Adaptation to a low carbohydrate high fat diet is rapid but impairs endurance exercise metabolism and performance despite enhanced glycogen availability

Abstract

Key points: Brief (5-6 days) adaptation to a low carbohydrate high fat diet in elite athletes increased exercise fat oxidation to rates previously observed with medium (3-4 weeks) or chronic (>12 months) adherence to this diet, with metabolic changes being washed out in a similar time frame. Increased fat utilisation during exercise was associated with a 5-8% increase in oxygen cost at speeds related to Olympic Programme races. Acute restoration of endogenous carbohydrate (CHO) availability (24 h high CHO diet, pre-race CHO) only partially restored substrate utilisation during a race warm-up. Fat oxidation continued to be elevated above baseline values although it was lower than achieved by 5-6 days' keto adaptation; CHO oxidation only reached 61% and 78% of values previously seen at exercise intensities related to race events. Acute restoration of CHO availability failed to overturn the impairment of high-intensity endurance performance previously associated with low carbohydrate high fat adaptation, potentially due to the blunted capacity for CHO oxidation.

Abstract: We investigated substrate utilisation during exercise after brief (5-6 days) adaptation to a ketogenic low-carbohydrate (CHO), high-fat (LCHF) diet and similar washout period. Thirteen world-class male race walkers completed economy testing, 25 km training and a 10,000 m race (Baseline), with high CHO availability (HCHO), repeating this (Adaptation) after 5-6 days' LCHF (n = 7; CHO: <50 g day^-1 , protein: 2.2 g kg^-1 day^-1 ; 80% fat) or HCHO (n = 6; CHO: 9.7 g kg^-1 day^-1 ; protein: 2.2 g kg^-1 day^-1 ) diet. An Adaptation race was undertaken after 24 h HCHO and pre-race CHO (2 g kg^-1 ) diet, identical to the Baseline race. Substantial (>200%) increases in exercise fat oxidation occurred in the LCHF Adaptation economy and 25 km tests, reaching mean rates of ∼1.43 g min^-1 . However, relative ${\dot{V}}_{O_2}$ (ml min^-1 kg^-1 ) was higher (P < 0.0001), by ∼8% and 5% at speeds related to 50 km and 20 km events. During Adaptation race warm-up in the LCHF group, rates of fat and CHO oxidation at these speeds were decreased and increased, respectively (P < 0.001), compared with the previous day, but were not restored to Baseline values. Performance changes differed between groups (P = 0.009), with all HCHO athletes improving in the Adaptation race (5.7 (5.6)%), while 6/7 LCHF athletes were slower (2.2 (3.4)%). Substrate utilisation returned to Baseline values after 5-6 days of HCHO diet. In summary, robust changes in exercise substrate use occurred in 5-6 days of extreme changes in CHO intake. However, adaptation to a LCHF diet plus acute restoration of endogenous CHO availability failed to restore high-intensity endurance performance, with CHO oxidation rates remaining blunted.

Keywords: athletic performance; ketogenic diet; sports nutrition.

Figure 1. Overview of 3 week training camp, testing blocks and dietary intervention

Figure 2. Oxygen uptake (A) and substrate utilisation (carbohydrate (CHO) (B) and fat oxidation (C)) during four‐stage economy test performed prior to and following adaptation to either a low CHO high fat (LCHF, n = 7) or high CHO (HCHO, n = 6) diet

Data are means ± SD. Significant differences within group relative to Baseline: ^* P < 0.05, ^*** P < 0.0001.

Figure 3. Blood metabolite concentrations

Blood glucose (mmol L⁻¹) (A), blood lactate (mmol L⁻¹) (B) and blood β‐hydroxybutyrate (mmol L⁻¹) (C) at rest, during four‐stage graded economy test and maximal aerobic capacity test (Max) prior to and following adaptation to either a low CHO high fat (LCHF, n = 6) or high CHO (HCHO, n = 6) diet. Data are means ± SD. Significant differences within group relative to Baseline: ^* P < 0.05, ^*** P < 0.0001.

Figure 4. Oxygen uptake (A and D) and rates of carbohydrate (CHO; B and E) and fat (C and E) oxidation during standardised 25 km long walk performed at baseline (Walk 1), following adaptation to either a low CHO high fat (LCHF, n = 7) or high carbohydrate (HCHO, n = 6) diet (Walk 2) and restoration where all athletes consumed a HCHO diet (Walk 3)

Data are mean ± SD. Significant change over the 25 km walking session (1 km < 25 km): †P < 0.01. Significantly different than Walk 1 and Walk 3: ^* P < 0.05, ^*** P < 0.0001.

Figure 5. Blood metabolite concentrations

Blood glucose (mmol L⁻¹) (A), blood lactate (mmol L⁻¹) (B), blood β‐hydroxybutyrate (mmol L⁻¹) (C) and serum free fatty acids (mmol L⁻¹) (D) at rest (0 km as well as fasted for FFA), and throughout standardized 25 km‐long walk performed at baseline (Walk 1), following adaptation to either a low CHO high fat (LCHF, n = 7) or high carbohydrate (HCHO, n = 6) diet (Walk 2) and restoration where all athletes consumed a HCHO diet (Walk 3). Data are means ± SD. Significant change over the 25 km walking session: ††P < 0.001, †††P < 0.0001. Significant differences between Walk 1 and Walk 3: ‡P < 0.05. Significant differences between Walks 1 and 3 compared to Walk 2: ^** P < 0.001, ^*** P < 0.0001.

Figure 6. Substrate utilisation (carbohydrate (CHO) (A and C) and fat oxidation (B and D)) at stages representing 50 km (Stage 2) and 20 km (Stage 4) race speeds during economy tests

Tests were performed prior to and following adaptation a low CHO high fat (LCHF, n = 7) diet as well as immediately prior to both Race 1 and Race 2. Both pre‐race tests were performed with a 1‐day HCHO diet to restore skeletal muscle glycogen content. Data are means ± SD, with results of individual athletes being shown by the circles (Baseline) and squares (Adaptation). Bars sharing a letter are not statistically different from one another (P < 0.05).

Figure 7. Relative maximum oxygen uptake (A), 10,000 m race performance (B) and performance changes between the two races (C) prior to and following adaptation to either a low CHO high fat (LCHF, n = 7) or high CHO (HCHO, n = 6) diet

Data are means ± SD with results of individual being shown by the circles (Baseline) and squares (Adaptation). Significant differences within group relative to Baseline: ^* P < 0.05. Significant difference in performance changes between groups: †P = 0.009.