“Hobbies are good for anxiety” is the kind of claim that gets repeated so often it stops being checkable. The version you see in lifestyle articles tends to lump aerobic running, sourdough baking, watercolor painting, and forest walks into one undifferentiated category called Self-Care, all of which supposedly help your nervous system in roughly equal measure. The actual research is messier and more interesting. Some hobbies have real, replicated effects on anxiety. Some have promising but smaller effects. Some are basically aesthetic, and the calmness people report is more about the absence of doomscrolling than about the activity itself.
Worth sorting them out. Here is, in my read, where the evidence actually lands.
What the data supports
Aerobic exercise has the strongest case. This one is not close. A 2025 meta-analysis in the International Journal of Mental Health Nursing pooled 11 RCTs and found a moderate effect of exercise on anxiety symptoms (standardized mean difference of −0.66). An older general-practice meta-analysis in BMC Health Services Research (Aylett et al., 2018) found a smaller but still real effect (SMD −0.41), with high-intensity programs outperforming low-intensity ones. The mechanism is partly direct (cardiovascular adaptation, HPA-axis modulation) and partly behavioral (you can’t catastrophize about your tax bill while running an interval). If you are picking a hobby explicitly to reduce anxiety, road running or any moderate-to-vigorous aerobic activity is the highest-evidence answer on the menu. Strength training has its own emerging literature and probably belongs in this bucket too.
Mindfulness-adjacent practices have moderate, well-replicated effects. Yoga in particular has been studied a lot, and the meta-analyses generally show real reductions in anxiety symptoms, though effect sizes are smaller and more heterogeneous than the aerobic-exercise literature. Same general story for seated meditation. If a hobby has a “pay attention to your breath while moving your body” component, there is decent evidence for that hobby helping. Yoga is the obvious case. Journaling, specifically the structured kinds (expressive writing in the Pennebaker tradition), has its own respectable literature for reducing rumination.
Nature exposure has measurable physiological effects. The Japanese shinrin-yoku (forest bathing) research is the most-cited body of work here. Park et al.’s 2010 study across 24 Japanese forests found salivary cortisol about 12% lower after forest walks compared to matched urban walks, with parallel reductions in pulse rate and blood pressure. A 2019 meta-analysis in the International Journal of Biometeorology confirmed cortisol reductions across most included studies. Important caveat: these are short-term physiological markers, not long-term anxiety-disorder outcomes, and the studies are heavily weighted toward Japanese samples and forest contexts. The honest version of the claim is: time in a forest measurably lowers your acute stress response that afternoon. Whether that translates to lower clinical anxiety over months is less established. Still, bird watching, gardening, and any regular hiking habit are getting some of that effect by default.
Rhythmic motor activity shows promise. The evidence here is smaller and earlier-stage than the categories above, but there’s a real signal. Repetitive, bilateral hand activity (knitting, crochet, drumming, some forms of pottery at the wheel) seems to produce something like a “flow-light” state that practitioners consistently report as calming. The peer-reviewed literature is thinner than I’d like and dominated by survey data rather than RCTs, so treat this one as “plenty of practitioners report” rather than “research shows.”
What’s mostly aesthetic
Now the part the genre tends to skip.
“Flow” is real, but it’s probably not the main mechanism. Csikszentmihalyi’s flow concept gets invoked constantly in hobby writing as the reason hobbies reduce anxiety. Flow is a real and well-described psychological state. The evidence that achieving flow during a hobby session translates into lower baseline anxiety the following Tuesday is much weaker than the popular framing implies. Flow probably matters for life satisfaction and for whether you keep doing the hobby. It is less clearly the lever for anxiety reduction specifically.
Sourdough as anxiety cure is mostly aesthetic. I love sourdough and I’m not knocking it. But the claim that baking bread is meaningfully anxiolytic, separate from the general effects of having a structured weekend ritual and getting off your phone, is not really supported. The studies people cite tend to be small, observational, and reliant on self-report. What’s probably happening with sourdough-as-therapy is twofold: it forces you offline for two hours, and it produces something concrete you made, which is good for self-efficacy. Those are real benefits. They aren’t specific to bread.
Coloring books, adult crafts as advertised, etc. Some of this works. Some of it is good marketing dressed up as neuroscience. The honest summary is that almost any absorbing manual activity probably helps a little, and the specific activity matters less than the marketing suggests.
The meta-point
Here is the thing I want you to walk away with, because it dominates everything above.
Consistency dominates modality. A hobby you do three times a week and aren’t catastrophizing about probably helps your anxiety more than a hobby with stronger research evidence that you do once a month.
Running has the strongest research backing of anything on this list. If you sign up for a half-marathon training plan, hate every minute of it, do it twice, and quit, you get approximately zero of the anxiety benefit. The person who knits for forty minutes most evenings while listening to a podcast is getting, in my estimation, a much larger real-world effect than the research-rankings would suggest, because they’re actually doing it.
So the right way to use this information isn’t to pick the highest-research-effect-size hobby and force yourself into it. It’s to use the research as a tiebreaker. Look at what fits your week (the friction match from our earlier piece on picking a hobby). Among the candidates that fit, lean toward the one with better evidence if anxiety reduction is a primary goal. But don’t override the consistency question to chase the effect size. You’ll lose.
If anxiety reduction is genuinely the thing you’re optimizing for, here’s my actual prescription: an aerobic activity three times a week (running, swimming, cycling, brisk hiking) plus a quieter daily anchor (journaling, knitting, yoga, time outside, whatever). Two channels, low-friction versions of each. That’s what the literature, read honestly, points at. The guides linked above are where to start on either side of that.
