Population-Based, Multigenerational Family Clustering Study of Obsessive-Compulsive Disorder.

Mataix-Cols D, Boman M, Monzani B, Rück C, Serlachius E, Långström N, Lichtenstein P.

Abstract

IMPORTANCE Controlled family studies have consistently found that obsessive-compulsive disorder (OCD) aggregates in families but have typically relied on samples recruited from specialist clinics. Furthermore, previous studies could not disentangle genetic from environmental factors contributing to the observed familiality. OBJECTIVE To provide unbiased estimates of familial risk for and heritability of OCD at the population level. DESIGN AND SETTING Population-based, multigenerational, case-control family and twin studies using the Swedish National Patient Register, Multi-Generation Register, and Twin Register. PARTICIPANTS All individuals diagnosed as having OCD between January 1, 1969, and December 31, 2009 (n = 24 768) and all their available first-, second-, and third-degree relatives, as well as nonbiological relatives and matched general population control subjects. Twins (n = 16 383) were included from the population-based Twin Register. MAIN OUTCOME AND MEASURE The risk for OCD among relatives of OCD probands. RESULTS The risk for OCD among relatives of OCD probands increased proportionally to the degree of genetic relatedness. The risk for first-degree relatives was significantly higher than that for second- and third-degree and nonbiological relatives. Second-degree relatives had higher risk for OCD than third-degree relatives. Relatives at similar genetic distances had similar risks for OCD, despite different degrees of shared environment. Separate twin modeling analyses confirmed that familial risk for OCD was largely attributable to additive genetic factors (47%; 95% CI, 42%-52%), with no significant effect of shared environment. Nonbiological relatives (spouses or partners who have at least 1 child together) also had an elevated risk for OCD (odds ratio, 2.61; 95% CI, 1.99-3.42). Early-onset probands (3907 individuals; mean age, 13.7 years) had slightly (nonsignificantly) higher familial risk than the total sample, although this was substantially lower than previously reported. There were no significant sex differences in the familial pattern or heritability estimates. CONCLUSIONS AND RELEVANCE Obsessive-compulsive disorder clusters in families primarily due to genetic factors. Nonshared environmental factors are at least as important. The finding of possible assortative mating in OCD is intriguing and should be investigated further.