Sugar has been suggested as a central risk factor in the development of noncommunicable diseases.
We assessed the evidence of the effects of free sugars compared with complex carbohydrates on selected cardiovascular disease risk factors.
DWe conducted a systematic review and meta-analysis of intervention trials to compare diets that provide a given amount of energy from free sugars with a control diet that provides the same amount of energy from complex carbohydrates. The primary outcomes were: blood pressure, total cholesterol, low–density lipoprotein (LDL) cholesterol, high–density lipoprotein (HDL) cholesterol, triacylglycerols, apolipoproteins A-I and B, or very low–density lipoprotein cholesterol. Body weight was also recorded but was not a primary outcome of the studies.
In all, 28 studies involving 510 volunteers were included. When free sugars were substituted for complex carbohydrates, no significant increases were detected in systolic or diastolic blood pressure, and no heterogeneity was observed. There were significant increases in HDL cholesterol, LDL cholesterol, and triacylglycerols, although for LDL cholesterol and triacylglycerols there was significant heterogeneity between studies and evidence of publication bias. After adjustment for missing studies, these increases lost significance. Subgroup analyses showed that diets providing the largest total energy intake and energy exchange enhanced the effect of free sugars on total and LDL cholesterol and triacylglycerols. The increase of triacylglycerols was no longer significant when studies with the highest risk of bias were excluded or when only randomized trials were considered. Free sugars had no effect on body weight.
In short- or moderate-term isoenergetic intervention trials, the substitution of free sugars for complex carbohydrates had no effect on blood pressure or body weight and an unclear effect on blood lipid profile. Further independent trials are required to assess whether the reduction of free sugars improves cardiovascular disease risk factors. This review was registered at http://www.crd.york.ac.uk/prospero as CRD42016042930.