Two twins decide to start training. One chooses the barbell. The other chooses the treadmill. After a while they meet again and ask the question that the fitness world never seems to tire of: who chose better?
A simple answer would have been convenient. But the body adapts to whatever task it is given. Training that develops one particular measure does not necessarily produce the greatest change in another.
The STRUETH project compared resistance and endurance training in twins. The results do not name a universal winner. They do something more useful: they help us ask "best for what?"
Quick answer
Which is more effective for body composition — strength training or cardio?
A twin study found both resistance and endurance training produced meaningful changes in body composition, but in different ways. Resistance training had a stronger effect on lean mass; endurance training on fat reduction. Individual responses varied widely, and neither training mode was universally superior to the other.
Key takeaways
- STRUETH used a randomised crossover design in which 84 untrained same-sex twins completed both resistance and endurance training.
- Resistance training produced a larger average change in lean mass in the body composition study.
- Strength and aerobic capacity responded differently to the two training modes, with substantial variation between individuals.
- The studies do not name a universal best programme and give no guarantees about muscle gain, fat loss, fitness, or health for any individual.
The eternal debate in fitness
Strength training or cardio? The question comes up in gyms, running groups, and comment sections because many people hope for a single choice that solves everything.
The problem is that "results" can mean entirely different things. Do we mean lean mass, maximum strength, aerobic capacity, body fat, or simply being able to do an activity that feels meaningful?
When the goal changes, the answer can change too. That is why a fair comparison must start with which outcome is actually being measured.
What happens with resistance training?
Resistance training means muscles working against load. In the STRUETH studies, this mode was used to examine changes in body composition and strength, among other things.
In the body composition study, resistance training produced a larger average change in lean mass than endurance training. That is a group result from this particular design, not a promise of a certain amount of muscle for everyone.
The concept of lean mass is also broader than visible muscle size. It should therefore not be translated into dramatic before-and-after promises.
What happens with endurance training?
Endurance training places different demands on the body than resistance training. The companion STRUETH study examined aerobic capacity, among other things, and showed that responses depended on training mode and individual.
This provides support for the relevance of endurance training when the measured goal is aerobic performance. The study is not used here, however, for broad claims about heart health, longevity, or disease risk.
Even within the same training mode, results varied. A group improvement therefore does not tell us exactly how a particular twin or reader will respond.
What the STRUETH study examined
In the two sources, 84 untrained same-sex twins participated in a randomised crossover design. They completed separate periods of resistance and endurance training.
The design made it possible to compare training modes within the same research project and to study both group patterns and individual responses for body composition, strength, and aerobic capacity.
The twin design is valuable, but the participants were not exclusively identical twins, and the study does not eliminate all biological variation, measurement uncertainty, or influences from life outside training sessions.
What the researchers found
The training modes produced different patterns depending on which outcome the researchers studied. Resistance training produced the larger average change in lean mass, while the other STRUETH source showed mode-specific responses for strength and aerobic capacity.
At the same time, there was clear individual variation. Some people classified as low-responders on one measure after one training mode showed a clearer response after the other.
That is not the same as saying everyone has a perfectly "matched" training mode. Response classification depends on measures, thresholds, and measurement error, so exact percentages require a clear definition of both outcome and cut-off.
So which training mode is best?
For lean mass in this study, the result pointed more clearly toward resistance training. For aerobic capacity, endurance training is the more relevant comparison. For strength, we need to look at the specific strength measures that were actually tested.
But the question cannot be extended to "best for health" using only these two sources. They are not sufficient for confident conclusions about longevity, cardiovascular disease, bone health, metabolism, or which programme a particular person should choose.
The most scientifically defensible answer is therefore: best depends on the goal, the measure, the individual, and the context.
What does this mean for twins?
If you and your twin prefer different training modes, neither of you has necessarily chosen wrong. You may be pursuing different goals, enjoying different activities, and responding differently on the measures you choose to track.
That makes the twin comparison interesting, but not a personal research experiment. Comparing body composition or performance measures in a meaningful way requires controlled methods, and the results should not be used for self-diagnosis.
It can still be valuable to talk about what each of you wants from training and what feels safe and sustainable to continue.
The TwinPare perspective
At TwinPare we like questions with a simple surface and a more interesting interior. "Strength or cardio?" sounds like a duel. Research turns it into a question about goals, variation, and how we measure progress.
The two training modes do not need to be enemies, but these sources are also not sufficient to write a universal combination recipe. What matters is distinguishing between what the research actually measured and what we wish it had answered.
Source and limitations
The primary source supports comparisons of body composition responses after resistance and endurance training, including a larger average change in lean mass after resistance training.
The secondary source is used for claims about mode-specific strength and aerobic capacity responses and variation between individuals.
The sources do not prove that one training mode is best for all goals or all people. They do not support article claims about longevity, lower disease risk, better heart health, stronger bones, or higher metabolism.
The studies' samples, training periods, response definitions, and measurement uncertainty limit generalisability. Results should not be used as individualised training advice.
Source notes
The sources have been verified and editorially reviewed for this article. The limitations below show which level of conclusion the sources support.
- [thomas-2021] Studies of Twin Responses to Understand Exercise Therapy (STRUETH): Body Composition. Hannah J. Thomas; Christopher E. Marsh; Benjamin A. Maslen; Katrina J. Scurrah; Louise H. Naylor; Daniel J. Green. Medicine & Science in Sports & Exercise, 2021. Evidence type: Randomised crossover trial in which 84 untrained twins completed periods of endurance and resistance training Limitation: Group and individual responses depend on the study's training design, measurements, and definitions. The study does not identify a universal explanation for every difference and does not prescribe a best programme for everyone. PubMed DOI
- [marsh-2020] Fitness and strength responses to distinct exercise modes in twins: Studies of Twin Responses to Understand Exercise as a THerapy (STRUETH) study. Christopher E. Marsh; Hannah J. Thomas; Louise H. Naylor; Katrina J. Scurrah; Daniel J. Green. The Journal of Physiology, 2020. Evidence type: Randomised crossover trial comparing 84 untrained twins after endurance and resistance training Limitation: Response classification is influenced by chosen measures and measurement uncertainty. Results show variation between training modes and individuals but do not prove that genetics is irrelevant or that every low-responder will respond to an alternative programme. PubMed DOI
Editorial source review
This section shows how the article's key factual claims are linked to the source.
Phrasings that require caution
- Always ask "best for what?" and tie the conclusion to a specified outcome.
- Describe the resistance training result for lean mass without promising muscle gain for every person.
- Describe the endurance training result for aerobic capacity without broad claims about heart health or longevity.
- Do not use 84 percent without a verified definition of population, outcome, threshold, and measurement uncertainty.
- Do not write that identical twins always respond differently or that the studies prescribe an individual programme.
| ID | Claim | Source support | Caution |
|---|---|---|---|
| C1 | Resistance training produced a larger average change in lean mass than endurance training in the STRUETH body composition study. | 2021 | Describe an average study result, not guaranteed muscle gain or a precise individual effect. |
| C2 | Aerobic capacity responded differently to endurance and resistance training in the companion STRUETH study. | 2020 | Keep the claim tied to the studied performance measure and draw no broad medical conclusions. |
| C3 | STRUETH used a randomised crossover design with 84 untrained same-sex twins. | 2021 , 2020 | Describe the population correctly and do not imply that all participants were genetically identical. |
| C4 | Training responses varied between modes and individuals, and some low-responders on one measure responded more clearly to the other training mode. | 2020 | Write "some", not "all", and explain that response depends on the study's definitions and measurement uncertainty. |
| C5 | The two STRUETH sources support goal- and measure-specific comparisons, not a universal winner for all health and performance. | 2021 , 2020 | Give no individualised training recommendations or universal health promises. |
| C6 | STRUETH results can describe studied group and individual responses but cannot predict every reader's outcome. | 2021 , 2020 | Keep the limitations visible close to the article's conclusion. |