- Click here for explanation of classification scheme.

To compare research results about high- and low-glycemic-index diets in the management of type 1 and type 2 diabetes.

The studies included in the meta-analysis met these criteria:

• published in English between 1981 and 2001
• randomized cross-over or parallel experimental design
• at least 12 days duration
• subjects had type 1 or type 2 diabetes
• HbA1c or fructosamine used as outcome measures of glycemic control
• modification of at least two meals per day

 

 

Excluded if not included above.

Recruitment:  Studies were identified by Medline and internet searches using the key words "glycemic" and "diabetes."

Design:  Meta-Analysis   

Blinding used (if applicable): not applicable

Intervention (if applicable):  not applicable 

Statistical Analysis: The steps used to execute the meta-analysis were drawn from Petitti, using a fixed-effects model.  

• The variance for each study was calculated under these assumptions:
  • the two values for each person were independent
  • the two values for each person were dependent with a correlation of 0.34
• The studies were weighted by the reciprocal of the variance.  The overall effect estimate and 95% CIs were then computed using these weights.
• The hypothesis of homogeneity of effect was tested, and where it was rejected, a sensitivity analysis was conducted by removing individual studies one by one.

 

 

Timing of Measurements  not applicable

Dependent Variables

• HbA1c
• fructosamine

Independent Variables

• the average GI of low-GI and high GI diets was 83 and 65 respectively, on the bread scale
• study duration ranged from 12 days to 12 months

Control Variables

• baseline values of HbA1c and fructosamine
• length of study

 

Initial N: 14 studies identified; number of subjects in each ranged from 6 to 104

Attrition (final N): not applicable

Age:  not given

Ethnicity: not given

Other relevant demographics: 356 subjects had type 1 diabetes and 153 subjects had type 2 diabetes

Anthropometrics: not given

Location: not applicable

 

 Mean difference and CI between low-GI and high-GI groups

Variables	Difference in endpoints between low-GI and high-GI groups	Difference in endpoints between low-GI and high-GI groups, controlling for baseline values
HbA1c, assuming dependence; change inabsolute percentage points	-0.40, -0.66 to -0.14	-0.33, -0.59 to -0.07
HbA1c, assuming independence; change inabsolute percentage points	-0.43, -0.72 to-0.13	-0.34, -0.64 to -0.05
Fructosamine, assuming dependence; change in mmol/l	-0.17, -0.30 to -0.04	-0.19, -0.32 to -0.06
Fructosamine, assuming independence; change in mmol/l	-0.18, -0.33 to -0.02	-0.2, -0.35 to -0.04
Combined HbA1c and fructosamine data, assuming dependence; [value for low-GI -value for high-GI) x 100/value for high-GI]	-6.8 %, -8.1 to - 5.5	-7.3%, -8.6 to -6.0
Combined HbA1c and fructosamine data, assuming dependence; value for low-GI -value for high-GI) x 100/value for high-GI]	-6.8%, -8.2 to-5.4	-7.4%, -8.8 to -6.0

Other Findings

Mean differences in outcomes between subjects with  type 1 and type 2 diabetes, using combined analysis of HbA1c and fructosamine:

• assuming dependence:  -10.6%,(-12.7 to -8.6)
• assuming independence:  -6.1% (-7.8 to -4.3)

 

The meta-analysis provides objective evidence that targeting postprandial hyperclycemia via choice of low-GI foods has a small but clinically useful effect on medium-term glycemic control in diabetes. 

University/Hospital:

University of Sydney, University of technology, Prince of Wales Hospital (all Australia)

Many of the studies included in the meta-analysis involved only small numbers of subjects and were of short duration.  Of the 14 studies identified, 6 had fewer than 10 subjects, and all but 2 studies were 12 weeks or less.