.
• Blood samples and hunger scale ratings were obtained every 30 minutes for 5 hours from the start of breakfast.
• After 5 hours, the intravenous line was removed and a second test meal (same GI content as breakfast on that day) was provided for lunch to be consumed completely in 20 minutes.
• After lunch, subjects were allowed out of bed and encouraged to request "ad libitum" snack meals (~22.7 mJ each) if "very hungry" (as many as desired) during the subsequent 5 hours. They could eat as much or as little as wanted to satiety. Timing before request of snack meals was recorded. Leftovers were weighed to determine amount eaten.
• Subjects were discharged and instructed to follow usual diet between admissions.

Test meals:

Alterations in GI content were obtained through differences in food structure and sugar types. Meals were not energy restricted. The nutrient breakdown for the 3 test meals were as follows:

• Low-GI: 40% energy from carbohydrate, 30% from protein, 30% from fat.
• Medium-GI and High-GI: 64% energy from carbohydrate, 16% from protein, 20% from fat. The 2 test meals differed in type of oatmeal ("steel cut" for Medium-GI and "instant oatmeal" for High GI); type of sweetener (dextrose for High-GI and fructose for Medium-GI); milk (lactaid drops added to milk to increase GI of milk sugar for high-GI).
• Energy density was similar for 3 meals (2.5 kJ/g).
• The size of the test meals were determined individually for each subject (18.5% of predicted RMR).

The snack meals were identical for each participant (~22.6 mJ each).

Statistical Analysis

• Repeated measures ANOVA with meal alone or meal and time as within-subjects factors.
• Tukey's tests to adjust for multiple meal comparisons, except for fatty acids, epinephrine, and growth hormones for which only the high- and low-GI meals were compared.
• Within-subject correlations coefficients between afternoon energy intake and hormonal and metabolic parameters.
• Areas under the glycemic response curves were calculated using the trapezoidal rule for values above baseline.

Significance estabished at P<.05.

Timing of Measurements

Anthropometrics: body weight obtained during preadmission physical examination. Method not described.

Dietary: For each admission (3 total):

• Standard low GI meal and bedtime snack after admission to the GCRC (6 pm).
• Next day, subjects consumed identical test breakfast and lunch meals that had a low, medium, or high-GI (randomly assigned). All meals were prepared under supervision of the research dietitians and were to be consumed completely within 20 minutes.
• 10-cm analog hunger scale rating ("not at all hungry" to "extremely hungry") was obtained before breakfast and every 30 minutes prior to lunch.
• Following lunch, until discharge, subjects were encouraged to request "ad libitum" snack meals if "very hungry." Timing to requests following test meals were recorded and leftovers were weighed to determine amount of energy consumed at each meal request.

Laboratory (for hormonal and metabolic changes after test breakfasts)

• Baseline: complete blood count, electrolytes, liver function tests, glycosylated hemoglobin, thryroid stimulating hormone and urinalysis.
• Blood samples obtained every 30 minutes (intravenous line) between breakfast and lunch test meals (plasma glucose, insulin glucagon, epinephrine, serum fatty acids, growth hormone levels). 

Dependent Variables

• Hormonal and metabolic responses: (mean plasma glucose, insulin, glucagon, epinephrine, fatty acids, growth hormone) from blood samples between breakfast and lunch
• Perceived hunger: hunger scale ratings
• Voluntary food intake: snack meals were used for measurement of voluntary food intake. Time between test meals and requests of snack meals and energy intake were recorded.

Independent Variables

• 3 test meals--low-, medium-, or high-GI

Control Variables--not reported

Initial N: 13 adolescent males

Attrition (final N): 12; one participant had recurrent difficulties establishing and maintaining intravenous access

Age: mean and SD: 15.7±1.4 y

Ethnicity: not provided

Other relevant demographics: not provided

Anthropometrics: all adolescent boys were obese (defined as >120% ideal body weight; NCHS, 1979)

Location: General Clinical Research Center (GCRC), Children's Hospital, Boston, MA

Hormonal and metabolic responses:

The hormonal and metabolic responses to the High-, Medium-, and Low-GI meals were markedly different as follows:

• The mean area under the glycemic response curve for the High-GI meals was twice that of the Medium-GI meal and nearly fourfold that of the Low-GI meal (p<.001).
• The mean plasma glucose concentration was lower after the High-GI meal than after the Medium-GI meal or the Low-GI meal (p=.02).
• The insulin level was greater after the High-GI meal than after the Medium-GI meal or Low-GI meal because of the rapid absorption of glucose (p<.01).
• Plasma glucagon level rose after the Low-GI meal but was suppressed after the Medium- and High-GI meals; likely because of the low protein content of the high-GI meal and inhibitory effects of high plasma glucose and insulin concentrations (p<.01).
• Serum fatty acids were suppressed to a greater extent after the High-GI meal compared with the Low-GI meal (p<.05).
• Higher concentrations of the hormones epinephrine (p<.05) and growth hormone (p=.05; not significant, respectively) after the High-GI meal relative to the Low-GI meal.

Perceived hunger and voluntary snack meal timing and energy intake:

• Perceived changes in hunger after test breakfasts (ratings of hunger) were greater at all 30 minute time periods for the High-GI breakfast, compared with the Low-GI breakfast, with Medium-GI yielding intermediate scores.
• On average, participants consumed 53% more energy after High-GI meals than after the Medium-GI meals (5.8 mJ and 3.8 mJ, respectively; p<.05) and 81% more total energy after High-GI meals than after Low-GI meals (5.8 mJ and 3.2 mJ, respectively; p=.01).
• Mean time to the first snack meal request was 2.6 hours for the High-GI meals, 3.2 hours for the Medium-GI meals, and 3.9 hours for the Low-GI meals (High- versus Low-Gi meals; p=.01; High-versus Medium-GI meals; not significant).

BMI

• Only baseline heights and weights were reported; no follow-up on weight status during study.

The results of this study demonstrated that meals containing varying GI content had markedly different effects on metabolism and hormones, perceived hunger, and subsequent food intake in overweight adolescent males. In comparison to Low-GI meals, High-GI test meals resulted in higher postprandial serum glucose and insulin levels, lower glucagon levels, and lower serum fatty acid levels. In addition, longer time to request voluntary meals after Low-GI meals (breakfast and lunch) was observed. Voluntary energy intake after High-GI meals was 53% greater than after Medium-GI meals and 81% greater than after Low-GI meals.

"We recognize that this study evaluated only the acute effects of Low-GI meals; the effectiveness of Low-GI diet in promoting long-term weight loss is unknown. Long-term dietary intervention studies are necessary to resolve this question. Moreover, reductions in dietary GI may also have beneficial effects on serum lipids, risk of diabetes mellitus, and other diseases associated with hyperinsulinemia."

Government:

NHLBI

Limitations:

• No blinding indicated
• Study only evaluated the acute effects of Low-GI meals; not long-term. may not reflect long-term effects of Low- or High-GI diets such as effects on body weight   
• Limited generalizability; only male adolescent subjects were included in study (females, different age groups?)