OBJECTIVE—The aim of this study was to compare the effects of a flexible lower- and higher-glycemic index (GI) Mexican-style diet on biochemical data and BMI during a 6-week treatment period.

RESEARCH DESIGN AND METHODS—This study was a randomized, crossover design of two 6-week periods with a 6-week washout period between treatments. Subjects with type 2 diabetes (n = 36) with a BMI >25 kg/m2 were selected. Fourteen subjects completed the study with eligible dietary records. Dietary instruction was provided on flexible diets with both a high and low GI. Fasting venous blood samples were taken at the start and finish of each dietary period, and biochemical data were analyzed. Multi- and univariate one-factor repeated-measures ANOVA were used to compare biochemical data.

RESULTS—Glycemic load and GI were lower during the low-GI diet, and dietary fiber was lower during the high-GI diet. The participants in the low-GI period consumed significantly fewer carbohydrates, such as white-wheat bread, white long-grain rice, potatoes, high-GI fruits, and carrots, and more carbohydrates, such as pinto beans, whole-meal wheat bread, and low-GI fruits than did participants in the high-GI period. There were no differences in the amount of carbohydrates consumed, such as corn tortillas and dairy products. At the end of the study periods, A1c was improved on the low- compared with the high-GI diet (P < 0.008).

CONCLUSIONS—We conclude that a low-GI diet, containing Mexican-style foods, may help to improve the metabolic control in type 2 obese diabetic subjects during a 6-week period.

In Mexico, the incidence of diabetes has been increasing, and this epidemic is likely to continue to escalate. The prevalence of diabetes increased from 8.2% in 1993 to 11.8% in 2000. In Mexico City, the central and southern states, and among underprivileged people (including the Mexican Indians), the prevalence of diabetes increased to 14% (1).

It is well established that different types of carbohydrates produce different glycemic responses (2–5). The glycemic index (GI) and glycemic load (GL) have been proposed as a method of ranking foods on the basis of the incremental blood glucose response they produce for a given amount of carbohydrates (6–8). Although the predictor effect of the GI and insulin responses to mixed meals is controversial (3,9–11), there are significant associations between the GI and/or GL and control of blood glucose, despite the imprecision of the measures (8,12). Results from two large prospective studies, the Nurses’ Health Study (7) and the Health Professionals’ Follow-Up Study (13), showed a positive association between a high-GI diet and risk of developing diabetes.

On the other hand, there have been studies examining the incidence of type 2 diabetes in African Americans (14) that showed no association of GI and GL with diabetes risk. In a recent study conducted in Australian overweight subjects, eating a low-GI diet showed a reduction of A1c, triacylglycerol, and LDL cholesterol after 12 weeks of treatment (15). Another study using a low-GI diet in healthy French men demonstrated a reduction of fat mass and triacylglycerol after a 5-week period (16). However, we have not found long-term studies showing the effect of low-GI diets containing Mexican-style foods on either Mexicans or Mexican Americans with diabetes. The typical Mexican diet includes beans (legumes) and corn tortillas (traditionally made), which are foods with a low GI. However, in Mexico and in the U.S., current dietary guidelines for people with diabetes focus on lowering dietary fat and increasing carbohydrate intake but do not mention the GI (17,18).

The aim of this study was to compare the effects of a lower and higher GI and flexible Mexican-style diet on biochemical data and BMI during a 6-week treatment period.

Subjects

Thirty-six subjects with type 2 diabetes (age, 53 ± 9 years; range, 35–75 years), and a mean diabetes duration of 8 ± 7 years participated in the study. Twelve subjects (four on the low-GI diet) dropped out of the study, and 10 did not complete the dietary records. Fourteen subjects completed the study with eligible dietary records (59 ± 9 years; 44–75 years); of these, six were men and eight were women. Their BMI was 29.6 ± 5.8 kg/m2, mean fasting glucose was 9.5 mmol/l, and mean A1c was 8.4%. The ethical committee of the Instituto de Nutrición de Baja California approved the study.

Study procedure

The study procedure consisted of a randomized, controlled, crossover design of two 6-week periods with a 6-week washout period between treatments. Dietary instruction was provided on flexible diets with both a higher and lower GI. Each participant entered a run-in period of 15 days before the start of the experimental diet, during which time they received their standard medications and were given detailed instructions, including a pamphlet, on lower- or higher-GI foods depending on randomization. Recommendations with respect to the amount of carbohydrates, which should be in the diet, were based on the number of portions of grain, fruit, vegetables, and dairy products recommended at the Apple of Health illustration guide (19). Subjects were randomly allocated to the lower- and higher-GI diets for a 6-week period in a crossover design (each subject received the two diets by random allocation to each diet) with a 6-week washout period. During the washout period, the habitual diet of each individual was used following the Apple of Health guide (19,20). Body weight was measured before and after dietary treatment periods.

Diets and dietary assessment

Typical lower-GI foods were oranges, beans (legumes), yogurt, pasta, and corn tortillas. Typical higher-GI foods were corn flakes, white bread, potatoes, and ripe bananas.

Participants completed unweighed dietary intake diaries for 1 day during the first, fourth, and sixth week of the two study periods, and computerized dietary analysis was done using Nutritionist Pro. The GI values are based on glucose as the standard taken from Foster-Powell’s international table of GI (5). The average dietary GL for each participant was calculated by summing the products of the carbohydrate content per serving for each food times the average number of serving of that food times the GI for that food (21). GI was computed for each subject by dividing the GL by the total carbohydrate intake per day (21).

Blood analysis

Fasting venous blood samples were taken at the start and finish of each dietary period and were analyzed for glucose, A1c, total serum cholesterol, HDL, and triacylglycerol. For the quantitative determination of glucose in serum, the glucose oxidase procedure, based on a modified Trindler method, was used. Enzymatic determination was performed for total cholesterol, triacylglycerol, and HDL. LDL cholesterol was calculated using the Friedwald formula (22).

Statistical analysis

General lineal model with repeated measures was used to compare BMI and biochemical data from baseline to after the diet, because the same treatments were compared on the same participants. The between-subjects factor was the order of the diet treatment, and the within-subjects factor was the diet treatment (low- and high-GI diets). The Bonferroni test was used to adjust for multiple comparisons. A P value <0.05 indicated statistical significance. To compare carbohydrate food sources between low- and high-GI diets, Wilcoxon’s rank-sum test was used.

Fourteen subjects completed dietary intake diaries for 3 days during the first, fourth, and sixth weeks of each of the two treatment periods. Most participants during the high-GI diet who dropped out of the study did so because of the perception of bad glycemic control or lack of acceptance of the diet. Compliance to the low-GI diet was high; only four participants dropped out of the study during this diet. Table 1 shows that the contribution to energy as protein, carbohydrate, sugar, and fat intake between the low- and the high-GI periods did not show statistical difference. Dietary fiber was significantly higher during the low-GI period (P = 0.003), and the GI and GL were significantly lower during the low-GI period (P = 0.0001).

BMI, A1c, and fasting serum glucose are shown in Table 2. BMI was lower during the low-GI period (P = 0.054), and A1c was also significantly lower after the low-GI period compared with after the high-GI period (P = 0.019). However, total, LDL, and HDL cholesterol, triacylglycerol, and fasting serum glucose did not show any significant changes after the dietary treatment periods (Table 3).

During the low period, participants consumed significantly less carbohydrates, such as white-wheat bread, white long-grain rice, potatoes, high-GI fruits (papaya, mango, banana, watermelon, and pineapple), carrots, and consumed more carbohydrates, such as pinto beans, whole-meal wheat bread, and low-GI fruits (pear, apple, apricot, orange, and nectarine), than did participants in the high-GI period. There were no differences in the amount of carbohydrate consumed as corn tortillas or dairy products (Table 4).

The study demonstrated that Mexican overweight and obese subjects with type 2 diabetes, who were given flexible dietary instructions based on cultural-based diets and low-GI choices achieved significantly better A1c levels after 6 weeks compared with those who received high-GI diets based on the Mexican dietary guidelines. Additionally, although there were differences in the consumption of different sources of food from grains, fruit, and vegetables, there was no statistical difference between the consumption of corn tortillas and dairy products in both diets.

The fact that a cultural-based flexible diet was given to the participants compensated the attrition, and therefore compliance was good when a high-fiber diet was given. The reported energy intakes are somewhat low, which might be the results of underreporting or to the low energy intakes in Mexican people with diabetes. Many low-GI foods, including beans, corn tortillas, pasta, and fruits, are part of the usual diet of Mexicans of a lower socioeconomic status. The fact that a flexible diet was given was particularly important for most of the participants, because they had already experienced a number of restricted diets. Other studies using flexible diets during a 12-month period with low-GI foods demonstrated beneficial effects on children with type 1 diabetes (23).

Additionally, this study showed a reduction of the BMI (P = 0.054) after the low-GI period. This is an important finding in a population with a high prevalence of diabetes (1) and child and adult obesity (24). A recent analysis of 26 studies showed that complex carbohydrates, regardless of their fiber content, have the greatest effect on satiety and also reduction in food intake (25). Buyken et al. (26) showed that low-GI food consumption was associated with low levels of A1c, independent of its fiber content.

One of the most interesting findings of this study was that subjects during both dietary periods consumed corn tortillas or dairy products the same during each dietary period. However, subjects in the low-GI period consumed less white bread, potatoes, white long-grain rice, and high-GI fruits (papaya, mango, banana, watermelon, and pineapple) and carrots, and consumed more pinto beans, whole-meal wheat bread, and low-GI fruits (pear, apple, apricot, orange, and nectarine) than did participants in the high-GI period. This may be the result of resistance toward changing specific foods, such as corn tortillas, which are the main staple food in a Mexican-style diet. The results on the consumption of dairy products were different than those observed in children with type 1 diabetes in Australia. In that study, subjects in the lowest GI quartile consumed more carbohydrates as dairy-based foods and whole wheat bread than did subjects in the highest GI quartile (27). As in our study, in the EURODIAB study dairy foods were not observed to be a major determinant of GI intake in the adult population (26).

Thus, despite the worldwide controversy of the advantages and disadvantages of using the low-GI diet criteria for the nutrition education in diabetes (18,28–30), this study provides evidence that more flexible Mexican-style instructions, with an emphasis on the use of low-GI foods, decreased BMI and improved the metabolic control in individuals with type 2 diabetes. Mexicans with type 2 diabetes and traditional food habits might benefit from using the low-GI foods criteria. Although this diet helped to increase the adherence to the diet, it also promotes Mexican-style dishes, such as “frijoles and tortillas,” increasing satiety as well as reducing body weight and improving metabolic control. Dietary recommendations for the treatment of Mexicans with type 2 diabetes may need to be reconsidered.

Table 1—

Reported energy composition, fiber, and GI of low- and high-GI diets (n = 14)

Low GIHigh GIP*
Energy (kJ/day) 5,949 ± 239 6,530 ± 232 0.1 
Carbohydrate (g) 214 ± 10.7 249 ± 10.7 0.02 
Total fat (g) 36 ± 2.8 35 ± 2.4 0.80 
Protein (g) 74 ± 3.1 69 ± 2.7 0.28 
Sugar (g) 64 ± 3.4 72 ± 2.5 0.04 
Fiber (g/day) 34 ± 2.1 25 ± 1.6 0.003 
GI 44 ± 0.9 56 ± 1.3 0.0001 
GL 86 ± 5.3 139 ± 7.3 0.0001 
Low GIHigh GIP*
Energy (kJ/day) 5,949 ± 239 6,530 ± 232 0.1 
Carbohydrate (g) 214 ± 10.7 249 ± 10.7 0.02 
Total fat (g) 36 ± 2.8 35 ± 2.4 0.80 
Protein (g) 74 ± 3.1 69 ± 2.7 0.28 
Sugar (g) 64 ± 3.4 72 ± 2.5 0.04 
Fiber (g/day) 34 ± 2.1 25 ± 1.6 0.003 
GI 44 ± 0.9 56 ± 1.3 0.0001 
GL 86 ± 5.3 139 ± 7.3 0.0001 

Data are means ± SEM.

*

Wilcoxon’s sum-rank test.

View Large
Table 2—

BMI and metabolic characteristics of subjects at baseline at week 6 after low- or high-GI diet (n = 14)

Low-GI period
High-GI period
P*
Baseline6 weeksBaseline6 weeks
Fasting serum glucose (mmol/l) 10.1 ± 0.99 8.9† ± 0.66 10.9 ± 1.43 10.0 ± 0.99 0.04 
A1c (%) 8.5 ± 0.28 8.1‡§ ± 0.24 8.6 ± 0.3 8.6 ± 0.24 0.02 
BMI (kg/m2) 32.4 ± 1.6 31.8‖ ± 1.6 32.3 ± 1.6 32.3 ± 1.6 0.05 
Weight (kg) 91.6 ± 6.5 90.1 ± 6.2 92.6 ± 6.8 92.0 ± 6.6 0.04 
Low-GI period
High-GI period
P*
Baseline6 weeksBaseline6 weeks
Fasting serum glucose (mmol/l) 10.1 ± 0.99 8.9† ± 0.66 10.9 ± 1.43 10.0 ± 0.99 0.04 
A1c (%) 8.5 ± 0.28 8.1‡§ ± 0.24 8.6 ± 0.3 8.6 ± 0.24 0.02 
BMI (kg/m2) 32.4 ± 1.6 31.8‖ ± 1.6 32.3 ± 1.6 32.3 ± 1.6 0.05 
Weight (kg) 91.6 ± 6.5 90.1 ± 6.2 92.6 ± 6.8 92.0 ± 6.6 0.04 

Data are means ± SEM.

*

General lineal model (repeated measures) global results. Bonferroni test

†

P = 0.03 (low GI vs. high GI);

‡

P = 0.04 (baseline, 6-week low-GI period);

§

P = 0.001 (low-GI vs. high-GI period);

‖

P = 0.04 (baseline, 6-week low-GI period);

P = 0.06 (baseline, 6-week low-GI period).

View Large
Table 3—

Serum lipids changes in response to 6 weeks of low- or high-GI diet (n = 14)

Low-GI period
High-GI period
P*
Baseline6 weeksBaseline6 weeks
Fasting total cholesterol (mmol/l) 5.4 ± 0.24 5.4 ± 0.23 5.6 ± 0.23 5.5 ± 0.24 0.5 
LDL (mmol/l) 3.2 ± 0.16 3.2 ± 0.17 3.3 ± 0.2 3.4 ± 0.19 0.4 
HDL (mmol/l) 1.22 ± 0.2 1.22 ± 0.4 1.24 ± 0.05 1.24 ± 0.06 0.78 
Triacylglycerol (mmol/l) 1.99 ± 0.26 2.1 ± 0.21 2.2 ± 0.23 1.9 ± 0.19 0.13 
Low-GI period
High-GI period
P*
Baseline6 weeksBaseline6 weeks
Fasting total cholesterol (mmol/l) 5.4 ± 0.24 5.4 ± 0.23 5.6 ± 0.23 5.5 ± 0.24 0.5 
LDL (mmol/l) 3.2 ± 0.16 3.2 ± 0.17 3.3 ± 0.2 3.4 ± 0.19 0.4 
HDL (mmol/l) 1.22 ± 0.2 1.22 ± 0.4 1.24 ± 0.05 1.24 ± 0.06 0.78 
Triacylglycerol (mmol/l) 1.99 ± 0.26 2.1 ± 0.21 2.2 ± 0.23 1.9 ± 0.19 0.13 

Data are means ± SEM.

*

Repeated measures ANOVA.

View Large
Table 4—

Differences of carbohydrate (in grams) food sources between the low-GI and high-GI diets (n = 14)

Carbohydrate source*Low GIHigh GIP†
Corn tortilla 24 (0–74) 20 (0–89) 0.96 
Atole (corn meal beverage) 0 (0–29) 0 (0–59) 0.009 
White-wheat flour bread 0 (0–78) 25 (0–130) 0.0001 
Whole-meal wheat bread 17 (0–64) 0 (0–43) 0.0001 
Oats (instant) 0 (0–66) 0 (0–33) 0.064 
White long-grain rice 0 (0–42) 10.5 (0–103) 0.0001 
Potato (baked, boiled) 0 (0–31) 5 (0–96) 0.0001 
Pinto beans 22 (0–66) 0 (0–31) 0.0001 
Nopales (prickly pear cactus) 0 (0–21) 0 (0–7) 0.026 
Carrots 0 (0–12) 3 (0–23) 0.002 
Fruit low GI 46 (0–131) 8 (0–47) 0.0001 
Fruit high GI 0 (0–75) 33 (0–86) 0.0001 
Milk- and dairy-based foods 14 (0–43) 13 (0–41) 0.89 
Carbohydrate source*Low GIHigh GIP†
Corn tortilla 24 (0–74) 20 (0–89) 0.96 
Atole (corn meal beverage) 0 (0–29) 0 (0–59) 0.009 
White-wheat flour bread 0 (0–78) 25 (0–130) 0.0001 
Whole-meal wheat bread 17 (0–64) 0 (0–43) 0.0001 
Oats (instant) 0 (0–66) 0 (0–33) 0.064 
White long-grain rice 0 (0–42) 10.5 (0–103) 0.0001 
Potato (baked, boiled) 0 (0–31) 5 (0–96) 0.0001 
Pinto beans 22 (0–66) 0 (0–31) 0.0001 
Nopales (prickly pear cactus) 0 (0–21) 0 (0–7) 0.026 
Carrots 0 (0–12) 3 (0–23) 0.002 
Fruit low GI 46 (0–131) 8 (0–47) 0.0001 
Fruit high GI 0 (0–75) 33 (0–86) 0.0001 
Milk- and dairy-based foods 14 (0–43) 13 (0–41) 0.89 

Data are median (range).

*

Median of total carbohydrate intake of food sources contributing 5% or more of carbohydrate intake.

†

Wilcoxon’s rank-sum test.

View Large

A table elsewhere in this issue shows conventional and Système International (SI) units and conversion factors for many substances.

This project was supported by Omnilife-Conacyt.

We sincerely thank Dr. Kenia Cazares for clinical assistance to the participants and Dr. Gudelia Rangel for statistical advice.

1
Secretaria de Salud: Programa de Acción: Diabetes Mellitus
. México, DF, SSA,
2000
2
Wolever TM, Bolognesi C: Source and amount of carbohydrate affect postprandial glucose and insulin in normal subjects.
J Nutr
 
126
:
2798
–2806,
1996
3
Ludwig DS, Eckel RH: The glycemic index at 20 y.
Am J Clin Nutr
 
76
:
264S
–265S,
2002
4
Jenkins DJA, Kendall CWC, Augustin LSA, Franceschi S, Hamidi M, Marchie A, Jenkins AL, Axelsen M: Glycemic index: overview of implications in health and disease.
Am J Clin Nutr
 
76
:
266S
–273S,
2002
5
Foster-Powell K, Holt SH, Brand-Miller C: International table of glycemic index and glycemic load values: 2002.
Am J Clin Nutr
 
76
:
5
–56,
2002
6
Jenkins DJ, Wolever TM, Taylor RH, Barker H, Fielden H, Baldwin JM, Bowling AC, Newman HC, Jenkins AL, Goff DV: Glycemic index of foods: a physiological basis for carbohydrate exchange.
Am J Clin Nutr
 
34
:
362
–366,
1981
7
Salmeron J, Manson JE, Stampfer MJ, Colditz GA, Wing LA, Willett WC: Dietary fiber, glycemic load, and risk of non-insulin-dependent diabetes mellitus in women.
JAMA
 
277
:
472
–477,
1997
8
Liu S, Willet WC, Stampfer MJ, Hu FB, Franz M, Sampson L, Hennekens CH, Manson JE: A prospective study of dietary glycemic load, carbohydrate intake, and risk of coronary heart disease in US women.
Am J Clin Nutr
 
71
:
1455
–1461,
2002
9
Hollenbeck CB, Coulston AM, Reaven GM: Glycemic effects of carbohydrates: a different perspective.
Diabetes Care
 
9
:
641
–647,
1986
10
Willet W, Manson J, Liu S: Glycemic index, glycemic load, and risk of type 2 diabetes.
Am J Clin Nutr
 
76
:
274S
–280S,
2002
11
Pi-Sunyer FX: Glycemic index and disease.
Am J Clin Nutr
 
76
:
290S
–298S,
2002
12
Wolever TM, Nuttal FQ, Lee R, Wong GS, Josse RG, Csima A, Jenkins DJ: Prediction of the relative blood glucose response of mixed meals using the white bread glycemic index.
Diabetes Care
 
8
:
418
–428,
1985
13
Salmeron J, Ascherio A, Rimm EB, Colditz GA, Spiellgerman D, Jenkins DJ, Stampfer MJ, Wing AL, Willett WC: Dietary fiber, glycemic load, and risk of NIDDM in men.
Diabetes Care
 
20
:
545
–550,
1997
14
Stevens J, Ahn K, Juhaeri, Houston D, Steffan L, Couper D: Dietary fiber intake and glycemic index and incidence of diabetes in African-American and white adults.
Diabetes Care
 
25
:
1715
–1721,
2002
15
Heillbronn LK, Noatkes M, Clifton PM: The effect of high- and low-glycemic index energy restricted diets on plasma lipid and glucose profiles in type 2 diabetic subjects with varying glycemic control.
J Am Coll Nutr
 
21
:
120
–127,
2002
16
Bouche C, Rizkalla SW, Luo J, Vidal H, Veronese A, Pacer N, Fouquet C, Lang V, Slama G: Five-week, low-glycemic index diet decreases total fat mass and improves plasma lipid profile in moderately overweight non-diabetic men.
Diabetes Care
 
25
:
822
–828,
2002
17
Secretaria de Salud: Norma Oficial Mexicana para la prevención, tratamiento y control de la diabetes mellitas en la atención primaria a la salud.
NOM-015-SSA2–1994 (updated
2000
)
18
American Diabetes Association: Clinical practice recommendations 2002.
Diabetes Care
 
25 (Suppl. 1)
:
S1
–S147,
2002
19
Bacardi-Gascon M, Jimenez-Cruz A, Jones E: An evaluation of two Mexican food guides.
Int J Food Sci Nutr
 
53
:
163
–169,
2002
20
Jiménez-Cruz A, Bacardí M, Rosales-Garay P, Herrera J, Willis OW: Motivating behavioral change in a Mexican diabetic population: the Si Se Puede intervention program.
Rev Biomed.
In press
21
Wolever TM, Nguyen P, Chaisson J, Junt JA, Josse RG, Palmason C, Rodger NW, Ross SA, Ryan EA, Tan MH: Determinants of diet glycemic index calculated retrospectively from diet records of 342 individuals with non-insulin dependent diabetes mellitus.
Am J Clin Nutr
 
59
:
1265
–1269,
1994
22
Friedwald WT, Levy RI, Frederickson DS: Estimation of the concentration of low-density lipoprotein cholesterol in plasma without the use of preparative ultracentrifuge.
Clin Chem
 
13
:
499
–502,
1972
23
Gilbertson HR, Brand-Miller C, Thorburn AW, Evans S, Chondros P, Werther GA: The effect of flexible low glycemic index dietary advice versus measured carbohydrate exchange diets on glycemic control in children with type 1 diabetes.
Diabetes Care
 
24
:
1137
–1143,
2001
24
Jimenez Cruz A, Bacardi-Gascon M, Spindler A: Obesity and hunger among Mexican-Indian children on the US-Mexico border.
Int J Obesity
 
27
:
740
–747,
2003
25
Stubbs RJ, Mazlan N, Whybrow S: Carbohydrates, appetite, and feeding behavior in humans.
J Nutr
 
131
:
2775S
–2781S,
2001
26
Buyken AE, Toeller M, Heitkamp G, Karamanos B, Rottiers R, Muggeo M, Fuller JH, EURODIAB IDDM Complications Study Group: Glycemic index in the diet of European outpatients with type 1 diabetes: relations to glycated hemoglobin and serum lipids.
Am J Clin Nutr
 
73
:
574
–581,
2001
27
Gilbertson HR, Thorburn AW, Brand-Miller JC, Chondros P, Werther GA: Effect of low-glycemic-index dietary advice on dietary quality and food choice in children with type 1 diabetes.
Am J Clin Nutr
 
77
:
83
–90,
2003
28
Wolever TMS: American diabetes association evidence-based nutrition principles and recommendations are not based on evidence.
Diabetes Care
 
25
:
1263
–1264,
2002
29
Franz MJ: Response to Wolever (Letter).
Diabetes Care
 
25
:
1264
,
2002
30
Food and Agriculture Organization/World Health Organization:
Carbohydrates in Human Nutrition: Report of a Joint FAO/WHO Report.
Rome, FAO Food and Nutrition Paper 66,
1998
DIABETES CARE
2003