Early versus late onset obsessive–compulsive disorder: Evidence for distinct subtypes

The distinction between early versus late onset is important for understanding many different kinds of disorders. In an effort to identify etiologically homogeneous subtypes of obsessive–compulsive disorder (OCD), numerous studies have investigated whether early onset OCD (EO) can be reliably distinguished from a comparatively later onset form of the disorder (LO). The present article presents a systematic review and evaluation of this subtyping scheme, including meta-analyses and re-analyses of raw data. Regarding the latter, latent class analyses of nine datasets, including clinical and community samples, consistently indicated that age-of-onset is not a unimodal phenomena. Evidence suggests that there are two distinguishable groups; EO (mean onset 11 years) and LO (mean onset 23 years). Approximately three-quarters of cases of OCD (76%) were classified as EO. Meta-analyses indicated that EO, compared to LO, is (a) more likely to occur in males, (b) associated with greater OCD global severity and higher prevalence of most types of OC symptoms, (c) more likely to be comorbid with tics and possibly with other putative obsessive–compulsive spectrum disorders, and (d) associated with a greater prevalence of OCD in first-degree relatives. EO and LO were also distinguishable on other psychosocial and biological variables. Overall, results support the view that EO and LO are distinct subtypes of OCD. Comparisons with other, potentially overlapping OCD subtyping schemes are discussed, implications for DSM-V are considered, and important directions for future investigation are proposed.

► Diagnostic subtyping is important for understanding OCD.
► Evidence is presented to support the distinction between early and late onset OCD.
► Empirically derived age cutoffs are presented for distinguishing the two groups.
► The groups differ in gender distribution, severity, comorbidity, and family history
► Implications for DSM-V are discussed.