A 13.7-year study of 87,648 UK Biobank participants reveals that targeting modifiable risk factors could prevent up to 70% of depression and 61% of anxiety cases.
During the follow-up period, 5.7% of women and 4.2% of men developed depression, while 6% of women and 3.6% of men developed anxiety. The study calculated the population-attributable fraction (PAF) for a range of modifiable risk factors.
Psychosocial Factors Drive the Largest Risk
Psychosocial factors—including neuroticism, adverse childhood experiences, and adverse adult experiences—had the single largest estimated contribution to both conditions.
- Combined PAF for depression: 61% in women, 67% in men.
- Combined PAF for anxiety: 60% in women, 56.8% in men.
Among individual factors, neuroticism symptoms showed the highest PAF: 49-60% for depression and 52-54% for anxiety.
Women with neuroticism had a 2.6-fold higher risk of depression; men faced a 3.5-fold higher risk. For anxiety, women had a 2.8-fold risk and men a 2.9-fold risk.
Adverse experiences (in both childhood and adulthood) contributed PAFs of 18-25% for depression and 11-14% for anxiety.
Physiological Factors and Obesity
Physiological factors—including obesity, chronic inflammation, and diabetes—combined for a 21-22% PAF for depression. Obesity alone had the largest individual physiological PAF at 15% for both sexes.
Female-Specific Reproductive Factors
Reproductive factors contributed a combined 19% PAF for depression in women, driven largely by:
- Hormone replacement therapy (HRT) use (13% PAF)
- Early menarche
- Pregnancy termination
- Early age at first birth
For anxiety, the reproductive domain PAF was negligible at just 0.13%.
Combined Impact of All Modifiable Factors
When all modifiable factors were considered together:
- Depression: 68% PAF in women, 70% in men
- Anxiety: 61% PAF in women, 58% in men
Study Limitations
The authors note several important caveats:
- Observational design prevents causal inferences.
- Potential misclassification from multiple sources for depression/anxiety.
- Retrospective self-reporting of adverse experiences may introduce recall bias.
- Incomplete data from a large proportion reduced sample size.
- Categorical variables prevented dose-response analysis.
- The UK Biobank cohort is not fully representative of the general population.
Conclusions & Recommendations
The authors state that targeting modifiable psychosocial, lifestyle, and physiological factors could theoretically reduce the population burden of depression and anxiety. They recommend sex-sensitive, life-course strategies that integrate psychological, metabolic, and reproductive health into prevention and clinical practice.