Course of depression symptoms between 3 and 8 months after delivery using two screening tools (EPDS and HSCL-10) on a sample of Sudanese women in Khartoum state

Depression is classified as postnatal depression (PND) by the 10th revision of the International Statistical Classification of Diseases and Related Health Problems (ICD-10) if symptoms occur within the first six weeks after birth, and by the Diagnostic and Statistical manual for Mental Disorders (DSM) if symptoms occur within the first four weeks after birth [1, 2]. Postnatal depression affects a mother’s capacity to form and maintain attachment with her newborn increasing the risk of attachment disorders during infancy and childhood [3]. Compared to a previous report by the World Health Organization (WHO) on PND in low and lower middle income countries (LLMIC) [4], recent reviews on prevalence of postnatal depression show higher estimates. Norhayati et al. [5] reported from LLMIC, based on screening results by the Edinburgh Postnatal Depression Scale (EPDS), prevalence estimates ranging from 12.9 to 50.7% at less than four weeks screening, 4.9–50.8% at four to eight weeks, 8.2–38.2% at six months after birth and 21.0–33.2% in the first postnatal year.

Prolonged postnatal depression is linked to maternal morbidity and to poor childhood cognitive and behavioral development [6]. Minkovitz et al. [7] followed children in the US up to age three and reported that mothers diagnosed with postnatal depression accessed well-baby and immunization visits at a lesser rate than mothers without depression, and they utilized emergency services for their infants at a higher rate. A 16 months US follow-up study on infant and maternal health-related quality of life by Darcy et al. [8] reported that maternal depressive symptoms at four months after birth predicted poorer infant health at 8, 12 and 16 months assessments. O’Brian et al. [9] examined two years old children in the UK and reported that those with faltering growth are more likely to have depressed mothers than children who are gaining weight appropriately. These findings elucidate the risk of continuing or “chronic depression” on health of children due to longer periods of maternal negative affects. The well-documented evidence on impact of PND on the quality of life of both mother and child makes it logical to identify and treat it. Wickramaratne et al. [10] reported that early remission of maternal depression leads to improvement in all child outcomes.

Many health systems worldwide advocate for universal screening for postnatal depression to increase detection rate, but the optimum time for screening or the frequency of screening needed during the postnatal period is still unclear. This issue has been raised by few scholars [11‐13]. The DSM 4 and 5, as well as the ICD-10, do not classify late onset postnatal depression [2, 14]. For that reason, exploration of the course of depressive symptoms after birth has become paramount.

A number of systematic reviews and longitudinal studies illustrated the course of depression symptoms after birth. Halbreich and Karkun [15] reviewed studies from high and low income countries. They showed that higher estimates of depression prevalence are commonly noticed during periods closer to birth than periods further away when brief instruments are used for screening, mostly EPDS [15]. Goodman’s [16] study reviewed PND beyond the early postnatal period (from 6 months up to 2 1/2 years after delivery) in high-income countries and reported that a significant proportion of mothers continue with depressive symptoms for months and years after giving birth. Vliegen et al. [17] reported on seven longitudinal studies that followed mothers up to three years after birth in high-income countries. The review showed a general reduction in severity of depression symptoms throughout pregnancy and the postnatal period, but not all reductions were statistically significant [17]. They also reported prevalence estimates from 20 longitudinal studies on community-based and clinical samples. They concluded that 30% of depressed mothers from community samples and 50% of depressed mothers from clinical samples remain depressed throughout and beyond the first postnatal year [17].

Studies from the US also reported continued symptoms of depression one and two years after birth [18, 19]. A study from Pakistan recruited 701 mothers during their third trimester of pregnancy and assessed them for depression at 3, 6 and 12 months after birth [20]. Fifty-six percent of mothers that completed follow-up exhibited depression symptoms at all points of assessment [20].

We have previously reported on the three months prevalence of PND symptoms using the EPDS [21]. We have little knowledge on how postnatal depressive symptoms evolve in a North African context: how many cases have a more prolonged course and will new cases appear after the three months period? In addition, will those new cases be “normal” depressive episodes or late postnatal depression? There is also a need to find out if other screening instruments for depressive symptoms used locally are useful in the detection of symptoms in the postnatal period.

The objective of this article is to investigate the course of depressive symptoms up to eight months after birth in a Sudanese sample using the Edinburgh Postnatal Depression Scale (EPDS). In addition, we want to explore correlations between the EPDS and a locally used tool for screening for depression outside pregnancy: the 10-items Hopkins Symptoms Checklist (HSCL-10) points.

Participants lost to follow-up at the eight months’ screening were not significantly different from women that were screened in terms of median age; 27 years old and 28 years old respectively, in parity (2 children and 1.8 children, respectively, p-value = 0.48) or in educational level (Pearson chi square p-value = 0.78).

Participants with complete follow-up at T1 and T2 were 223 (74.3% of sample) and 77 participants did not complete screening either at T1 or at T2 (see Fig. 1). Mean scores of EPDS at T1 and T2 for the entire sample were compared with mean scores for the 223 participants. Differences in mean scores (between 0 and 0.06 points on a 30-point scale) were insignificant between participants that had complete follow-up and participants that were lost at either T1 or T2. Table 1 illustrates the characteristics of women that completed follow-up.

The study illustrates that depression symptoms can exist well beyond the early postnatal period on a subset of women. The study also showed a general reduction of severity of symptoms for most mothers further away from birth. The decrease in the level of depression symptoms was probably not due to active treatment of clinically diagnosed cases, as very few proactively sought our referral services. We also showed that depression at three months constitutes higher risk and could predict depression later on in the first postnatal year. These findings were paralleled by both screening tests: all mothers reporting high depressive symptoms also showed a similar pattern of heightened distress Heron et al. [36] assessed the pattern of symptoms in a large community sample in England (n = 8323) during pregnancy up to 8 months after birth by self-reporting tools [37]. They reported a mean decrease in depression and anxiety across the period. McMahon et al. followed 100 postnatal mothers in Australia and measured depression symptoms by self-reporting tools one year after delivery [37]. They reported that 30% of all mothers and 60% of those depressed at 4 months continued to report significant depression at one year [37]. They also reported on factors predicting the persistence of depressive symptoms. Horowitz et al. [38] followed mothers that had high depression symptoms when screened 2 to 4 weeks after birth. Two years after birth, 30% of these mothers continued with depressive symptoms [38]. Women with a chronic course of depression were more likely to have suffered from previous depression, have high parental distress and limited current partner support [38]. Depression symptoms with atypical time of onset or course have raised the argument whether symptoms are in fact PND or an undiagnosed bipolar disorder that was triggered by the birth experience. Recent studies have explored the possibility of PND being an indicator for underlying bipolar depression. Liu et al. [39] conducted a Danish register-based cohort study and reported that the risk of bipolar depression among women with postnatal depression symptoms was higher than in women with depression outside the first year after birth. Sharma et al. [40] reported that 21.4–54% of women with PND in different studies have a diagnosis of bipolar depression. Having a mood disorder during such sensitive periods has tremendous impact on the mother, child, partner, and family. A chronic course of depression will involve more impairment and role limitations for depressed mothers. The period during and after pregnancy in low resource settings is a time of highest contact between mothers and healthcare services and it is a missed opportunity to detect and manage depression symptoms, whether it is unipolar or bipolar.

This article adds to existing research from North Africa by taking a prospective approach, measuring, and monitoring the course of depression and distress symptoms after the customary period for postnatal screening. This analysis highlights to healthcare personnel that they should not rule out PND if symptoms occur after the early postnatal period [41]. Women may not recognize the symptoms and may not seek help in the immediate weeks following birth. There are still limits in existing classifications of postnatal depression by the Diagnostic and Statistical Manual for Mental disorders (DSM 4 and 5) and in ICD-10 in spite of clear evidence from studies on late postnatal onset of depression [41‐43]. Nonetheless, a depression screening strategy should take into account potential relapse or chronic course of depression symptoms. Women with relevant risk factors are more likely to have a consistent level of depression symptoms throughout the first postnatal year [44]. Hence, it will be cost-effective, in low resource settings, to redirect repeated screening and follow-up care to women at higher risk of depression, as the number of women may be small. Low risk mothers will be reached through health promotion and awareness raising, even if they are not targeted for repeated screening for depression. That way they are equipped with knowledge if they do suffer from symptoms during pregnancy or after birth. To prevent a chronic course of postnatal depression, it may be necessary to identify symptoms and risk factors as well [44]. As our study showed, a negative early screening test may predict a continued negative test later on and this reduces the need for repeated screening for all women.

The study suggests that existing screening tools used to screen for depression can correlate with tools developed to detect women at risk for PND, especially in low resource settings where the possibility of adopting new tools is difficult. Further studies to compare performance of HSCL-10 against a diagnostic tool are needed to optimize cut-off points suitable for the perinatal period. We also recommend further qualitative studies on why women are reluctant to respond to referrals to mental health services.

Our study had several limitations. We relied on self-reporting instruments, not on clinical interviews, to measure severity and course of symptoms. In addition, the study did not investigate onset of depression symptoms during pregnancy. We missed the opportunity to assess if antenatal depression predicted the prevalence and course of postnatal depression. Due to this limitation, we illustrated our results as new onset late postnatal depression. Also, repeated screening by the investigator may have provided participants with some supportive talk therapy that could have biased (underestimated) our eight months screening results. Studies have shown that the most desired method of treatment by women feeling depressed is talking about their feelings with a sympathetic listener who understands the nature and extent of their feelings [45]. On the other hand, the concept of repeated screening may have raised feelings of stigma or intrusion among our respondents. This may have increased the number of participants who ignored or refused our second screening, although this risk was investigated statistically and was proved minimal. In addition, we do not consider our estimates generalizable to all women in Sudan. Our estimates may underestimate the magnitude of PND among mothers, for example, in the Eastern and Southern regions of Sudan, which are states suffering from political unrest and other psychosocial risk factors.

The ICD-10 and DSM-5 PND onset classification does not reflect consistent evidence brought about from longitudinal studies. Hence, postnatal depression will remain underdiagnosed with consequent policy implications. Screening strategies should advocate repeated screening during the first postnatal year as symptoms could persist or relapse beyond the early postnatal period. In low resource settings, repeated screening could be directed to mothers with relevant risk factors for depression.