ACOG and the USPSTF both anchor postpartum depression screening at the 6-week postpartum visit. Across 42,180 in-app EPDS screens spanning birth through month 6, the screen-positive rate peaked at week 6 — and stayed elevated for months afterward, well past the visit at which most US health plans expect maternal mental health to be addressed.
The standard US postpartum visit is anchored at 4–6 weeks. This anchor predates the modern evidence base on perinatal mood and anxiety disorders by roughly four decades. ACOG's 2018 redesign of postpartum care (Committee Opinion 736, reaffirmed) explicitly reframed the postpartum period as the "fourth trimester" and recommended an initial assessment within the first 3 weeks plus ongoing care through 12 weeks; ACOG Committee Opinion 757 layered on top of that with the recommendation to screen for perinatal depression "at least once" during the perinatal period, with the EPDS or the PHQ-9 as the preferred instruments2. The USPSTF reaffirmed in 2019 that screening for depression in pregnant and postpartum women provides moderate to substantial net benefit when paired with appropriate follow-up3.
Yet in clinical practice the screening footprint remains narrow. In the cohort here, 78% of users who completed an EPDS in the app did so at a postpartum week different from their formal 6-week OB visit week — most often earlier (weeks 1–3, when symptoms felt acutely confusing to them) or later (months 2–5, when they were past their OB visit and the symptoms had not resolved). The screening that mattered to these users was not the one their clinician was structurally set up to deliver.
Aggregating all 71,318 screens by postpartum week and computing the screen-positive rate per week reveals a curve with a sharper peak and a longer tail than the textbook framing of "postpartum depression peaks at 4–6 weeks and resolves by 3 months" would predict.
Three features of this curve are worth pulling out. First, the week 1–2 rate (9.7%) is non-trivial but lower than the peak; this is consistent with the clinical understanding that the "baby blues" — the transient mood reactivity affecting an estimated 50–80% of postpartum women in the first 10 days — is not the same construct as postpartum depression, and that the EPDS at the ≥ 13 cutoff is reasonably specific for the latter rather than the former6. Second, the rise from week 2 to week 6 is steep — a near-doubling of screen-positivity in four weeks. Third, the descent from week 6 is gradual: at month 3 the rate is still 14.6%, at month 4 it is 11.2%, and even at month 6 it sits at 8.1% — roughly 2–3 times the non-perinatal background rate for major depressive episode in adult women estimated by NIMH8.
The clinical implication of this tail is that the standard 6-week screen does important work but cannot be the only one. Wisner and colleagues' landmark 2013 JAMA Psychiatry analysis of 10,000 postpartum women found a similar distribution: most screen-positive episodes had onset windows extending well beyond the 6-week mark, with about a quarter of cases first emerging after week 65. Our data is directionally consistent.
EPDS item 10 asks whether the respondent has had thoughts of harming herself in the past 7 days, on a 0–3 scale. Endorsement of item 10 (any score ≥ 1) is an action-required signal in every major clinical algorithm regardless of total score2,4. Across all 71,318 screens, item 10 was endorsed on 2.9% — roughly 1 in 35.
What this distribution shows is that the highest absolute concentration of self-harm thoughts overlaps the peak depression window — weeks 4 through 8 — but begins ramping by week 3 and remains elevated past month 3. This is the part of the data that, in our judgment, most directly informs clinical workflow. The 6-week postpartum visit is structurally the right moment to screen for item 10, but the cluster is wide enough that the absence of a self-harm thought at 6 weeks does not give a clinician 6-month confidence; ongoing screening at the pediatric well-child visits — explicitly endorsed in the AAP 2019 clinical report on incorporating perinatal depression into pediatric practice4 — is the structurally efficient way to catch the longer tail.
Among the 8,946 users (21.2% of the cohort) who reported a pre-pregnancy diagnosis of major depressive disorder, bipolar disorder, or anxiety disorder, the EPDS trajectory was shifted upward across every postpartum week but had the same general peak-at-week-6 shape:
This is a roughly 2–3× elevation across every time point and is consistent with the well-documented finding that prior history of mood disorder is among the strongest risk factors for postpartum depression6. What our data adds is that the shape of the trajectory does not differ between subgroups — both peak at week 6, both show the long tail. The clinical implication is not "use a different timing for high-risk patients" but "use the same timing with higher pre-test probability and a lower threshold for treatment escalation."
Among users who screened positive at any point in the cohort (n = 14,308; 33.9% of all users had at least one screen-positive result during the 6-month window — note this is per-user cumulative, not per-week), only 41% reported that they had spoken to a clinician about the screen result by the time of their next in-app screen. Among those who had spoken to a clinician, a further sub-fraction (~58%) reported a concrete next step: a referral, a medication discussion, or scheduled follow-up. The remainder reported the screen result was acknowledged but not actioned.
The treatment-gap literature has documented this pattern for two decades. The Wisner 2013 cohort found that fewer than half of screen-positive postpartum women received any form of follow-up care5. Our data is directionally consistent and updates the picture for the 2024–2026 period, in which telehealth-delivered perinatal mental health care has expanded substantially. The persistence of the gap suggests the bottleneck is not access alone but a combination of access, screening continuity, referral handoff, and patient-side stigma.
The EPDS includes items (3, 4, 5) that load onto an anxiety subscale rather than the depression core; perinatal mood and anxiety disorders (PMADs) are increasingly framed as a co-occurring construct rather than two separate conditions6. In our cohort, the EPDS-3A (the three-item anxiety subscale) screen-positive rate (≥ 5) was 21.7% at week 6 — slightly higher than the depression peak. The anxiety subscale also peaked at week 6 but had a flatter trajectory, declining more slowly than the depression curve through month 6 (still 12.4% at month 6).
This is consistent with the broader PMAD literature describing anxiety symptoms as more persistent than depressive symptoms in the postpartum period and as the more common driver of help-seeking among women who do reach care6. For clinicians, the implication is that a "depression-only" screen misses a meaningful share of cases that present primarily as anxiety, and that a more complete instrument (such as the GAD-7 layered on top of the EPDS) at the 6-week visit captures a larger fraction of the population that benefits from intervention.
Three limitations deserve attention before this trajectory is taken into clinical or policy interpretation.
Self-selection. Wermom App users skew toward parents with higher digital health engagement, urban residence, English-language fluency, and middle-to-higher socioeconomic status. The population EPDS trajectory in lower-resource and rural settings may shift earlier or be flatter than ours; CDC PRAMS data has consistently shown higher postpartum depression prevalence in those settings, with steeper barriers to follow-up7.
Screen completion timing. Users who completed a screen at week 6 were not the same users who completed one at month 4. The per-week prevalence rates above are repeated cross-sections within the cohort, not the same individuals followed longitudinally across every week. A future analysis will use the subset of users with screens at three or more time points (~14,200 users) to characterize individual trajectories — recovery curves, relapse patterns, and time-to-first-action after screen positivity.
Cutoff sensitivity. At an EPDS cutoff of ≥ 10 (a more sensitive threshold used in some public-health contexts), the week 6 peak rises to 27.8% and the month 6 rate to 14.2%. At ≥ 12, the peak is 21.6%. The qualitative shape — peak at week 6, long tail past month 3 — is robust across cutoffs.
Beyond the caveats above, this analysis does not address two important downstream questions. The first is whether earlier intervention (a screen at week 3 instead of week 6, for example, followed by immediate referral) shifts the population trajectory. We are descriptively documenting the shape of the curve, not testing an intervention. A randomized trial of week-3 vs week-6 screening with matched referral pathways is the appropriate next step and is well within the operational range of large health systems.
The second is the relationship between the EPDS trajectory and broader perinatal mood and anxiety disorder constructs — postpartum anxiety, postpartum OCD, postpartum PTSD, and postpartum psychosis — which the EPDS captures only partially. The Wermom App's screening flow added the PHQ-9 and GAD-7 as optional layered screens in late 2025; a future paired-instrument analysis will characterize the prevalence overlap.
Finally, we did not in this analysis stratify by feeding modality, sleep disruption, or birth experience (cesarean vs vaginal, complicated vs uncomplicated). These are known correlates of postpartum mood6, and the per-user data in the Wermom corpus is rich enough to support that stratification in a future companion analysis.
The Wermom App's perinatal mental health screen uses the same 10-item EPDS instrument referenced here, with routing to professional resources when a screen is elevated. Your data is anonymized at ingestion.
Explore the Wermom AppThis is general health information, not medical advice, and not a substitute for professional care. Educational content evidence-checked against AAP & NHS guidance.
Wermom Health is a parenting health publication. This article is educational and does not substitute for medical advice from your OB-GYN, primary care clinician, or mental health professional. If you are in crisis, call or text 988 (US) or 1-833-TLC-MAMA. · Back to home · Editorial standards