Role of dietary carbohydrate quality and quantity in visceral obesity reduction : Results from a randomized controlled trial (CARBFUNC)

Abstract

Background: Increased visceral adipose tissue (VAT) volume is associated with common lifestyle diseases and poses an increasing health challenge worldwide. There is a great need for effective approaches that result in successful long-term VAT loss. Carbohydrate amount may be of importance as low-carbohydrate high-fat (LCHF) diets have been suggested to suppress increases in hunger feelings following diet-induced fat loss. Also, beyond established markers of carbohydrate quality, the food matrix and “cellularity” of carbohydrate-rich foods may influence both VAT volume and appetite.

Aims: The overall aim of this thesis was to investigate the effect of long-term dietary intervention approaches differing in carbohydrate quality (degree of cellularity) and amount on visceral adiposity and measures of appetite in adults with obesity.

Methods: All three papers included in this thesis are based on findings up to 12 months from a three-armed randomized clinical trial (RCT) of adults with obesity/central adiposity. In Paper I, we examined the habitual meal and dietary patterns, using exploratory hierarchical cluster and k-means cluster analyses, respectively, on cross-sectional preintervention dietary data using 6-day weighed food records. In Paper II, we compared changes in VAT volume, measured by abdominal computed tomography, after 1 year on three isocaloric eating patterns based on “acellular” carbohydrate sources (e.g., flour-based whole-grain products; comparator arm), “cellular” carbohydrate sources (minimally processed foods with intact cellular structures such as fruits, potatoes/tubers and rice), or LCHF principles. Finally, in Paper III we compared changes in fasting plasma concentrations of total ghrelin, the ketone body β-hydroxybutyrate (βHB) and subjective appetite (visual analog scale, VAS) after 3 and 12 months on the three diets. Outcomes were compared by an intention-to-treat (ITT) analysis using constrained linear mixed modeling (cLMM).

Results: In Paper I, we identified five clear temporal meal patterns among 192 (females: 53%, males: 47%) participants completing 99.1% of the planned preintervention dietary recording days. The reported energy intake (mean ± SD) was highest and lowest among “midnight-eaters” (2551 ± 554 kcal/d) and “dinner-eaters” (2064 ± 546 kcal/d), respectively. Despite a reported difference of 490 kcal/d, there were no significant differences between these meal patterns in anthropometric measures or physical activity levels. In total, 57 participants (30%) completed 12 months with similar intakes of energy (females: 1820−2060 kcal, males: 2480−2550 kcal) and protein (16−17 energy percent, E%) throughout the intervention, and only modest reductions in energy from baseline. Reported mean dietary intakes were 11−15 (59−86 g/d) and 41−44 E% carbohydrate and 66−70 and 36−38 E% fat on the LCHF and HCLF diets, respectively. In Paper II, there were no significant between-group differences in VAT volume after 12 months (cellular vs. acellular [95% CI]: −122 cm³ [−757, 514]; LCHF vs. acellular [95% CI]: −317 cm³ [−943, 309]). VAT volume decreased significantly within all groups by 14−18% after 12 months. In Paper III, no significant between-group differences were seen in change scores of neither fasting ghrelin nor hunger feelings, despite a significant difference in βHB after 3 months. Fasting concentrations of ghrelin increased significantly from baseline to 3 months (mean [95% CI]) on the acellular (46 pg/ml [11, 81]) and cellular (54 pg/ml [21, 88]) diets, but not on the LCHF diet (11 pg/ml [−16, 38]).

Conclusions: Our data revealed five clearly distinct meal patterns among people with obesity, which may be important to consider for more effective individual follow-up. Results from the intervention suggest that the previously observed appetite suppression of carbohydrate restriction on very low-energy diets may be of less clinical relevance on more modest energy-restricted diets unless carbohydrate restriction is sufficient to induce greater ketosis. Overall, our findings indicate that similar and clinically relevant VAT reduction may be achieved on isocaloric dietary approaches differing in carbohydrate cellularity and amount in people with intra-abdominal obesity.