We tell patients to exercise. We're usually vague about the dose.

A 2025 UK Biobank study put real numbers on it — using accelerometers, not self-report.

Yu K et al, Am J Geriatr Psychiatry 2025;33(5):512–523

https://lnkd.in/eVAh4Dtq

71,556 adults. Mean age 62. No depression or anxiety at baseline. Followed a median of 8 years.

How the study worked:

▸ Moderate-to-vigorous physical activity (MVPA) was tracked by wrist accelerometer

▸ Participants were grouped by weekly MVPA minutes

▸ Sedentary time was measured separately as daily inactive hours

What 150–300 min/week of MVPA was associated with (vs <75 min/week):

▸ 29% lower risk of depression 

↳ risk plateaued around 150 minutes

▸ 20% lower risk of anxiety 

↳ kept improving past 300 minutes

▸ Biggest gains came from moving out of the very-low group 

↳ "some is much better than none"

Light activity helped too — but less, and mostly for depression.

Sedentary time mattered on its own:

▸ >10.6 hours/day of sitting

→ 19% higher depression risk

▸ No clear effect on anxiety

▸ Findings held across age, sex, and BMI

The truth?

We can stop being vague. Aim for 150 minutes/week of MVPA — brisk walking 30 minutes most days works. And ask patients not just "Do you exercise?" but "How many hours a day are you sitting?"

Good advice for patients. Good advice for us.

Follow Daniel Carlat for more evidence-based psychiatry updates.