Table 1.
Studies on exercise and gut microbiota conducted in athletes, physically active individuals and sedentary population.
| Subjects | Training Regimen, Exercise Protocol | Dietary Intake | Main Results | Reference |
|---|---|---|---|---|
| Athletes: | ||||
| Rugby players vs. BMI-matched sedentary controls n = 86, males Age 29 ± 4 y |
Habitual training and exercise | Self-reported intake by FFQ In athletes, higher total energy, macronutrient and fiber intake. Protein intake 22 E% in athletes, 16 E% in low-BMI and 15 E% in high-BMI controls |
In athletes, higher α-diversity and Akkermansia spp. abundance vs. sedentary controls. Protein intake was positively correlated with microbial diversity. | [12] |
| Rugby players vs. BMI-matched sedentary controls n = 86, males Age 29 ± 4 y |
Habitual training and exercise | Self-reported intake by FFQ In athletes, higher total energy, macronutrient and fiber intake. Protein 22 E% in athletes vs. 16 E% in low-BMI and 15 E% in high-BMI controls |
In athletes, fecal SCFAs, microbial pathways for antibiotic biosynthesis, and amino acids and carbohydrate metabolism were increased. | [30] |
| Professional cyclists vs. amateur cyclists n = 33 (22/M, 11/F) Age 19–49 y |
Habitual training | Dietary intake data collected by questionnaire, reported and analyzed as overall dietary patterns. | Prevotella spp. abundance was positively correlated with the amount of exercise and branched chain amino acid and carbohydrate metabolism pathways. Professional cyclists had increased Methanobrevibacter smithii transcripts and upregulated genes involved in the production of methane compared with amateur cyclists. No correlations between overall diet and gut microbiota clusters. | [13] |
| Cross-country runners n = 18, males Age: Control group 35.4 ± 9.0 y Protein group 34.9 ± 9.5 y |
Habitual endurance training | Habitual diet by FFQ No differences in habitual dietary intake within or between groups, at baseline or after the intervention. Dietary intervention: habitual diet and whey isolate (10 g) + beef hydrolysate (10 g) or maltodextrin (control) for 10 weeks |
After the intervention, higher Bacteroidetes and lower Firmicutes abundance in the protein group. Bifidobacterium longum was reduced after intervention in the protein group. No changes in microbiota composition in the control group, from pre- to post-intervention. No differences within or between groups in fecal SCFA, before or after the intervention. | [32] |
| Bodybuilders, long-distance runners vs. sedentary subjects n = 45, males Age: Bodybuilders 25 ± 3 y, distance runners 20 ± 1 y, sedentary 26 ± 2 y |
Habitual training and exercise | Self-recorded 3-day food diary Bodybuilders had a high-protein and distance runners had a low-dietary-fiber dietary pattern. Dietary fiber intake was below recommendation in all groups. |
Compositional differences in bodybuilders and runners associated with exercise type and diet. No difference in microbial diversity between groups. In distance runners, protein intake was negatively correlated with microbial diversity. | [33] |
| Highly trained ultra-endurance rowers n = 4, males Age 26.5 ± 1.3 y |
ca. 5000 km rowing race over 34 days | Self-reported intake (FFQ), detailed daily record pre-race and during the race No fresh produce consumed during race. Pre-race fiber intake: 21.45 g/day, intra-race 23.1 g/day. Only small changes in intra-race macronutrient intake compared with pre-race |
After the race, increased diversity and butyrate-producing species including Roseburia hominis and changes in microbial composition were observed. | [34] |
| Elite race walkers n = 21, males Age 20–35 y |
3-week structured program of intensified training | Dietary intervention for 3 weeks with planned and individualized menus. Subjects allocated into High-carbohydrate diet (HCHO) Periodized-carbohydrate diet (PCHO), or Low-carbohydrate, high-fat diet (LCHF) (ketogenic) group |
At baseline, microbiota profiles could be separated into Prevotella- or Bacteroides-dominating enterotypes. HCHO and PCHO resulted in minor changes, whereas LCHF resulted in stronger changes in microbial composition. LCHF was associated with reduced Faecalibacterium, Bifidobacterium, and Veillonella spp. Increased Bacteroides and Dorea spp. in the LCHF group was associated with decreased performance. | [35] |
| Marathon runners: n = 15 (4/M, 11/F) Mean age 27.1 y; Non-runners: n = 11 (5/M, 6/F) Mean age 29.2 y; Ultramarathon and rower athletes: n = 11 (5/M, 6/F) Age not reported |
Habitual training and a marathon Type of exercise not reported for the cohort of ultra-marathon and rower athletes |
Dietary intake data collected by questionnaire | In marathon runners, the relative abundance of Veillonella spp. increased post-marathon. In ultramarathon and rower athletes, the relative abundance of the methylmalonyl-CoA pathway (degrading lactate into propionate) in the gut microbiome increased post-exercise. No correlations between dairy, protein, grains, fruits, or vegetables and Veillonella spp. abundance was observed among marathon runners. | [14] |
| Non-athletes and sedentary subjects: | ||||
| Healthy subjects n = 39 (22/M, 17/F) Age 18–35 y |
VO2Peak test to assess CRF and to allocate subjects into groups (low, average, and high CRF) | 24-h dietary recall interview No significant differences in dietary intake between groups. |
CRF correlated with microbial diversity and butyrate production. | [36] |
| Active vs. sedentary women n = 40 Active: 30.7 ± 5.9 y, BMI 24.4 ± 4.5 kg/m2; Sedentary: 32.2 ± 8.7 y, BMI 22.9 ± 3.0 kg/m2 |
Habitual physical activity measured by accelerometer. | Self-reported food intake (FFQ) Fiber, fruit, and vegetable intake significantly higher in the active group. |
Higher abundance of Faecalibacterium prausnitzii, Roseburia hominis and Akkermansia muciniphila in active women. Physical activity was not associated with differences in microbiota richness. | [37] |
| Lean and obese sedentary subjects n = 32 Lean: n = 18 (9/M, 9/F), mean age 25.10 y; Obese: n = 14 (3/M, 11/F), mean age 31.14 y |
Exercise intervention study: 6 weeks of moderate-to-vigorous intensity aerobic exercise and 6 weeks without exercise | Maintenance of habitual diet during the intervention. A designed 3-day food menu, based on previous reported habitual diet, before fecal sample collection. | At baseline, the composition of gut microbiota differed between lean and obese subjects, but after exercise training, no difference was observed between lean and obese subjects. Exercise increased fecal SCFA and SCFA producing bacteria in lean subjects. |
[15] |
| Children and teenagers n = 267 (178/M, 89/F) Age 7–18 y |
Self-reported physical activity | Type of diet reported as omnivore or vegetarian. | Gut microbiota composition was affected by BMI, exercise frequency, and diet type. Firmicutes were significantly enriched in subjects with more frequent exercise. | [38] |
| Overweight sedentary women n = 17 Age 36.8 ± 3.9 y BMI 31.8 ± 4.4 kg/m2 |
Habitual physical activity. Exercise intervention study: 6-week control period without exercise, 6-week programmed endurance exercise, on a bicycle ergometer |
Habitual diet Self-reported 3-day food record No changes in intake of total energy, macronutrients or fiber from baseline, after control or exercise period. A modest increase in energy from starch |
Exercise did not affect α-diversity. Exercise increased Akkermansia spp. and reduced Proteobacteria abundance. No significant changes in BMI or total fat mass after exercise. Significant reduction in android fat mass. | [16] |
| Healthy subjects n = 37 (20/M, 17/F) Age 25.7 ± 2.2 y |
VO2max test to assess CRF | Habitual diet recorded for 7 days | CRF correlated with Firmicutes/Bacteroidetes ratio. No correlation between dietary factors or BMI and Firmicutes/Bacteroidetes ratio. | [39] |
| Elderly community-dwelling men n = 373 Age 78–98 y |
Habitual physical activity, measured by activity sensor, for 5 days. Step count as primary physical activity variable | Self-reported food intake (FFQ) Step count was not associated with food or alcohol intake. |
Physical activity was not associated with α-diversity but was positively associated with β-diversity. Increased physical activity was associated with greater Faecalibacterium and Lachnospira spp. prevalence. | [40] |
| Elderly sedentary women n = 29 Age 65–77 y |
Exercise intervention study: resistance training (trunk muscles) or aerobic exercise (brisk walking) for 12 weeks | Self-reported food intake (FFQ) No changes in energy or nutrient intake after interventions. |
Brisk walking increased the relative abundance of Bacteroides spp. Bacteroides spp. abundance was positively associated with improved CRF after aerobic training but not with improved CRF after resistance training. | [17] |
BMI, body mass index; y, years; FFQ, food frequency questionnaire; E%, percentage of total energy intake; SCFA, short-chain fatty acid; M, males; F, females; VO2Peak/VO2Max, maximum rate of oxygen consumption; CRF, cardiorespiratory fitness.