Depression in pregnant non-diabetic women and women with gestational diabetes in Bangladesh—a comparative study based on multiple logistic regression

Abstract

Background

Depression and gestational diabetes mellitus (GDM) pose significant challenges during pregnancy. Limited literature exists on depression in women with GDM, with most studies focusing on pre-pregnancy diabetes or postpartum depression. This study fills a crucial gap by specifically investigating and comparing antenatal depression among subjects with and without GDM in Bangladesh, utilizing data from the gestational period.

Methods

A cross-sectional study with a convenient, purposive sampling technique was undertaken among 111 pregnant women with (66) and without (45) GDM from September 2017 to March 2018 in the BIRDEM-2 GENERAL HOSPITAL, Dhaka. A semi-structured interview schedule was designed with items relevant to socio-demographics, obstetric history, diabetes, and depression.

Results

Different degrees of antenatal depression were identified in 61.3% of the subjects (i.e., 27% had mild, 4.5% had moderate, and 29.7% had severe depression, respectively). Out of 45 non-diabetic mothers, 11 (24.4%) had depression whereas out of 66 GDM mothers, 57 (86.4%) have depression. The exploratory analysis revealed that age group, menstruation history, and presence of GDM significantly affected depression but the multiple logistic regression model supported only GDM as a significant factor of depression. All the socio-demographic factors in this study were statistically insignificant to explain depression.

Conclusion

The risk of developing depressive symptoms increases with the presence of GDM. Therefore, it is important to screen for depression and provide treatment if necessary in women who are diagnosed with GDM.

Introduction

Pregnancy brings high hopes for the future; nevertheless, pregnant women may encounter unexpected challenges due to significant physiological and psychosocial changes [1]. Depression is the predominant psychological condition connected with it. Pregnancy elevates inflammatory response as well as stress hormones such as cortisol [2] which is also observed in individuals with depression [3]. A systematic review and meta-analysis conducted in 2023 found that the overall mean prevalence of antenatal depression was approximately 28.5% [4]. Other independent studies have recorded higher depression rates i.e., 30%, 45% [5–7]. Pregnant women in Bangladesh also face similar types of difficulties. According to a recent study covering the rural region of Bangladesh, 18% out of 671 mothers were found to be depressed during their ante-partum phase [8]. Although globally, depression being the prime reason for disability, affecting approximately 350 million people [9], diabetes affects almost 451 million [10]. Anderson et al. showed a substantial connection between type two diabetes and depression in both men and women. They found that 28% of women and 18% of men with type-two diabetes were positively diagnosed with depression [11]. Now, there is a special situation when high blood sugar (glucose) develops for the first time during pregnancy and usually disappears after giving birth. This particular form of diabetes is termed gestational diabetes [12].

In many regions of the world, gestational diabetes mellitus (GDM) is becoming an increasingly serious health issue. Globally, GDM causes sufferings to approximately 12% of women [13]. The manifestation of GDM in Asia ranges from 1 to 3% e.g., northeast of Turkey at 1.23% [14], Japan at 2.9% [15], and China at 2.31% [16]. An upsurge in non-insulin-dependent diabetes is causing growth in the GDM occurrence [17]. Among all the women with diabetes mellitus during their pregnancy, 61% are having diabetes for the first time (GDM) and as a whole GDM occurs in 2% to 9% of pregnancies [18, 19]. At the time of pregnancy, women face unprecedented changes. The increase in maternal adiposity and insulin resistance correspondingly in early pregnancy and late pregnancy are physiological causes for a woman forming gestational diabetes [20]. Lustman et al. suggested a possible link between elevated insulin resistance and depression by showing that people with depression had lower insulin sensitivity and treatment for depression decreased insulin resistance [21]. There is also evidence linking depression to hyperglycemia, a consequence of gestational diabetes. Additionally, depression has been found to be substantially associated with hyperglycemia in both type 1 and type 2 diabetes [22].

So, there is plenty of literature indicating the fact that type-2 diabetes may be more likely to develop in people with depression. [21–23]. Although depression and type 2 diabetes are linked, not much is known regarding the relationship between depression and gestational diabetes. Only a few studies were available that examined as well as compared depression with the presence and absence of gestational diabetes. To find out if depression and maternal–fetal attachment were linked to high-risk pregnancy, Chazotte et al. [24] performed a cross-sectional, descriptive study on 90 high-risk pregnant women between 34 and 36 weeks gestation. Thirty women with gestational diabetes, thirty women at risk of preterm birth, and thirty women with an uncomplicated pregnancy were compared using the CES-D. Depression was observed in 33.3% of women with straightforward pregnancies, 70% of women at risk for preterm birth, and 57% of women with gestational diabetes (> 16 on the CES-D). The frequency of depression was higher in pregnant women with gestational diabetes than in women with a normal pregnancy, although the difference was not statistically significant. The limited sample size and consequent lack of power required to detect meaningful differences between the groups, according to the researchers, may have contributed to this lack of difference. It is uncertain whether these results would apply to White women because minority women comprised 94% of the sample. According to a different recent study, pregnant women with diabetes had a higher risk of depression [25]. Using ICD-9 codes and healthcare data from New Jersey's Medicaid administrative claims database, a retrospective cohort design with a large sample (n = 11,024) identified women with diabetes and depression. 15.2% of pregnant women with diabetes experienced depression, compared to 8.5% of pregnant women without diabetes (OR 1.85, 95% CI 1.45–2.36), after adjusting for age, race, and preterm birth. Additionally, this study found that postpartum depression was more common among women with diabetes who did not have depression during pregnancy (OR I.69, CI 1.27–2.23). The sufficient sample size of this research was one of its benefits. The study's grouping of women with type 1 diabetes, type 2 diabetes, and gestational diabetes constituted one of its weaknesses, though. The literature only contained these studies, which were the only ones to compare and measure depression in pregnant women who had or had not gestational diabetes. Nevertheless, the findings of all four studies had insufficient power and small sample sizes, thus it is not certain if pregnant women with gestational diabetes are more likely to experience depression [24, 26, 27].

Moreover, there exists only one literature related to the comparison of depression among GDM and non-diabetic mothers in Bangladesh. Natasha et al. [28] executed a study to evaluate and compare the incidence of depression between expected mothers with or without Gestational Diabetes Mellitus (GDM). They discovered that 18.32% of individuals had depression overall. Individuals with GDM had a higher rate of depression (25.92%) than individuals without GDM (10.38%), with a mean age of 28.34 and 27.17 years, respectively. At both extremes of age, the prevalence of depression was alarming. Past history of GDM (P < 0.018) and dwelling place (P < 0.009) were significantly correlated with depression. While the risk of gestational diabetes mellitus (GDM) increased with parity, primipara was related to a higher prevalence of depression.

Thus, there is a significant gap in the literature and a demand for a more recent investigation with sufficient power to ascertain whether women who have gestational diabetes have an increased risk of depression. The only existing study which was executed from August 2011 to September 2012 is pretty old relative to the current time and context of the disease. Considering these factors, the current study intended to compare the depression level of gestational diabetes mellitus mothers with non-gestational diabetes mellitus mothers in an urban hospital in Dhaka City. This study also tried to identify the variables that contribute to depression in expecting mothers.

Methods

Study design and population

A cross-sectional study design with convenient, purposive sampling technique was adopted for this study. Our study samples were enrolled from the BIRDEM-2 GENERAL HOSPITAL, Dhaka. It is a private Diabetes Endocrine and Metabolic Hospital in Dhaka for women and children and focused on the treatment of endocrine diseases. It is located in the Segunbagicha area. This hospital provides comprehensive healthcare delivery to a vast number of GDM MOTHERS all over the country. This was an 8-month-long study that started on 1st September 2017 and ended on 31st March 2018. All pregnant mothers who were admitted to the hospital with manifestations of GDM and non-GDM during the study period were included in the study. Non-pregnant, type-1 diabetic, type-2 diabetic, physically/mentally unfit, known alcohol/drug abusers as well as women with a previous history of depression, anxiety, disabilities and chronic illness were excluded from the study.

Sample size

The following formula was used for sample size calculation in this study.

$$n=\frac{Z^{2}P(1-P)}{d^2}$$

where,

n = Sample size.

Z = the statistic corresponding to a 95% level of confidence (1.96).

P = Expected prevalence or proportion (0.18).

d = precision/ effect size/ margin of error (0.05).

For this study, we will consider a 95% level of confidence. Hence, the desired sample size is n = 230. However, due to a shortage of time and other limitations, we enrolled 111 participants for our study. As for the comparative study, 66 subjects were taken for GDM mothers and 45 were for non-GDM mothers.

Measurements of outcomes

A questionnaire was prepared in English comprising questions related to (1) socio-demographics, (2) obstetric history, and (3) diabetes and depression. Below is a brief explanation of each variable used in this investigation.

Socio‑demographic factors

The core information of the participants, such as age, education, employment status, and monthly family income of unemployed mothers were documented under this category.

Obstetric history

Information on factors such as the menstruation history of the respondents, total number of children, gravida of the respondents, and trimester of pregnancy of the respondents were collected in this segment of the questionnaire.

Diabetes and depression

Gestational diabetes mellitus of the respondents was measured clinically by identifying whether a patient is pregnant, new onset of diabetes in any gestational age and also using WHO and ACOG criteria. GDM was controlled for some patients and some of them were left uncontrolled. Depression status of the respondents was measured using the longer version of the Depression, Anxiety and Stress Scale (DASS-42) which was encompassed in the self-report questionnaire booklet. Each of the 42 items in the DASS-42 is scored on a four-point scale (0–3). The total score on the scale ranges from 0 to 42 points for depression (similar scoring is also applicable for anxiety and stress). The DASS-42 scores are categorized into 5 groups, normal (0–9 points), mild depression (10–13 points), moderate depression (14–20 points), severe depression (21–27 points), and extremely severe depression (28 +).

Development of research instrument

The semi-structured questionnaire was pretested on 10% of the total sample size. Response and feedback from the pretest were accommodated in the questionnaire and the finalized questionnaire was translated into Bengali for respondents’ convenience. Some other instruments such as pens, paper boards, pencils, erasers, rulers, and other required instruments in hospital settings were also used for collecting and recording the data.

Data collection

A questionnaire that participants self-administered in Bangla, the native tongue of both the participants and the researcher, was used to gather data from the respondents. Participants gave verbal consent to fill out the questionnaire to participate in the study. Before data collection, permission as well as written approval from the Institutional Review Board (IRB) was obtained from the concerned authority. Data collection was continued for up to five months. Approximately 12–13 interviews were conducted every month. Respondents were explained about the nature and purpose of the study. Adequate time was given for each individual to answer.

Data analysis

Individual questionnaires were amended for consistency and completion following the data collection. Subsequently, the data entry and analyses were done by SPSS version 20. Data analysis was done according to specific objectives. Descriptive statistics such as frequency, percentile, mean, median, mode, and standard deviation were performed. For association and relationship, chi-square tests were conducted. To see the strength of the association multiple logistic regression was done. A P-value of < 0.05 was considered statistically significant.

Results

This study was carried out to assess the depression level of gestational diabetes mellitus mothers and non-gestational diabetes mellitus mothers. A total of 111 (66 GDM and 45 non-diabetic) mothers were selected for the study. The findings of the study are presented by graphs and tables. The distribution of our main variable of interest level of depression is presented in Fig. 1. It indicates that 30 (27.03%) mothers had mild depression, 5 (4.5%) mothers had moderate depression, 33 (29.73%) mothers had severe depression, and the rest of the respondents (38.74%) were normal. No patient had an extremely severe level of depression in the sample. The distribution of socio-demographic, obstetric history, and gestational diabetes mellitus-related variables is presented in Table 1. Additionally, the association of each of these variables with depression is provided in this table using the P-value of the chi-square test. The age range was 15–42 years and the majority of the women were aged from 26 to 42 years old (53.2%). Amongst them (78.4%) studied up to higher secondary and the rest.

Fig. 1.

Frequency distribution of level of depression among respondents

Table 1.

Distribution of variables under study along with association with depression measured by chi-square test

Variables	n (%)	P-value (χ2)
Age group
15–25 years	52 (46.8)	0.05
26–42 years	59 (53.2)
Educational status
Primary	19 (17.1)	0.42
Secondary	36 (32.4)
Higher secondary	32 (28.8)
Graduate	24 (21.6)
Employment status
Employed	12 (10.8)	0.68
House-wife	99 (89.2)
Family income of unemployed mothers
≤ 30,000 taka	53 (53.5)	0.37
≥ 31,000 taka	46 (46.5)
Menstruation history
Regular	68 (61.3)	< 0.01
Irregular	43 (38.7)
Parity
One	19 (17.1)	0.17
Multiple	92 (82.9)
Gravida
Primigravida	65 (58.6)	0.38
Multigravida	46 (41.4)
Pregnancy trimester
1st trimester	4 (3.6)	0.62
2nd trimester	47 (42.3)
3rd trimester	60 (54.1)

Bold indicates statistically significant at 5% level of significance

were graduates (21.6%). Respondents were mostly housewives 99 (89.2%) and 53.5% informed that their family income was less than 30,000 BDT. Obstetrical history exhibited 68 (61.3%) of the respondents had regular menstrual history, mostly multiple parity 92 (82.9%), primigravida 65 (58.6%), and more than half of the total respondents were in 3rd trimester of Pregnancy 60 (54.1%).

Table 2 describes the level of depression among the categories of the factors GDM, age group and menstruation history. It indicates that 45 (40.5%) pregnant mothers had gestational diabetes mellitus and 66 (59.5%) of them were unexposed to gestational diabetes mellitus. Out of these 45 non-diabetic mothers, 11 (24.4%) have depression whereas out of 66 GDM mothers, 57 (86.4%) have depression. Again, higher prevalence of depression is visible in the age group 26–42 years (69.5%) and in mothers with irregular menstruation histories (83.7%). So, according to this table GDM, age group and menstruation history are statistically significantly associated with the level of depression.

Table 2.

Association between level of depression and categories of GDM, age group and menstruation history

	Level of depression			P-value
Gestational diabetes mellitus	No depression	Depression	Total
	n (%)	n (%)	n (%)
Non GDM mothers	34 (75.6)	11 (24.4)	45 (100.0)	0.001
GDM mothers	9 (13.6)	57 (86.4)	66 (100.0)
Total	43 (38.7)	68 (61.3)	111 (100.0)
Age group
15–25 years	25 (48.1)	27 (51.9)	52 (100.0)	0.05
26–42 years	18 (30.5)	41 (69.5)	59 (100.0)
Total	43 (38.7)	68 (61.3)	111 (100.0)
Menstruation history
Regular	36 (52.9)	32 (47.1)	68 (100.0)	0.001
Irregular	7 (16.3)	36 (83.7)	43 (100.0)
Total	43 (38.7)	68 (61.3)	111 (100.0)

Bold indicates statistically significant at 5% level of significance

Multivariate logistic regression

Multivariate logistic regression was performed to find out the factors associated with depression and shown in Table 3. The final model contained three independent variables (gestational diabetes mellitus, maternal age, and menstrual history). The categories “mild’, “moderate’,’severe’, and ‘extremely severe’ of the dependent variable are merged into “depression’ for this analysis. Also, the ‘normal’ category is renamed as ‘no depression’.

Table 3.

Multiple logistic regression model along with adjusted odds ratio (OR), 95% confidence interval and statistical significance with the level of depression as a dependent variable

Variables	Adjusted OR	95% CI	P-value
GDM
No	–	–	–
Yes	32.3	6.4–162.2	< 0.001
Age
15–25 years	–	–	–
26–42 years	1.0	0.4–2.9	0.957
Menstruation
Regular	–	–	–
Irregular	0.5	0.1–2.6	0.399

According to Table 3, maternal depression was about 32 times more likely in GDM mothers (OR = 32.3, 95% CI 6.4–162.2) than in non-diabetic mothers after controlling for all other factors in the model (Table 3). Interestingly, age as well as menstruation were statistically significant (Table 1) when the association was with depression alone without controlling any other variables but came out as insignificant with the presence of other variables. So, it can be clearly said that gestational diabetes mellitus is a significant factor for depression in pregnant women. Additionally, depression can occur in all age groups irrespective of menstruation status of women.

Discussion

Finding out if women with GDM experienced higher levels of depression than women without GDM was the main goal of this study. Not only was the prevalence of depression significantly different between GDM and non-GDM women, but the average depression scores were also significantly different (86.4 vs. 24.4). The result is analogous to the study from New Jersey [29]. Similar to the current study, another study [27] found that women with GDM had mean scores that were higher (M 7.55) than those of women without GDM (M = 6.41). There was a statistically significant relationship between GDM and depression.

This study's second goal was to determine how prevalent depression is during pregnancy. In this study, the prevalence of depression was significantly higher—61.3% of the respondents reported having depression. The study's frightening conclusion was in line with another Bangladeshi study [8]. It is obvious that the time gap between the two studies is quite large and the level of depression is very high among pregnant women nowadays.

This study also aimed to identify the contributing factors. Certain studies have revealed that among GDM individuals, “increasing age” [5, 6, 30] and the “third trimester” [31] are related to increased levels of depression. In this study, logistic regression showed that women with GDM were 32 times more likely to experience depression than women without GDM, after adjusting for age and menstrual history. When age, income, and parity were taken into account, a similar analysis found that women with GDM had a 2.3—fold increased risk of depression. However, these results were not statistically significant [32]. In one study, there was a noteworthy distinction between individuals with a diploma or higher education and those with depression ratings [33]. However, in our study, factors like age and menstruation were significant predictors of depression when compared with depression alone. However, in the multiple logistic regression, they became insignificant.

Few reviewed literature have indicated that age, [34, 35] and income [5, 34] have an influence on depression. Housewives, who do not receive a wage from outsourcing, made up 89.2% of the participants in our study and were more likely to suffer from both diseases. However, this relationship did not hold strong and may not replicate the real situation as the ‘employed’ category was very small. A greater share of primigravida were found depressed whereas multipara subjects were at high risk of GDM. These results are consistent with a distinct Jordanian study [33]. According to our study's findings, women with GDM scored higher on the depression scale than women without the condition. This is in line with the findings of Natasha et al. [28], who discovered that depression was more common in GDM participants (25.92%) than in non-GDM subjects (10.38%). Additionally, compared to women without GDM, Burgess et al. [36] discovered that women with GDM exhibited higher depression scores. Further recent study focusing on the solution to this problems is greatly needed and we propose the lifestyle intervention iRT-ABCDEF program combining optical medical treatment (OMT) and healthy E(e)SEEDi lifestyle as an interesting way of remission from GDM and depression [37, 38].

Conclusion

The exploratory analysis of this study demonstrated that extreme age, irregular menstruation, and GDM mothers are at more risk of developing depression than others but the multiple logistic regression rejected all the factors except GDM. Higher GDM is an unchangeable risk factor. The study's identification of this unchangeable risk factor may aid in the development of viable treatments for the early diagnosis and treatment of depression in pregnancy. Furthermore, given that it has been demonstrated that women with GDM are more likely than those without to experience depression, healthcare professionals may wish to test GDM women for depression more frequently during prenatal care visits. The prevalence of depression in our nation is concerning. These estimates for depression and GDM could be useful in developing new guidelines for managing and preventing both conditions.

Recommendations

In our study, we found that the presence of GDM greatly increases the risk of depression during pregnancy. It is recommended to understand and act on the need to treat pregnant mothers with depressive symptoms effectively. Appropriate follow-up is essential to identify and resolve patients’ problems early on. Psychological counseling and psychotherapy may be necessary. Our final recommendation is to conduct similar studies at the national level on the largest sample to assess depression among gestational diabetes women. Educate all government and non-government institutions to conduct periodic tests for early detection of diabetes.

Limitation

The study faced several limitations, including a small sample size, which may reduce the generalizability of the findings, and a limited timeframe that restricted the ability to observe long-term trends. Additionally, the lack of funding constrained the scope of data collection and analysis, potentially impacting the depth and thoroughness of the results.

Declarations

Conflict of interest

The author declares that there is no conflict of interest.

Human rights statement

All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (American International University of Bangladesh (AIUB) ethical board) and with the Helsinki Declaration of 1964 and later versions. Informed consent was obtained from all patients for being included in the study. Respondent was given the right to participate in the study or discontinue at any time. The study's confidentiality was upheld at all times. The respondents were given full assurances that all information would be kept private and that neither their names nor anything that could be used to identify them would be disclosed or published. They were also given clear information about the confidentiality of their data. They were also informed about the fact that the collected data would only be used for study purposes.