Will cardio hurt your strength gains? It depends on your training level!

What are the effects of concurrent resistance and endurance training, compared to resistance training alone, on the development of maximal strength in untrained, moderately trained, and trained individuals?

Endurance training and resistance training promote distinct and, to a degree, opposing adaptations. Specifically, endurance training activates molecular pathways^[1] that promote increases in mitochondrial mass and the formation of new blood vessels, thus stimulating metabolic adaptations^[2] that lead to improved endurance capacity. In contrast, resistance training mainly activates the protein-kinase B mammalian target of rapamycin (Akt-mTOR^[3]) pathway, stimulating protein synthesis in muscle fibers, and promoting increases in muscle strength, power, and hypertrophy. Performing resistance- and endurance-based exercises within the same training plan (in the same or separate training sessions), is known as concurrent training^[4].

Although trainees often perform concurrent resistance and endurance training in order to stimulate the adaptations of both training modes, concurrent training may result in an interference effect, which negatively affects the gains in muscular strength. However, while some^[5] studies^[6] give support to the notion of the interference effect, other^[7] studies^[8] do not.

The training status of the participants is one potential explanation for these divergent findings. It is also one of the few putative confounders that has not been examined in previous meta-analyses. As such, the primary aim of the meta-analysis under review was to fill this research gap. The secondary aim was to determine whether the length of the recovery period between the endurance and resistance training sessions can influence the potential interference effect.

Although some studies suggest that concurrent resistance and endurance training may produce inferior strength gains compared to resistance training only, other studies have reported conflicting results. While these discrepancies may relate to differences between studies in the training status of the participants, the meta-analysis at hand is the first to examine this possibility.

The authors conducted meta-analyses of 27 randomized and non-randomized controlled trials that looked at the effects of lower-body concurrent resistance and endurance training, compared to resistance training only, on lower body maximal strength (1RM squat or leg press) in healthy untrained (7 trials), moderately trained (10 trials), and trained (10 trials) adults 18–40 years old.

In each study, the participants in the intervention and control groups were at an equal baseline training status. The number of participants in the studies ranged between 4 and 16, with a total of 750 participants (523 men and 227 women), though all trained participants were men. The researchers used the following criteria to classify participants as untrained, moderately trained, or trained:

• Untrained: participants classified as untrained or sedentary by the author, or who reported no involvement in regular physical activity for at least three months before the trial.
• Moderately trained: participants classified as recreationally or physically active by the author, but not involved in a regular structured training program for at least three months before the trial.
• Trained: participants classified as athletes by the authors, or who participated in regular structured training programs for at least three months before the trial.

The endurance sessions consisted of running or cycling at an intensity of at least 70% of maximal heart rate, and included both steady-state exercise and high-intensity interval training (HIIT). The frequency of endurance training ranged between two and six sessions per week, with most studies using training frequencies of two to three sessions per week. The resistance sessions included the exercises used for the main outcome. Unlike for participants in trials using intensities of more than 60% of a participant’s 1RM, failure was obligatory for participants in trials prescribing RT with less than 60% 1RM. The frequency of strength training ranged between two and five sessions per week. The length of the studies ranged from 6 to 21 weeks, with an average length of 11 weeks.

The order of training for during concurrent training wasn’t completely clear, since many of the included studies didn’t report the order. In the studies that did report the order, endurance training was usually done before resistance training when the participants were either untrained or moderately trained, while trained participants most often completed resistance training first.

The researchers performed subgroup analyses based on whether the resistance and endurance training activities were performed during the same session (less than 20 minutes between activities) or on separate sessions (more than 2 hours between activities).

The researchers used the Physiotherapy Evidence Database (PEDro^[9]) scale to assess the methodological quality of individual studies. Only moderate- and high-quality studies (PEDro scores of 5–10) were included in the meta-analysis.

To analyze the data, a random-effects model was used to pool the weighted estimation of standardized mean differences (SMD) across studies. The SMD is used to determine the size of the effect, with 0.2 defined as small, 0.5 as moderate, and 0.8 as large. Heterogeneity was tested using the I² statistic. The meta-analysis followed PRISMA guidelines, but was not preregistered.

This meta-analysis compared the effects of concurrent training to resistance training, alone, on maximal strength in the squat or leg press in healthy untrained, moderately trained, and trained young adults. In addition to the main analyses, the researchers also conducted subgroup analyses based on whether the resistance and endurance training activities were performed in the same session (less than 20 minutes between activities) or in separate sessions (more than 2 hours between activities).

In the main analyses, concurrent training did not affect maximal lower body strength in untrained individuals. However, it resulted in small (SMD: -0.2), but statistically non-significant (p=0.08), reductions in maximal lower body strength in moderately trained individuals, and in statistically significant (p<0.01) small–moderate (SMD: -0.35) reductions in maximal lower body strength in trained participants. These results are depicted in Figure 1.

Figure 1: Main analyses (with 95% confidence intervals)

According to the subgroup analyses, the negative effects observed with concurrent training in trained participants occurred when resistance and endurance training were performed within the same training session (SMD: -0.66), but not when performed in separate training sessions. No heterogeneity was detected in any of the analyses.

According to the study quality assessment, 14 trials were assessed as being of moderate quality, and 13 trials were assessed as being of high quality.

Concurrent training resulted in small, statistically non-significant reductions in maximal lower body strength in moderately trained participants, and in statistically significant small to moderate reductions in maximal lower body strength in trained participants. The negative effects observed with concurrent training in trained participants occurred only when resistance and endurance training were performed within the same training session.

The study under review is the first meta-analysis to investigate whether training status can influence the development of maximal dynamic strength during concurrent training. The results of the main analyses suggest that concurrent training can negatively affect maximal lower body strength in trained individuals, with potential smaller effects in moderately trained individuals, and no effects in untrained individuals.

One potential explanation^[10] for these results is that in untrained individuals, any overload stimulus from exercise (resistance- or endurance-based) may be large enough to stimulate training adaptations. In fact, it’s reasonable to speculate that, rather than interfere with each other, the response to the two exercise modes may be additive, promoting a generic adaptation without the need for true training specificity. As training progresses and an individual gains more training experience, there is a need for progressively greater training loads and training specificity (i.e., a greater focus on one mode of training) to stimulate further adaptations, resulting in an impaired response to concurrent training, compared to resistance training alone.

That said, there are a number of potential confounding variables that may affect the results. One such variable is the length of the recovery period between the endurance and resistance training sessions. Specifically, some research^[11] suggests that concurrent training may impair strength gains when resistance and endurance training are performed within the same session, but not when performed during separate sessions. In several cases, this was true even if those sessions took place only a few hours apart (such as morning and afternoon, for example). While the results from the subgroup analyses in the study under review support the findings of the aforementioned research, these results were largely driven by two interventions from the same study^[6] as shown in Figure 2.

Figure 2: Same-session results in trained participants

The modality of training also represents a potential confounding variable. Because running was more frequently used in trials with trained participants (14 trials used running, 1 trial used cycling) than trials with untrained (4 trials used running, 3 trials used cycling) and moderately trained participants (5 trials used running, 9 trials used cycling), it is possible that this may have affected the results by enhancing the size of the interference effect in trained participants. That said, a 2018 meta-analysis^[12] reported a trend for lower body strength to be negatively affected by cycling HIIT, and not by running HIIT, which does not support the idea that the modality of training could have confounded the results of the meta-analysis under review.

Another potential confounding variable is endurance training intensity. Since high-intensity training (HIT) and HIIT were more frequently used in the training programs of participants in the trained and moderately trained categories than of those in the untrained category, it is possible that endurance training intensity may have affected the results. This point is supported by the results of the aforementioned 2018 meta-analysis^[12], which found that concurrent HIIT and resistance training resulted in a smaller increase in lower body strength compared to resistance training alone.

The generalizability of the results also merits discussion. Since the outcome of the meta-analysis was the maximal strength of the squat and leg press, these findings cannot be generalized to other performance outcomes, or to measures of upper body strength. Moreover, it is worth pointing out that the meta-analysis did not include highly trained participants or elite athletes in the trained category, which means that the results are not transferable to this population. Furthermore, the lack of trials employing trained women means the results of the trained category cannot be generalized to this population, either.

Overall, the inclusion of only moderate and high-quality trials in the meta-analysis under review, and the absence of heterogeneity in all of the analyses increase confidence in the findings. However, keep in mind that, in addition to randomized trials, the meta-analysis also included non-randomized trials, which are more prone to systematic and confounding biases. Moreover, it’s worth noting that the results of the main analyses may have been confounded by the more frequent use of HIT and HIIT in trials with trained participants. Lastly, the results of the subgroup analysis in trained participants were largely driven by two interventions from the same study, which used a non-randomized design.

The good quality of the trials and the absence of heterogeneity increase confidence in the findings. However, the inclusion of non-randomized trials in the meta analysis, the possibility for endurance training intensity to have confounded the results of the main analyses, and the issue with the outlier non-randomized study largely driving the results of the subgroup analysis in trained participants suggest that the results should be interpreted cautiously.

The results of this meta-analysis suggest that concurrent resistance and endurance training has negative effects on maximal lower body strength in trained individuals, with potentially smaller effects in moderately trained individuals, and no effects in untrained individuals. Moreover, the results of the subgroup analyses suggest that, in trained individuals, maximal lower body strength is negatively affected when resistance and intense endurance exercises are performed within a short time of each other (less than 20 minutes), but not when performed separately (more than 2 hours apart). However, the inclusion of non-randomized trials in the meta analysis, and the issue with the outlier non-randomized study largely driving the results of the subgroup analysis in trained individuals, suggest that the results should be interpreted cautiously.