The Effect of Fructose‐Containing Sugars on Glycemic Control: A Systematic Review and Meta‐Analysis of Controlled Trials
Conferences
Overview
Research
Identity
Additional Document Info
View All
Overview
abstract
BACKGROUNDFructose‐containing sugars have become a focus of widespread concern regarding their contribution to the development of diabetes.AIMSTo assess the effect of fructose‐containing sugars on glycemic control, we conducted a systematic review and meta‐analysis of controlled trials.DATA SOURCESWe searched MEDLINE, EMBASE, CINAHL and the Cochrane Library (through October 3, 2015).DATA EXTRACTIONWe included controlled trials ≥7 days. Two independent reviewers extracted relevant study data and assessed risk of bias (Cochrane Risk of Bias Tool). Glycemic control outcomes included glycated blood proteins (HbA1c, fructosamine, glycated albumin), fasting glucose, and fasting insulin. Data were pooled using the generic inverse variance method and expressed as mean differences (MDs) or standardized mean differences (SMDs) with 95% confidence intervals. Heterogeneity was assessed (Cochran Q statistic) and quantified (I2 statistic).RESULTSEligibility criteria were met by 126 trials (N=2186). We identified 3 types of trials: 98 substitution trials (N=1625), in which fructose containing sugars were compared with other macronutrients matched for energy; 23 addition trials (N=492), in which fructose‐containing sugars supplemented diets with excess calories compared to the same diets alone without the excess calories; and 6 ad libitum trials (N=69), in which fructose‐containing sugars freely replaced other macronutrients without any strict control of calories. Fructose‐containing sugars improved glycated blood proteins compared with other macronutrients in substitution trials (SMD −0.15 [95%CI, −0.29 to −0.01]). Provision of excess calories from fructose‐containing sugars, however, increased fasting insulin (MD 6.14 [95% CI, 1.30 to 10.97]) in addition trials. None of the other analyses were significant.LIMITATIONSThere was no serious risk of bias, but most trials were small, short and several analyses were complicated by significant unexplained heterogeneity. Very few trials compared fructose‐containing sugars with macronutrients other than starch.CONCLUSIONSPooled analyses suggest that fructose‐containing sugars are no worse in their effects on glycemic control than other macronutrients (mainly starch) in energy matched comparisons. Fructose‐containing sugars supplementing diets with excess energy, however, show an increasing‐effect on insulin, an effect which appears more attributable to the excess energy than the sugars. Larger, longer, higher quality trials with a range of macro‐nutrient comparators are required.Support or Funding InformationCanadian Diabetes Association, PSI Foundation, CIHR Banting and Best Graduate Student Award, Ontario Graduate Scholarship, Banting and Best Diabetes Centre Novo Nordisk Studentship
