Carbohydrate Intake Affects Weight Loss Related Improvements in Glycemic Control: Results from a Systematic Review and Regression Analysis

Abstract

ObjectiveWeight loss is shown to improve glucose control and extensive research exists on the effect of carbohydrate (CHO) intake on weight loss. However, whether or not diets varying in CHO content elicit superior improvements in glycemic control when the degree of weight loss is comparable has yet to be systematically investigated.MethodsThis systematic review and regression analysis used randomized controlled trials to assess the effects of different quantities of CHO intake during energy restriction‐induced weight loss or weight maintenance on glycemic control among overweight and obese adults. Carbohydrate intake was categorized as normal (≥45 to ≤65% of total energy intake), low (≥20 to <45% of total energy intake), or restricted (utilization of a ketogenic diet, initial CHO intake <50 g/d). Two researchers independently screened 1882 articles from PubMed, CINAHL, Cochrane, and Scopus databases and 44 articles (with 98 intervention groups) were included in this review. Data extracted from each intervention group included the mean percentage change of body weight and mean change of serum or plasma fasting glucose (mg/dL) and insulin (μU/mL) concentrations. A mixed model ANCOVA (SAS, version 9.3) was performed on changes in glucose or insulin separately, with percent body weight change, CHO intake level (normal, low, or restricted), and their interactions as primary predictors of interest. Other information such as mean baseline body mass index, age, percentage of female subjects, continent of study origin, study duration, and dietary fat and protein intakes were also extracted and included in the model unless multicollinearity existed among the predictors.ResultsThe mean (and range) body weight changes was −5.5% (−13 to 0%), −5.7% (−13.6 to 0.3%), and −8.5% (−12.3 to −5.9%) for the normal CHO (n=48 groups), low CHO (n=40 groups), and restricted CHO (n=10 groups) diet groups, respectively. ANCOVA analyses were only performed using results from the normal and low CHO diets. This was necessary because the range of weight change for the CHO restricted diet groups was less than half of those of the other two groups which would force extrapolation of the regression results. Among all normal and low CHO diet groups, greater energy restriction‐induced weight loss was associated with greater reductions in fasting glucose (β=0.42 mg/dL, p=0.002). Each 1% weight loss predicted a 0.42 mg/dL reduction in glucose. This association was independent of normal versus low CHO intake. The impact of weight loss on changes in insulin was dependent on dietary CHO intake. Each 1% weight loss predicted 0.21 and 0.43 μU/mL reduction of insulin in the normal and low CHO groups, respectively (p=0.04), a difference of 0.22 μU/mL.ConclusionThese findings support that for a given percentage weight loss, consuming low CHO diets predicts a greater reduction in fasting insulin concentration, consistent with improved insulin‐mediated glycemic control.Support or Funding InformationUSDA‐NIFA #2011‐38420‐20038 (WWC)