Skip to content

Kendler KS, et al. Major depression and generalized anxiety disorder. Same genes, (partly) different environments? Arch Gen Psychiatry. 1992;49:716-22. PMID 1514877

One-paragraph summary

Bivariate twin modelling applied to lifetime diagnoses of major depression and generalized anxiety disorder, assessed at personal interview in a population-based sample of 1,033 female same-sex twin pairs. Three definitions of GAD were used, varying minimum duration (1 vs 6 months) and the presence or absence of a diagnostic hierarchy. For every definition the best-fitting model was the same: familial environment played no role in either condition; genetic factors were important for both and were completely shared between them; a modest proportion of non-familial environmental risk factors was shared. The conclusion, stated in the paper's own terms, is that in women the liability to major depression and to GAD is influenced by the same genetic factors, and whether a vulnerable woman develops one or the other is a result of her environmental experiences. A companion paper from the same sample estimated GAD heritability at around 30% and showed that it was not explained by GAD occurring only during episodes of depression or panic (Kendler 1992, PMID 1558460).

Key findings

  • Genetic correlation between MDD and GAD indistinguishable from 1.0 in women.
  • Robust to the duration criterion (1 vs 6 months) and to hierarchy — i.e., not an artefact of how GAD is bounded.
  • Shared individual-specific environment only modest; familial environment absent.
  • GAD heritability ~30%, with the remainder non-shared environment (companion paper, PMID 1558460).

Limitations

  • Women only; the later Swedish national twin study put the genetic correlation at +1.00 in women but +0.74 in men (Kendler 2007, PMID 17121688), so the sex restriction matters.
  • Sample size limits power: the paper says so explicitly ("within the limits of our statistical power").
  • DSM-III-R-era diagnoses by interview, with the reliability limitations of that era.
  • A genetic correlation of 1.0 constrains etiology, not phenomenology: it is compatible with two clinically separable syndromes, which is the counter-argument the field has made ever since (Hettema 2008, PMID 18412057).

Why it matters

This is the paper that made the GAD–depression boundary a genuine scientific dispute rather than a taxonomic preference. Its finding has survived thirty years of replication and extension: multivariate modelling puts GAD and major depression on the same genetic factor, separate from phobia/panic/bulimia (Kendler 1995, PMID 7726718); the Swedish national twin sample reproduces the near-unity correlation (Kendler 2007, PMID 17121688); and molecular work now finds an overarching internalizing liability driving comorbidity across anxiety disorders and depression (Ter Kuile 2026, PMID 42374129). Every argument in this knowledge base about whether GAD is a distinct disorder, and every decision to cross-link depression rather than absorb it, runs through this result.

Cited by wiki pages

  • the-diagnostic-boundary.md
  • mechanism-and-models.md