Antenatal depression and anxiety and early pregnancy BMI among White British and South Asian women: retrospective analysis of data from the Born in Bradford cohort

Bradford is the seventh largest city in the UK with a population of 537200 in 2018 and a growth rate of 0.06% [15]. This growth is largely due to the higher number of births than deaths; in 2017 there were 3600 more births than deaths. Bradford is one of the most deproved areas in the UK [15]. Around 20% of the popualtion are of South Asian origin, with higher fertility rates compared with the White British population. Almost half of the babies born in the city have parents of Pakistani origin and form a significant proportion of the individuals with poor health outcomes in the city [15]. BiB is a longitudinal multi-ethnic community birth cohort of 12453 women (including 6900 South Asian, 50.1%) established in Bradford in 2007 to understand the reasons behind poor health outcomes within the city [15]. The cohort aimed to examine how environmental, genetic, behavioural and social factors impact on the health of the mother and child, and development from childhood through to adult life [15]. Many research studies using the BiB cohort put particular emphasis on investigating differences between White British and South Asian individuals to explore important health inequalities, and to provide culturally appropriate healthcare within this population [15].

This study was a retrospective analysis of data from the BiB cohort. White British and South Asian pregnant women were the populations of interest. Women from other ethnic groups, or where ethnicity data were missing, were excluded from the analysis. The BiB questionnaire was administered to women in a standardised manner during 26-28 weeks gestation when they attended for an oral glucose tolerance test [15]. The original questionnaire was translated into Urdu and Mirpuri for women of South Asian origin who could not read or speak English and this was verbally administered [15]. This was carried out in a standardised process to ensure accurate translation using bilingual translators [15]. No incentives were offered for participation reducing self-reporting bias [15].

The dependent variables were the General Health Questionnaire (GHQ) scores indicating antenatal depression and anxiety. The GHQ is a validated self-report questionnaire with 28 items relating to the respondent’s current mental state, ability to carry out functions and daily activities and appearance of new and distressing phenomena [16]. The 28 items are grouped into four categories each used to identify symptoms of certain psychiatric disorders. Since this study focused on depression and anxiety as the most common mental health conditions during pregnancy, the total Likert score of the subscales for depression and anxiety (D, items 22–28; B, items 8–14) were used in the analysis (see Additional File 1) [16]. There are no agreed thresholds for the subscales to indicate depression or anxiety, but studies suggest that the cut-off score should be based on the mean/median of the sample of interest [17, 18]. In this study, the median was used due to the non-normality of distribution. This was 0 for depression and 6 for anxiety. Therefore, a score of > 0 was used to indicate depression and > 6 indicated anxiety.

The main independent variable was maternal early pregnancy BMI. Data from the BiB cohort includes information on the mother’s booking BMI calculated using measured height and weight between 10 and 12 weeks of pregnancy [19]. A realistic lower limit of 11 kg/m² was set as this has been shown to be the lowest BMI for survival in women [20]. An upper limit of 80 kg/m² was based on the frequency distribution in the data from the BiB cohort and a published study [21]. Women with a booking BMI outside this range were excluded from analysis (n = 720, 6.5%). BMI was analysed as a categorical variable due to the inclusion of underweight which is also associated with increased depression and anxiety [22], therefore, a continuous analysis may skew the results. BMI was stratified by the WHO’s classification. For White British women, the categories were: underweight, < 18.5 kg/m²; recommended weight, 18.5–24.9 kg/m²; overweight, 25–29.9 kg/m²; and obese class 1, 30.0–34.9 kg/m²; class 2, 35.0–39.9 kg/m²; class 3, ≥40.0 kg/m² [2]. The categories used for South Asian women were: underweight, < 18.5 kg/m²; recommended weight, 18.5–22.9 kg/m²; overweight, 23–27.49 kg/m²; and obese class 1, 27.5–32.49 kg/m²; class 2, 32.5–37.49 kg/m²; class 3, ≥37.5 kg/m² [23]. A secondary analysis was performed using the general population BMI criteria (see Additional File 2) for South Asian women due to the current lack of guidance in the UK for using Asian-specific criteria in pregnancy.

Additional variables included in the adjusted models were maternal age, maternal education, area of residence deprivation (based on postcode) and maternal smoking. Maternal age (years) was analysed as a continuous variable. Maternal education was defined as mother’s highest educational qualification (equivalised) with the following categories: none, GSCE equivalent, A-level equivalent and higher than A-level (used as reference group). The Index of Multiple Deprivation (IMD) was used to categorise area of residence deprivation. The IMD is a measure of relative deprivation for small areas in England and is the most widely used tool to measure deprivation in health-related research in the UK [24]. Since BiB was carried out in Bradford, the national IMD quintiles would be of limited use for this study because Bradford has a higher level of deprivation compared to most areas of the UK [15]. Therefore, the deprivation data were skewed towards the most deprived quintiles for this population. A binary variable was created with quintiles 2–5 combined to represent lower levels of deprivation (used as the reference group) and quintile 1 represented the highest level of deprivation. Maternal smoking during pregnancy was a binary variable (yes/no). Full details for all variables can be found online in the BiB Data Dictionary [25].

Primarily, data were summarised using descriptive statistics (proportions and percentages for categorical variables and means and standard deviations for continious variables). Following this, univariate independent t-tests were used to test for significant associations for normally distributed continuous variables and the chi-squared and Fishers exact test were used for categorical variables. Statistical hypotheses were tested using two-tailed 95% confidence intervals (95% CI), with a signficance level of 0.05 and critical value of 1.96 set. Regression assumptions were met, and data were analysed using multiple logistic regression modelling with maternal BMI, maternal age, maternal education, deprivation level and smoking during pregnancy as independent variables, and GHQ depression > 0 and GHQ anxiety > 6 as the dichotomous dependent variables. Finally, interaction analysis was carried out to determine if there were differences in the regression associations between White British and South Asian women i.e. associations between ethnicity and antenatal depression and anxiety within each BMI category. The analyses were performed using STATA version 16.

This study identified little evidence that early pregnancy BMI is a risk factor for antenatal depression or anxiety within South Asian and White British women in this population. There was an association between ethnicity and antenatal depression and anxiety with a significantly higher proportion of South Asian women with depression compared with White British women, but lower proportion with anxiety, independent of early pregnancy BMI.

Analysis of other socio-demographic variables identified smoking during pregnancy was a more important risk factor for both antenatal depression and anxiety in White British and South Asian women than early pregnancy BMI. Lack of education was a risk factor for antenatal depression in White British women but reduced the odds of antenatal anxiety in South Asian women. Older age was a risk factor for antenatal anxiety in South Asian women but reduced the odds of antenatal depression in White British women. However, effect sizes for maternal age were very small compared to smoking and education.

Our results indicate there were no significant associations between antenatal depression and anxiety and early pregnancy BMI within this population. There is limited literature on the association between maternal BMI and depression and anxiety in the UK, with one study also showing no significant association in either mental health condition [26]. However, previous studies across international, mainly high-income, settings showed significantly increased odds of depression during pregnancy among women with overweight and obese BMIs, compared with women of recommended BMI [4‐6]. The meta-anlaysis [4] also found significantly increased odds of anxiety in women with obesity, compared with women of recommended BMI, but no significant association for overweight.

The difference in statistical significance of the results between this study and previous literature could be due to the outcome measurement tool used. The GHQ was not used in any of the studies which mostly used either the Centre for Epidemiological Studies Depression Scale [5‐7] for depression and The State-Trait Anxiety Inventory for anxiety [7]. This can introduce variation in the classification of depression and anxiety between this study and other literature.

In this study, depression rates are higher in South Asian women compared with White British, whereas, anxiety rates are lower. Gater et al. [27] found that depressive disorder was more common in Pakistani women compared with White women. Mckenzie et al. [28] found higher rates of suicide among older South Asian women, compared to White women. In relation to anxiety, there is a lack of published literature examining rates of anxiety within South Asian women in the UK. However, Weich et al. [3] found that rates of common mental disorders (anxiety and depression) were higher in Pakistani women compared to White women of similar age. The contradicting results for anxiety may be partly methodological, since Weich et al. [3] used the Revised Clinical Interview Schedule, whereas, this study used the GHQ which is not recommended for epidemiological assessment in South Asian women [29]. It may also be explained by the study settings; this study was set in Bradford which has a high density of South Asian’s. These large social networks have been shown to be beneficial for mental health due to feelings of shared identity and provision of support [30]. Therefore, in this study population, general anxiety symptoms may be managed by these large social networks among South Asian women.

Smoking during pregnancy has been associated with increased antenatal depression and anxiety in the literature [31, 32]. This may be due to the negative impact of nicotine on psychiatric symptoms [32] but also related to increased guilt for not being able to quit and health concerns for the baby [32]. Although this study found that less South Asian women smoke during pregnancy, the impact of smoking tends to be higher among South Asian women due to cultural and religious stigma attached to it [33]. This could play a role in higher depression rates found in South Asian women and should be effectively managed during antenatal care for South Asian women.

Lack of education has been shown to be a risk factor for antenatal depression in many studies [34‐36]. However, in a study of South Asian women, it reduced antenatal anxiety and depression [37] which is also highlighted in this study. This could be explained by gender roles in South Asian culture in which women are expected to be more involved in taking care of the family than earning money or getting educated [38]. Therefore, South Asian women with less education tend to be more accepted by their in-laws as they fulfil their roles and responsibilities within the family. This may reduce the amount of stress and conflict within the family.

This study used data from a large prospective birth cohort, which minimises selection bias. However, following exclusions and categorisation of data, sample sizes were reduced. Although this study had a large sample size of 3514 White British women and 4310 South Asian women, the numbers of women with early pregnancy obesity were much smaller (903 and 1312 respectively). This study applied the Asian-specific BMI criteria as well as carrying out a secondary analysis using general population BMI criteria, which adds to the pregnancy evidence-base for South Asian women. We also used mother’s booking BMI which is measured in early pregnancy rather than self-reported pre-pregnancy weight and height. This approach reduces self-reporting bias since self-reported BMI tends to be under-estimated and provides a more accurate measure of BMI [39].

The South Asian ethnic group encompassed Pakistani, Indian and Bangladesh women which may introduce heterogeneity. Pakistani women have higher rates of obesity compared to Indian and Bangladeshi women [40]. Rates of depression are also different between these sub-groups, with Indians showing higher rates (61%) compared with Pakistani and Bangladeshi people (55%) [41]. Nevertheless, 88% of the included South Asian group were Pakistani, therefore, inclusion of women from India and Bangladesh may not have resulted in much bias.

The use of the GHQ in a South Asian sample comes with limitations as screening instruments may perform differently in different populations due to cultural and social differences [29]. This is especially important to consider when assessing differences between populations as was the case in this study. In a study examining the psychometric properties of the subscales of the GHQ in a multi-ethnic maternal sample from the BiB cohort [29], results showed that there was variation in the concepts measured by the GHQ between groups of different language and ethnic heritage. This may be due to the artefacts of translation and administration bias. The meaning of underlying concepts for some of the GHQ items differ according to language of administration [29]. Nevertheless, these issues of subjectivity and translations are found within most measurement tools, hence, more studies should be carried out in South Asian women using measurement tools such as the Edinburgh Postnatal Depression Scale which has validated Urdu [42] and Bengali [43] versions.