About 3 in 4 depressed adults had insecure attachment. No style led.

Meta-analysis of observational studies. 61 studies. literature searched from database start to 20 June 2026; most included studies measured attachment at one time point.

Many studies, added up. People were watched, not told what to do. A clue, not proof.

Who: Adults (18+) with a standardised clinical diagnosis of major depressive disorder, including postpartum depression: 5,450 patients (MDD 5,105; PPD 345; mean age 40.7; 68.1% female) and 2,344 healthy controls (in the 33 case-control studies). Studies with specific comorbidities (psychosis, bipolar, PTSD, chronic medical illness) excluded. Mostly female and Western samples

• Securely attached1 in 4in depression
• Secure, vs healthylowerodds, 21 studies
• Fearful, vs healthyhigherodds, 12 studies

More to know, from our read of the paper:

• Observational and mostly cross-sectional: the authors say this cannot show whether attachment insecurity is a vulnerability before depression, a consequence of it, or both. Say 'linked, not proven'
• Healthy control groups were screened to exclude any psychiatric history, so the depression-vs-control contrasts 'may be larger' than against an unselected community sample (authors' limitation)
• Continuous fearful attachment was not significant (g = 0.35, p = 0.062) until one outlier study was removed
• Substantial heterogeneity (e.g. prevalence I2 72-92%; secure prevalence 95% prediction interval 8.6%-51.5%); categorical classification depends on cut-offs; publication bias detected for secure and fearful categorical analyses (trim-and-fill kept both significant)
• Postpartum subgroup small (345 patients, 7-9 studies); race/ethnicity rarely reported; mostly female, Western samples

This systematic review and meta-analysis pooled 61 studies of 5,450 adults diagnosed with depression, including depression after giving birth; 33 of the studies also compared them with healthy people. Only about 24% of the people with depression were securely attached, so about three in four had an insecure style, and the fearful, avoidant and anxious styles were each roughly a quarter. Compared with healthy controls, they had much lower odds of secure attachment and higher odds of anxious, avoidant and fearful attachment. Most studies measured everything at one time, so this is linked, not proven: it can't say which came first.

We read the abstract and the full text of this paper and checked every number on this page against it (18 facts, checked on 2026-10-29). The study design is named above. Cohort and other observational studies are described as linked, not proven. Found an error? Email theis@onestudyaday.com.