Edinburgh Postnatal Depression Scale (EPDS): why standard scales fail in the perinatal period
Why a standard depression scale fails for mothers
Fatigue, fragmented sleep, appetite changes, reduced libido, psychomotor slowing — these are DSM-5 symptoms of major depressive disorder. They are also normative phenomena for a woman with a newborn. PHQ-9 contains 4 somatic items out of 9. BDI-II has roughly 7 out of 21. In the perinatal context, these items systematically inflate scores and produce false positives.
The Edinburgh Postnatal Depression Scale was developed by John Cox, Jenifer Holden, and Ruth Sagovsky (1987) precisely to address this problem. The authors conducted pilot interviews with postnatal women and deliberately excluded somatic items. The result: a 10-item questionnaire that, in five minutes, separates emotional distress from the normal physiology of early motherhood.
According to a systematic review by Gavin et al. (2005), up to 19.2% of women experience a depressive episode in the first 3 months postpartum — and most episodes begin after delivery, not continuing from pregnancy. One in five mothers. Without the right screening instrument, these cases are lost in the noise of "normal tiredness."
What EPDS measures: 10 items, no somatics
All 10 EPDS items ask about the past 7 days. Each is scored 0 to 3, total range 0–30. Content: anhedonia, anxiety, guilt, ability to cope, dysphoria, crying, suicidal thoughts. Not a single item about physical fatigue, appetite, weight, libido, or psychomotor activity.
The one exception that may raise a careful reader's question is item 7: "I have been so unhappy that I have had difficulty sleeping." This is not a vegetative sleep item in the classical sense. The wording deliberately frames insomnia as a consequence of mood, not as a neutral sleep disturbance. The item measures mood-induced insomnia — an emotional state, not a physiological change.
- Anhedonia (2 items): ability to laugh, looking forward with enjoyment
- Anxiety and fear (2 items): unprovoked worry, panic
- Guilt and self-blame (1 item): blaming oneself without reason
- Feeling overwhelmed (1 item): "things getting on top of me"
- Dysphoria (3 items): sadness, crying, mood-induced insomnia
- Suicidal ideation (1 item): thoughts of self-harm — standalone safety trigger
"The EPDS is a 10-item self-report scale developed to screen for postnatal depression in the community. It avoids somatic symptoms of depression that commonly occur in women after childbirth and could confound screening results."— Cox, Holden & Sagovsky, 1987, British Journal of Psychiatry
Severity thresholds: why ≥13 is outdated
The original recommendation by Cox et al. (1987) was a cutoff of ≥13 for "probable depression." This was based on a validation sample of N=84 and remained the standard for three decades. In 2020, that changed: Levis and colleagues published an individual participant data (IPD) meta-analysis in the BMJ — 58 studies, 15,557 pregnant and postpartum women, 2,069 cases of major depression. The most precise diagnostic-accuracy estimate for the EPDS ever produced.
The key finding: the ≥13 cutoff is too specific for universal screening. With 66% sensitivity, it misses roughly one in three cases of major depression. Levis et al. (2020) recommend ≥11 as the new standard: 81% sensitivity with 88% specificity. The ≥13 cutoff remains useful for identifying more severe cases, but as a primary screen it is obsolete.
Screening across the perinatal timeline
A one-time screen at the hospital is a weak strategy. Gavin et al. (2005) showed that most postpartum episodes begin after delivery, not continuing from pregnancy. This means: a single measurement at discharge will miss women whose depression develops at 2, 4, or 6 months. Perinatal depression requires longitudinal screening — this is an MBC task by definition.
Recommended measurement points: third trimester of pregnancy, 6 weeks postpartum, 3 months, 6 months, 12 months. The USPSTF (2016, Grade B recommendation) was the first to explicitly name pregnant and postpartum women as a target population for universal depression screening. NICE (CG192) recommends the Whooley questions for initial triage and the EPDS for full screening. This is the international standard.
The EPDS is not a one-off test but a series of measurements along the perinatal timeline. One score is a snapshot. A series is a trajectory. It is the trajectory that reveals when "normal tiredness" becomes clinically significant suffering. Harel et al. (2024, N=16,813) confirmed: digital administration of the EPDS is psychometrically equivalent to paper — meaning automated monitoring does not sacrifice accuracy.
Item 10: suicidal ideation as a standalone signal
The EPDS is one of few depression screening scales with a direct item about self-harm. Item 10: "The thought of harming myself has occurred to me." Any score above 0 on this item requires immediate clinical assessment — regardless of the total score. This is a rule, not a suggestion.
Zhong et al. (2014, N=1,517 pregnant women) showed that item 10 identifies suicidal ideation in 8.8% of women in the perinatal period. A critical finding: ~49% of women endorsing item 10 did not meet criteria for probable depression on the total EPDS score. In other words, suicidality in the perinatal period can exist independently of the depressive syndrome. The total score may be in the normal range, but the risk is not.
Limitations and complementary use
The EPDS is a screening tool, not a diagnostic one. A positive screen requires confirmation through a structured diagnostic interview. The original Cox (1987) validation sample was small — N=84. This is why the Levis 2020 meta-analysis (N=15,557) was such an important event: it redefined cutoffs on the basis of a database three orders of magnitude larger. False positives remain possible, particularly in women with anxiety disorder without depression — 2 of the 10 EPDS items directly concern anxiety.
- For fathers, use a ≥6 cutoff (Matthey et al., 2001) — men systematically endorse items at lower rates, especially the crying item
- If EPDS screens positive, conduct a structured diagnostic interview (MINI, SCID) — EPDS does not diagnose
- For anxiety assessment, add GAD-7 or STAI — EPDS contains anxiety items but not quantitatively
- If suicidality is present, conduct immediate clinical assessment regardless of total EPDS score
- In LMIC contexts, consider cultural adaptation — EPDS is translated into 60+ languages, but norms may vary
The EPDS remains the most widely used perinatal depression screening instrument in the world precisely because its design solves a specific problem — the phenomenological mismatch between standard scales and a new context. In MBC practice, its strength emerges not from a single test but from a series of measurements along the perinatal trajectory. Not "how bad is it right now," but "when exactly will intervention be needed" — that is the question the EPDS answers over time.