Table 1.

Summary of data extracted from selected studies and short evaluation of the correlation between longer DUI and worse outcomes.

First Author (Publication Year)	Type of Study	Sample	Methods	Outcome Assessment	Mean DUI (Years)	Results	Longer DUI Worse Outcome
Dell’Osso et al. (2010) [30]	Longitudinal study	66 patients with primary diagnosis of OCD according to DSM-IV-TR. Two groups: DUI ≤ 24 months and DUI > 24 months. Two subgroups: monotherapy and polytherapy.	SCID-I and SCID-II at baseline for clinical assessment. Y-BOCS at baseline and after 12 weeks of pharmacological treatment to measure outcome.	Response to treatment: Y-BOCS decrease > 25% Remission: Y-BOCS score ≤ 10	7.75 * (±9.24)	DUI considered as a continuous variable does not influence treatment response. DUI ≤ 24 months is predictive of treatment response (OR = 0.27; p = 0.03) but not of remission (OR = 0.41; p = 0.12). It suggests the existence of a time-dependent effect of the DUI, that, after a certain period of time, may vanish.	YES
		Anti-obsessive treatment consisted of monotherapy (SSRI) or polytherapy (combination of SSRI with benzodiazepines, mood stabilizers, antipsychotics, or clomipramine).
Jakubovski et al. (2013) [31]	Longitudinal study	196 patients with primary diagnosis of OCD according to DSM-IV (only 75 continued 2 years of follow-up). Two groups: SSRI (n = 108) and GCBT (n = 88) treatment.	SCID-I and Y-BOCS at baseline for clinical assessment. Y-BOCS, BDI, BAI at baseline and after 3,6, 12, 18 and 24 months.	Response to treatment: Y-BOCS decrease > 35% Remission: Y-BOCS score ≤ 8	20.87 (±11.25)	Patients who suffered from OCD for a period of 30 years or longer had consistently higher Y-BOCS scores and did not further improve over time. Early onset of symptoms and longer duration of illness seem interconnected.	YES
		Patients allocated to pharmacological treatment received fluoxetine up to 80 mg/day. Patients allocated to GCBT attended 12 weekly therapy sessions. Subsequent treatment options for non-responders were: CGBT + SSRI; switching SSRI; SSRI + clomipramine; SSRI + quetiapine/risperidone; combination of pharmacologic add-on therapy + CGBT
Dell’Osso et al. (2015) [32]	Cross-sectional study	114 patients with primary diagnosis of OCD according to DSM-IV-TR. Four subgroups based on clinical phenotypes: checking/aggressive, contamination/cleaning, symmetry/order, and multiple phenotypes.	SCID-I for clinical assessment; Y-BOCS to define OCD severity; Y-BOCS Symptom Checklist to identify clinical phenotypes.	Y-BOCS scores	7.27 * (±0.97)	DUI and DI were significantly higher in the aggressive/checking subgroup compared to the other subgroups (p < 0.01). Y-BOCS scores were significantly higher in the aggressive/checking subgroup. This result may indicate a greater severity for this phenotype, but it may also be related to longer DUI and DI per se.	YES
		All patients were on a stable pharmacological treatment for at least 4 weeks.
Poyraz et al. (2015) [33]	Cross-sectional study	96 patients with primary diagnosis of OCD according to DSM-IV-TR. Two groups: DUI ≤ 4 years and DUI > 4 years.	SCID-I and SCID-II for clinical assessment; Y-BOCS to define OCD severity; Y-BOCS Symptom Checklist; a questionnaire to identify reasons for delaying treatment.	Remission: Y-BOCS score ≤ 10	7.02 (±8.52)	Patients with early onset (<12 years) of symptoms had a significantly longer DUI (p = 0.001). DUI was not predictive of remission when DUI was considered as a continuous variable or as categorical variable. In logistic regression, DUI was not predictive of remission (OR = 1.1; p = 0.074), but p-values indicated a distinct trend toward significance.	NO
		50 patients were on SSRIs and/or clomipramine, 44 patients were on different augmentation strategies including SSRIs and/or clomipramine and antipsychotic mood stabilizers.
Dell’Osso et al. (2017) [34]	Cross-sectional study	124 patients with primary diagnosis of OCD according to DSM-5. Two groups: Y-BOCS score ≤ 24 and Y-BOCS score > 24	SCID-I and SCID-II for clinical assessment; Y-BOCS to define OCD severity; Y-BOCS Symptom Checklist to identify clinical phenotypes; GCI score.	Y-BOCS scores	7.29 * (±9.06)	The group with increased severity received first pharmacological treatment earlier than the other group, consequently reporting a shorter DUI (p < 0.01). This could possibly be due to a worse clinical presentation leading to an earlier seeking of treatment.	NO
		Pharmacological treatment based on antidepressant drugs.
Albert et al. (2019) [17]	Retrospective study	251 patients with primary diagnosis of OCD according to DSM-IV (only 240 had a baseline and a 12-week Y-BOCS to determine response rate). Two groups: brief DUI (≤24 months) and long DUI (>24 months). Two different groups: DUI below median (≤60 months) and DUI above median (>60 months)	SCID-I and SCID-II for clinical assessment. OCD severity assessed by Y-BOCS, Y-BOCS Checklist, HAM-A, HAM-D.	Response to treatment: Y-BOCS decrease ≥ 25%	8.84 * (±9.84)	Long DUI (>24 months) reduces response rates (41% vs. 69%) as well as above the median DUI (>60 months) (40% vs. 61%). Mean DUI is significantly longer in subjects not responding to the first adequate SRI treatment. In individuals with long/above median DUI, Y-BOCS scores at 12 weeks were higher and percentage changes in Y-BOCS scores lower. In regression analyses, DUI > 24 months predicted response and 12-week Y-BOCS scores, but not using DUI as a continuous variable.	YES
		All patients treated with clomipramine and/or SSRIs for at least 12 weeks at adequate doses.
Perris et al. (2021) [35]	Longitudinal study	83 patients with primary diagnosis of OCD according to DSM-IV (59 completed 3 years follow-up).	SCID-I, SCID-II and BABS at baseline for clinical assessment. Y-BOCS and HADRS administered at baseline and monthly (for the first year of follow-up) or every two months (for the remaining 2 years of follow-up).	Response to treatment: Y-BOCS decrease > 35%. Partial remission: Y-BOCS < 15 for at least 8 weeks. Full remission: Y-BOCS < 8 for at least 8 weeks.	7.3 (±5.8)	Patients with “good outcome” (defined as fulfilling criteria for partial remission for more than 40% of the follow-up period) showed a shorter DUI than patients with “poor outcome” (4.5 ± 3.1 years versus 10.1 ± 5.7 years; p < 0.001). In the logistic multivariable model, a short DUI was the only significant predictor of “good outcome”.	YES
		First-line treatment: 25 individual ERP session + SSRI. Add-on strategy in resistant patients: venlafaxine; mirtazapine; imipramine. Second-line treatment: low dosages of antipsychotics as add-on therapy. Benzodiazepines to manage sleep disorder and/or panic attacks.
Zheng et al. (2021) [36]	Longitudinal study	207 patients with primary diagnosis of OCD according to DSM-5. Two groups: DUI ≤ 3 years and DUI > 3 years.	SCID-I at baseline for clinical assessment. GAF at baseline to evaluate overall functional impairment in the past month. Y-BOCS at baseline and after 8, 12, 24, and 48 weeks of pharmacological treatment to measure outcome.	Partial response: Y-BOCS decrease > 25%. Full response: Y-BOCS decrease > 35%	4.07 (±3.49)	In the brief DUI subgroup response rate was significantly increased and Y-BOCS score percentage changes higher after 48-week follow-up (p < 0.001). In a logistic regression analysis, a shorter DUI was predictive of a better response (p = 0.003). DUI was positively associated with DI but not with age of onset; this revealed that longer DUI indicates a longer clinical course.	YES
		Patients were treated with selective serotonin reuptake inhibitors or venlafaxine for 48 weeks in open-label conditions.

BABS = Brown Assessment of Beliefs Scale; BAI = Beck Anxiety Inventory; BDI = Beck Depression Inventory; CGI scores = Clinical Global Impression Scale; DI = Duration of Illness; DSM-IV-TR = Diagnostic and Statistical Manual of Mental Disorders-IV-Text Revision; DUI = Duration of Untreated Illness; DY-BOCS = Dimensional Yale–Brown Obsessive–Compulsive Scale; ERP = Exposure and Response Prevention; GAF = Global Assessment of Functioning Scale; GCBT = Group cognitive–behavioural therapy; HAM-A = Hamilton Anxiety Rating Scale; HAM-D = Hamilton Depression Rating Scale; HDRS = Hamilton Depression Rating Scale; OCD = Obsessive–Compulsive Disorder; SCID I = Structured Clinical Interview for DSM IV Axis I Disorders; SCID II = Structured Clinical Interview for DSM IV Axis II; SSRI = Serotonin Reuptake Inhibitors; Y-BOCS = Yale–Brown Obsessive–Compulsive Scale; SD= Standard Deviation; OR= Odds Ratio. * DUI originally reported in months.