Prenatal obsessive beliefs predict postpartum obsessive-compulsive symptoms: a prospective study

Abstract

This longitudinal study examined whether specific cognitive patterns during pregnancy predict postpartum obsessive-compulsive (OC) symptoms. A diverse sample of 256 women was assessed at 20 weeks of pregnancy (Time 1), 6 weeks postpartum (Time 2; n = 233), and 6 months postpartum (Time 3; n = 231). At each point, participants completed interview and/or self-report measures of OC symptoms, psychological distress, and obsessive beliefs (i.e., cognitive patterns related to OC symptoms). Postpartum OC symptoms were common, reported by 87.1 % at Time 2 and 74.5 % at Time 3. Stronger prenatal obsessive beliefs during pregnancy predicted greater postpartum OC symptom severity and higher likelihood of an OCD diagnosis at Time 3, even after controlling for baseline OC symptoms and distress. Findings highlight the clinical importance of prenatal cognitive risk factors and support cognitive-behavioral models of postpartum OCD.

1. Introduction

Obsessive-compulsive disorder (OCD) is marked by distressing intrusive thoughts (obsessions) and repetitive behaviors (compulsive rituals) aimed at reducing distress (American Psychiatric Association, 2013). Common symptoms include: (a) fear of contamination with a compulsion to clean, (b) excessive checking and reassurance-seeking in response to obsessions about harm or mistakes, (c) obsessions about taboo subjects (e.g., sex, cruelty, blasphemy), and (d) concerns about order and symmetry (e.g., Abramowitz et al., 2010; McKay et al., 2004). Although lifetime OCD prevalence is 1–2 % (Ruscio et al., 2010), subclinical obsessions and compulsions are nearly universal (e.g., Muris et al., 1997; Radomsky et al., 2014).

Pregnancy and the postpartum period heighten vulnerability to OCD. Prevalence estimates reach 10.9 % in the postpartum (Salari et al., 2024), and subclinical OC symptoms are even more common (e.g., Abramowitz et al., 2003, 2006; Abramowitz, Meltzer-Brody, et al., 2010; Fairbrother and Woody, 2008). Postpartum obsessions often center on infant safety and well-being, with rituals involving checking, seeking reassurance, and avoidance (Fairbrother et al., 2021; Starcevic et al., 2020). Notably, many women experiencing postpartum OC symptoms have no prior history of OCD (e.g., Lord et al., 2011).

Cognitive-behavioral models emphasize that obsessions emerge when normal intrusive thoughts are catastrophically misappraised, prompting compulsions that reinforce obsessional fear (Salkovskis, 1985, 1989). The Obsessive-Compulsive Cognitions Working Group (OCCWG; 2005) identified three “obsessive belief” domains contributing to these misappraisals: (a) overestimation of threat and responsibility, (b) beliefs about the importance and need to control intrusive thoughts, and (c) perfectionism and intolerance of uncertainty. These beliefs amplify distress and maintain OC symptoms (e.g., Abramowitz, 2006).

Fairbrother and Abramowitz (2007) adapted this model to the postpartum, proposing that increased responsibility and threat perceptions during early parenting heighten susceptibility to OCD. While most new parents experience intrusive thoughts, those with strong obsessibe beliefs may interpret these thoughts as danger signs, provoking compulsions.

Two prospective studies have provided initial support for this model. In both, higher obsessive beliefs during pregnancy predicted more severe postpartum OC symptoms, even after accounting for general distress (Abramowitz et al., 2006; Fairbrother et al., 2018). However, these studies were limited by small, homogeneous samples that primarily included subclinical individuals, thereby limiting generalizability. They also featured brief follow-up periods and did not differentiate among specific cognitive domains.

To address these gaps, the present study prospectively examined obsessive beliefs as predictors of postpartum OC symptoms in a large and diverse sample of women followed from mid-pregnancy through 6 months postpartum. We assessed specific obsessive belief domains, controlled for psychological distress and adverse experiences, and examined OC symptom dimensions to clarify patterns of risk. In doing so, this research offers clinically relevant insights into OCD vulnerability during a key developmental transition and highlights opportunities to develop targets for prevention programs for those at greatest risk.

We hypothesized that (a) most participants would report postpartum obsession-like intrusive thoughts and ritualistic behaviors, (b) the severity of these phenomena in the postpartum would be associated with prenatal obsessive beliefs, general distress, and adverse experiences, and (c) prenatal obsessive beliefs—particularly overestimation of threat/responsibility—would predict greater postpartum OC symptoms (especially obsessions and checking), even after controlling for baseline symptoms and distress. This research offers clinically relevant insights into OCD vulnerability during a key developmental transition and supports cognitive-behavioral conceptual models of symptom emergence

2. Method

2.1. Participants

The sample consisted of 256 women enrolled in a longitudinal study of the course and predictors of postpartum OC symptoms conducted at two U.S. sites selected to enhance cultural, racial/ethnic, and socioeconomic diversity: the University of North Carolina at Chapel Hill and Johns Hopkins University in Baltimore, Maryland. Participants were recruited between 2020 and 2023 through medical practices, ongoing studies, and targeted social media outreach.

Inclusion criteria were: (a) being 15–20 weeks pregnant, (b) at least 18 years old, and (c) providing written informed consent. To ensure a wide range of obsessive beliefs, we stratified recruitment based on baseline scores on the Obsessive Beliefs Questionnaire (OBQ; OCCWG, 2001), selecting participants with high (>130) and low (<96) scores.

Exclusion criteria were: (a) medically complicated pregnancy or birth, (b) recent suicidal behavior, (c) DSM-5 diagnosis of psychosis, substance use, or eating disorder within the past two years, (d) BMI > 35 or < 18, (e) chronic or acute medical conditions, (f) recent surgery or traumatic injury, and (g) substance use or significant tobacco consumption during pregnancy. These criteria minimized confounding influences on psychological and biological data.

2.2. Procedure

All participants provided written informed consent, and all procedures were approved by the IRBs at both study sites. Eligible participants were screened by phone after completing the OBQ and scheduled for a baseline session during their 20th week of pregnancy (Time 1), which included a review of procedures, clinical interviews via Zoom, and online self-report measures via REDCap. Follow-up assessments occurred at 6 weeks (Time 2) and 6 months postpartum (Time 3) to capture early and later stages of postpartum adjustment

2.3. Measures

2.3.1. MINI international neuropsychiatric interview (MINI)

The MINI (Sheehan et al., 2015) is a diagnostic interview assessing the presence of most DSM-5 disorders. It was administered to all 256 participants at Time 1 and 202 participants at Time 3. It has excellent inter-rater reliability, sensitivity, and specificity values for most diagnoses (Sheehan et al., 2015).

2.3.2. Postpartum thoughts and behaviors checklist and severity scale (PTBC)

The PTBC (Abramowitz et al., 2006) is a semi-structured interview assessing postpartum OC symptoms. It includes: (a) definitions to normalize intrusive thoughts and compulsive behaviors during the postpartum; (b) a checklist of 32 intrusive thoughts (e.g., fear of infant suffocation) and 14 compulsive-like behaviors (e.g., excessive checking); and (c) a 10-item severity scale modeled after the Yale-Brown Obsessive-Compulsive Scale (YBOCS; Goodman et al., 1989), assessing frequency, distress, impairment, and control over the past week (score range: 0–40). PTBC data were collected from 233 participants at 6 weeks postpartum (Time 2) and 231 at 6 months (Time 3)

2.3.3. Obsessive-compulsive inventory-12 (OCI-12)

The OCI-12 (Abramovitch et al., 2021) is a 12-item self-report measure of OC symptoms that excludes the hoarding and neutralizing subscales of the 18-item OCI-Revised (Foa et al., 2002) to better reflect the four primary OCD dimensions: washing, checking, ordering, and obsessing. Items are rated from 0 (not at all bothered) to 4 (extremely bothered) based on past-month experience, yielding total scores from 0 to 48 and subscale scores (corresponding to the four symptom dimensions) from 0 to 12. In this study, the OCI-12 was completed by 256 participants at Time 1, 218 at Time 2, and 200 at Time 3. The measure shows strong internal consistency, test-retest reliability, and convergent validity.

2.3.4. Edinburgh postnatal depression scale (EPDS)

The EPDS (Cox et al., 1987) is a 10-item self-report questionnaire assessing perinatal depression and generalized distress over the past week. Each item is rated on a 4-point scale and items are summed to produce a total score from 0 – 30. The EPDS has good internal consistency and is a sensitive and specific tool for detecting depression in pregnancy and the postpartum (Martin and Redshaw, 2018).

2.3.5. Perceived stress scale (PSS)

The PSS (Cohen et al., 1983) is a 10-item self-report measure assessing the degree to which life situations are appraised as stressful, focusing on feelings of unpredictability, uncontrollability, and overload. Items are rated on a 5-point scale. The PSS has strong internal consistency and good convergent validity with measures of anxiety (e.g., Roberti et al., 2006).

2.3.6. Adverse childhood experiences (ACE) scale

The ACE (Felitti et al., 1998) is a self-report measure assessing exposure to 10 types of childhood adversity (e.g., abuse, neglect, household dysfunction). It yields a cumulative score (0–10) linked to adult mental health outcomes. The ACE demonstrates good test-retest reliability and strong construct validity, with significant correlations to health risk behaviors and chronic conditions.

2.3.7. Obsessive beliefs questionnaire (OBQ)

The OBQ (OCCWG, 2005) is a 44-item self-report instrument with good reliability and validity as a measure of obsessive beliefs. Items are rated on a scale from 1 (disagree very much) to 7 (agree very much) and load onto three empirically derived subscales: (a) overestimates of responsibility and threat (OBQ-RT), (b) beliefs about the importance of, and need to control, intrusive thoughts (OBQ-ICT), and (c) need for perfectionism and certainty (OBQ-PC).

2.4. Missing data, power, and data analyses

All 256 participants completed baseline measures (Time 1). At Time 2 (6 weeks postpartum), 233 participants remained, and 231 at Time 3 (6 months). Follow-up response rates varied by measure. Chi-square tests showed that missing data were unrelated to eligibility group, supporting the use of multiple imputation. We used the MICE package in R to generate 10 imputed datasets (van Buuren and Groothuis-Oudshoorn, 2011). Power analyses based on prior studies (e. g., Abramowitz et al., 2006; Fairbrother et al., 2018) indicated 0.75–0.92 power to detect small to large effects.

We examined sample characteristics and postpartum OC symptom frequency, intensity, and content. Pearson correlations tested associations between Time 1 predictors (OBQ, EPDS, PSS, ACE, OCI-12) and outcomes (PTBC, OCI-12) at Times 2 and 3. Regression models then tested Time 1 predictors of (a) OCD diagnosis at Time 3, (b) PTBC severity at Times 2 and 3, and (c) OCI-12 subscale severity at both follow-ups. Data are publicly available via the National Data Archive.

3. Results

3.1. Demographic characteristics

At Time 1 (pregnancy), the sample had a mean age of 32.9 years (SD = 4.35; range = 20–45 years). All participants identified as cisgender female. The sample was racially and ethnically diverse: 73.7 % identified as White, 14.9 % as Black/African American, 7.1 % as Asian, 1.2 % as Native American or Alaska Native, 7.8 % as Hispanic or Latino, and 3.1 % as other racial backgrounds.

Most participants (93.8 %) were in a relationship with the baby’s father, and 2.7 % had become pregnant via donor sperm. The majority (85.2 %) were married, and most (84.3 %) held at least a bachelor’s degree, with over half (54.9 %) holding graduate degrees.

3.2. Pregnancy characteristics

For 33.5 % of the sample, this was their first pregnancy. 74.5 % underwent a vaginal delivery (the remainder underwent C-section). 35.1 % reported at least one medical complication during pregnancy, the most common being high blood pressure (10.2 %). Infant health complications were reported by 17.5 %. 13.5 % reported using psychotropic medications during pregnancy.

3.3. Clinical characteristics

3.3.1. OCD diagnostic status

At Time 1, 50 of the 256 participants (19.5 %) met DSM-5 criteria for OCD and 68 (26.6 %) met lifetime criteria. At Time 3 (6 months postpartum), 44 of the 202 participants (21.8 %) met OCD diagnostic criteria, including 17 who did not meet criteria at Time 1. Twelve participants received a diagnosis at Time 1 but not at Time 3.

3.3.2. Postpartum intrusive thoughts and compulsions

As assessed by the PTBC, most participants reported distressing intrusive thoughts about their infant and use of compulsive strategies during the past week at Time 2 (87.1 %) and Time 3 (74.5 %). The most common intrusions concerned suffocation/SIDS and accidents, with all types except illness-related intrusions declining over time. Avoidance and checking (including reassurance-seeking) were the most common compulsive behaviors (see Table 1).

Table 1.

Frequency and percent of participants reporting each type of intrusive thoughts and compulsive behaviors on the PTBC at Time 2 (6 weeks postpartum) (N = 233) and Time 3 (6 months postpartum) (N = 231).

	Time 2	Time 3
PTBC category	n (%)	n (%)
Intrusive thoughts
Suffocation/SIDS	192 (82.4)	123 (53.2)
Accidents	180 (77.3)	161 (69.7)
Contamination	119 (51.1)	78 (33.8)
Losing the baby	84 (36.1)	73 (31.6)
Intentional harm	60 (25.8)	31 (13.4)
Sexual	49 (21.0)	26 (11.3)
Illness	11 (4.7)	14 (6.1)
Compulsive behaviors
Avoidance	209 (89.7)	195 (84.4)
Checking	149 (64.9)	103 (44.6)
Thought suppression	93 (39.9)	79 (34.2)
Cognitive distraction	73 (31.3)	76 (32.9)
Behavioral distraction	67 (28.8)	66 (28.6)
Religious/prayer	45 (19.3)	27 (11.7)

Note. PTBC = Postpartum thoughts and behaviors checklist; SIDS = sudden infant death syndrome.

3.3.3. Severity of OC, depression, and stress symptoms

Table 2 shows mean scores across assessments. Mean OCI-12 scores remained in the subclinical range but showed wide variability (range = 0–46). At Time 1, 20.7 % of the sample scored above the clinical cutoff of ≥ 12 for a likely diagnosis of OCD, compared to 15.9 % at Time 2 and 17.3 % at Time 3. No significant changes in OCI-12 total or subscale scores were found across time points, but ordering symptoms were consistently highest (all ps < 0.05).

Table 2.

Means and standard deviations on study measures at each time point.

	Time 1	Time 2	Time 3
Measure	M (SD)	M (SD)	M (SD)
OCI-12 total	6.60 (7.19)	6.94 (8.11)	7.02 (7.98)
Washing	1.15 (2.16)	1.26 (2.45)	1.39 (2.52)
Checking	1.43 (2.29)	1.55 (2.42)	1.48 (2.28)
Ordering	2.57 (3.05)	2.61 (3.03)	2.63 (3.15)
Obsessing	1.46 (2.09)	1.52 (2.26)	1.51 (2.15)
PTBC		8.78 (6.57)	7.67 (6.69)
EPDS	5.28 (4.43)
PSS	14.54 (6.82)
ACE	1.66 (1.93)
OBQ total	127.80 (52.50)
Responsibility/threat overestimation	47.89 (22.30)
Perfection/certainty	54.50 (23.35)
Importance/control of thoughts	25.41 (12.57)

Note. OCI-12 = 12-item obsessive-compulsive inventory; EPDS = Edinburg Postnatal Depression Scale; PSS = Perinatal Stress Scale; ACE = Adverse childhood experiences scale; OBQ = Obsessive beliefs questionnaire; PTBC = Postpartum thoughts and behaviors checklist severity scale.

Mean PTBC scores indicated mild OC symptom severity, but with a range from 0–33. At Time 2, 51.5 % of participants scored in the subclinical range (0–7), 27.0 % in the mild range (8–15), 11.2 % in the moderate range (16–23), 3.0 % in the severe or extreme range (24–40). The distribution was similar at Time 3. A small but significant decrease in PTBC scores was observed between Time 2 and Time 3, t (183) = 3.31, p < .001, d = 0.24.

EPDS and PSS scores at Time 1 reflected mild depressive and stress-related symptoms. ACE scores indicated some exposure to childhood adversity, though below clinical risk thresholds (i.e., ≤ 4).

3.4. Pregnancy characteristics predicting postpartum OC symptoms

Pearson correlations, one-way ANOVAs, and chi square tests were computed as appropriate to examine whether characteristics of the pregnancy and health of the infant were related to Time 2 or Time 3 scores on the PTBC, OCI-12, or OCD diagnostic status. Results indicated no significant relationships (ps > 0.05). Accordingly, these variables were not controlled for in subsequent analyses.

3.5. Correlations between time 1 measures and postpartum OC symptoms

The OBQ subscales, OCI-12, EPDS, and PSS at Time 1 were moderately to strongly associated with the OCI-12 and PTBC at Time 2 and Time 3. The ACE was the only predictor that showed relatively weak (and non-significant) associations with later OC symptoms.

3.6. Time 1 obsessive beliefs predicting postpartum OCD diagnostic status

We used logistic regression to test whether the three OBQ subscales at Time 1 (pregnancy) predicted OCD diagnostic status at Time 3 (6 months postpartum) (Table 3). Control variables included Time 1 OCD diagnostic status (MINI), EPDS, PSS, and ACE scores. The model was significant, χ² (7) = 59.77, p < .001, explaining 45 % of the variance (Nagelkerke R² = 0.45). OBQ-RT and prior OCD diagnosis were significant individual predictors: each point increase on the OBQ-RT was linked to a 4 % increase in the odds of a postpartum OCD diagnosis, and participants with a history of OCD were over 6 times more likely to receive a postpartum OCD diagnosis.

Table 3.

Regression analyses of Time 1 obsessive belief domains predicting postpartum OC symptom severity at Times 2 and 3.

Time 1 predictors	B	SE B	Exp(B) or β	Wald or t	LMG
Predicting OCD Diagnostic Status at Time 3
History of OCD	1.86	0.479	6.40	15.06**
OBQ-RT	0.04	0.016	1.04	6.45*
OBQ-PC	0.00	0.016	1.00	0.00
OBQ-ICT	0.00	0.022	1.00	0.00
EPDS	−0.01	0.082	0.99	0.02
PSS	0.05	0.060	1.05	0.74
ACE	0.01	0.110	1.01	0.00
Predicting Time 3 PTBC
OCI-12 total	0.095	0.074	0.14	1.27	0.18
OBQ-RT	0.115	0.037	0.38	3.10*	0.35
OBQ-PC	−0.013	0.030	−0.07	−0.43	0.10
OBQ-ICT	−0.014	0.048	−0.01	−0.30	0.10
EPDS	−0.124	0.155	−0.16	−0.80	0.09
PSS	0.152	0.114	0.25	1.33	0.18
ACE	−0.031	0.284	−0.03	−0.11	0.01
Predicting Time 3 OCI-12 obsessing
OCI-12 total	0.052	0.024	0.18	2.20	0.20
OBQ-RT	0.031	0.010	0.31	3.00*	0.36
OBQ-PC	−0.010	0.009	−0.14	−1.12	0.12
OBQ-ICT	0.026	0.015	0.23	1.77	0.17
EPDS	−0.025	0.051	−0.13	−0.50	0.04
PSS	0.027	0.031	0.19	0.84	0.08
ACE	−0.052	0.065	−0.04	−0.81	0.01
Predicting Time 2 OCI-12 ordering
OCI-12 total	0.251	0.029	0.62	8.56**	0.54
OBQ-RT	−0.030	0.013	−0.21	−2.34	0.09
OBQ-PC	0.039	0.011	0.28	3.60*	0.17
OBQ-ICT	0.018	0.016	0.07	1.13	0.09
EPDS	−0.045	0.049	−0.07	−0.92	0.03
PSS	0.007	0.035	0.01	0.19	0.04
ACE	0.178	0.080	0.12	2.22	0.04
Predicting Time 3 OCI-12 ordering
OCI-12 total	0.253	0.031	0.62	8.11**	0.50
OBQ-RT	−0.031	0.014	−0.21	−2.29	0.09
OBQ-PC	0.042	0.011	0.31	3.81*	0.17
OBQ-ICT	0.017	0.018	0.03	0.93	0.09
EPDS	−0.074	0.60	−0.12	−1.24	0.04
PSS	0.018	0.043	0.03	0.41	0.05
ACE	0.162	0.105	0.11	1.55	0.06

Note.

^*p < 0.01.

^**p < 0.0001.

OCD = Obsessive-compulsive disorder; OBQ = Obsessive beliefs questionnaire; RT = Responsibility/threat overestimation subscale; PC = Perfectionism/certainty subscale; ICT = Importance and control of thoughts subscale; EPDS = Edinburg Postnatal Depression Scale; PSS = Perinatal stress scale; ACE = Adverse childhood experiences scale; PTBC = Parental thoughts and beliefs checklist severity scale; OCI-12 = 12-item Obsessive-Compulsive Inventory.

3.7. Time 1 obsessive beliefs predicting postpartum OC symptom severity

We conducted linear regressions to examine whether obsessive beliefs (OBQ subscales) assessed during pregnancy (Time 1) predicted postpartum OC symptom severity, as measured by the PTBC and OCI-12 subscales at Times 2 and 3. Models included baseline OCI-12, EPDS, PSS, and ACE scores as covariates. To reduce Type I error, a Benjamini-Hochberg correction (α = 0.01) was applied. No multicollinearity was detected. Table 3 presents the results for models in which one or more OBQ subscales were significant individual predictors.

3.7.1. PTBC scores

The model predicting Time 2 PTBC scores was significant, F(7, 203) = 10.62, p < .001, accounting for 19.6 % of the variance (R² = 0.196). The Time 1 OCI-12 total was the only significant predictor (β = 0.28, t = 3.52, p < .001), contributing 37 % of the explained variance (LMG = 0.37).

The model predicting Time 3 PTBC scores was also significant, F(7, 162) = 7.03, p < .001, accounting for 17.4 % of the variance (R² = 0.174). OBQ-RT was the only significant predictor, contributing 37 % of the explained variance.

3.7.2. OCI-12 obsessions

The model predicting Time 2 OCI-12 obsessions was significant, F(7, 206) = 16.59, p < .001, accounting for 32.2 % of the variance (R² = 0.322). The OCI-12 total, which contributed 27 % of the explained variance (LMG = 0.33), was the only significant predictor (β = 0.29, t = 4.07, p < .001).

The model predicting Time 3 OCI-12 obsessions was also significant, F(7, 167) = 9.50, p < .001, accounting for 26.3 % of the variance (R² = 0.263). OBQ-RT was a significant predictor, contributing 36 % of the explained variance.

3.7.3. OCI-12 ordering

The model predicting Time 2 OCI-12 ordering was significant, F(7, 206) = 32.91, p < .001, accounting for 51.9 % of the variance (R² = 0.519). OCI-12 and OBQ-PC were significant predictors, contributing 54 % and 17 % of the explained variance, respectively.

The model predicting Time 3 OCI-12 ordering was significant, F(7, 167) = 20.34, p < .001, accounting for 47.2 % of the variance (R² = 0.472). OCI-12 and OBQ-PC were significant predictors, respectively contributing 50 % and 17 % of the explained variance.

3.7.4. OCI-12 washing

The model predicting Time 2 OCI-12 washing was significant, F(7, 206) = 24.91, p < .001, accounting for 40 % of the variance (R² = 0.400). The OCI-12 total, which contributed 74 % of the explained variance (LMG = 0.74), was the only significant predictor (β = 0.72, t = 9.33, p < .001).

The model predicting Time 3 OCI-12 washing was significant, F(7, 167) = 18.70, p < .001, accounting for 40.5 % of the variance (R² = 0.405). The OCI-12 total, which contributed 74 % of the explained variance (LMG = 0.74), was the only significant predictor (β = 0.67, t = 8.31, p < .001).

3.7.5. OCI-12 checking

The model predicting Time 2 OCI-12 checking was significant, F(7, 206) = 34.63, p < .001, accounting for 48.4 % of the variance (R² = 0.484). The OCI-12 total, which contributed 60 % of the explained variance (LMG = 0.60), was the only significant predictor (β = 0.65, t = 6.35, p < .001).

The model predicting Time 3 OCI-12 checking was significant, F(7, 167) = 24.34, p < .001, accounting for 49.7 % of the variance (R² = 0.497). The OCI-12 total, which contributed 63 % of the explained variance (LMG = 0.63), was the only significant predictor (β = 0.63, t = 9.15, p < .001).

4. Discussion

Using a diverse sample and extended follow-up, this study provides clinically meaningful evidence that extends our understanding of vulnerability to OCD during the perinatal period. Intrusive thoughts about the infant were common, even six months postpartum, and often resembled clinical obsessions in their irrational content, distressing nature, and association with compulsive ritual-like behaviors. While many participants reported minimal impairment, a substantial subgroup experienced clinically significant symptoms.

As hypothesized, prenatal obsessive beliefs about inflated responsibility and threat were robust predictors of postpartum OC symptom severity and OCD diagnosis, even after accounting for prior OCD, general distress, and adverse experiences. These findings support cognitive-behavioral models (e.g., Fairbrother and Abramowitz, 2007), suggesting that misinterpretation of naturally occurring intrusive thoughts, particularly under heightened perinatal responsibility, leads to obsessional anxiety and compulsive responses in attempt to reduce the obsessional anxiety.

Analyses of OC symptom dimensions revealed meaningful patterns. Overestimation of threat and responsibility predicted obsessional symptoms at six months postpartum, highlighting a potential cognitive target for prevention and early intervention. Contrary to expectations, ordering symptoms were the most prominent subtype. Beliefs about perfectionism and need for certainty significantly predicted ordering at both time points—findings not reported in previous perinatal OCD research. This may reflect a broader drive for control as new mothers strive to meet internalized caregiving standards, particularly in our sample enriched for obsessive beliefs. These findings underscore the importance of assessing for less traditionally emphasized OC symptom types, such as ordering, in postpartum populations.

Surprisingly, obsessive beliefs did not significantly predict checking or washing symptoms, despite their frequency in our sample. These behaviors may be more influenced by normative caregiving routines, vigilance, and context-specific anxieties rather than distorted beliefs. This raises the possibility that some postpartum behaviors that resemble compulsions serve adaptive functions or reflect situational concerns rather than psychopathology. Future research should explore how cultural norms and safety expectations influence the expression of OC-like behaviors in the postpartum.

In concert, our findings demonstrate the heterogeneity of postpartum OC symptoms and suggest that specific cognitive vulnerabilities contribute differentially to symptom dimensions. The absence of a clear cognitive basis for some symptoms, and the emergence of unexpected patterns (e.g., elevated ordering), point to the need for refined conceptual models that incorporate contextual, developmental, and social factors. In addition, much of the variance in symptom severity remained unexplained, indicating that additional cognitive, biological, interpersonal, or environmental processes are also critical.

Clinically, these findings offer guidance for early identification and prevention. Assessing obsessive beliefs—particularly the overestimation of threat and responsibility—during pregnancy may help identify individuals at risk for developing postpartum OCD. Cognitive-behavioral interventions targeting these beliefs could be integrated into prenatal care or early postpartum services. Moreover, by demonstrating the predictive role of obsessive beliefs over time, this study contributes to the broader understanding of OCD onset and maintenance, not only in perinatal populations but also as a model of how cognitive vulnerability shapes the development of obsessions and compulsions.

Our study also advances the literature by drawing on a racially and ethnically diverse sample with a broad range of symptom severity. This diversity enhances the generalizability of findings and strengthens the clinical relevance. Our recruitment closely mirrored U.S. Census demographics (US Census Bureau, 2021), allowing us to explore symptoms in groups often underrepresented in OCD research. Future work should further examine how cultural and demographic factors shape symptom expression and response to treatment, with an eye toward inclusive, tailored interventions.

Several limitations should be noted. While racially and ethnically diverse, the sample was largely composed of highly educated and partnered individuals, which may limit generalizability to populations with fewer resources or less social support. These factors could influence both symptom vulnerability and detection. In addition, we did not assess the mental health of partners, whose symptoms or behaviors may impact maternal functioning. Future studies should consider dyadic or family-level influences, as well as socioeconomic and structural variables, in understanding postpartum OCD trajectories.