Quality and quantity of carbohydrates, faecal short-chain fatty acids and gastrointestinal symptoms – results from a randomised, controlled trial (CARBFUNC)

Abstract

Background & aims

Diet is a key determinant of gastrointestinal (GI) health, in part in association with microbiota-derived short-chain fatty acids (SCFAs). However, we need more knowledge of the relative impact of dietary carbohydrate amount and quality on GI symptoms, GI-associated quality of life (QoL) and faecal SCFAs.

Methods

193 males and females with obesity were randomly allocated to follow one of three isocaloric, iso-proteinic dietary patterns (2000 kcal/day for females, 2500 kcal/day for males): a higher-carbohydrate lower-fat diet with refined carbohydrate sources (acellular diet, A-HCLF, comparator arm), a higher-carbohydrate lower-fat diet with minimally refined carbohydrate sources (intact cellular structures, cellular diet, C-HCLF), or a low-carbohydrate high-fat diet (LCHF), all low in added sugars. Secondary outcomes of this randomised controlled trial (CARBFUNC) were assessed, i.e., changes in abdominal symptoms (irritable bowel syndrome severity scoring system (IBS-SSS)), reflux symptoms (gastro-oesophageal reflux disease questionnaire (GerdQ)), GI-related QoL (Short-Form Nepean Dyspepsia Index (SF-NDI)) and fatigue (Fatigue Impact Scale), and faecal SCFAs concentrations after following the diets for 3 and 12 months. Group differences were analysed by constrained linear mixed effect modelling (cLMM).

Results

118 and 57 participants completed 3 and 12 months of the dietary intervention, respectively, with no significant group differences in weight loss at 12 months (5–7%). At 12 months, the mean daily fibre intake was 34 ± 7 g/day, 41 ± 14.3 g/day, and 18.5 ± 6 g/day on the A-HCLF, C-HCLF and LCHF diet, respectively, compared to 21 ± 7, 21 ± 7 and 20 ± 6 g/day at baseline. We observed no significant between-group difference in IBS-SSS sum score after 3 and 12 months. We found significant improvement in GerdQ score (change score [95 % CI]: −0.62 [−1.18, −0.048], p = 0.034), and SF-NDI sum score (−1.88 [−3.22, −0.52], p = 0.007) after 3 months on the LCHF diet compared to the A-HCLF diet, and GerdQ remained significant at 12 months (−1.03 [−1.88, −0.19], p = 0.017). Compared to the A-HCLF diet, the concentration of faecal butyric acid increased significantly more after 3 months on the C-HCLF diet (4.97 [1.71, 8.23] p = 0.003) and faecal acetic acid decreased more (−6.41 [−12.8, −0.047]. p = 0.048) on the LCHF diet. At 12 months the greater reduction in faecal acetic acid on the LCHF diet remained significant (−9.82 [−19.0, −0.67], p = 0.036), along with significantly greater reductions also in total SCFAs (−21.3 [−38.0, −4.56], p = 0.013), propionic (−4.42 [−7.79, −1.05], p = 0.010), and butyric acid (−5.05 [−9.60, −0.51], p = 0.030).

Conclusion

In this sample of adults with obesity and mild GI symptoms at baseline, modest improvements were observed only for the LCHF diet in QoL (at 3 months) and reflux symptoms (at 3 and 12 months), which was significantly different from the acellular carbohydrate diet, and independent of total fibre intake. Concomitantly, compared to the acellular carbohydrate diet, the cellular diet significantly increased the faecal concentration of butyric acid, whereas the LCHF diet lowered acetic acid after 3 months and all the major SCFAs after 12 months.

Clinical trials identifier

NCT03401970. https://clinicaltrials.gov/ct2/show/NCT03401970.