Effects of Exercise Training on Mitochondrial and Capillary Growth in Human Skeletal Muscle: A Systematic Review and Meta-Regression

Abstract

Background: Skeletal muscle mitochondria and capillaries are crucial for aerobic fitness, and suppressed levels are associated with chronic and age-related diseases. Currently, evidence-based exercise training recommendations to enhance these characteristics are limited. It is essential to explore how factors, such as fitness level, age, sex, and disease affect mitochondrial and capillary adaptations to different exercise stimuli.

Objectives: The main aim of this study was to compare the effects of low- or moderate intensity continuous endurance training (ET), high-intensity interval or continuous training (HIT), and sprint interval training (SIT) on changes in skeletal muscle mitochondrial content and capillarization. Secondarily, the effects on maximal oxygen consumption (VO₂max), muscle fiber cross-sectional area, and fiber type proportion were investigated.

Methods: A systematic literature search was conducted in PubMed, Web of Science, and SPORTDiscus databases, with no data restrictions, up to 2 February 2022. Exercise training intervention studies of ET, HIT, and SIT were included if they had baseline and follow-up measures of at least one marker of mitochondrial content or capillarization. In total, data from 5973 participants in 353 and 131 research articles were included for the mitochondrial and capillary quantitative synthesis of this review, respectively. Additionally, measures of VO₂max, muscle fiber cross-sectional area, and fiber type proportion were extracted from these studies.

Results: After adjusting for relevant covariates, such as training frequency, number of intervention weeks, and initial fitness level, percentage increases in mitochondrial content in response to exercise training increased to a similar extent with ET (23 ± 5%), HIT (27 ± 5%), and SIT (27 ± 7%) (P > 0.138), and were not influenced by age, sex, menopause, disease, or the amount of muscle mass engaged. Higher training frequencies (6 > 4 > 2 sessions/week) were associated with larger increases in mitochondrial content. Per total hour of exercise, SIT was ~ 2.3 times more efficient in increasing mitochondrial content than HIT and ~ 3.9 times more efficient than ET, while HIT was ~ 1.7 times more efficient than ET. Capillaries per fiber increased similarly with ET (15 ± 3%), HIT (13 ± 4%) and SIT (10 ± 11%) (P = 0.556) after adjustments for number of intervention weeks and initial fitness level. Capillaries per mm² only increased after ET (13 ± 3%) and HIT (7 ± 4%), with increases being larger after ET compared with HIT and SIT (P < 0.05). This difference coincided with increases in fiber cross-sectional area after ET (6.5 ± 3.5%), HIT (8.9 ± 4.9%), and SIT (11.9 ± 15.1%). Gains in capillarization occurred primarily in the early stages of training (< 4 weeks) and were only observed in untrained to moderately trained participants. The proportion of type I muscle fibers remained unaltered by exercise training (P > 0.116), but ET and SIT exhibited opposing effects (P = 0.041). VO₂max increased similarly with ET, HIT, and SIT, although HIT showed a tendency for greater improvement compared with both ET and SIT (P = 0.082), while SIT displayed the largest increase per hour of exercise. Higher training frequencies (6 > 4 > 2 sessions/week) were associated with larger increases in VO₂max. Women displayed greater percentage gains in VO₂max compared with men (P = 0.008). Generally, lower initial fitness levels were associated with greater percentage improvements in mitochondrial content, capillarization, and VO₂max. SIT was particularly effective in improving mitochondrial content and VO₂max in the early stages of training, while ET and HIT showed slower but steady improvements over a greater number of training weeks.

Conclusions: The magnitude of change in mitochondrial content, capillarization, and VO₂max to exercise training is largely determined by the initial fitness level, with greater changes observed in individuals with lower initial fitness. The ability to adapt to exercise training is maintained throughout life, irrespective of sex and presence of disease. While training load (volume × intensity) is a suitable predictor of changes in mitochondrial content and VO₂max, this relationship is less clear for capillary adaptations.

Fig. 1

Flowchart of research articles included in the systematic review and meta-regression (PRISMA diagram). PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses

Fig. 2

The change (Δ) in mitochondrial volume density (MvD), citrate synthase (CS), cytochrome c oxidase/complex IV (COX), succinate dehydrogenase (SDH), and hydroxyacyl-CoA dehydrogenase (HADH) from before to after training. In 2A, the estimated marginal means with 95% confidence limits from model 1 is presented, which included 353 studies, 506 training groups, 5650 participants, and 943 observations of mean changes (MvD: N = 61; CS: N = 416; COX: N = 177; SDH: N = 77; HADH: N = 212). In 2B, the individual raw values are presented divided into training intensity categories (ET, endurance training; HIT, high-intensity interval training; SIT, sprint interval training) and mitochondrial content markers

Fig. 3

The effects of intervention weeks by training intensity category (A), training frequency (B), initial fitness level (C), amount of recruited muscle mass during exercise (D), disease status (E), sex (F), and age (G) on training-induced changes (Δ) in mito_pooled (model 3). The arrows in panel (A) indicate significant increases between 2-6 weeks and 6-10 weeks of training. Mito_pooled, a pooled measure of the five mitochondrial content markers mitochondrial volume density, citrate synthase, succinate dehydrogenase, cytochrome c oxidase/complex IV and hydroxyacyl-CoA dehydrogenase; ET, endurance training; HIT, high-intensity interval training; SIT, sprint interval training. Values are estimated marginal means with 95% confidence limits [N studies = 343, N training groups = 496, N participants = 5511, N observations of mean changes = 928 (MvD: N = 60; CS: N = 409; SDH: N = 76; COX: N = 177; HADH: N = 206)]

Fig. 4

The interaction effect of initial fitness level and training intensity category on changes (Δ) in mito_pooled per training hour (model 6). Mito_pooled, a pooled measure of the five mitochondrial content markers mitochondrial volume density, citrate synthase, succinate dehydrogenase, cytochrome c oxidase/complex IV, and hydroxyacyl-CoA dehydrogenase. ET, endurance training; HIT, high-intensity interval training; SIT, sprint interval training. Values are estimated marginal means with 95% confidence limits. [N studies = 345, N training groups = 491, N participants = 5488, N observations = 918 (MvD: N = 59; CS: N = 404; SDH: N = 75; COX: N = 174; HADH: N = 206)]

Fig. 5

Change (Δ) in capillary-to-fiber ratio (model 7), capillary density (model 8), and muscle fiber cross-sectional area (model 9) in response to exercise training. In the upper panel, the effects of training intensity on training-induced changes in capillary-to-fiber ratio (A), capillary density (B), and muscle fiber cross-sectional area (C) are shown. In the second panel (D–F), the effects of initial fitness level are displayed, while in the third panel (G–I), the impact of the number of training intervention weeks on the same variables are presented. In J, the individual changes (Δ) in capillary-to-fiber ratio (C/F), capillary density (CD), muscle fiber cross-sectional area (CSA), and type I fiber distribution from before to after a period of ET, HIT and SIT. For the number of training groups and participants, see Table 2. Values are estimated marginal means with 95% confidence limits. Significant change (*P < 0.05) and tendency to change (^#P < 0.10) from baseline. ET, endurance training; HIT, high-intensity interval training; SIT, sprint interval training

Fig. 6

A Presents the average type I fiber distribution per training group before and after training, and B presents the mean change with 95% confidence limits (model 12). For the number of training groups and participants, see Table 2

Fig. 7

In A, the unadjusted individual raw changes in maximal oxygen consumption ($\dot{V}$O₂max) are presented divided into training intensity categories (ET, endurance training; HIT, high-intensity interval training; SIT, sprint interval training). The remaining figures show the effects of intervention weeks by training intensity category (B), training frequency (C), initial fitness level (D), disease status (E), sex (F), and age (G), on training-induced changes (Δ) in $\dot{V}$O₂max after adjusting for covariates (model 14). Values are estimated marginal means with 95% confidence limits (N studies = 202, N training groups = 298, N participants = 3524)

Fig. 8

The interaction effect of initial fitness level and training intensity category on changes (Δ) in $\dot{V}$O₂max per training hour (Fig. 8; model 17). Values are estimated marginal means with 95% confidence limits. (N studies = 202, N training groups = 295, N participants = 3487)

Fig. 9

Correlation plots between relative changes (Δ) after exercise training in maximal oxygen consumption ($\dot{V}$O₂max) and muscle mitochondrial (A, B) and capillary measures (C, D). Only training interventions using exercises recruiting large muscle mass were used. E shows correlation plot between relative changes (Δ) in citrate synthase content and capillaries per muscle fiber after exercise training (both exercises recruiting small and large amounts of muscle mass were used). COX, cytochrome c oxidase/complex IV; CS, citrate synthase; ET, endurance training; HADH, hydroxyacyl-CoA dehydrogenase; HIT, high-intensity interval training; MvD, mitochondrial volume density; SDH, succinate dehydrogenase/complex II; SIT, sprint interval training