- Top of page
Common mental disorders (CMDs) describe non-psychotic mental health conditions including depressive, anxiety, adjustment and somatoform disorders that compromise day-to-day functioning (Andersson et al. 2004), the most common of which in pregnancy are depression and anxiety (Bailey & Cole 2009). A meta-analysis of studies published up to 2011 found that the average prevalence of maternal antenatal CMD was 15.6% in low- and lower middle-income countries, which is at least 50% higher than in high-income countries (Beck et al.). Among women of childbearing age (15–44 years) (Beck et al. 2010), depression is the global leader in terms of disease-related disability (Bergman et al. 2010). In Ho Chi Minh City, Vietnam, the prevalence rate of post-partum depression (33%) is higher than the mean prevalence rate in low-income countries (19.8%) (Beck et al.; Bodnar & Wisner 2005). A recent study in north Vietnam showed a 29.9% prevalence rate of perinatal common mental disorders, with higher rates in rural regions (Bonari et al. 2004). In contrast, a small exploratory study performed at an urban antepartum clinic in Hanoi, Vietnam, with socio-economically advantaged women documented a low prevalence rate of 1.6% of antenatal depression (Boyd et al. 2005).
Preterm birth and low birthweight (LBW) are major determinants of infant mortality and morbidity (Chuc & Diwan 2003), and among the almost 8 million children that die each year, 4.6% come from South-East Asia (Cox et al. 1987). A LBW infant, that is, below 2500 g (Dayan et al. 2002), is at higher risk of facing a number of health problems, including adverse cognitive development (Dayan et al. 2006), chronic diseases (Diego et al. 2009) or inhibited growth (Dossett 2008). Preterm birth, that is, birth before 37 completed weeks of gestation, is associated with increased rates of cerebral palsy, sensory deficits, learning disabilities, respiratory illnesses and death (Field et al. 2006).
Depression during pregnancy can harm the mother and the child in serious ways concerning morbidity and mortality (Field et al. 2009), ultimately leading to suicide or infanticide (Fisher et al. 2004). Also, there is growing evidence that maternal anxiety in pregnancy can compromise the neurobiological development of the foetus (Fisher et al. 2007, 2010). Antenatal depression negatively affects pregnant women regarding attendance at antenatal care, substance abuse and lower weight gain during pregnancy (Fisher et al. 2012), which in turn increase the likelihood of preterm birth, LBW and neonatal mortality (Giang et al. 2010). While a number of studies in high-income countries have shown an association between antenatal depression/anxiety and preterm birth (Goldberg & Huxley 1992; Goldenberg et al. 2000), there is a paucity of equivalent studies performed in low-income countries. Some studies from India (Hendrick et al. 1998), Pakistan (Hoffman et al. 2008) and Bangladesh (Jehan et al. 2010) have shown an association between antenatal depression/anxiety and LBW, whereas the results from high-income countries have been more mixed (Kessler 2003; Klingberg-Allvin et al. 2010), suggesting that the association may only be apparent under certain contextual circumstances. CMD may also be a marker for another, yet undefined variable present in only certain populations (Goldberg & Huxley 1992). Thus, there is need for further research on these associations in various contexts.
The aim of this study was to examine the association of LBW and preterm birth with clinically significant symptoms of antenatal common mental disorders during the third trimester of pregnancy in a semi-rural area in Vietnam.
- Top of page
This study found a prevalence of clinically significant symptom levels of antenatal common mental disorders of 37.4% when estimated with the Edinburgh Depression Scale. ACMD was significantly associated with LBW and preterm birth. None of the other examined risk factors were found to be statistically significant for ACMD. None of the other examined risk factors, apart for maternal age over 25 years, were statistically significant for LBW and preterm birth.
We are aware of some limitations in our study. We were not able to recruit participants according to a sample size calculation due to resource restrictions. A larger sample size may also have shown association with additional risk factors. As a screening instrument, the EDS does not compensate for a clinical diagnosis. However, it is the most widely used screening scale for post-natal depression and common mental disorders in the world, and its validity in various settings has been assessed (Patel & Prince 2006). A recent validation study among women in the perinatal period in Vietnam found that it has a sensitivity of 69.7%, a specificity of 72.9% and a correct classification rate of 69.7% in detecting perinatal common mental disorders (Nasreen et al. 2010).
We did not have the possibility to personally weigh the newborns, but had to rely on the mothers' verbal account of the weight taken in the delivery room. Although the procedures for weighing newborns are standardised in the Vietnamese healthcare setting, we cannot be sure whether standard procedures were followed for each individual case. Thus, there was possibly some rate of measurement bias in the birthweight measures. We are not able to estimate the error in ascertainment of birthweight in the study, but we do not expect this form of bias to have been systematic in terms of the exposure (ACMD) status. Intimate partner violence may be an additional important confounder of the association between ACMD and preterm birth (Beck et al. ; Patel et al. 2003), and not including this risk factor in our study can be regarded a weakness.
The reliability of LMP measures for estimating length of parturition in both high- and low-income country settings is a topic of continued discussion (Rahman et al. 2007; Rosenberg et al. 2009), although its main weakness is in predicting post-term birth (Rahman et al. 2007; Sanchez et al. 2012). LMP is widely used to estimate the duration of pregnancy (Rosenberg et al. 2009; Sanchez et al. 2012). Factors decreasing the reliability of LMP include recall bias and vaginal bleeding in early pregnancy, which tends to be more common for women with antenatal stress and anxiety. This could be a possible confounder of the association between preterm birth and ACMD, although such confounding was not observed in our study.
The rates of clinically significant symptoms of ACMD found in our study were high compared with global rates, but similar to rates found in three other studies in Vietnam, which varied from 20% to 33% in both rural and urban populations (Bonari et al. 2004; Bodnar & Wisner 2005; Santos et al. 2007). One Vietnamese study found very low rates of antenatal depression among socio-economically advantaged urban residents (Boyd et al. 2005). Another study in the same setting as the present one found a 6.8% prevalence rate of mental distress among women (Stewart 2005), which is much lower than our result. These variations may reflect both methodological factors and the variance in the contextual distribution of risk and protective factors. In addition, the validated EDS cut-off score used in our study was much lower than that in the other studies that used the EDS (Bodnar & Wisner 2005; Boyd et al. 2005).
The association found in our study between ACMD and preterm birth is consistent with a number of studies in high-income countries (Goldberg & Huxley 1992; Goldenberg et al. 2000; Kessler 2003). This link may be due to a number of reasons such as poor self-care among depressed/anxious mothers (Hendrick et al. 1998), which may increase the occurrence of other risk factors such as intrauterine infection (Suri et al. 2007), and effects mediated through the hypothalamic–pituitary axis (Tran et al. 2011). We also found an association between ACMD and LBW, which was a replication of findings in other low-income country settings including India (Hendrick et al. 1998), Pakistan (Hoffman et al. 2008) and Bangladesh (Tuan et al. 2003). Results from high-income countries on such an association have been more mixed (Kessler 2003; Klingberg-Allvin et al. 2010), and positive associations have been reported mainly in studies of disadvantaged populations where socio-economic status has acted as a modifier (Dossett 2008). Thus, it is of interest for future studies to examine why such associations exist mainly in conditions of socio-economic adversity. The association may be mediated through psychoneuroendocrine function (UNDP 2008) as well as maternal behaviour factors such as reduced attendance at antenatal care, increased substance use and lower weight gain during pregnancy (Wadhwa et al. 2004). Another plausible explanation for the association may be that ACMD and preterm birth have some risk factors in common. A recent review of risk factors for perinatal common mental disorders in low- and middle-income countries (Beck et al. ) indicated rural residence and intimate partner violence as important – both of which may also act as risk factors for preterm birth/LBW (Patel et al. 2003; Wardlaw et al. 2004).
The rate of preterm birth in our study was high (19.8%), but corresponds to previous findings in the same setting (WHO 2008). A recent meta-analysis has found a rate of 11% for preterm births in South-East Asia (Chuc & Diwan 2003). A possible reason for the high rate is that using LMP as a measure may cause overestimation of preterm birth rates (Rosenberg et al. 2009). However, we would not expect an eventual overestimation to have changed the association between ACMD and preterm birth. On the other hand, the rates of LBW (8.4%) found in our study were similar to those found in a previous study: In 2004, the rate of LBW for Vietnam was estimated at 9% (You et al. 2011). Surprisingly, we did not find any association between almost any of the examined outcomes and risk factors, although the evidence base for such associations is vast. This may reflect low statistical power of the study.
Despite its weaknesses, this study also has strengths. The study population was selected to be representative of a typical rural district in Vietnam – 70% of the population in Vietnam live in rural settings (Zscheck & Murray 1990). The use of an internationally widely used rating scale for depression allows comparison with other studies worldwide. Also, the association found in our study between ACMD and preterm birth and ACMD and LBW was strong.