What the Research Says About Exercise
A synthesis of 5 studies on exercise — what actually works, what doesn't, and how to test it yourself.
The Surprising Truth: Exercise Alone Doesn't Significantly Improve Adolescent Anxiety or Depression
We often hear that exercise is a panacea for mental health, especially for young people. But a rigorous meta-analysis of 13 randomized controlled trials (RCTs) found that physical activity interventions produced no statistically significant improvement in anxiety (Standardised Mean Difference, SMD = -0.07, 95% CI -0.21–0.08) or depression (SMD = -0.10, 95% CI -0.21–0.02) in adolescents aged 10–19. This isn't to say exercise is useless, but it challenges the simple narrative and forces us to look closer at what actually works, for whom, and under what conditions – especially if you're trying to run your own experiment.
What the research actually shows
When we dig into the data, a clearer, albeit more complex, picture of exercise's effects emerges. It's not always about a universal "good for everything" but rather specific benefits for specific outcomes and populations.
For instance, if you're interested in brain health, particularly boosting levels of brain-derived neurotrophic factor (BDNF) – a protein linked to neuroplasticity and mood – the type of exercise matters. A meta-analysis of 29 studies, involving 910 healthy adults, found that aerobic exercise training reliably increased resting BDNF concentrations by a small but statistically significant amount (SMD = 0.28, 95% CI 0.16–0.40). However, resistance training did not produce a statistically significant change in BDNF (SMD = 0.09, 95% CI -0.11–0.29). This suggests that if you're targeting this specific biomarker, your personal experiment should lean heavily on cardio.
Similarly, when it comes to cognitive function, how you integrate movement into your day is key. A 2024 meta-analysis of 25 RCTs, involving 1,289 participants, revealed that taking multiple short physical activity breaks during prolonged sitting significantly improved cognitive function (Hedge's g = 0.24, 95% CI 0.13–0.35, a small effect). Crucially, a single physical activity break did not produce a statistically significant cognitive benefit (g = 0.04, 95% CI -0.09–0.17). For the n=1 experimenter, this means breaking up long periods of sitting with frequent, brief bouts of movement is more effective for acute cognitive boosts than one longer break.
Beyond specific biomarkers and acute cognitive effects, exercise plays a foundational role in overall health, especially for populations facing significant health disparities. A meta-analysis focusing on people with severe mental illness (SMI) – schizophrenia, bipolar disorder, and major depressive disorder – highlighted a stark reality: these individuals spend nearly 8 hours a day sedentary (7.9 hours/day) and get roughly half the moderate-to-vigorous physical activity (MVPA) of healthy peers (29.9 minutes/day vs. 57.1 minutes/day). Only 25.3% meet the WHO's recommended physical activity guidelines. While this was an observational study, it underscores that for many, the primary challenge isn't optimizing a specific exercise type, but simply increasing overall activity from a very low baseline.
Finally, for broader mental health and wellbeing, combining approaches shows promise. A systematic review of 35 trials found that interventions combining physical activity with mindfulness significantly improved mental health and wellbeing compared to passive controls (e.g., for depression, SMD = -0.56, 95% CI -0.80 to -0.32). This indicates that a multi-pronged approach is better than doing nothing, offering a robust foundation for improving mental health.
The nuance most people miss
The headline finding about adolescent mental health is a prime example of nuance often overlooked. While the idea of exercise as a universal mental health booster is appealing