Metabolic recovery from heavy exertion following banana compared to sugar beverage or water only ingestion: A randomized, crossover trial

Abstract

Objectives and methods: Using a randomized, crossover, counterbalanced approach, cyclists (N = 20, overnight fasted state) engaged in the four 75-km time trials (2-week washout) while ingesting two types of bananas with similar carbohydrate (CHO) but different phenolic content (Cavendish, CAV; mini-yellow, MIY, 63% higher polyphenols), a 6% sugar beverage (SUG), and water only (WAT). CHO intake was set at 0.2 g/kg every 15 minutes. Blood samples were collected pre-exercise and 0 h-, 0.75 h-,1.5 h-, 3 h-, 4.5 h-, 21 h-, 45 h-post-exercise.

Results: Each of the CHO trials (CAV, MIY, SUG) compared to water was associated with higher post-exercise plasma glucose and fructose, and lower leukocyte counts, plasma 9+13 HODES, and IL-6, IL-10, and IL-1ra. OPLS-DA analysis showed that metabolic perturbation (N = 1,605 metabolites) for WAT (86.8±4.0 arbitrary units) was significantly greater and sustained than for CAV (70.4±3.9, P = 0.006), MIY (68.3±4.0, P = 0.002), and SUG (68.1±4.2, P = 0.002). VIP ranking (<3.0, N = 25 metabolites) showed that both CAV and MIY were associated with significant fold changes in metabolites including those from amino acid and xenobiotics pathways. OPLS-DA analysis of immediate post-exercise metabolite shifts showed a significant separation of CAV and MIY from both WAT and SUG (R2Y = 0.848, Q2Y = 0.409). COX-2 mRNA expression was lower in both CAV and MIY, but not SUG, versus WAT at 21-h post-exercise in THP-1 monocytes cultured in plasma samples. Analysis of immediate post-exercise samples showed a decrease in LPS-stimulated THP-1 monocyte extracellular acidification rate (ECAR) in CAV and MIY, but not SUG, compared to WAT.

Conclusions: CHO ingestion from bananas or a sugar beverage had a comparable influence in attenuating metabolic perturbation and inflammation following 75-km cycling. Ex-vivo analysis with THP-1 monocytes supported a decrease in COX-2 mRNA expression and reduced reliance on glycolysis for ATP production following ingestion of bananas but not sugar water when compared to water alone.

Trial registration: ClinicalTrials.gov, U.S. National Institutes of Health, identifier: NCT02994628.

Fig 1. Subject flow diagram.

Fig 2. Changes in plasma glucose during 45-h recovery from 75-km cycling across four treatments (water only, Cavendish and mini-yellow bananas, and 6% sugar beverage) in 20 cyclists.

Interaction effect, P<0.001. * P<0.017 compared to the change from pre-exercise in the water condition.

Fig 3. Changes in total blood leukocytes during 45-h recovery from 75-km cycling across four treatments (water only, Cavendish and mini-yellow bananas, and 6% sugar beverage) in 20 cyclists.

Interaction effect, P<0.001. * P<0.017 compared to the change from pre-exercise in the water condition.

Fig 4. Changes in serum myoglobin during 45-h recovery from 75-km cycling across four treatments (water only, Cavendish and mini-yellow bananas, and 6% sugar beverage) in 20 cyclists.

Interaction effect, P = 0.002.

Fig 5. Changes in plasma 9+13 HODEs during 45-h recovery from 75-km cycling across four treatments (water only, Cavendish and mini-yellow bananas, and 6% sugar beverage) in 20 cyclists.

Interaction effect, P<0.001. * P<0.017 compared to the change from pre-exercise in the water condition.

Fig 6. Changes in plasma IL-6 during 45-h recovery from 75-km cycling across four treatments (water only, Cavendish and mini-yellow bananas, and 6% sugar beverage) in 20 cyclists.

Interaction effect, P<0.001. * P<0.017 compared to the change from pre-exercise in the water condition.

Fig 7. OPLS-DA and the overall pattern in the data.

The centroid was calculated for the samples in each treatment and time point combination. Water trial is in blue; 6% sugar beverage trial is in red; Cavendish bananas trial is in green; Mini-yellow banana trial is in yellow. This analysis passed permutation based validation, which indicates that the observed pattern did not occur by chance. However, the low Q2Y score (0.049) indicates poor prediction ability, especially for the time points that have similar metabolomics profiles, such as pre-exercise, 21-h and 45-h post-exercise. It is not a concern in this analysis, because complete separations of all of the treatment and time point combinations were not expected. The goal of this analysis was to examine the overall pattern in the data.

Fig 8. OPLS-DA for the separation between water trial (blue), 6% sugar beverage trial (red), Cavendish bananas trial (green), and mini-yellow banana trial (yellow).

R2Y = 0.848, Q2Y = 0.409. This analysis passed permutation based validation. The ratio between immediately post-exercise and pre-exercise for each metabolite in each subject was calculated and used as input data for this analysis. Metabolites important for the separation are listed in S5 Table.

Fig 9. COX-2 mRNA expression in THP-1 monocytes cultured for six hours in plasma obtained from study participants pre- and immediately-post-exercise, and 1.5 h- and 21-h post-exercise across four treatments (water only, Cavendish and mini-yellow bananas, and 6% sugar beverage).

Data are mean ± SE expressed as mRNA fold change relative to GAPDH. * P≤ 0.05 compared to the change from the water-only trial.

Fig 10

A) Changes in oxygen consumption rate (OCR) (basal state, after LPS stimulation, after injection of mitochondrial complex inhibitors) in THP1 monocytes immediately post 75-km cycling across four treatments (water only, Cavendish and mini-yellow bananas, and 6% sugar beverage) in cyclists. The bar chart (B) summarizes the LPS phase for OCR. C) Changes in extracellular acidification rate (ECAR) (basal state, after LPS stimulation, after injection of mitochondrial complex inhibitors) in THP1 monocytes immediately post 75-km cycling across four treatments (water only, Cavendish and mini-yellow bananas, and 6% sugar beverage) in cyclists. The bar chart (D) summarizes the LPS phase for ECAR. Data are mean ± SE. * Significantly different from the water-only trial, P≤0.05, respectively.